HDX Latitude and Low Light Performance

[reservoir]

I can't remember if I read it here at dvxuser or on another forum, but I thought I remember someone saying that *if* the new HDX camera only has 1/3 CCD's that it will have less latitude and have poorer low light performance. Is there any truth to this. I think the dvx100a has...like what.... about 6-7 full stops that it can resolve? And the low light performance is pretty good as well? If this is true, what would help curve the problem? Using 1/2 CCD's or bigger? Just curious about this. I just wanted to see if I could get a more *explained* answer from you DVX-PERTS!! Ha, Ha, I made a funny!! ~reservoir~

[NewBee]

If a new camera is released that uses 1/3" CCDs *and* provides a native 16x9 Aspect Ratio, then your assumption is correct. It is the AR ( Aspect Ratio ) that is the key.

Canon, and Sony are now selling ProSumer Cams with Natvie 16x9 AR. They accomplish this by playing with the Geometry of the CCD ( due to the fact that a 1/3" CCD is too small to accomplish this any other way ). By playing with the geometry of the CCD, you effectively reduce the lattitude and LUX of the camera. So the theory is that if Panny is going to release a camera that will also provide this capability and it also uses a 1/3" CCD, then this camera will experience the same limitations ...

It is also possible that this will impact the quality of any 4:3 work that you might do with the camera. Because of the geometry manipulation of the CCD, Canon's 4x3 video is of lesser quality/resolution than it would otherwise be ...

Hope this Helps ... 8)

[Isaac_Brody]

Wait till NAB. Nobody know anything besides what's been mentioned, and what's been mentioned already is pretty amazing.

Let the camera speak for itself before we judge it.

[reservoir]

I wasn't judging....I was just asking *WHAT IF* questions. I'm sure whatever limitations the new camera WILL HAVE (for under 10k there will be a few guaranteed) that it will still be *THE CAMERA* to get. Until then we are all just speculating really. ~reservoir~

[xander76]

If a new camera is released that uses 1/3" CCDs *and* provides a native 16x9 Aspect Ratio, then your assumption is correct. * It is the AR ( Aspect Ratio ) that is the key.

I had always been told that pixel density was the key; the more dense the pixels on the CCD, the worse the latitude and low light performance. Under this explanation, the reason the XL2 has issues is that it uses a smaller area of the CCD for the same amount of pixels and thus increases pixel density.

I am not an expert on this at all, and I could be totally wrong. Under this theory, though, increasing the number of pixels (in order to shoot HD) and still using a 1/3" sensor should, all things being equal, reduce latitude and low light performance. Of course, we won't really know until we see the camera.

[Barry_Green]

I'd second what xander76 says. *Now, the thing is, this is dependent on all other things being equal. *We don't know if all other things will be equal -- what if the new camera uses CMOS sensors? *Or bigger chips? *Or has some hyper-HAD-style technology in it that helps get better performance?

It is my guess that if all other things were equal, then yes, more pixels = smaller pixels which typically means less latitude and less sensitivity. *But it's exceptionally premature to say that the new camera would have those restrictions, or anything like them, until we see the actual functioning product. I mean, an XL2 CCD has 960 pixels across, and an FX1 CCD has 960 pixels across, but watch 'em on an HDTV and the pixel-shifted FX1 beats the pants off the XL2 in horizontal resolution.

So the proof will be in the actual product, not in theoretical speculation.

[taubkin]

Hey, but a lot of the Latitude of the CCD's is already lost in DV compression. Check The Andromeda tests. We are not using the full latitude of the CCD's with 12bit A/D conversion. Maybe with DVCPRO50 or 100, the latitude could be even higher than the DVX's...

[NewBee]

Hmmm ... that's and interesting point Taubkin. I hope you are right.

It is way too early to pass judgement ... I am looking forward to testing the new camera myself when it is released ...

[reservoir]

I have a question about CCD's and CMOS. I know what they are and what they stand for (Charged Coupled Device) & (Complimentary Metal Oxide Semiconductor) but Exactly what would be the difference if the new HDX cam uses standard CCD's or if it uses CMOS instead? Better quality I'm guessing. Canon's highest end Digital SLR's (EOS-1DS Mark II - 16.7 Megapixel CMOS - retail $8,000) uses a CMOS chip so I'm guessing they are great for high end picture quality. Please fill me in on anything I'm missing? I'd love to hear a detail description about the pros and cons of CMOS. Thanx all!! ~reservoir~

[Robert_Niemann]

Just found this one: http://www.ccd-sensor.de/english/html/ccd_today.html. Maybe that helps.?

[Sumfun]

The main advantage of CMOS is that they are cheaper to manufacture because most semiconductor companies produce millions of CMOS chips every year (not necessarily CMOS sensors). CMOS also allows for larger size and faster data scan rate.

The traditional advantages of CCD are that is has better light sensitivity and better color reproduction. But Canon has been using CMOS sensors in its high end SLR cameras for several years now with excellent results. So good that Nikon announced that it will be using a CMOS sensor for the first time in its new D2X pro-level camera.

Regarding the earlier discussion regarding low light performance, I believe that light sensitivity is relative to the area of each pixel. For example, say that you have two 1/3 inch CCD's, the first with 500k pixels and the 2nd with 1M pixels (just to use round numbers). Well in order to fit 1M pixels into the 2nd CCD, the size of each pixel must be roughly half of the pixel size in the 1st CCD. Because the pixel is half the size, then only half the light falls on it - which makes it less sensitive to low light. I'm sure there's some signal processing you can do to increase sensitivity, but that's the basic theory.