Hands-on with the new Sony FX1

179 replies [Last post]
tom hardwick
Offline
Joined: Apr 8 1999

Well I've seen the light chaps. I've just spent all afternoon with the Sony FX1 cradled in my hands, and I'd say this - if you're about to buy a PD170, DVX100A, XL2 and so on - don't. DVC had the new Sony at their open day last Saturday.

This baby is such a delight. It is bathed in good ideas and has buttons and switches scattered all over it. A wonderful 12x Zeiss lens (f1.6 to f2.8 4.5mm to 54mm) feeding big 1/3" 16:9 chips, an excellent top mounted screen
that doesn't obstruct any of the controls, no silly memory sticks or sepia bevel multi segmented fades. (I made that up). I loved the balance, the weight, the price, the facilities, the way so many aspects can be configured
to suit you and the way you film. You could even switch the Steadyshot (not VAP BTW) into 4 different modes - one designed for wide-angle converter use.
You can set zoom speeds with gentle acceleration/deceleration curves built in. You have an aperture control wheel that's light years better than the silly wonky wheel we have on the VX/170.

Oh I love this beast. I simply can't imagine why (when they have no competition) Sony are selling it so cheaply - it's the price of the PD170 a year ago. The pictures on a HiDef JVC monitor were breath-taking, and switching it to DV mode was interesting too. The pictures looked a lot more 'liney' and the superb lens is able to push the DV envelope to bursting.

I was shown A4 prints taken from the 50i timeline that looked superb. And it runs off normal NPF batteries and records onto bog standard Sony Premium Mini DV tape - for 62 minutes in either mode.

Amazing - I just love it. Those pictures are wonderful to behold, and Richard Payne had captured some strikingly good shots out in the sunshine. There's no going back chaps.

tom.

PaulD
Offline
Joined: Aug 31 2002

Hi
Did you find a dark corner to check out the low-light performance?

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Tom, the prints were from a Thomson Viper, a "proper" 1920x1080p 4:4:4 camera, I showed them as a reminder that stills from video needn't look like tv pictures. I'm pretty sure I can get a profile to optimise the FX1 performance, but I haven't tried to do that yet and there's no way it can be as good as the Viper). Professionally, I've not yet been asked, and I'm not sure I will be until the pro version launches (although I don't expect picture quality to be very different).

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003
Quote:
Originally posted by PaulD:
Hi
Did you find a dark corner to check out the low-light performance?

We all saw on HD monitors "live" output from the camera in the meeting room - not too dark, but "nasty" fluorescent lighting, and just the sort that previous consumer cameras would not cope with very well. The FX1 seemed in a different league, though a side by side test is really needed.

tom hardwick
Offline
Joined: Apr 8 1999

Thanks for the update on the stills Alan - I was just coming in through the door as they were being passed around, so missed your intro. As to the low light capabilities (which of course sells the PD170 and VX2100 in droves at the moment), Sony claim 3 lux. As a reference they claim 1 lux for the PD170 and 7 lux for the PDX10 (TRV950).

My guess is we'll have to accept more gain-up in the gloom, but my hope is that the chip output will be quieter and grain will be no more of a problem. It's going to make a fascing side-by-side comparison though. But then as Richard Payne said, if it gets dark on goes the Pag light, and I'm right with him there. As photographers the first thing we need is light. Cameras and inspiration come second and third on the list.

tom.

cstv
Offline
Joined: Jul 26 2002

I couldn't agree more strongly with Tom about the images that this camera pumps out, stunning! I'd been using a single chip Sony TRV20 for about 4 years when (back in june) i got my first propper play with a PD150 and was amazed! That said, this thing looks even better! :D

However... i wouldn't buy this because it lacks XLR inputs and by the time Sony release the version with XLR inputs i'd struggle to justify investing in something that still records to tape.
Then there's the problem of the zoom controls. They're all far too slow! Yes, slow zooms are useful, but not at the expense of being able to jump in and out really quickly. Even the manual zoom on the lens barrel is seriously damped. It's better than the PD150's zoom controls though, in my opinion...

It also has some fantastic programmable controls for zoom, focus and exposure. You set an A and B point for any of those three, press a button and it does the rest! wonderful! I've no doubt that pro camera ops are very good at doing 17 things at once, but i just can't do it, so this is a gift!

mark.

tom hardwick
Offline
Joined: Apr 8 1999

You say that you'd struggle to justify investing in something that still records to tape Mark, yet the very backwards compatability of this product is just what it needs to gain market acceptance. And just this week I had an experienced video man come up to me to say he'd bought a DVD camcorder because the quality was better than from tape. Sigh.

When Sony introduced the VX1000 and 700 models in 1995 they were heralding in a whole new tape format, and they've not always been good at this (Betamax and El Cassette and Micro DV to name but three). So the Mini DV tape had to be damn good, and was - lifting the standard definition format by a whole quantum wallop.

This FX1 uses the tried and tested Mini DV tape drive mechanism, and such decks have been pumped out in their million these past 9 years. Interestingly the tape drive is now back on the LHS of the camcorder - where it started off in the VX1000 all those years ago.

tom.

cstv
Offline
Joined: Jul 26 2002

well, firstly, you couldn't pay me enough to use a DVD camcorder. I really don't see optical discs as a good aquisition format, particularly the 12cm falvour, and hard drives are probably an accident waiting to happen. Solid state on the other hand...

my coments weren't ment as a dig at MiniDV, i love it! but it's the part of editing that i hate! It's just so frustrating knowing that i've got all of the data i want on 5 MiniDV tapes and a MiniDisc, all of which has to be captured in real time... if it had all been shot on solid state then i could just dump it straight on to my PC for editing.

You say that this product needed backwards compatibility, but why? Are the sort of people who are going to buy this really going to have a stack of MiniDV tapes laying around that they just can't do without? Even then, it would only be a problem if they were replacing a camcorder rather than just adding to a collection ;) And if you were buying as a replacement, i think the advantages of solid state over tape would be worth the £300 or so for a cheapo MiniDV camcorder to use as a deck for a MiniDV archive. Although as Richard said, the R&D costs are greatly reduced by using the existing MiniDV tape mech.

I didn't see where the tape access was, was it that bit where the LCD would've been? or do you mean the other left as if you were looking into the lens?

mark.

PaulD
Offline
Joined: Aug 31 2002
Quote:
Originally posted by tom hardwick:
....But then as Richard Payne said, if it gets dark on goes the Pag light, and I'm right with him there....

Sigh...
Sure you can video the signing of the register in the vestry with a PAG light, but it will do çod all to light the main body of the church....

tom hardwick
Offline
Joined: Apr 8 1999

I didn't see where the tape access was, was it that bit where the LCD would've been?

That's right Mark.

it will do sod all to light the main body of the church....

That's right Paul. Every time I'm using f1.6 and +6dB of gain-up with the VX2000 I thank my lucky stars I'm not shooting with anything else at this price point.

tom.

infocus
Offline
Joined: Jul 18 2003
Quote:
Originally posted by cstv:
well, firstly, you couldn't pay me enough to use a DVD camcorder. I really don't see optical discs as a good aquisition format, particularly the 12cm falvour, and hard drives are probably an accident waiting to happen. Solid state on the other hand...

There are optical discs and optical discs..... UK broadcasters are starting to buy XDCAM blue laser equipment, and it looks like becoming the new Beta SP for the lower end broadcast market, with HDCAM for the higher end. The reasons are precisely to get over the irritatons you mention regarding transfer to the PC - they may be irritations to you, in news they may be the difference between getting a breaking story on air or not! It looks like at the moment P2 is losing out in the broadcast market, and that doesn't really surprise me, since for a big organisation "what-the-pictures-are-recorded-onto" needs to be a consumable item, rather than a capital item. On the other hand, solid state may well be the way to go for amateurs and small businesses who would have the control over where their shooting media goes.

When the XDCAM was demoed to me, I also expressed some reservations over the size of the media - a 12cm disc in caddy and case is a lot bigger than a 30 min DVCAM tape. I understand Sony are aware of the issue and there will be/are 8cm versions, hopefully of pocketable size!

As far as HDV goes, I don't think it should be seen as anything like the end of the game. My feeling is that it was a convienient way of getting a relatively cheap HD camcorder onto the US and Japanese markets quickly, and with not too much initial investment. In those countries, with HD well established, it was being screamed for, in the UK it seems to have just appeared out of nowhere, and could well play a big part in HD adoption.

What we may well see eventually is another physical format appearing capable of recording the HDV bitstream. You may have three cameras (tape, blue laser and solid state), but to an edit system they all appear identical - except perhaps for download speed.

To Tom - I agree wholeheartedly with everything you've said in this topic, with one exception, and that's to do with the balance. Here I feel very strongly that Sony could have learnt a lot from the Canon XL series, and having the eyepiece viewfinder at the rear of the camera is a bad mistake. Apart from being less comfortable, it would make it unnecesarily difficult to use in confined spaces (compared to an XL2 type design) - the FX1 is a very long beast. If you can help do anything to bring this thought to Sonys attention in time for the development of the next generation, you'll be doing us all a favour!

That moan over, the XL2 doesn't have a fold out screen at all, let alone being SD only, and if I was intending to buy a camera in this price range tomorrow, I don't really think the FX1 has a serious competitor. I'll be interested to hear how it compares side by side with other cameras in low light, and think it's worth emphasising the wide angle coverage of its lens - 4.5mm equates to a 9mm for 2/3in chips, very respectable for a camera of this class.

tom hardwick
Offline
Joined: Apr 8 1999

Ah, so now I know who Infocus is. And yes, I could follow your logical argument to have the v/f up alongside the fold-out screen, where the XL2 has it. But as I invariably use my VX2k with the side-screen open and a big and hefty L bracket to stabilise my hand-held work, I'd envisage using the same bracket under the FX1. Used like this the FX1 can be pressed into tight corners, used at ground level and held up on high over spectator's heads.

The focal lengths chosen by Sony exactly match those chosen by Panasonic for their Leica-badged DVX100A. Of course the latter feeds 4:3 1/3" chips, so the Sony may well see wider. The Panasonic wide-angle relates to a respectable (though not dramatic) 32.5mm in 35mm still film camera terms.

tom.

cstv
Offline
Joined: Jul 26 2002
Quote:
Ah, so now I know who Infocus is

i was thinking the very same thing...

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

The tape compartment is on the left, where the lcd usually is. Top loading.

I reckon all the manufacturers are missinjg a few tricks, like being able to move bits of the camera around. I mentioned one trick on Saturday, having the lcd removable so that you can walk away from the camera with it, a few buttons and you get remote tape control as well (e.g. by IR link). Why not? Sharp have gone part way but not far enough. And why not have the monocular finder shoe-mounted, so you can have it at the back or the front of the camera? Big, proper cameras have separate finders that you can demount and with a bit of fiddling use remotely, why not on consumer kit? The consumer monocular finder is already a separate item (hinged but attached), it wouldn't take much to make it removable.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003
Quote:
Originally posted by Alan Roberts:
And why not have the monocular finder shoe-mounted, so you can have it at the back or the front of the camera?

As long as you can get the front mounting position correct and sturdy, that's not a bad idea, though will it involve a bulk/weight penalty? Tom, would you see anything wrong with the monocular where I suggest, and the LCD screen in the usual place? In general, I find the LCD little use in bright light, and with this type of camera would generally use the monocular for handheld use, the screen when on a tripod - in which case having it towards the rear makes sense, more like big viewfinders on OB/studio cameras.

Theres a lot the consumer market could learn from the pro, and vice versa. Why has it taken so long for flip out screens to appear on pro cameras, even if only for quick playbacks on site? Why is it so difficult to get to the end of recorded material on a DSR500 after playing back, when a consumer camera has a simple "end search" button?

The FX1 sits very interestingly between these two markets - engineering wise they seem to have seen the light with features of real benefit rather than "silly memory sticks or sepia bevel multi segmented fades", and I can only hope that come the next generation a similar thing happens ergonomics wise. The FX1 is larger than most other consumer cameras and needs to be styled far more along pro lines.

But overall it really is stunning, currently without competition for the price, and now it's been proved that it is editable in a mainstream editing application, then as Tom says, "There's no going back chaps."

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Yep, that's my view as well, the FX1 is the first of a new breed. For years now we've been getting new features on consumer kit first, it's taken ages for the flip-out to get into the pro game, possibly because pro-programme-makers would never use it except for reviewing. But I think they're wrong, they should try everything out, any new trick is worth a look, the flip-out lcd was revolutionary when it happened, it converted consumer cameras from boys toys into useful tools, I rarely use the monocular finder these days.

In Sony, though, there is a lot of interia because the individual factorties/divisions are autonomous, they don't talk to each other. With luck, that'll change, because products from each divisison are now being used in the "wrong" arena, and they're starting to notice. The fact that the FX1 is aimed at the consumer market didn't stop them talking to broadcasters about what it should do (and they listened, even if they ignored some of it). That's why I'm keen to se the pro version when it gets to us.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003
Quote:
Originally posted by Alan Roberts:
I rarely use the monocular finder these days.

Even on bright sunny days????

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

I don't meet that many bright sunny days :rolleyes: My DX100 lcd's quite bright anyway; the only problem I have in summer is that my glasses go really dark so I have to take them off to see what the camera's doing, and then I have to use the monocular vf because I'm very long sighted. But I don't oftenh want to take pictures from eye height anyway, I much prefer to be able to change the height quickly, and I can't do tyhat using monocular vf.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

tom hardwick
Offline
Joined: Apr 8 1999

Yes, even on bright sunny days.
Not long ago I was asked to shoot a video of a 150 guest surprise birthday party and it's in these 'unplanned' situations that you need eyes
to left and right, in the back of your head, as well as on the viewfinder. Small children ran under my feet, adults sang and got drunk, unexpected things happened at odd times and I was expected to capture it all.

Lots of you will know of my love for the side screen on modern camcorders. With a conventional viewfinder I'd have been as effective as a one eyed man with tunnel vision. I'd have had the viewfinder bashing against my eye, I'd have tripped over wriggling kids and I'd have missed a lot of what was going on around me.

The big and sharp side screen on modern camcorders is a Godsend to anyone in such a situation. I can track forwards and backwards through the crowds with the camera while making sure I don't trip over small people; I can look left and right to ensure I'm not missing interesting action happening out of frame and I can get much more fluid 'steadicam' footage by having my slightly bent arms take the shudder out of my walk. In such a crowded situation it's also a lot safer not having a 1.5kg lump of machinery being an extension of a valuable eye.

With the camera at my eye (Super8, Sony TR2000 Hi8) people were very aware that they were being filmed and many reacted accordingly, becoming unnatural and even loosing the thread of what they were saying as I approached. Using the side screen and having the tally lamp permanently off I'm able to film with what appears to be a very much more casual attitude and this encourages those being filmed to remain calm and behave naturally. Of course I'm far from filming casually in such a situation, it's just an impression I've perfected to enable me to film the real world without disrupting it too much. The very act of carrying a camera tends to distort reality and in my view is a big plus point for the modern compact Mini DV over a shoulder mount, for instance.

But in sunlight all is not well and the side screen struggles. Pumping up the brightness using the little screen switch or menu option is not the answer. You lose the screen's ability to tell you about the image exposure and sharpness, and this is too great a sacrifice. I'm so keen on using the side screen that I've invested in a flexible Hoodman and can highly
recommend it. They are available to fit a huge variety of screen sizes and some come with built-in magnifiers, too. For those that use anamorphics for wide-screen work there are hoods with built-in cylindrical elements to correct the screen distortion. It's such a simple idea yet so effective, and just recently I used it while filming a school play. Why use such a device indoors when the screen contrast needs no enhancement? Well, for purely selfish reasons. I didn't want the audience sitting around me to be distracted by my bright viewfinder screen, and I didn't want them seeing that it wasn't their beautiful child I had in close-up. I wanted the finished DVD to be a surprise.

So to the Hoodman. This is a fold flat, near weightless rectangular tube that when opened up is very stiff laterally. This is held on by elasticised Velcro straps that quickly fit around the back of the side screen and grip the underside of the hood. Four self adhesive Velcro strips stick around the perimeter of the screen and these immediately grab and hold onto the hood as it's offered up. Assembly and adjustment is a matter of seconds, as it should be. There's no need to adjust the brightness of the screen so no mental adjustment need be made as you film. The Hoodman is stiff yet soft,
beautifully made, well thought out and in my view worth every penny. Why? Because it enables me to continue shooting using the side screen, and I now know that I make better movies when I don't have a camera strapped to my forehead.

Sorry, I've rambled on a bit.

tom.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Guys, I just got this from my usual sources, it's Son'y press release on the professional version.

==================================================

Sony confirms pricing, delivery date & features for first professional HDV camcorder and VTR

Basingstoke, 10 November 2004 - Sony has confirmed that first European deliveries of the new HVR-Z1E HDV Professional Camcorder - priced at SRP €5,500 in EU Countries - will commence from February 2005, with over 40 additional benefits (compared with consumer version) for professional users.

Joining the existing DVCAM range, the HVR-Z1E offers users an affordable migration path from Standard Definition whilst retaining the qualities of the popular DVCAM range such as ease of use and i-LINK (IEEE1394) connectivity. Designed for the professional user, the HVR-Z1E incorporates a number of significant features over consumer oriented products and includes:
* Switchable recording and playback of HDV, DV and DVCAM formats, providing the professional user with the flexibility to choose the format that suits the need of their production.
* 50/60Hz (PAL/NTSC) switchable recording and playback
* 2-channel balanced audio XLR inputs, with independent adjustment of input levels for each audio channel
* Colour and B/W switchable viewfinder
* Simultaneous operation of both LCD panel and viewfinder, giving professional users the opportunity to focus shots accurately via the viewfinder while simultaneously framing images via the fold-out LCD panel
* And many more!

In addition to the professional HDV camcorder, Sony will also introduce a compact Video Tape Recorder - The HVR-M10E. Like the camcorder, the VTR supports the recording and playback of DV, DVCAM and HDV formats and incorporates a 16:9 colour LCD viewing screen. The compact VTR also benefits from both mains power and battery operation and is priced at SRP €3,600 in EU Countries.

Emphasising the format's suitability for low-cost HD migration, both the HVR-Z1E and the HVR-M10E feature HD/SD switchable operation (offering full 1080i resolution), plus on-board downconversion to output SD or HD signals. For even greater production flexibility, the professional camcorder also supports switchable 50/60Hz PAL/NTSC recording and playback.

HDV is attracting fast-growing industry backing. The format has already gained the support of leading software developers including Adobe, Apple, Avid, Canopus, Pinnacle Systems, Sobey, Sony Pictures Digital Networks and Ulead.

The HVR-Z1E and HVR-M10E are aimed at professional videographers, documentary makers and feature film makers working to tight budgets, HDV is also an attractive option for mainstream broadcasters and corporate programme makers needing to acquire HD content in restricted or hard-to-reach locations and HDV also serves as an ideal source for contributing HD content into an HDCAM production environment. The format's quality, versatility and low operational costs also extend its appeal to educational establishments and hire companies.

"Putting the HDV format to one side for a moment, my first impressions of the Z1E is that it's a great camera. I don't see the need to shoot anything in SD anymore. The variations between the consumer and professional versions of the camera are such that it really is worth waiting the few extra months - and paying a premium." Kevin Cook, Executive Administrator, Institute of Videography.
ENDS

The World's First Ever* HDV1080i consumer camcorder was announced (by press release) on 7th September 2004. Find out more information by visiting: www.sony-europe/presscenter
* According to Sony research, September 2004

For more information please contact:
Claire Thatcher, PR Manager, Sony Europe
Tel: +44 (0) 1256 483 156
Mob: +44 (0) 7786 115 072
Email: claire.thatcher@eu.sony.com
www.sonybiz.net/HDV

About Sony:
Sony manufactures audio, video, communications and information technology products for the global consumer and B2B markets. With its music, pictures, game and online businesses, Sony is uniquely positioned to become a leading personal broadband entertainment company in the 21st century.

In Europe, its business division is now a leading supplier of broad horizontal communications solutions as well as magnetic and optical storage solutions. Amongst others, Sony Business Europe's market segment solutions and service & support packages are targeted at organisations in the Healthcare, Media, Retail, Large Venues and Transport sectors. Device components such as camera sensors and modules, displays, batteries and semiconductors are targeted at OEM manufacturers and system integrators. For more information please visit www.sonybiz.net/media.

Full Review by Kevin Cook, Executive Administrator of the Institute of Videography:

"Putting the HDV format to one side for a moment, my first impressions of the Z1E is that it's a great camera. A true 16:9/4:3 switchable solution at this level would be a good enough news, but the added benefit of being able to record everything in stunning HD quality - and the ability to easily down-convert this to SD out of the back of the camera or CTR is going to be a winner amongst the pro-video community.

There is a growing feeling that shooting everything in HDV mode, and then down-converting to SD produces better pictures than if the material was shot in SD in the first place. I was dying to try this out and have to say that, without the benefit of measuring the output, to the naked eye the down-converted images do look richer and more detailed. I don't see the need to shoot anything in SD anymore.

The variations between the consumer and professional versions of the camera are such that it really is worth waiting the few extra months - and paying a premium. Features such as DVCAM and DV recording, Hyper Gain, XLR audio connections and individual audio channel control might have been enough. But, the additions of PAL or NTSC recording mode, filmic effects settings, audio limiter and noise reduction - plus a more varied choice of component outputs when you are able to fully appreciate the quality offered by HDV - are what will make this camera a firm favorite in the pro-video community.

It's a HDV future - in a product for today."
3
=================================================

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

RayL
Offline
Joined: Mar 31 1999

Alan, thanks for that. A little 'milestone' in the description is that the Z1E (only one step above a consumer camera) is PAL/NTSC switchable, which means that Sony can supply one camera design to the whole world and we, the users, can overcome another of life's little editing hurdles. Hopefully the power supply will be 'universal' (any mains voltage from 100-250v).

Ray Liffen

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Yep, that's my view as well. This is a "world" model, so we should get the benefits of manufacturing scale. Sony's policy on power supplies is equally enlightened, it'll be 100-240v, just like the power supply for my laptop, you can depend on that.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003

After reading Toms last post, I am left thinking that Alan Roberts has the right idea with his "modular approach to viewfinding" - the user has the option where to mount it. I respect Toms arguments, and if his methods work for him, fine. Certainly there are times when I would agree the side screen is the best option, but in a hazardous or hostile environment (and young children is certainly that......) I would prefer a side monocular, keeping the left eye open whilst viewfinding with the right. To me, with a correctly setup eyepiece, this has the advantage that both eyes are focused in the distance, with the camera picture "superimposed" on the real world. Using the side screen means both eyes are focused close, and I suggest this impairs your awareness of whats happening around, rather than enhances it?

Whatever, I think it's healthy that debates on the future of camera design are on issues like this, rather than on the number of "sepia bevel multi segmented fades". And if Alan talks to the right people about "modular viewfinding", we may both be happy come the next generation!

At the end of the day, it's an incredible camera, and even if the XL2 had come out with HDV, then the lack of a sidescreen AT ALL I would perceive as more of a disadvantage than the FX1's viewfinder positioning.

Jim Bird
Offline
Joined: Sep 15 2000

Hi,

In real terms, if pre-recorded HDV footage from this new Sony camcorder(s) was to be broadcast on national TV by the BBC, how would the image quality compare to the existing higher quality offerings that we receive today.

I’m of course assuming the footage was shot professionally using good studio lighting and at the other end, view on a half decent bog standard TV.

Jim Bird.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Sony say that SD downconversions from HDV recordings are not as good as, say, a PD170. But, you should be able to get pictures easily as good as a 170 by recording DV on the FX1. Of course, then you wouldn't get the HD recording. Sony say that the difference will not be visible on the majority of material, only on specific, testing subjects (and I can imagine a few test sequences that woudl show it). Inherently, I have to agree with this view because it makes sense, but in practice the difference is going to be small, and could easily get lost in the MPEG stages that take place before broadcast happens. We really don't know yet, no broadcast programme has been made with an FX1 in the UK (yet, but it won't be long now )

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Alan McKeown
Offline
Joined: May 9 2001

“Sony say that SD downconversions from HDV recordings are not as good as, say, a PD170.”

“There is a growing feeling that shooting everything in HDV mode, and then down-converting to SD produces better pictures than if the material was shot in SD in the first place. I was dying to try this out and have to say that, without the benefit of measuring the output, to the naked eye the down-converted images do look richer and more detailed. I don't see the need to shoot anything in SD anymore.”

Those two quotations appear to contradict each other??

Alan

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Indeed so, that's why I said "Sony say that...".

It all depends on what you mean by "better pictures". Shooting with an HDV camera gets you a better lens, and a sampling system that nicely oversamples in the camera, so you should be able to fill the DV bandwidth more neatly than by using an SD camera (i.e. not needing so much detail enhancement). The downside is that DV is designed to be fed with an SD-sourced signal, not with a top-quality HD picture downconverted (I've got a tape showing what happens when you feed really good HD into DV through a really good downconverter, it looks horrid, I showed it on Saturday at DVC). However, the HDV cameras are priced above SD equivalents and it shows, they have extra controls and abilities that let you tweak them to get pictures that look better anyway (i.e. better highlight/lowlight handling). So I can handle the argument either way.

I'd expect a sequence shot using an HDV camera to have better pictures, looking nicer, less ringy, better contrast. But, in some circumstances with very "busy" pictures, the downconverted pictures may not look so nice because the DV compressor is being "overdriven"; that's the circumstance that Sony are worried about, and so am I.

All that make sense now?

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Gavin Gration
Offline
Joined: Jul 29 1999

PAL NTSC Switchable - nice!

Dragonslayer
Offline
Joined: Jan 27 2004

Just found this, much more comprehensive then the press release above, tons more features listed and ALL of the differences between the consumer and this new pro Sony cam, great pics too:

http://www.dvuser.co.uk/Main%20pages/what%27s%20new%20pages/sony%20HVR-Z1E.html

This is my Signature so don't take offense:

"The box said Windows 98 or better - so I bought a Mac"

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

The features list has pretty well all we asked for.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Christian Lett
Offline
Joined: Apr 26 1999

Is that a proper focus ring on the HVR-Z1E (rather than the infinite rotate ones we get on every other camcorder)?

Christian Lett After Effects and Maya Artist www.quarterlightpictures.com

tom hardwick
Offline
Joined: Apr 8 1999

It's an infinite-rotate for the focus but the zoom has limit stops. You can twist that zoom ring as fast as you like, but all you'll do is engage the highest LANC protocol speed, and it'll lag somewhat behind you.

tom.

tom hardwick
Offline
Joined: Apr 8 1999

I'm so pleased to see that the Z1 version of the FX1 retains the on-board stereo microphones (like the DVX100A). It's something that I've really appreciated when using my VX2000 + Sennheiser ME66 ~ the ability to pull out that little mini-plug and let the on-board stereo mics take over for background (stereo) chit-chat.

tom.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

It's the infinitey rotating one.

There's a good reason why we get controls like this on consumer cameras, power consumption. The servos required to move the optical components are quite small because they're light and don't have to move far. But the external controls have to be human size, and that means big. So, if you want a physical focus control you need a much bigger servo bevcause it has to move the external control as well as the innards. Not only that, but it needs to have a torque limit so that it doesn't break your wrist if you try to stop the auto, and the barrel has to be stringer to accommodate all this. So they take the natural step of using small, light, low power servos, and making the manula control supply a guiding voltage to it, rather than shifting the components directly.

This way, it all gets cheap enough, light enough, simple enough, and sufficiently low poer to run on the small batteries we're used to.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Christian Lett
Offline
Joined: Apr 26 1999

Shame. I see your point Alan but auto-focus SLR still cameras have been doing this for ages using either an on-camera servo or an in-lens mechanism (Canon).

I guess on pro-pro-cameras there's no auto-focus (?). It's just really hard to do a decent pull-focus with an infinitely rotating focus ring, and when you're shooting scripted programmes I always use manual focus :(

Thanks,

C

Christian Lett After Effects and Maya Artist www.quarterlightpictures.com

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

All agreed. It's simply down to power consumption, thatr's what all the manufacturers say every time we ask for real controls. And I've never seen auto-focus on any broadcast-quality lens.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Alan McKeown
Offline
Joined: May 9 2001

“The downside is that DV is designed to be fed with an SD-sourced signal, not with a top-quality HD picture downconverted”

I don’t understand why DV coding or decoding need be involved.

For example, Jim’s question:

“In real terms, if pre-recorded HDV footage from this new Sony camcorder(s) was to be broadcast on national TV by the BBC, how would the image quality compare to the existing higher quality offerings that we receive today.”

There is surely no need for DV coding to be involved in this case?

Is the problem, if there is one, that the down-conversion to SD should preferably precede any compression process rather than come after it? That would not be possible if the HDV were first recorded to tape via MPEG-2 compression.

That said, the MPEG-2 decoded HDTV pictures from my set-top-box satellite receiver look fine when down-converted to SD. Better picture quality than BBC SD MPEG-2 broadcasts!

Alan

Richard Payne
Offline
Joined: Sep 15 2000

I would have thought if there was 'too much' information supplied to a compression codec the quality would suffer and wouldn't this be the case with HDV-DV down-sampling ? Its just like when you digitise from poor, noisy VHS, all the noise just acts like more detail to the compressor.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

OK, I should have said "SD compression". Stuffing high levels of high frequency detail into any SD compressor will stress it. If the down-conversion doesn't limit the hf content, you can end up with more detail than the next compressor can handle, and that's what I showed last Saturday.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Alan McKeown
Offline
Joined: May 9 2001

I think it would be a shame if anyone were to use this camcorder in anything other than its genuine HDTV MPEG-2 recording mode.

If SD compressors are not capable of handling legal high frequencies, that’s their fault - the camera is not to blame.
It is up to the SD compression to cope with a high quality input signal, not for the acquisition system to be “dumbed down”.

As far as SD broadcasting is concerned, this problem of “over stressing” a compressor must also arise when HDTV cameras are used for acquisition and the programme is broadcast in SD (as was this year’s “last night of the proms”).

Alan

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

My point is that the industry-standard DV compressor is dumbed down; I've got a tape that proves it. But that doesn't imply that the FX1 overloads the compressor in down-conversion mode, only that it is possible to overdrive DV.

All I'm pointing up is that when dealing with differing standards, you have to be careful with the content, because it can all go pear shaped. That does not mean the FX1 gets it wrong, only that getting it wrong is a possibility. I've evry confidence that the FX1 gets it right, most of the time. I have not seen any evidence that the FX1 gets it wrong.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

steelej
Offline
Joined: May 11 2001

How does the quality of the FX1 compare to something shot using a pro DV cam like the DSR-390 or DSR-570?

John.

Alan McKeown
Offline
Joined: May 9 2001

If a standard definition (SD) camcorder’s DV recorded output was up-converted to 1920 * 1080 , so as to make a fair comparison with the FX1’s MPEG-2 recorded output, then I would be surprised if the FX1 was not the clear winner for picture quality.

The pictures being displayed on an HDTV monitor, of course.

Alan

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Alan, fully agree. Because the FX1 starts with an HD picture while the SD camera doesn't. Bit obvious, that one. But the real issue is whether the FX1 makes better SD pictures than an equivalent SD camera, either recording SD origination or down-converting HD recording. I don't know the answer to either question yet. The point I'm labouring is that it is possible to overdrive any compression system, and I've got a tape showing that happening when a HDTV signal was down-converted and saved on DV tape.

John, don't know.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003
Quote:
Originally posted by steelej:
How does the quality of the FX1 compare to something shot using a pro DV cam like the DSR-390 or DSR-570?

John.

Theres rarely a simple answer to a simple question..... and just what is meant by "quality"? If we compared material shot in ideal conditions (plenty of light, not too much contrast etc, etc) I think the FX1 would perform very creditably, especially if the HD mode could be taken advantage of. But put them both into dim lighting, such that even the DSR570 would need full aperture and gain, and I suspect the FX1 would noticeably show signs of struggling. At the end of the day you can't change the laws of physics, and 2/3" chips will be more sensitive than 1/3", all else being equal. And do you class depth of field issues as a "quality" argument? It's easier to throw a background out of focus with a 2/3" camera - can that be classed as a "better quality picture"?

But, yes, I'd love to do a side by side comparison of them both......... and lets not forget the relative costs.

Alan McKeown
Offline
Joined: May 9 2001

Quote:
“It's easier to throw a background out of focus with a 2/3" camera”

SDTV is best viewed from about 6 picture heights (6H) and HDTV from about 3H. Consequently the diameter of the “disc of confusion” defining the depth of field limits for HDTV should be about half that for SDTV.

Does this not effectively mean that 1”/3 chips in an HDTV camera (used with an HDTV system), will give a similar depth of field to 2”/3 chips in an SDTV camera (used with an SDTV system), for the same F stop?

Alan

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

What's important to the viewer is the overall magnification of the image. That's why HD with double resolution can be viewed from half distance of SD.

All other things being equal, a camera with 1"/3 sensors will have double the DoF of one with 2"/3 sensors, because the magnification is higher. If you now view the image from the 1"/3 sensors at half the viewing distance, the DoF doubles again because you can resolve the smaller pixels. So a 1"/3 HD sensor provides 4 times the DoF of a 2"/3 SD sensor when viewed at the "correct" viewing distances.

At least, that's the way I understand it.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Alan McKeown
Offline
Joined: May 9 2001

“All other things being equal, a camera with 1"/3 sensors will have double the DoF of one with 2"/3 sensors”

Agreed.

“If you now view the image from the 1"/3 sensors at half the viewing distance, the DoF doubles again because you can resolve the smaller pixels.”

So, on that basis, if we were to view the image from 100 times the original viewing distance (600H), the depth of field would be very, very, small.... because we can hardly resolve anything at all???

Surely the depth of field, in the 600H case, is very large and the focusing accuracy of the camera scarcely matters since we are viewing the image from so far away?

Alan

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Oops, yes, of course you're right. Sorry, wasn't thinking straight.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Nigel Longman
Offline
Joined: Apr 28 1999
Quote:
Originally posted by infocus:
...At the end of the day you can't change the laws of physics, and 2/3" chips will be more sensitive than 1/3", all else being equal...

That's an interesting thought. At first glance it seems perfectly reasonable that doubling the dimensions of the sensor should increase the sensitivity. But if you are forming an identical image of the same scene on both sensors, won't the light energy/unit area on the larger sensor be lower than that on the smaller sensor, thereby effectively taking you back to where you started from, albeit with greater resolution being available from the larger sensor?

I wonder :confused: .

Regards NL

Alan McKeown
Offline
Joined: May 9 2001

Quote:
“But if you are forming an identical image of the same scene on both sensors, won't the light energy/unit area on the larger sensor be lower than that on the smaller sensor, thereby effectively taking you back to where you started from, albeit with greater resolution being available from the larger sensor?”

The 2”/3 sensor will require a lens of twice the focal length of that for the 1”/3 sensor, for a similar picture. For the same F-stop in both cases, the 2”/3 sensor lens will have an aperture of twice the diameter of the 1”/3 sensor lens, giving the larger aperture an area four times that of the smaller.

Hence four times the light energy per unit time will fall on the four times larger area of the larger sensor. So the light energy per unit time falling on unit sensor area is the same for both sensor cases.

Alan

womblingfree
Offline
Joined: May 16 2001
Quote:
Are the sort of people who are going to buy this really going to have a stack of MiniDV tapes laying around that they just can't do without? Even then, it would only be a problem if they were replacing a camcorder rather than just adding to a collection ;) And if you were buying as a replacement, i think the advantages of solid state over tape would be worth the £300 or so for a cheapo MiniDV camcorder to use as a deck for a MiniDV archive. Although as Richard said, the R&D costs are greatly reduced by using the existing MiniDV tape mech.[/QB]

I've got a stack of of miniDV's I've accumulated over the last eight years. I have a limited budget and am upgrading.

Sounds like this is the perfect camera for me.

I go to my local Bic Camera store to gaze in amazement at the projected FX1 image on an almost daily basis.

I would like to see the death of tape, not because of quality issues but because I've had so many broken tape drives, mangled footage etc, over the years. Rare, but can cause irreversible loss of hard shot footage.

The less moving parts the better. No better alternative on the horizon though as far as I can see.

You'd need a flash memory card as big as Blackpool Tower to record HD footage.

Maybe in the future some kind of Blu Ray, Hard Drive combination?? More likely some new technology.

Mini DV it is then for now.

cstv
Offline
Joined: Jul 26 2002

broadcast must be interested in the Z1 if only for it's 16:9 capability, which is important here and now with our existing tv system, before they even think about archiving for the future or selling to 3rd parties. I think that given the choice, very few people would choose a PD150/170 over the Z1 if the FX1 is anything to go by...

mark.

infocus
Offline
Joined: Jul 18 2003
Quote:
Originally posted by cstv:
broadcast must be interested in the Z1 if only for it's 16:9 capability...........I think that given the choice, very few people would choose a PD150/170 over the Z1 if the FX1 is anything to go by...

Well - I feel the same way you do, but wonder if people will be quite as logical about it..... To me the Z1 technically has three core arguments in it's favour over the cameras you mention - HD, true 16:9 and a degree of controllability thats previously not been available outside true pro cameras. But I can see potential purchasers thinking of no current need for HD, can't be bothered with 16:9, and.... maybe I can save a few hundred quid?

But there's one aspect that hasn't been considered, and that's resale value. I expect the new offerings to hold their value far better than new PD150/170s in the coming few months, and think that that PD170 "bargain" might not seem so good come the time to trade up. Of course, SECONDHAND PD170 etcs might become VERY good buys if HD etc isn't needed?

Regarding the discussion about the DOF point I made, I intended the point quite generally, rather than as cross comparisons between SD/HD. It was an attempt to query just what exactly is meant by the term "quality". Alan McKeown posted "Does this not effectively mean that 1”/3 chips in an HDTV camera (used with an HDTV system), will give a similar depth of field to 2”/3 chips in an SDTV camera (used with an SDTV system), for the same F stop?" - and technically I suspect he's right.

But I feel AESTHETICALLY what matters is how well objects are distinguishable in a background (and hence distract). In this case, something just "out of focus" in HD may be noticeably less sharp than a foreground item, but be as equally noticeable (and objectionable) as in SD, all else being equal. Hence 2/3" chips are still aesthetically preferable.

Alan McKeown
Offline
Joined: May 9 2001

Quote:
“Regarding the discussion about the DOF point I made, I intended the point quite generally, rather than as cross comparisons between SD/HD. It was an attempt to query just what exactly is meant by the term "quality". Alan McKeown posted "Does this not effectively mean that 1”/3 chips in an HDTV camera (used with an HDTV system), will give a similar depth of field to 2”/3 chips in an SDTV camera (used with an SDTV system), for the same F stop?" - and technically I suspect he's right.

But I feel AESTHETICALLY what matters is how well objects are distinguishable in a background (and hence distract). In this case, something just "out of focus" in HD may be noticeably less sharp than a foreground item, but be as equally noticeable (and objectionable) as in SD, all else being equal. Hence 2/3" chips are still aesthetically preferable.”

I would speculate that what correlates with a viewer’s sense of the degree to which an image appears “out of focus” is the magnitude of the plane angle that each blur circle diameter subtends at the viewer’s eye.

If this assumption is correct, the closer the viewer is to the screen (the distance being expressed as a multiple of the picture height), the larger will be this angle and the more the image will appear blurred.

In other words, the HDTV 1”/3 sensor would be subjectively the equivalent of the SDTV 2”/3 sensor, not just for the depth of field region but also for the more out of focus background region.

[The angle determining the depth of field limit is about 300 micro-radians in my case (wearing spectacles)].

Alan

infocus
Offline
Joined: Jul 18 2003

I agree that for a given size sensor, for HD the depth of field will be shallower than for SD, all else equal - that must be true, since we're working to a smaller circle of confusion. An object just behind the plane of focus may appear in focus in SD, but noticeably soft in HD, compared to an object on the plane of focus.

But imagine it's (say) a sheet of text, and if readable it's distracting, if illegible it merges nicely into the background. In HD it may well look soft compared to the foreground, but (because of the improved resolution) still be legible - and hence distracting. All else equal again, the only way to throw it far enough out of focus to render it illegible (and hence aesthetically satisfying) is to use bigger sensors.

I believe for cameramen coming from 35mm, the greater depth of field of 2/3" is one the things they least like about HD. Of course, for fast action etc the opposite may hold true, when big DOF is a positive advantage.

From your post on the 18th, this would seem to indicate a four fold increase in light per pixel for 2/3" against 1/3" (each with the same resolution), and for the fewer number of pixels/chip of a DSR500 v an FX1, I'd expect roughly another doubling per pixel due to the larger pixels. Hence, if other aspects of the technology are equal, and the same aperture lens, I'd predict a DSR500 to have something like a 3 stop advantage over an FX1. It would indeed be interesting to do a practical test.....

tom hardwick
Offline
Joined: Apr 8 1999

That would probably equate as the Sony spec for the DSR500 is something like 0.31 lux as against the FX1's 3 lux.

Jim Bird
Offline
Joined: Sep 15 2000

Hi,

Am I correct in assuming that, someone needs to do a side by side comparison test with the new FX1 and some other Sony cameras which produce excellent low light pictures before we will know for sure how good the FX1 will perform in low light situations?

Jim Bird.

Nigel Longman
Offline
Joined: Apr 28 1999
Quote:
Originally posted by tom hardwick:
That would probably equate as the Sony spec for the DSR500 is something like 0.31 lux as against the FX1's 3 lux.

Ah yes Tom but that figure for the DSR500 (actually the spec sheet for the DSR570 quotes a minimum illumination of 0.4 lux at f/1.4) includes a whopping 42dB of gain switched in doesn't it?

If we're just comparing chips, rather than the performance of the whole video chain, shouldn't we leave the gain at 0dB? Even then there's no guarantee you're comparing like for like as the reference video gain of each camera could be different.

Tricky subject.

Regards NL

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

It's not even that simple, just setting the gain to zero doesn't flatten the playing field. The quoted sensitivity is dependant not only on maxiumum aperture and shutter speed, but on gamma and other video mprocessing as well. I can set cameras up to have as much as two stops different sensitivity, accoring to the type of picture s the DoP wants (in some cases that goes up to 4 to 4.5 stops different).

Add to that, the noise level. In professional measurements, we specify the light level required to produce 100% video signal (that eliminates gamma usually) from an 89% reflectance card, with 1/50th shutter speed, and video noise at a prescribed level (used to be about 45dB, can be as high as 63dB these days). If camera A makes 100% video at 45dB signal-to-noise level, and camera B does the same at 65dB s-n, then camera A is 20dB less sensitive than camera B (about 3.2 stops), just bgecause it makes the same signal level doesn't mean that the sensitivity's the same.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003

All agreed, but a ballpark assessment could be quite useful. I've used the phrase "all else equal" a few times, and that said would you agree 3 stops is a fair ball park figure? I have more hope comparing lux figures from the same manufacturer (but lets not forget the "0 lux" nightshot cameras!) - even if the FX1 figure is not at 42dB gain, the noise level may well correspond to a DSR500 at that gain.

Alan McKeown
Offline
Joined: May 9 2001

Quote:
“then camera A is 20dB less sensitive than camera B (about 3.2 stops)”

In case anyone is puzzled by the equating of 20 dB and 3.2 stops:

It implies that if we increase the light power by a factor of: 2^3.3219 = 10 times, which expressed in decibels is 10 dB

then the video signal power increases by a factor of 100 times, which, expressed in decibels, is 20 dB

(taking 3.3219 stops rather that 3.2 stops, to round up to 10 dB)

The reason is that the camera sensors are what are termed square-law detectors. The light Photons (proportional to light energy) "produce" electrons. So the incident light energy per unit time (light power) produces a proportional quantity of electrons per unit time (electric current).

Hence the video signal Current (or Voltage) is proportional to the incident light Power.

Which makes the video signal power proportional to the SQUARE of the incident light power.

This is quite different to a radio antenna, for example, where the electric signal power at the antenna’s output is proportional to the incident radio-wave power.

Alan

Alan McKeown
Offline
Joined: May 9 2001

I should have mentioned that this linear relationship between light power and video signal current or voltage, is usually distorted in practice by the deliberate introduction of a non-linear “gamma” function at the camera end and a complimentary non-linearity at the display end.

Alan

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Indeed. Gamma-correction magnifies camera noise near black by a factor of up to 5 times (linear) and crushes noise near white by a factor of up to 1/3 (linear), precisely the opposite of negative film. But, it renders channel noise linearly because the gamma-corrected signal is a notional match to the response of the eye. So compression artefacts are equally visible at all signal levels, but camera noise is most visible near black.

It's also distorted by a filter, intended to mimic the response of the eye, the so-called "weighted" signal-to-noise ratio.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Alan McKeown
Offline
Joined: May 9 2001

Would I be correct to assume that the video system overall transfer characteristic (from light input to camera lens to light output from display device) should be linear?

If it were non-linear, would that not introduce harmonics of the spatial frequency components of the televised scene, not present in the original?

Alan

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

No, not linear overall. Conventional wisdom, based on the photographic industry, is that the overall characteristc should be a power law of about 1.25, to compensate for viewing with a dark surround. However, it isn't, and canm never be, a real power law.

Traditionally, we assign a power law to the crt display, and the value is 2.2 if you're in NTSC-land, 2.8 +-0.3 in PAL-land. Those values are in the system specifications. That always worried me, because we all use crts from the same factories, and the drive electonics is fairly standard, so I measured lots of displays about 12 years ago. I got values of 2.35 +-0.05 for a lot of studio monitors, with slightly lower values for consumer tv sets. In theory, the value is 2.5 for the electrode configuration of a cathode-driven crt, but the actual value gets lowered somewhat when the output amplifiers don't have zero output impedance. Typical values for consumer kit is about 10kohms, rather less for studio monitors.

Very early on, it was decided that correction for this power law made sense only if it was done in the camera, because of the "happy accident". THis value of about 2.5 just happens to match the law of the eye, very closely, so the non-linear signal-domain of gamma-corrected signals is perceptually linear. So, the camera needs a gamma-corrector.

Applying a pure power law was never possible (still isn't) and wouldn't be a good idea if you could. There are two flavours of equation widely used, the BBC one and the others. The BBC gamma-corrector curves (in TV2248) have a mximum slope at black of 5 and the input signal is shifted and scaled such that it is tangential to this slope where it joins:

BBC: Vout = [(Vin-0.02262)/(1-0.02262)]^0.4, above Vin=0.037703

The other flavour offsets the output instead, to the same effect. The popular curve for HD is that in ITU-R BT-709, which has a slope of 4.5 near black:

709: Vout = (1-0.099)V^0.45-0.099, above 0.018

Now, if you raise the output of either of these gamma-correctors to the power 2.35 (or anything else for that matter), you don't get a power law, you get an approximation. The BBC curve does a better job but produces more noise near black. Both curves, when square, produce a near fit to an overall linear transfer curve.

So, the overall power law, if you really wany one, is about 1.1 in studios and about 1.0 in consumer tv sets.

This does indeed produce harmonics in the non-linear domain, which are more-or-less cancelled when shown on a crt. Or so it seems. The problem is that the channel bandwidth-limiting filters are in the non-linear domain, so they generate big harmonics on the crt. You can see them if you know what to look for on a test signal. They're the subject of much work on companies like Philips and Snell&Wilcox, I often talk to them about nthings like this.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

cstv
Offline
Joined: Jul 26 2002

i saw a Sony DSR200 at work today, i know this jumps back a few posts, but is that the sort of shoulder mount infocus would be looking for? doesn't stick out the front of your head too much, quite small and light. Stick the gubbins from an FX1 in that and it's looking good!

mark.

Alan McKeown
Offline
Joined: May 9 2001

Thanks for your explanation, Alan.

Quote:
“The problem is that the channel bandwidth-limiting filters are in the non-linear domain, so they generate big harmonics on the crt.”

Presumably that is because the low-pass channel filters remove harmonics falling in their stop-band, whereas we need to preserve these harmonics so that they are present to cancel the corresponding harmonics generated by the non-linear CRT transfer function.

Quote:
“THis value of about 2.5 just happens to match the law of the eye, very closely, so the non-linear signal-domain of gamma-corrected signals is perceptually linear. So, the camera needs a gamma-corrector.”

I didn’t understand the significance of the 2.5 law matching the eye.

I would have thought that the displayed picture should match the characteristics of the original scene, irrespective of what law the eye follows. The eye should then see both the original and the TV picture in the same way?

Alan

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

The eye responds approximately logarithmically to light level, with a power law of about 1/2.4. The CIE has an equation for it, the CIELUV L* equation:

L* = 116 V^(1/3)-16

with a limiting slope of about 9 near black. That curve looks very like a power law of 1/2.4 when you plot it.

So, since the crt has a law of about 2.4, if we generate non-linear signals to drive the crt, we end up with signals that are perceptively matched to the human eye. This means that, say doubling the signal, doubles the percieved brightness. These gamma-corrected signals form a visually-uniform signal space. And that means if we add any noise or quantising/compression effects to the signal, they will be equally visible at all levels of the signals, just what we want. So we end up using the least number of quantising levels that we can get away with.

Provided the overall characteristic of the system is near unity, we get accurate colour rendering, it doesn't matter what we do in-between. But, by a happy accident, we've been forced into using non-linear signals that closely match the performance of the human eye.

Filtering in the gamma-corrected domain is essential in an analogue system because we can't cope with infinite bandwidths. The filters are needed to stop channels interfering with each other. In a digital system, the filters are needed so that the codec doesn't have to cope with excessive frequencies. Both those reasons are sound and can't be avoided. The filters result in harmonics being generated in the signals that are visible when viewed, and the same filters remove some of the harmonics generated in the gamma-correctors that would have been cancelled if it weren't for the filtering in gamma-corrected signals.

The precise shapes of the filter characteristics are chosen to minimise both these effects, not just to achieve maximum image sharpness. The filters are defined (or rather, templates for filters are defined) in the ITU recs (601 for SD, 709 for HD).

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Alan McKeown
Offline
Joined: May 9 2001

Thanks Alan.

That makes sense to me now.

Alan

sleepytom
sleepytom's picture
Offline
Joined: Sep 15 2000

Progressive scan?

Do these new HDV cameras do any form of progressive scan recording?

as most of the video i produce is for display on computers progressive scan is an important feature to me (for ease of workflow and quality of output, not for any silly "film look" reasons!) - i've been attracted to the DVX100 because it has full progressive scan at 25 fps - do these new sonys do any form of progressive? (all the other sonys are limmited to 12.5 fps in progressive mode which makes it useless) of course i could deinterlace after capture but this is timeconsuming and never really produces the best results.

You can contact me at http://tombassford.org
People interested in live production might like to check out http://atemuser.com 

tom hardwick
Offline
Joined: Apr 8 1999

I don't think it's true PS mode as the 12.5fps is - I think it's more a Canon-type frame-movie mode, where there may be some loss of vertical resolution.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

The first tests I did at a hire company in North London clearly showed that it's a 12.5 fps mode. However, that was before I found out how to tweak the profiles (had no manual at the time). When I saw the one DVC had, I knew how to do that so shot test sequence which we then had a quick look at. It looked like 25. So now I'm confused. I suspect that I hadn't got it into "movie" mode in Hove.

I'll ask Richard to repeat that test and see what he gets.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Des
Offline
Joined: Apr 7 1999

I had heard reports, at least on the pro model, that it could not shoot progressive, at least not in the strict sense BUT that it could shoot 24 frames and by using interpolation achieve an effective 24p rate. Because it's HD there's enough info to approximate the effect you might have got if you shot it true 24p.

Sony Z1 / A1E / PD100 - Avid Liquid 7.1 - 2.66GHZ Core 2 Duo / ATI 950 Pro / 2 x Iiyama 17" flat screens
Storage: 1 x 80GB / 2 x 400GB / Offline Firewire 1 x 2TB and 3 x 400GB

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

I seriously doubt that. I don't yet know whether the movie mode (FX1) shoots 25p or 12.5p, my information is conflicting, so I've asked Richard to investigate. BUT, there's no way you can interpolate to produce a decent proscan look without using a lot of expensive processing, the S&W ARC for instance does the best job there is at SD, but it ain't cheap. The fact that the images are HD has nothing nto do with the frame rate, it only leads to having much more data to process. Temporal interpolation is very difficult at such low speeds.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Des
Offline
Joined: Apr 7 1999

Looking at Sony's press release at:
http://news.sel.sony.com/pressrelease/5331
under CineFrame Mode, it says:

The HVR-Z1U offers several options for "cine-like" shooting in three modes: 30 frames, 25 frames and 24 frames.

It doesn't state if it is inter or prog but elswhere it implies interlaced with iterpolation to bring it back to single frames. I fully realise that this is from Sony's own factsheet but it does clearly state 30 frames, 25 frames and 24 frames and even with Sony ''public porkies'' are just not on. It says ''cine-like'' I presume this to mean the effect however the delivery is described in unambiguous terms.

Sony Z1 / A1E / PD100 - Avid Liquid 7.1 - 2.66GHZ Core 2 Duo / ATI 950 Pro / 2 x Iiyama 17" flat screens
Storage: 1 x 80GB / 2 x 400GB / Offline Firewire 1 x 2TB and 3 x 400GB

Rookie
Offline
Joined: Sep 27 1999
Quote:
Originally posted by Alan Roberts:
...there's no way you can interpolate to produce a decent proscan look without using a lot of expensive processing, the S&W ARC for instance does the best job there is at SD, but it ain't cheap.


Having never heard of S&W ARC i decided to do some Googling. What do you know, I found a BBC pdf titled "How to recognise video image sources" by a certain A. Roberts.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Gosh, there's a thing :D

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Richard Payne has confirmed that he'll do the experiment and let me see the results early next week, when he's feeling better.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003

I’ve been away for the last week, and whilst no FX1, I had a PD150 with me for some shots when a big camera couldn’t be used – mainly in car. Checking it over on the first night I was thinking of the posts in this topic, and took the opportunity to do a couple of tests (albeit not too scientific). In my rather dim hotel room, the PD150 just managed a properly exposed picture, albeit wide open and with 18dB gain. A DSR500 managed the same with 12dB of gain. When 36dB of “hypergain” was switched in, noise was VERY apparent in the viewfinder, but it produced a fully exposed picture stopped down to a staggering f8 – over 4 stops!

It was later looking at the results on a broadcast monitor that it got interesting. The picture from the DSR500 at 12dB was clean, with noise only just visible. From the PD150, the noise was much more noticeable, but what I hadn’t been expecting was how much softer the image had gone, with deterioration of the colorimetry. I suspect adding gain to that camera does more than just up the gain, I think it introduces other circuitry to keep the noise down at the expense of other attributes. I suspect the FX1 would behave similarly to the PD150 – does anyone have quoted lux figures for these two cameras?

Steelej’s original question was “How does the quality of the FX1 compare to something shot using a pro DV cam like the DSR-390 or DSR-570?”, and I think the above backs up my original view that there’s not a simple answer. In dim light I’m sure the DSR570 will beat the FX1 hands down, in other conditions….. well, I’d like to see!

Technical issues aside, I also thought of last week’s posts about handling and balance during the time using the PD150. I now thank my lucky stars I don’t have to use one all the time, mainly due to the way it’s weight and balance were acting to twist my wrist during normal hand holding, to say nothing of it being difficult to get far enough back in a confined space. To cstv: yes, the style of the DSR200 is something like I’d like to see, though I don’t confess to know that model very well - isn't it somewhat bigger than an FX1? Just imagine an FX1 exactly as it is, except with the monocular roughly where the LCD screen is now, and then move the right hand grip as far forward along the lens barrel as possible. That would make the camera naturally sit out of the way by the side of the head, much further back, and much of the weight nicely balanced on the right shoulder. Please, Mr Sony!

tom hardwick
Offline
Joined: Apr 8 1999

The PD150 was always quoted as 2 lux, the PD170 as 1 lux and the FX1 as 3 lux (all Sony figures).

infocus
Offline
Joined: Jul 18 2003

Thanks Tom. It now begs the question relating to the 170 of whether they achieved the improvement by new technology, or just more processing/accepting higher noise levels at the expense of picture quality, as with the 5oo's hyper gain? The 150's performance was good considering the conditions, but when I saw the comparison it really ran home to me the difference between 1/3" and 2/3".

paultv
Offline
Joined: May 16 2002

Well there's a lot of technical stuff here, but
the proof is in the pudding as they say.

Having been using multi camera rigs of Digibetas and 570's for live and studio music shoots, I've
become used to their results, quality always acceptable plus or minus the odd grumble.

But no more.

My FX1 is in another galaxy. Comparing even the DV
downconversion output of this camera makes a 570
look like a muddy, foggy jittering wreck of an image. Looking at this stuff converted to CFHD codec avi's in After effects - wow!- more tests to do but I know what I'm shooting with from now on - roll on the pro version!

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

The comments about noise quailty are exactly right. In the quest for ever more impressive light-level specification, tricks are played. One is to increase the shutter duration beyond the 1/field frequency (called "slow shutter") which smears motion but avoids having to use extra gain, another is to introduce spatial filtering to knock out high frequencies. Both have merits and detractions. It all depends on what you like.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Jim Bird
Offline
Joined: Sep 15 2000

Paul,

How does the FX1 compare with those other cameras in low light?

Thanks.

Jim Bird

infocus
Offline
Joined: Jul 18 2003

According to the specs, not quite as well as the other 1/3" models, and considerably worse than 2/3" models (hardly surprisingly). Do we have anyone with an FX1 (DVC?) that could compare one to another camera and answer this practically? The PD150 seems to average pixels at high gain to reduce noise (but make the image soggy) - if the FX1 did this it may loose its HD capability, but still give a reasonable SD image in low light - maybe as good as a 150, anyway?

paultv
Offline
Joined: May 16 2002

I'm doing a shoot in December with three 570's and the FX1, at Abbey Road, with a classical orchestra, no extra lighting is permissable in the studio so we are stuck with the house lights.

This should prove a very interesting practical test, particularly comparing sensitivity and "look".

I'm finding that the worst effects to the FX1 quality is "capture". There is a significant increase in noise in the image after capture, compared to direct monitoring of footage. (downconvertion to DV in camera, captured using DVStorm firewire) The best results so far are Connect HD import as TS stream and then convert to CFHD codec avi, these are extremely clean, detailed but smooth, no edge twittering from overshoots (picture profile 1)
- this is of course HD so it should be good!
(this is output from After Effects thru Storm card to Normal res Monitor)

The other point about gain factors, we always set gains to be 3dB and 6db (or 6 and 12) on
broadcast cameras, very unusual to use 18, setting the FX1 to do this produces excellent images without apparent reduction in detail or significant increases in noise.

It all depends on what you view the results on, a broadcast monitor will make everything look fab, if not sometimes a little flat, whereas your average telly has massive detail correction which adds noise and peaking errors.

I must admit I prefer those dark corners to look dark, without loads of forced gain and dancing pixels, you can do loads more with it in post that way.

I'm just cutting something which contains footge I shot with a PD150 last month - looks a bit sad now I have to say!

Richard Payne
Offline
Joined: Sep 15 2000

More and more convinced that the 'in camera dv downsampling' is not a good way to work.

Downsample through Edius SP for HDV seems much better. Software render in Edition 6 Pro also gives excellent result.

Must try that progressive test for you Alan as soon as our Edit system is back in one piece as we are testing differnt MB/Graphics Combos at the moment.

ChrisH
Offline
Joined: Jul 31 2004

Hi
Great Thread.
I have been trying to get my head round all that the HD changes mean. And i'm a little confused by the FX1/ Z1E. I had understood that true 16:9 HD was 1920 x 1080. The FX1 is 1440 x 1080 ... 4:3
so why does it have a 16:9 screen? does it have an anamorphic lens? what does switchable mean in these terms?
Thanks
Chris

StevenBagley
Offline
Joined: Aug 14 2000

It's 16:9 but with 1440 wider pixels as opposed to 1920 square pixels, similar to how SD uses anamorphic 16:9 images.

Interestingly, Sony's pro-format HDCam resamples the 1920 down to 1440 on tape as well, Panasonic's DVCProHD takes it down to 1280 I believe.

Steven

ChrisH
Offline
Joined: Jul 31 2004

Thanks Steven
It all makes sense now. My other questions concern the Mpeg2 compression. I have read that HDTV 1080i delivers just over 19mbit/s, is this the rate at which firewire captures from the FX1/ Z1E? and does it remain in Mpeg2 codec for edit? ..

and what does this mean for DVD? at the moment DVD's are under 9mbits/sec Mpeg2, does that mean HDDVD is set to arrive in the next few years?

infocus
Offline
Joined: Jul 18 2003
Quote:
Originally posted by ChrisH:
I had understood that true 16:9 HD was 1920 x 1080. The FX1 is 1440 x 1080 ...

If you're interested, some fairly definitive quotes from Alan Roberts on the subject here. In brief the two main HD pixel resolutions are 1920x1080 and 1280x720 - the former normally interlaced, the latter progressive. Note that unlike most other digital video standards these are square pixel (hoorah!). To save data, the HDV format (and HDCAM, though not HDCAM SR) use non square pixels for recording - hence the 1440 horizontzal figure.

infocus
Offline
Joined: Jul 18 2003
Quote:
Originally posted by ChrisH:
My other questions concern the Mpeg2 compression. I have read that HDTV 1080i delivers just over 19mbit/s, is this the rate at which firewire captures from the FX1/ Z1E? and does it remain in Mpeg2 codec for edit? ..

I think 19Mbit/s is the transmission rate for broadcast HDTV in the States, and also for the 720 flavour of HDV that JVC have two cameras for in the States. I believe the FX1 and it's 1080 flavour of HDV are 25Mbit - the same as DV. For editing, see the answers at http://www.dvdoctor.net/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=10;t=013594 and the Canopus website.

Quote:
and what does this mean for DVD? at the moment DVD's are under 9mbits/sec Mpeg2, does that mean HDDVD is set to arrive in the next few years?

There are two rival systems heading our way - both using blue laser, HD DVD and Blu-Ray. Think VHS and Betamax again....... I've heard that we may expect to see these drives built into edit systems as early as next year. Yes, expect big, big change in the next year or two.

tom hardwick
Offline
Joined: Apr 8 1999

Yes, expect big, big change in the next year or two. THis little sentence has been true in the world of movies every year since 1994.

ChrisH
Offline
Joined: Jul 31 2004

Thanks for your answers, they have been a great help. Just one more question... will reality TV be any more real in HD?

infocus
Offline
Joined: Jul 18 2003

Probably...... oh no! Do you realise you've just put me off the whole idea!

ChrisH
Offline
Joined: Jul 31 2004

Sorry I've taken this off topic. Looking forward to hearing more about how the FX1 performs and the results of down-sampling during the transition from SD.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Results just in from Richard at DVC.

He's recorded a quick pan in the "movie" mode, which the menus describe as "25fps". From the evidence, yes it's 25 and not 12.5, they've got it right this time. That doesn't make it a good emulation of film because there's far more to it than just the jerky motion, but at least they've got that part right.

So I withdraw my reservations about tyhe movie mode.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Ed Stradling
Offline
Joined: May 18 1999

Richard, you were saying you're unconvinced about recording in HD and then downconverting "in camera". Presumably most of us will have no alternative as our editing software doesn't support HD yet.

So given the choice between shoting in HD and downconverting, or simply shooting widescreen in DV mode, which would you recommend?

thks

ed

Richard Payne
Offline
Joined: Sep 15 2000

The Jury is still out but I think an in-camera downconverted FireWire output from HDV is not as good quality as stuff shot originally on DV in the same camera.

Need more time to be sure.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

That's what I'd expect though.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

cstv
Offline
Joined: Jul 26 2002

yeah, the numbers alone suggest that would be the case because of the subsampling... long term though, if you wanted to have the option of HD in the future then downconverting may be worth the loss in quality...

paultv
Offline
Joined: May 16 2002

I'm not sure its a quality loss here,
my tests confirm what Richard was
thinking, the down converted HD signal
appears to "push" the limits of IEE1394
- this produces added noise on digitising
which may be due to the 1394 bandwidth
limits or more likely, capture cards never
having to face this kind of response in a
signal, oddly, the S-video output is actualy
more pleasing to look at, no extra noise or
"ringing" actually looking cleaner than firewire.

( this all thru a Storm 2 )

Will test the difference between HD
- downconvert to Dv and Dv direct
tomorrow - both firewire and s-video.

DVdoctor
Offline
Joined: Apr 1 1999

Great discussion. I have to say that I have found that when I go from DV to Mpeg 2 in SD (using a variety of set top DVD recorders I consistantly get better results using the ANALOG conversion and NOT the direct firewire connection.
John

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Yes, Paul, that's what I'd expected. I showed a tape with exactly this effect at DVC; downconvetred and onto DV from an HDCAM tape, the compression artefacts are huge.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Bruce
Offline
Joined: Apr 20 2001

So is it right to think in the short term that an HDV edited programme will look best if dowloaded from the edit machine in s-video mode?

cstv
Offline
Joined: Jul 26 2002

DV to Mpeg 2 in SD (using a variety of set top DVD recorders I consistantly get better results using the ANALOG conversion and NOT the direct firewire connection

John, I have a sneeking suspicion that this could be related to the cheapo DACs in DVD recorders being even cheaper than those in consumer camcorders... any thoughts Mr Roberts?

which may be due to the 1394 bandwidth

anyone know why Sony didn't opt for IEEE1394b (aka FireWire800) on the FX1? I assmue it's a backwards compatability issue, especially now that 1394a is everywhere...

So is it right to think in the short term that an HDV edited programme will look best if dowloaded from the edit machine in s-video mode?

That depends on what you're going to and from, Bruce. The problems Paul's talking about are on the downconverted output from the camera. If you're coming out of the edit machine then it depends on whether you've edited in HD or in SD...

mark.

StevenBagley
Offline
Joined: Aug 14 2000
Quote:
anyone know why Sony didn't opt for IEEE1394b (aka FireWire800) on the FX1? I assmue it's a backwards compatability issue, especially now that 1394a is everywhere...

Why bother? The HDV signal has a bitrate of 25Mbit/s which is less than IEEE1394a limit of 400Mbit/s by a factor of 8. IEEE1394a silicon is also cheap and plentiful (no doubt Sony just loaded the source file for their implementation into the chip when creating it). Adding 1394b would up the cost of the unit for no benefit (except perhaps for people wanting to daisy chain the camera off a 1394b chain).

The only reason that a DV downconvert done in the camera looks worse than one done on a PC is because the algorithm being used on the PC is doing a better job. This could either be down to a lousy implementation on Sony's part or more likely a lack of CPU cycles to do the downconvert.

SD recording from the camera is better than a downconvert because it presumably is scaling the data off the CCD (ie before it has been interlaced) to 720x288@50 rather than to 1440x540@50.

Of course that leaves the question of why the downconvert to the Svideo looks better, most likely because the lower resolution of the svideo link smooths out the high frequencies that disrupt the DV compressor.

But that doesn't mean realtime DV conversion is a gimmick -- it makes a very nice method to get your footage into your computer to edit offline. Then you can batch capture the material used in your final edit as the HDV bitstream and conform it to match.

Steven

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

MPEG2 builds in artefacts, as high frequency alias patterns. Output through the DAC will attenuate them by up to 4dB (because they don't have proper filters), so the pictures look cleaner in DV. This is exactly the effect I showed on the tape at DVC. Proper filtering happens only in high-end broadcast kit, I've seen even professional kit that doesn't do it properly.

And the FX1 uses the HDV format, which specifies the data rates and interconnect formats. There's no need to go faster.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Bruce
Offline
Joined: Apr 20 2001

I have just had one to play with and have transferred via YUV to Digibeta. I will edit on Avid Symphony uncompressed. is this a good route to take in these early days?

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Yes, that's a very good way to treat it now.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

StevenBagley
Offline
Joined: Aug 14 2000

Just found this link, side by side comparison of the VX2100 (with anamorphic lens) and the FX1

Some nice piccies

http://www.supervideo.com/shtoutsvFXVX.htm

Steven

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

I have to point out a mistake in his text. He reports that there's no sharpening etc added in either camera, the menus were left on the Sony default setting. After many years of experience in this game, I can confirm that the default settings will always leave some detail enhancement in, usually too much and at too low a frequency. The FX1 has fairly significant enhancement, it's a balancing trick between supplying resolution through having a better camera and through processing. The FX1 is a very nice compromise but it certainly doesn't deliver resolution to its limit without help from fiendish picture processing.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003

Point taken Alan, I still commend the author for his work, and (as I've said before) using enlarged frame grabs to illustrate the points made - otherwise you just don't know if you're evaluating the camera or your monitor. Oh that more magazine reviews would follow the example. This sort of thing is commonplace in the stills world, even using the same test for many varieties of camera. In these cases tests are posted for differing ISO settings, and it's possible to see how they compare sensitivity wise. (And boy - that's what shows up the DSLRs from the compacts!)

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Agreed, it's a good comparison. To get a better one, you need lab conditions, the sort of thing I did professionally for many years.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

StevenBagley
Offline
Joined: Aug 14 2000
Quote:
He reports that there's no sharpening etc added in either camera, the menus were left on the Sony default setting. After many years of experience in this game, I can confirm that the default settings will always leave some detail enhancement in, usually too much and at too low a frequency.

It's also visible in the enlarged image of the text on the side on the bus of the FX1

Steven

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Exactly. It's very difficult to get to the native performance of a camera, you really need to know how to tweak menus and even that often doesn't undo all the tricks they've played inside. Comparison of cameras on the same format is hard, comparing cameras of different scanning standards is a nightmare.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003

Whats useful about these tests is that they do show quite clearly the differences between two cameras "out of the box" - which is what is important for the average user for this type of camera. They also graphically show the differences between HDV and mini DV, given input from comparably priced cameras, (and if anybody still needs to be convinced that HDV is a quantum leap forward....) in a reasonably scientific fashion, given limited resources.

For still cameras the sort of thing I'm talking about is typified by such as http://www.steves-digicams.com/hardware_reviews.html . (Look at sample pictures within a review.) The still life tableau is repeatable camera to camera and whilst not totally scientific is a huge leap forward from many "reviews".

teeUK
Offline
Joined: Aug 1 2003

A question for those who have used the HDR FX1E:

Will this camera shoot stills to tape or memory card?

If shooting in high definition is anything needed to be able to edit in premiere 6.5 in normal dv? (I used pinnacle pro-one card which does not support premiere pro & does not have HD plugin)

cheers

trev

tom hardwick
Offline
Joined: Apr 8 1999

It shoots exclusively to tape, no flash memory card on board as far as I'm aware.

Richard Payne
Offline
Joined: Sep 15 2000

The FX1 does not shoot stills to tape or card.

You should be fine for normal DV with Premiere 6.5 using the in camera down-conversion. Please note that should be is not will be.

Alan McKeown
Offline
Joined: May 9 2001

Quoted from supervideo.com:
“The HDR-FX1 has three recording modes: HDV (high-definition video) mode, which has 1080 horizontal interlaced lines with 1440 vertical lines resolution at widescreen 16:9 aspect ratio; a DV mode, which also features a 16:9 aspect ratio, and has 480 horizontal interlaced lines and 720 vertical interlaced lines; and a DV mode with 4:3 aspect ratio has 480 horizontal interlaced lines and 720 vertical interlaced lines.”

http://www.supervideo.com/2004i.htm

I bet you didn’t know that the Sony HDR-FX1 has DV modes which produce pictures with interlacing in both vertical and horizontal directions!

Alan

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

It just goes to prove that you shouldn't believe all you read on the web

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Alan McKeown
Offline
Joined: May 9 2001

A few lines further down the same report on the FX1 we are “informed”:
“Three batteries are available; a 4.6 POUND battery provides 65 minutes of recording time on both HDV mode and DV mode; a 5.9 Pound version provides 130 minutes of recording time on HDV mode, and 240 minutes of recording time on DV mode. The 7.1 pound NP-F970 provides a maximum of 205 minutes of recording on HDV mode and 215 minutes of recording time on DV mode.”

Judging by the masses quoted above, I think these must be Plutonium-ion batteries!

Alan

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Quite :D

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003

Careful, careful - before you know it the web will be full of reports of "the new Plutonium-ion batteries", including people who have actually used them and are reporting sharper pictures. I jest, but.........!

I just wonder what Panasonic have in the pipeline. Of the main manufacturers, I believe they're the only ones not to have signed to HDV, but I can't believe they will just sit back and let Sony and Canon (in 2005?) corner the con/pro-sumer Hi-Def market. Especially relevant in the States and Japan.

Ed Stradling
Offline
Joined: May 18 1999

Would be very interested to see a similar screen grab comparison but using the FX1's HD downconverted image against a the same image shot in standard DV.

teeUK
Offline
Joined: Aug 1 2003

thanks ed
i'm not bothered about the memory card however will it shoot a still to mini dv tape?

people have talked about downgrading the output to standard 720x576 which should edit fine in premiere 6.5 however is there a noticeable picture degradation?

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

In a perfect world, it would not matter whether you standards-convert down to SD before or after recording. But this isn't a perfect world. The camera has access to all the lines of the image and so can do a good downconversion, but...

If you record HD, and then standards-convert, you have access to only the lines of one field at a time, unless there's a frame store (or more) in the converter plus a pretty smart converter. Plus, there will be the artefacts of the recording system confusing any motion-sensitive processing.

If you record SD, you have access to all the lines of the HD picture to make the camera output, so the conversion should be better (you don't have to de-interlace, i.e. invent missing lines). Plus, there are no recording compression artefacts to mess it up. BUT, you have no HD recording.

Oh, and it will record 25 stills per second, each of 1440x1080 pixels, to tape. Call it HDTV.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

tom hardwick
Offline
Joined: Apr 8 1999

129 posts over 4 pages. Is this some sort of record?

Chirpy
Offline
Joined: Sep 7 2000

It may be in this forum Tom but it's not the longest on all the forums by a long chalk. In the Chatter forum Tennis, Elbow is restarted when it reaches the 250 mark (7 pages) which happens almost every week. It's now on Round 60. So how about an 'Over 50's' page? - Second Anniversary Edition is up to 330 replies (9 pages) and the current thread has been running since 29th May 2003. However the longest thread is It's the 'Tennis, elbow, foot ' game...Round 5 which has 522 replies so far (it's still an active thread) and runs for 14 pages!

Keep posting Tom. ;)

Chirpy

Chirpy's Big Breakfast can be heard on Radio England International. These are repeat shows (he's retired now) played Monday to Friday 8am-12 noon and repeated in the evening from 8pm-midnight. Also, Sunday 8am-12 noon. (Click link to listen) www.onlineradio5.com/2013/06/radio-england-international.html

tom hardwick
Offline
Joined: Apr 8 1999

I'm humbled by your statistics, Chirpy.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

But this is easily the longest on a single technical topic.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Chirpy
Offline
Joined: Sep 7 2000

Not quite...

Desktop Video General forum, 321 replies - 9 pages

(Thanks Tom )

Chirpy's Big Breakfast can be heard on Radio England International. These are repeat shows (he's retired now) played Monday to Friday 8am-12 noon and repeated in the evening from 8pm-midnight. Also, Sunday 8am-12 noon. (Click link to listen) www.onlineradio5.com/2013/06/radio-england-international.html

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

OK, but that's a pretty general catch-all, this one's about a single camera.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Richard Payne
Offline
Joined: Sep 15 2000

Have I just filmed the first Hi Def nursery nativity play?

Seriously - the FX1 was impressive in 'murky' lighting conditions.

tom hardwick
Offline
Joined: Apr 8 1999

I'd still like to see its performance in the gloom side by side with the VX2100. I've just done a school play Richard and I switched the VX2k into the 'spotlight' mode. Wonderful! Perfect exposure when the scenes were evenly lit and highlights when it got contrasty (however small the highlights, apparently) were never allowed to burn out.

tom.

Richard Payne
Offline
Joined: Sep 15 2000

Your lucky Tom, my Hi Def pictures revealed one of the toddlers at the back had a Nit. And the nit was wearing a tiny tiny medalian that kept catching the light from a single candle in the next room.

tom hardwick
Offline
Joined: Apr 8 1999

Hate you.

infocus
Offline
Joined: Jul 18 2003

Richard, when in years to come the offspring bring their girl/boyfriend round to meet the parents for the first time, your Hi Def version of the Nativity play will be even more embarrassing for them than my SD versions for my offspring! :D

StevenBagley
Offline
Joined: Aug 14 2000

Found this article on Sony's site this evening -- though I can't for the life of me place where I recognise one of the names from.

BBC predicts bright future for HDV

“I honestly can’t believe what I’m seeing. A full resolution HD camera at this price is just unbelievable. It could take on everything that’s currently being shot with DSR-170Ps — it’s that concept, but better.”

The new HDV camera is causing quite a buzz in the broadcast production world. At the BBC, it’s at the beginning of a round of tests, but there’s already a sense that this format could make a big impact. “It gives us the chance to produce affordable HD programmes at all levels of programming,” says Andy Quested, at the BBC’s HD department. “It’s as important as DV is to the current standard definition format.”

Of course, it’s early days. The consumer version of the camera is only just on the market and the production community is keen to see how the professional version of the camera will look when it comes out in early 2005. But hopes are high and the BBC is putting the current HDV camera through its paces.

Indeed, the camera has already won over the BBC’s classical music department, where it has shot behind-the-scenes documentary footage for a triple bill celebrating the life of Sir Frederick Ashton, renowned choreographer for the Royal Ballet.

Andy King-Dabbs is the producer and documentary director of this 140-minute special for BBC4, called Ashton at 100, Fred’s Steps. He says, “It’s very exciting. It cost me no more to use than a DV camera, but the results I’ve got look spectacularly good. I’m hoping the results of this test will meant that very early in 2005 we’ll be able to rely on the HDV for broadcast television.”

“The goalposts might be changed with this new camera”
Meanwhile, the Corporation’s internal hire operation DV Solutions has been sounding it out for all-round usefulness and independent consultants are assessing its technical advantages.

Fiona Macbeth is the BBC’s Self-Op Development Manager for People Development. She sees great promise if the HDV is all it seems to be, most importantly because it’s a wide-screen format. “The goalposts might be changed with this new camera: It’s the idea of a small camera at a lower cost with true widescreen, enhanced picture quality and a number of other advantages,” she says.

Macbeth argues that the camera could lead to editorial and cost benefits in other programme genres than those currently benefiting most from DV. “It’s not just the technology, it’s the way of working that it offers. It allows for greater access and filming in real situations.” As well as areas such as news and current affairs, HDV could be used for drama or entertainment.

First, the camera needs to get the thumbs up from seasoned operators. At DV Solutions, mentor Andy Benjamin has given the camera a good going over and — even with just a short trial — he’s very positive. “I honestly can’t believe what I’m seeing,” he says. “A full resolution HD camera at this price is just unbelievable. Who could have though four years ago that it was possible for £4,000?”

The BBC has already publicly declared that it is committed to phasing in HD gradually over the next few years, so that all productions will be produced in HD by the year 2010.

“We’ve been asking the manufacturers for years for a true 16:9 camera and now they’ve done that and they’ve done it in HD as well, which is a bonus,” says Benjamin. It isn’t going to replace the more serious professional cameras, but he does see it doing an important job.

“It’s definitely a lightweight solution that one person could take on. It could take on everything that’s currently being shot with DSR-170Ps — it’s that concept, but better.” He suspects that it could even, in some cases, take on the bigger, higher-end cameras.

Although Benjamin has one or two niggles with the consumer camera – in particular with the audio –hopes are high that the professional version of the camera will have tackled this.

“A startlingly good camera for its price”
The BBC has also involved outside consultants in the initial stages of assessing the format. Alan Roberts is an expert in colour science and has specialised in HDTV and colour cameras. “I think it’s a nice camera,” he says, talking about its picture output. “It will be a second camera and while it’s not as good as the HDW-750P or HDW-F900, nevertheless it’s a startlingly good camera for its price.”

The BBC is happy to see that the camera uses 1080-lines HD, given the Corporation’s goal to see all HD production in the highest quality as soon as possible. “It makes a very acceptable 1080-line pattern,” says Roberts. “It’s interlaced, but quite a significant proportion of lower-end programmes are going to be interlaced because that’s the sort of look we’re used to, so this sort of camera will drop into things perfectly.” Of the camera’s six potential profiles, he feels that two are right for professional use and he notes that there’s a “very nice flip-out LCD” at the front of the camera.

One man who is taking a technical overview of the trials is Chris Price. It’s his job as mentor in the BBC’s Production Modernisation division to direct technological decisions. “I’m very interested in HDV cameras because they have the potential to replace everything we do with DSR-PD170Ps — our workhorse cameras in terms of self-op,” he says, “but they offer much better picture quality.”

The interest is primarily in the wide-screen capability, although he sees that “as HD scales up, on probably quite a steep curve, in terms of future use it offers the potential for a reliable high definition picture.”

There are more tests in the pipeline and the most crucial of those will be in early 2005, with the arrival of the professional version of the camera. The BBC would also have to discuss use of the format with co-production partners, who are involved in many of the HD productions. But Price is looking ahead: “There’s potential for this camera to affect all areas currently self-op DV and that’s something like ten per cent of the BBC’s total output.”

Quested is also looking into the future, specifically the BBC’s future working with HD. His main issue with the camera is ensuring that the HDV pictures can be easily downloaded for an HD edit, but he expects the professional camera to sort this out. And he shares his colleagues’ excitement at what this type of camera could do for broadcast production.

The HDV camera’s switchability from standard definition to high definition means that it could become a natural replacement camera, with broadcasters gradually phasing it in to take over from current alternatives. It would be used more and more for HD as that becomes the norm.

For the next few months, HDV will continue to be tested, but the camera has clearly already impressed. “We need to see how and where it sits in the hierarchy of cameras,” says Quested. “But it’s already looking quite high up that hierarchy.”

Steven

Marcaaroni
Offline
Joined: Apr 22 1999

I have been reading this thread and trying to dicide if I should slash out on the new HD Sony cam. I have learned that the in camera conversion of HD footage down to SD footage gives poor results. I know that Adobe is going to release a HD plugin (for free). What will be the results if you capture the footage in HD and then export to DVD? I know the resoltion will be decreased but surely the fact that their is more original detail in the footage should make for really good DVD viewing. I suppose that this will depend on the quality of the Encoding software.
Has anybody be able to capture HD footage and then encode it to DVD? is the quality any better than SD footage (after conversion)?
All of my jobs now go out on DVD format VHS is dead (but I suppose every family still has at least one member that hasn't got to the DVD stage yet, ie my Gran)

Awaiting feedback before I make my decision to buy.

Mark

infocus
Offline
Joined: Jul 18 2003
Quote:
Originally posted by Marcaaroni:
I have learned that the in camera conversion of HD footage down to SD footage gives poor results.

I've no personal experience, but the way I read the posts, thats not the conclusion I draw. What I do understand is that technically the HDV quality is superb, and that the SD derived from it in camera is also good, BUT not as good as if the downconversion is done in a computer, or if it had originally been recorded to tape as DVCAM. It is also currently the only true widescreen prosumer camera, regardless of HDV, which is relevant if your market is mainly DVD. Just because something is "not as good" does not mean it's "poor".

From what I have seen, technically it's currently without equal in the price range, by a long way. My reservations are solely about handling issues when used handheld, when I'd consider it a pig. If you look back over previous posts not everyone agrees, but it does seem to be the only serious negative issue raised, and not just by me. Another contributor begged for "an FX1's innards in an XL2's body", and that for me just about summed it up.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Re. the BBC views on HDV, you should be aware that the opinions that Andys Quested and Benjamin give are to a large extent based on my advice to them. For several years, I did the technical testing of new HD camera products before I retired, and I still do the same now as a freelance. The only difference is that I get to set the charge rate for the job, and I do it for anyone who wants it and will pay the rate (which means that you guys are getting expensive advice for free)

So, the BBC's view is, to a very large extent, similar to what I've been saying here since the biginning of this thread. HDV is a revelation and will open up more production genres to HD, simply because of the cost drop.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

tom hardwick
Offline
Joined: Apr 8 1999

It sounds as if the introduction of the Z1 in the new year will bring an awful lot of second-hand PD170s onto the market pretty much all at once, and one wonders how long Sony will continue to produce it. I suppose it is an interim model between the PDX10 and the Z1, but the sort of person eyeing up a 170 right now must also be salivating at the thought of the backwards compatible Z1.

tom.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

The only problem with the Z1 is editing, tyhere aren't many systems running yet. It will take about 6 more months before decent systems are routinely available, then it should take off. At that stage, 150/170s should start to appear 2nd hand.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003

True Alan, but if I was intending buying a camera in this range soon, I'd still go for the FX1. You can still use it DVCAM for the next few months (as you would have to with any competitor) then really reap the rewards in the next 6 months. In the meantime it is still a better 16:9 camera than the PD170.

What I really wonder is what will happen to the market in between the FX1 and Z1 and the top end HDCAMs. Can we expect something along the lines of the DSR570, but Hi Def and HDV with 2/3 inch chips? Or even a camera along the same lines, but recording the same 25Mb signal to disc, as XDCAM? Best of all, recording a less compressed "S-HDV" 50 Mb stream to XDCAM!!

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

I don't think you'll see a 2"/3 HDV camera, that would compete directly with the HDW range, Sony aren't that stupid. HDV isn't a professional format, it's aimed squarely at the consumer (that's the views of the HDV consortium, not mine), the Z1 is a real risk for Sony because it has the potential to subvert sales of the HDWs, and they're only just starting to turn real profits on that range. The HDV format is a "standard", there won't be much development of it, but I suspect there might be another range, intermediate and aimed at the professional. Panasonic are notably absent from HDV, they have a history of not appearing at the forefront in this game, and are clearly more interested in junking tape than in getting sales at this point. I reckon the next stage in this game will be a P2 based HD product from the Panasonic stable. But in the meantime, there will be other HDV products, Canoin first.

That's my 2 pen'orth anyway.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Bruce
Offline
Joined: Apr 20 2001

Anyone want a PD170 with Century zoom through anamorphic lens and 3 Digibetas? I think HDV will be the start of higher picture quality assistant producer crap.

infocus
Offline
Joined: Jul 18 2003
Quote:
Originally posted by Alan Roberts:
......but I suspect there might be another range, intermediate and aimed at the professional.

Well, for workflow reasons XDCAM is starting to make its mark in the quick turn round broadcast market (News, Sport etc) and for such as these companies new kit is unlikely to be tape based. They're unlikely to be tempted either by HDCAM or the Z1, but are increasingly likely to be tempted by the thought of HD capable cameras for future proofing if the price is right. Hence the thought of recording an HDV data stream onto XDCAM disc - as editing for HDV is likely to become commonplace.

Quote:
But in the meantime, there will be other HDV products, Canon first.

And from a cameramans point of view, thats the one I really await with interest! I also wonder what the future holds for much cheaper, smaller cameras, the true consumer market. How long before say 50% of that market is Hi-Def, be it HDV, solid state or whatever? I think it may come sooner than many may think, since many cameras in that category already have higher resolution chips to satisfy the stills capability. I really do think the FX1 is comparable to the VX1000 in terms of being revolutionary, and think we may get an idea of timescales from that and standard DV.

RayL
Offline
Joined: Mar 31 1999

Interesting to note that Prestons are already offering the Z1 for sale on their website for £2977 + VAT
http://www.shoppersworld.co.uk/templated/midlands/prestons/dept.asp?s=80&d=198

Has it been released or are Prestons taking orders ahead of stock?

Ray Liffen

tom hardwick
Offline
Joined: Apr 8 1999

Put in an order to test it out Ray. Then you can do a road test on it and tell us all about it. Look on it as a New Year's present because you didn't get any socks.

tom.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Sony haven't released it yet, they must be taking orders.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

Nigel Longman
Offline
Joined: Apr 28 1999

It's on offer here too but a tad more expensive (if you compare the VAT inclusive prices - there seems to be a pricing error on Preston's site between the VAT-free and VAT-inc prices unless I'm suffering the effects of too much sherry).

I wonder what will be on offer at Video Forum?

Regards NL

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

The Z1 will be out by then, and there may well be news of other impending products.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

elbow
Offline
Joined: Dec 21 2000

I've heard the progressive mode isn't really a true progressive mode but some sort of fudge - apparantly you need a progressive chip to get true progressive mode. I was wondering if you could tghrow any light on this Alan - I can't quite understand how the chip would make a difference - I understand the true 16:9 chip which seems quite straightforward but not the progressive element.

Dual eyes and ears, single nose and one very large mouth.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

It's down to how much storage you can put in the camera. To read the ccd(s) interlaced, you simply read two adjacent lines at a time and call them field lines, you just have to make sure to start the process one ccd line offset on each field. But to read it as a progressive frame, you have to read it into a store first so that you read it all at the same time, you can't just read all the odd lines then all the even lines.

So, a fudge is applied. I guess that there's a field store, so that it read say odd lines out to make one field and even lines out into the store, which is then read to make the other. It's cheaper than a frame store. Knowing how clever some of the Sony chaps are, I'd be prepared to bet that it's a lot cleverer than that though.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

elbow
Offline
Joined: Dec 21 2000

Thanks Alan - so does that mean in a true progressive chip the ability to record interlaced is the opposite fudge?

Either way I think you are implying there would be no noticible degredation from what you may expect from a true progressive chip - is that right?

Dual eyes and ears, single nose and one very large mouth.

StevenBagley
Offline
Joined: Aug 14 2000

Surely there is no differenece between the CCD used in a progressive or interlaced camera, it's just the controlling electronics that are the problem. Though presumably a progressive camera will need a CCD with a lower noise floor (as the averaging of adjacent scanlines in an interlaced camera will reduce the appearance of the noise) I guess this is how the Sony 730 works out at being cheaper than the 750.

Steven

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

There's a big difference. Interlaced reading of the ccd can happen at real-video speed, but proscan has to get the entire frame out before it can start delivering the video. This is because, for proscan, all the information has to be taken at the same time (or at least in one field period). It's delivered as two fields, interlaced, even though the fields both belong to a single exposure. So it needs a frame store at best, a field store at least, unless So ny have a better trick.

The 730's cheaper than the 750 for lots of reasons: for example, the 730 has only one filter wheel, the 750 has two; the 730 can read directly from the ccds, the 750 has to have an intermediate store; and so on.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

PaulD
Offline
Joined: Aug 31 2002

Hi
More than 'hands-on'..!
www.eidomedia.com/hdv/

mooblie
mooblie's picture
Offline
Joined: Apr 27 2001

Take that big grin off your face!

(I am only jealous.) :(

Martin - DVdoctor in moderation. Everyone is entitled to my opinion.

simon ferrigan
Offline
Joined: Jan 8 2003

Can anybody help clear up my confusion over the the fx1 downconversion to SD?

There are artefacts when using the cameras own cable but is it there or will there be a way of downconverting by computer to stop this?

Or would I have to film in SD mode until I can afford a hi-def set?

simon

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

All downconverters produce artefacts, even the £100k+ Snell & Wilcox boxes. The problem is simply that the converter is trying to invent the information that would have been there for each pixel, estimating the value using the contents of spatio-temporal neighbours. Much of the time it will do a good job, on some pictures it will be perfect, on others it can be horrendous.

In theory, the FX1 has all the information it needs to get it 100% right all the time, because it has access to the ccd signals. However, we suspect that the camera operates always as an HD camera, and downconverts to get SD, rather than reading the ccds in such a way as to deliver SDTV (which is surprisingly hard to do, it looks easy on paper but the hardware gets to be a bit nightmarish).

Given that a software downconverter does not have to do it in real-time, like the camera does, it ought to get it more right, but it all depends on exactly how good that design is, and we don't know. Yet.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

simon ferrigan
Offline
Joined: Jan 8 2003

Well is it worth buying one of these camera without a hi-def tv?

I had saved enough to buy either a vx2100 or pd170 but it seems stupid to buy one now with this camera on the market.

It would be a while before I would get a hi-def TV and don't want to be cursing buying the camera between now and then. I know its earlier days with this camera but any advice would be helpful.

simon.

tom hardwick
Offline
Joined: Apr 8 1999

Might help us if you could describe the sort of projects you undertake Simon. The PD170 could well turn out to be better in low light (faster lens, bigger pixels) but better SD is probably worse than less good HD if you see what I mean.

Assuming you want to use XLR mics with the FX1 you might have to buy a Beechtek box. Now the FX1 is starting to look seriously bulky, so does that matter to you? How do you feel about 16:9?

The deal of the moment has to be the PD170. Comes with a wide-angle converter, proper mics plug straight in and it's tried, tested, and hewn from the solid.

tom.

paultv
Offline
Joined: May 16 2002

I have to add that I ran a working test with 3 Sony DSR 570 cameras shooting DVCAM and my FX1 alongside.

The shoot was inside a recording studio with a 100 piece orchestra and "studio" lighting (these are strip lights and some kind of mercury lamps, but very high up producing an odly greenish tinge)

The DSR 570's ran around F4, the FX1 about a stop wider.

Looking at the results, (Sony 570's tapes played out thru DSR11 firewire Storm 2, FX1 internal downconversion to DV firewire Storm 2) I would prefer to shoot with 4 FX1's.

The main differences are a much smoother image with the FX1, all the diagonal "stepping" errors on chrome mic stands and violin bows, have vanished, the overall detail is much higher, but smooth.

The colour purity looks similar, but there's lots of edge buzzing and "twinkling" on moves that I now see on the DSR's.

On the downside, the zoom control on the FX1 has a jumpy start and stop, ( fixed on the Z1 ) and why oh why does the expanded focus facility (really useful) not work when you're in record???

As a user of both big DSR's, PD 170's and my Cannon XM1, the FX1, really seems to make quick work of the lot of them, not bad for £2099 incl.
(Sony Centre Tottenham Court Road London!)

Can't wait to get my new HD monitor to see what it really looks like!

paul

simon ferrigan
Offline
Joined: Jan 8 2003

The size of the camera doesn't bother me and I am looking to add beachtek box.

I will be mainly shooting outdoors in good light so the most important thing is simply image quality and the fx1 does seem to be one for the future and I feel the vx2100 and pd170 could quickly become outdated. £2000 is alot of money and I want the camera to last a while.

simon

infocus
Offline
Joined: Jul 18 2003

I really wouldn't entertain a new PD170 now. Either buy an FX1 (which should hold reasonable resale value) or buy a used PD170. With the advent of the FX1/Z1 there are likely to be an increasing number of PD170s etc coming on the market, and hence price competition. (All the more reason not to buy a new PD170.)

rbarry
Offline
Joined: Mar 27 1999

..........getting back to PaulD's posting; that was pretty brave of you Paul! How do you find this setup of benefit to you, over the original lens supplied with the Sony FX1 ? I have seen your web site still shots, but how do you feel now that you are "living" with this creation, like Dr Frankenstein or Dr Christiaan Barnard?

Rick

bcrabtree
Offline
Joined: Mar 7 1999

PaulD,

Utterly gob-smacked!

I presume you will be bring this along to have with you when you spend time on the DVdoctor stand at VideoForum.

Pretty please!

I presume you'll also be selling off the stand a DIY kit of plans for the conversion, plus a DVD of images.

;)

Bob C

Nigel Longman
Offline
Joined: Apr 28 1999

That camera conversion looks very impressive - but is it actually Paul's or the work of a chap called Marco, who is a colleague of Matteo Ricchetti, who in turns runs a company called Eidomedia (which engineered the work?) who's website Paul's post links to?

And is the man in the photo Paul, Marco or Matteo?

I'm a bit confused :confused: .

Regards NL

bcrabtree
Offline
Joined: Mar 7 1999

Nigel,

This is what comes of posting in the early hours of the morning when you are in no fit state to do so.

What's especially stupid is that I know Paul Dickin and what he looks like so I can't think what made me think any of the pics could be of him.

Mind you, one one thing I did get right, I think, is that PaulD will be on the DVdoctor stand at VideoForum, along with PD!

I've now emailed Matteo asking if he wants to spend some time on the stand, too!

Bob C

cstv
Offline
Joined: Jul 26 2002

that's fabulous!!! :D

i'm so glad that there are people out there with the money, balls and technical know-how to do this sort of thing!

If canon don't get on with an HD version of the XL2 then this mod could be the solution... will the gents involved be offering their services?

mark.

Matteo Ricchetti
Offline
Joined: Jan 9 2005

Hello everyone...
it is a little strange to find here my face.
Btw I am Matteo Ricchetti (the one in the pic)
And i mede the conversion of the FX1...

bye

Nigel Longman
Offline
Joined: Apr 28 1999

Ah! All now becomes clear.

Matteo, congratulations on what looks like a very impressive piece of engineering .

Probably not, I should think, the first camera you've done something like this to?

Kind regards NL

rbarry
Offline
Joined: Mar 27 1999

Ah ha, like Bob, I was burning the midnight oil last night. As cstv said, it takes some bottle to do this sort of makeover on a brand new piece of kit.

bcrabtree
Offline
Joined: Mar 7 1999

Thought I'd just comment to bring this up to the top again.

Bob C

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Matteo's done mods on many cameras, including diving in and doing things that Sony said were impossible. I was impressed.

By the way, Matteo, I talked to Phil Smith (the Sony man in the Q&A session) and to Neil Thompson (Product Manager, Image Creation Group, Sony UK) about getting the service manual. They haven't got one in the UK yet, but Neil said he can propbably get it as a PDF download from Japan. If you don't hear from them in the next week, email me and I'll see what I can do.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

bcrabtree
Offline
Joined: Mar 7 1999

Commenting to, hopefully, make this one appear (on account of I moved it to the HDV&HD forum and it disappeared!).

Bob C