Has anyone seen comparisons of the low light performance between the Sony HVR-V1U and the Canon HV20 Camcorders? (Since both have CMOS sensors).
I'm not talking manufacturers specs, because they can seldom be realistically compared.

I haven't seen anything to compare, but you are talking about 2 cameras in completely different classes. The V1U has 3 CMOS chips while the HV20 has 1. I would be very suprised if the HV20 beat the V1U.

Thanks,
I realize that they are totally in different classes, but I am curious about how they compare on that one specification.

I may be very wrong on this, but wouldn't a single CMOS chip that is 1/2.7" in size capture more light than 3 1/4" chips since the light in the V1U (and other 3 chip cameras) goes through the lens and the has to be split in a prism (thus loosing some brightness) to the 3 RGB chips?

I may be very wrong on this, but wouldn't a single CMOS chip that is 1/2.7" in size capture more light than 3 1/4" chips since the light in the V1U (and other 3 chip cameras) goes through the lens and the has to be split in a prism (thus loosing some brightness) to the 3 RGB chips?

You are not wrong; the HV20 has better low light performance than the V1, probably 2 stops, but not because of a "split in a prism." The HV20 actually uses a color filter to seperate the RGB to give it a 3 CCD look, so the HV20 would also lose some brightness with this filter. The chip size is the key; the larger the chip, the better low light performance. Of course, other factors like the number of pixels and the lens also make a difference. That's why the HV20 and HC7 only have a 10x lens.

The 3 sensors in cameras such as the V1 will give you brighter, more vivid and life-like colors, but not necessarily better low light performance. Now if the V1 had 3, 1/3 inch sensors, then I believe it would be better in low light than the HV20 or HC7.

You are not wrong; the HV20 has better low light performance than the V1, probably 2 stops, but not because of a "split in a prism." The HV20 actually uses a color filter to seperate the RGB to give it a 3 CCD look, so the HV20 would also lose some brightness with this filter. The chip size is the key; the larger the chip, the better low light performance. Of course, other factors like the number of pixels and the lens also make a difference. That's why the HV20 and HC7 only have a 10x lens.

The 3 sensors in cameras such as the V1 will give you brighter, more vivid and life-like colors, but not necessarily better low light performance. Now if the V1 had 3, 1/3 inch sensors, then I believe it would be better in low light than the HV20 or HC7.

Thanks for the informative discussion. I appreciate the feedback.

The light in the HV20 is split three ways with a color filter. Also, the overall light-gathering area of the V1 is .75" while the 1/2.7" chip is .37" at about half the size. Also, I have not seen footage that shows me that the HV20 is better than the V1 in low light. If it's two stops better than the V1, it will rival the PD170 and I do not believe that at all. Has anyone directly compared these two cameras in the same lighting conditions and location?

The light in the HV20 is split three ways with a color filter. Also, the overall light-gathering area of the V1 is .75" while the 1/2.7" chip is .37" at about half the size. Also, I have not seen footage that shows me that the HV20 is better than the V1 in low light. If it's two stops better than the V1, it will rival the PD170 and I do not believe that at all. Has anyone directly compared these two cameras in the same lighting conditions and location?

I don't believe it is that good, and I have VX2000. Grain gets nasty real quick, too, on the HV20.

I found a German site that tests most of the cameras. Some can be understood, and the screen grabs of various tests are very good. You have to use the FX7 as a reference for the V1U.

http://www.camcorder-test.com/

I downloaded the low light (200 lux) screen grabs from both the HV20 and the FX7 and lightened them up to simulate post production correction; I then layered them with the XH-A1 shot. I tried to correct them so they were close to the XH-A1. I also layered the resolution test charts. I then viewed the images on a Samsung 56" 1080p TV. (Note that these are all 1440 x 1080 pixels, so I resampled them to 1920 x 1080 pixels) I was surprised that both the V1U and the FX7 looked better than the XH-A1 (at least to me), and were also visibly sharper on the test chart. The FX7 and the HV20 looked very close in performance.

The light in the HV20 is split three ways with a color filter. Also, the overall light-gathering area of the V1 is .75" while the 1/2.7" chip is .37" at about half the size. Also, I have not seen footage that shows me that the HV20 is better than the V1 in low light. If it's two stops better than the V1, it will rival the PD170 and I do not believe that at all. Has anyone directly compared these two cameras in the same lighting conditions and location?

.75"? The V1 doesn't have 2/3rd inch chips. .25" is the light gathering surface of the V1. You can't add the three chips because each chip has a job of gathering information for each color. That is why I said the colors are brighter and more vivid. One thing I forgot to consider, though, was the angle of the V1 chips, making the surface area appear larger. To be specific, the V1 also has a larger lens aperature (1.6 vs 1.8) and larger pixels. Both of these factors help its low light performance. However, even with these improvements, the V1 is rated at 4lux minimum (F1.6 and 18db gain) and the HV20 is rated at 3lux minumum (not including their marketing 0.2lux with slow shutter). The Sony HC7 is rated even better at 2lux minimum, but that's with their auto slow shutter on which is 1/30 second.

I know I was over exergerating the light performance difference. I thought that the V1 was a 7lux minimum. Still, the HV20 is about a 1/2 stop better. Of course the V1 is a better overall camera. That was not the question. The question pertained to better in low light. The HV20 is slightly better than the V1 in low light.

I was surprised that both the V1U and the FX7 looked better than the XH-A1 (at least to me), and were also visibly sharper on the test chart. The FX7 and the HV20 looked very close in performance.

I'm not surprised if this was a low light test because of the nature of CCDs vs CMOS chips in high definition. CCDs get significantly noisier as the resolution climbs. The newer CMOS chips don't suffer the same fate. I won't get into the technical jargon of why this is so, but That's why you will see more and more CMOS chips in high definition cameras in the future.

".75"? The V1 doesn't have 2/3rd inch chips. .25" is the light gathering surface of the V1. You can't add the three chips because each chip has a job of gathering information for each color."

For the same reason the three 1/4" chips on the V1 can't be added up to 3/4", the HV20 can't consider it's entire surface area toward it's light-gathering capability. The HV20 and V1 both have their chips split up between the three colors so the HV20 doesn't have full 1/2.7" for each color. It doesn't matter how you divide the chip, the red and blue sensitivity will always be reduced by either a color filter or a beam splitter.

Thanks for that link, Robert. The test is helpful and the HV20 and FX7 look to have similar abilities in low light. Unfortunately, the lighting is different for each camera so there is a big glare on the FX7 contrast chart. Regardless, it is fairly obvious that the two cameras are both in the same ballpark in low light.

".75"? The V1 doesn't have 2/3rd inch chips. .25" is the light gathering surface of the V1. You can't add the three chips because each chip has a job of gathering information for each color."

For the same reason the three 1/4" chips on the V1 can't be added up to 3/4", the HV20 can't consider it's entire surface area toward it's light-gathering capability. The HV20 and V1 both have their chips split up between the three colors so the HV20 doesn't have full 1/2.7" for each color. It doesn't matter how you divide the chip, the red and blue sensitivity will always be reduced by either a color filter or a beam splitter.

Thanks for that link, Robert. The test is helpful and the HV20 and FX7 look to have similar abilities in low light. Unfortunately, the lighting is different for each camera so there is a big glare on the FX7 contrast chart. Regardless, it is fairly obvious that the two cameras are both in the same ballpark in low light.

Incorrect. You are confusing light gathering capability with color. The color filter on the HV20 is just that, a filter to divide the colors. Yes, it factors in the overall light sensitivity, but it has nothing to do with the light gathering surface area of the chip. The chip as you pointed out has a .37 inch surface in which to gather light. The V1 has 3 chips .25 inch dedicated to each color, but the surface area to gather the light is .25 inch. The surface area appears larger than a standard 1/4 inch because of the 45 degree angle of the chips, but still the V1 has less area to gather light than the HV20. Larger pixels and a larger lens aperature as I pointed out helps the V1, but the fact is this camera is not quite as light sensitive as the HV20. You can argue this all you want, but you cannot deny the science that cameras with a larger chip have a better low light performance.

I also pointed out if you read my comment that if the V1 had 1/3 inch chips, it would beat the HV20 in low light, but that is not the case. Again I emphasize that I believe the V1 is a better camera overall than the HV20, but the HV20 is more light sensitive. That is a fact and not my opinion.

Agreed. Three 1/4 sensors will not compare to one 1/2.7" sensor in light gathering abilities.

I was surprised that both the V1U and the FX7 looked better than the XH-A1 (at least to me), and were also visibly sharper on the test chart. The FX7 and the HV20 looked very close in performance.

Not to me. In my 24" 1920x panel the A1 looks better in all regards. Sharper, no moire as in the hv20 and also brighter. FX7 has a lot noise reduction that makes the resolution drop down.

Each camera has their light-gathering surface divided by color. On third of the V1's chips are dedicated to each color and roughly one third of the HV20's chip is dedicated to each color. I say roughly because green likely has more area devoted to it than the red or blue. Regardless, the pixels won't be larger on the HV20 because three times as many need to be put on the same chip to get pixels the same size as one .25" V1 chip. You can't put larger pixels on the HV20 chip if it isn't three times the size of the V1's chips combined. It doesn't matter how you divide the total surface area.

I have yet to see any test or footage from the HV20 that looks brighter than the V1. It looks like a very nice camera for the money and form-factor, but it doesn't have larger pixels than any of the 3-chip HDV cameras.

Each camera has their light-gathering surface divided by color. On third of the V1's chips are dedicated to each color and roughly one third of the HV20's chip is dedicated to each color. I say roughly because green likely has more area devoted to it than the red or blue. Regardless, the pixels won't be larger on the HV20 because three times as many need to be put on the same chip to get pixels the same size as one .25" V1 chip. You can't put larger pixels on the HV20 chip if it isn't three times the size of the V1's chips combined. It doesn't matter how you divide the total surface area.

I have yet to see any test or footage from the HV20 that looks brighter than the V1. It looks like a very nice camera for the money and form-factor, but it doesn't have larger pixels than any of the 3-chip HDV cameras.

I never said the pixels were larger on the HV20. In fact I said, "larger pixels and a larger aperature help the V1..." The V1 looks brighter in low light because the colors look better.
How can I explain this, so you understand? Let me try this:

Sensors do not know color. The light is gathered on a .37 inch grid in the case of the HV20 and then passes through a color filter to produce the color image. Color suffers, but brightness does not, except for smaller pixels. The V1 has 3, .25 inch sensors that are used for each color. They are still gathered by sensors that do not know color, but since they are split by a prism, the light gathering surface is still .25 inch (all three are not gathered to one larger grid). The result is sharper, more vivid and more real colors, but not better light perfomance. The angle of the chips and larger pixels give the V1 an advantage, but remember the advance optics used to produce the color image hurts its low light capability much more than a color filter.

The fact is if you put both cameras in a room with only 3 lux of light, the HV20 will produce an image; whereas, the V1 won't. From the factory, the V1 is rated at 4 lux. Sony is not going to underestimate the light capability of their camera. In fact, they're going to pad it. We're talking low light capability, and even with Sony's tricks, the V1 will not win this battle. By the way, I don't have a HV20, I have an HC7 which is my backup camera. The V1 is one of the cameras I'm considering for my main camera. If the V1 had 1/3 inch sensors, it would be an easy decision; the V1 would be my main camera.