Canon XH A1 actual capture?

[John Huling]

I read that the Canon A1 captures moving images at 1440 x 1080. Just to clarify from stills.
Sony says their new V1 is 1920 x 1080.
1) Do I understand this correctly?
2) Does it make a difference in image quality in the final product? If so how.

[Chris Hurd]

Hi John, the short answer is that the recorded resolution going to tape for any camcorder in the 1080i HDV2 format, including both of the cameras you mention, is 1440 x 1080. It has to be, or else it wouldn't be HDV.

Now to get technical, the Canon 1/3" CCD image sensors are each 1440 x 1080 on the chip with an additional resolution boost from horizontal Pixel Shift. Their advertising literature doesn't make a big deal out of Pixel Shift because most 3-chip camcorders already use some form of it anyway. But the benefit is real, and the information going into the camera's Digital Signal Processor (the Canon DSP is their own proprietary Digic DV HD processor) is something "significantly greater than" 1440 x 1080.

Meanwhile the Sony 1/4" CMOS image sensors are each 960 x 1080 on the chip with a substantial resolution boost from their patented ClearVid layout (described here). Their advertising literature *does* make a big deal out of this because ClearVid is new and unique, and different from Pixel Shift. The benefit is real, and the information going into the camera's Digital Signal Processor (the Sony DSP is their own proprietary Enhanced Imaging Processor) is also "significantly greater than" 1440 x 1080... which they state as 1920 x 1080.

Again, it's important to note that both cameras downconvert to 1440 x 1080 for recording the HD signal to tape. Your question is does it make a difference in image quality in the final product (between Canon's 1440 x 1080 CCDs plus Pixel Shift vs. Sony's 1920 x 1080 derived from 960 x 1080 ClearVid CMOS)? Hard to tell, since the Sony V1 isn't shipping just quite yet. My take on it is that the HDV tapes they produce will most likely be very close to each other in the quality of images they're recording. Once again I hate to slip into broken record mode, but if you ask me, the biggest difference between them will be neither their sensor sizes, sensor types, pixel counts, resolution boosting methods nor image processors... but rather, as always, ergonomics, form factor, specific feature sets, and subjective determination of the flavor of images they produce.

Honestly I don't think you can go wrong either way. It'll probably have to come down to a hand-holding test to find out which one smiles back at you the most brightly. Hope this helps,

[John Richard]

I would add to Chris' evaluation that the importance to many of the lens differences between the 2:

- Quality of results from the H1
- Options afforded by removable lenses for the H1

[John Huling]

Wow! A great explanation. thanks! I do have the XH A1. So far so good. Shot the stuff I wanted the other day. I lose track which days is which. But I slowed it down in my editing program Vegas 7. With the help from this great forum of course. I have to say here Chris as I may be repeating my self. But this is one great place to find info with very helpful people. Thanks!

[Harm Millaard]

Head scratching going on....????

I thought this was about the A1 vs. V1. Both have fixed lenses. Why the sudden H1 reference?

[Michael Mann]

Addition to Chris' explanation:
The difference in image sensor size (1/3'' Canon vs. 1/4'' Sony) should result in a (visible!) difference in depth of field for a given aperture.

[Bill Pryor]

That would be true--you will have a much broader depth of field with the smaller chips. Of course, with a 1/3" chip camera you have a much broader depth of field than with a 2/3" or 1/2" chip camera. I shoot a lot of documentary style interviews, and I've found I can control the depth of field adequately with 1/3" chips if I shoot at wide open or nearly wide open apertures. I can't totally blur the background as I can with a bigger chip camera, but can soften it enough so the subject pops, which is usually all I want. Most of the interviews are head and shoulders, with some ECUs cut in. With a 1/4" chip camera wide open, the background is going to be fairly sharp even wide open. However, to make the subject pop, you can light the shot so the background goes down between one and two stops.

It's interesting to me how smaller chips are gaining more acceptance. It wasn't that long ago that I only used smaller chip cameras for pro bono work I do and things that didn't need to be really good. But now, with HDV, I'm getting just as good quality with the XH A1 (and also with the Z1 I've used) as I do with the 2/3" chip DSR500 in DVCAM. By the time it's all edited and output as SD, the HDV-originated stuff still looks better. However, you still get that good depth of field control with bigger chips.

[Matthew Nayman]

On a totally unrelated top,

Michael Mann... fitting for someone into digital cinema :)

But yes... the bottom line is, both cams output to HDV with m2T compression and 4:1:1 colour sampling (or 4:2:0 playback), so we have to wait and see about optics and other subjective characteristics.

[Jerome Marot]

Originally Posted by Chris Hurd

if you ask me, the biggest difference between them will be neither their sensor sizes, sensor types, pixel counts, resolution boosting methods nor image processors... but rather, as always, ergonomics, form factor, specific feature sets, and subjective determination of the flavor of images they produce.

...and low-light capabilites, I think.

[Michael Mann]

Originally Posted by Matthew Nayman

Michael Mann... fitting for someone into digital cinema :)

I am not responsible for my American alter ego :)

[Bill Pryor]

Yeah, low light capabilities too. The smaller the chip, the less the low light capability, with the same number of pixels. And the more difficult it is to make a wide angle lens that's wide enough.

[Matthew Nayman]

I think the DOF will be a quantitative differnece too. Might subtly suggest to viewers this is cheap video... we have all seen 1/4" endless DOF off of digi 8 cameras...