New HDV Cameras from Sony.

[steve]

So it seems that HDV is not dead as a consumer, (as opposed to prosumer) format.

HC3 replacement:
http://news.sel.sony.com/documents/consumer/digital_imaging/camcorders/high_definition/HDR-HC5_Specs.pdf

HC1 replacement?:
http://news.sel.sony.com/documents/consumer/digital_imaging/camcorders/high_definition/HDR-HC7_Specs.pdf

These are of course US releases, but as its only a firmware change for 'E models then I'm sure they'll be along soon.

The new HC5 & HC7 cameras must be seen as fusion between the SR1 and HC3 models with an improved Clearvid sensor and OIS on the HC7.
Not withstanding the limitations of HDV, The HC7 looks like a handy pocket camera.

Steve

[Alan Roberts]

I reckon HDV to be a decent recording-compression format, it does well when it has good material to deal with. And the tape's cheap. It'll survive for quite a while, and we'll see a range of prosumer camcorders using it, with decent cameras on the front (just like DVCAM does well in SD, because the cameras are good).

When HDV shows its warts, it's usually because the camera's feeding it with relatively poor video (like poor interpolation from the sensor(s), or poor filtering that allows aliasing through). Give it decent material and it stands up quite well.

[Ray Maher]

How about these.

The HDRUX1E disc recorder and the HDRSR1E hard disc recorder.

[JMCP]

Ray,

it looks like the HDRUX1E has already been superceded. Have a read here
http://www.camcorderinfo.com/content/AVCHD-Expands-to-Two-New-Models-in-the-Sony-DVD-Line.htm


Cheers John

[Gavin Gration]

HC 7 has mic/headphone - if it's good in low light I might be tempted.

[steve]

The interesting thing on the HC7 spec is that it calims a sensitivity of 2 lux! That's even better than the FX1.
That coupled with OIS should give some improvement over the HC1/3 models (assuming that signal processing has not gone backwards over the last year).
There's a lot of froth about the relative sensor cell count on some of the US forums, but at the end of the day, it's what goes on the tape that matters.

Steve

[Alan Roberts]

[ruminating mode]

I always take these specs with a pinch of salt. Simply claiming 2 lux tells me nothing about the quality of the picture it makes at that level, it probably means that it's just recognisable as a picture. In broadcast cameras, I'm used to reading specs saying that the sensitivity is F/x at 2000lux, which means the camera makes peak white from a white card when it's ullimunated at 2000lux and the lens is at F/x. That all makes sense, because the noise performance is then as specified. But simply quoting a light level tells us nothing about what sort of picturfe comes out of it (my A1 claims 0lux, meaning a picture with no light at all, and it's true if you're using it with infrared, but it tells me nothing about performance in normal lighting).

[/ruminating mode]

Now back to the sums :(

[steve]

Alan,

Shouldn't it be:

[/illuminatring mode]

Actually, that rumination was very informative. Have you any idea how the A1 would look if specified the pro way?

Steve

[Alan Roberts]

I'd guess it would be around F/6 at 2000 lux; modern studio cameras are around F/11 at 2000. You can see the noise measurements on my report (see my sig).

[infocus]

I always take these specs with a pinch of salt.
Absolutely. Lower end cameras seem to employ "noise reduction" techniques (ie soften and smear the picture) at lower light levels. They can give the same brightness of image for a given scene as a far more expensive camera (so have the same lux rating? ;)), but with a far greater drop in quality compared to their relative performances in decent light.

Same with resolution. I've finally been able to compare a DSR500 and a Z1 side by side (the latter in DVCAM mode) and was pretty shocked to find out how big the difference was - far more than I'd been expecting. To be honest I don't really understand why. The same recording resolution, and the Z1 downrezzing from more pixels. I can only suspect that if you measured system mtfs the DSR has a high value all the way virtually to the limit, whereas the Z1 starts tailing off early. (Even though it may have the ability to keep going beyond the limits of the SD system.)

[Alan Roberts]

Yep, that's about right. But don't put the DSR500 in the same category as the Z1, the DSR range are good, pro, cameras while the Z1 is a prosumer (actually, I still regard it as a consumer camera with pretentions).

[infocus]

Yep, that's about right. But don't put the DSR500 in the same category as the Z1, ............
This all really goes back to when the Z1 first launched, when many people saw it's resolution (in HDV mode) and said "better than Digibeta". My point at the time was that whatever the resolution, a 2/3" camera would be better in low light, have better connectivity, be more controllable etc etc. But I've never had the chance to compare the two side by side until now.

And that's what surprised me. All the above WAS true, but the Z1 was far softer cf the DSR500 than I'd expected. (That said, the Z1 is far cheaper, smaller and lighter!) Which just proves Alans point - be very careful what conclusions you draw from specs.

[StevenBagley]

It's worth remembering that the CCDs in the DSR500/570/450 is almost certainly has as many (if not more) sensors horizontally as the Z1 has and that they each sensor is 4 times the size of the equivalent on the Z1 (2/3" vs 1/3" chips).

Even vertically, the Z1 has to resample the interlaced HD signal to SD 576i and from what Alan's tests suggest it isn't using the best downconverter to do this, so I suspect that the signal from the 5x0s CCDs gives a clean sampling of actuality even though they aren't oversampled.

Couple this with a much better lens (the Z1's lens show chromatic aberration at times) and I'm not at all surprised that the DSR570-class produce much nicer pictures than the Z1.

If you want to get pro results, you need to use pro equipment.

Steven

[Alan Roberts]

That's about it. You get what you pay for, there are no free lunches in this business.

[infocus]

It's worth remembering that the CCDs in the DSR500/570/450 is almost certainly has as many (if not more) sensors horizontally as the Z1 has and that they each sensor is 4 times the size of the equivalent on the Z1 (2/3" vs 1/3" chips).
The manual for the DSR500WSPL (ie PAL) states 980x582 pixels/sensor (so not quite all active?), so pretty much the same horizontally as for the Z1. Yes, they're each bigger individually, so the sensitivity will be much better, but these numbers would indicate similar resolutions. And it's the scale of the difference that has surprised me - much more than I'd anticipated, and much more than the specs would suggest.

For my test I set up what is for me a "bread and butter" location. Simulated interview in average room lighting, with key and backlight of intensity to suit. Maybe the difference wouldn't have been so great in higher light levels? Maybe the Z1 is playing the consumer trick of trading off resolution to conceal noise? Not that it really matters - the scenario was a very relevant one for me.

[Alan Roberts]

The 500's 980x582 sounds right to me; it needs 960x576 for the image, so there's 20 pixels/line spare, and 16 lines. Those spare lines and pixels form a border all around the picture and are covered by a metallised layer so that they are "blind". They are ready as a normal part of the picture and the values in them are averaged and the result is subtracted from the normal video, the value accumulated is the leakage due to temperature. This makes the pictures automatically black balanced, including the effects of thermal drift. It's a standard practice in all but real top end cameras where you're expected to get it right as an operator (although I think even in those cameras this happens, so that there's no drift).

While I accept your reports on this, I'd still be happier with a lab comparison using proper test charts. That's the only way to get to the bottom of it. My problem is that I only ever get hold of one camera at a time, unless something special's happening.

[infocus]

While I accept your reports on this, I'd still be happier with a lab comparison using proper test charts. That's the only way to get to the bottom of it.
Absolutely. The more I look at cameras, the more difficult the whole business of valid comparisons seems to become - so many variables. And that's before we even think about line up differences and individual unit variations.

But I don't think you'd be unhappy with the lineup of the DSR500, and the Z1 was a hired model in day to day use - in isolation it seemed pretty reasonable. And I did make most of the variables constant in this case - same subject, same lighting, etc, and a good example of a common setup. Slight differences I'd put down to experimental error, but the differences here were pretty obvious.