low light xl h1 vs. xl2

[Gregory S. Ouellette]

this is such an important question to me that i wanted to separate it from my previous post.
i like to shoot a lot of very low light work for effect. (picture: confessional interior). my used xl1, being my first 3 chipper, does quite well compared to what i am used to being able to do.

question 1: is the sd quality for this kind of work better on the xlh1 vs. xl2?

question 2: is the hd quality of the xlh1 for this kind of work far better than the sd on either unit? i seem to recall an post but cant find it.

thx

[Chris Hurd]

Typically when you're talking about the same size CCDs (one-third inch in the XL2, one-third inch in the XL H1), the camera with more pixels is the camera with smaller pixels. They've gotta be smaller, in order to cram more pixels into roughly the same physical area. The smaller the pixels, the worse the low-light performance is. In other words, HDV camcorders tend to be less sensitive in low-light situations than DV camcorders. This is just a generalization though. There are always those exceptions which prove the rule, but this ain't one of them.

[Jonas Nystrom]

...according to previous threads and my own experience, after some "tweaking" of the presets, H1 have great low-light performance.

[Jon Bickford]

i have both the H1 and the xl2 and they are very close in low light the xl2 just barely winning out, chris is right but somehow a year of R&D and double the price made those little fellas in the h1 scream

[Tony Davies-Patrick]

Greg - I find the low light performance of the XL1 quite poor in natural light(as far as maintaining quality without grain), and although the XL1s is slightly better, I still find it cannot cope very well in low light without giving a lot of noise. If you just keep the gain to -3 or 0 it helps keep noise at bay, but then you just get no detail at all in the shadows when working in low light levels when the subject is moving a lot. I was hoping the H1 would vastly improve matters, but what I've seen and heard so far has not given me hope...although the H1 would at least possibly provide a 'cleaner' image at low light levels (anybody know?).

For a lot of my work I am shooting around water, so more often need a polarizer over the lens to cut down on harsh reflections, and will use an f/stop of around f/5.6-f/8 to maintain enough depth of field with some subjects. So I'm already placing my camera at a disadvantage of extra stops by adding the filter, even before light levels begin to drop. Sometimes I need to film fish in the wild from above the water and need that polarizer to provide enough details in the subject. Pointing a lens downwards eats light, and with the added filter it eats more, and without using some gain you'll just be filming dark shadows instead of subjects with colour and detail.

If the XL-H1 is not going to improve on the XL1s & XL2, then it is another negative point for the H1. I'd really like to see some footage of the H1 & Xl2 side by side used outdoors in low light levels without lights.

Jon's mention that the H1 is slightly worse than the XL2 has made me hold back my thoughts about buying the H1 just yet. With the added points that the H1 is also slightly worse than the Xl1s/Xl2 with audio, then the positives about the H1 are beginning to looking far less rosy.

[Chris Hurd]

Originally Posted by Tony Davies-Patrick

Jon's mention that the H1 is slightly worse than the XL2 has made me hold back my thoughts about buying the H1 just yet.

I guess I didn't explain this very well. It's not a matter of the H1 being "slightly worse" than the XL2. The issue is simply high definition vs. standard definition. In the one-thid inch chip size, high definition camcorders *in general* suffer more in low light than do standard definition camcorders. That's across the board. It's the same way with the Sony FX1 vs. the VX2100. In HDV, low light performance is a format issue is a format issue, not a camera issue, for the reasons I explained above (smaller pixels don't perform as well as bigger pixels in low light).

I don't think it's fair to focus on one aspect -- low light performance -- and make a sweeping generalization that the H1 is "worse than" the XL2. Because in sooo many other ways, it's so much better. Besides, did you read the part where Jon says the XL2 just *barely* wins out? Don't make a mountain out of a molehill.

Same thing applies to audio... you can argue either way:

XL2: PCM 4 ch. audio, XL H1: 192kbps audio; so XL2 is better than the H1.
XL2: 12 bit 4 ch. audio, XL H1: 16 bit 4 ch. audio; so XL H1 is better than XL2.

See what I mean? It's a mistake to focus on one aspect and overlook the big picture.

[Tony Davies-Patrick]

I'm not making sweeping statements, Chris, but just explaining that low light performance is very important to me. A camcorder costing three times more is expected to perform better in many aspects (especially build & image pixels) ...but if it doesn't also improve dramatically in some other very important fields - such as sound and low light - then I will think twice about upgrading.

I use 12 bit 4 channel a lot...but in reality, would the H1 in 4 ch modes actually provide a better quality audio to the ear? Reading other posts I had thought that actual audio was not as good as with the XL2 (if similar mics were connected). If there is no actual 'real' difference in audio performance between the XL2 & H1, then my apprehension about audio problems is unfounded.

[Chris Hurd]

Originally Posted by Tony Davies-Patrick

A camcorder costing three times more is expected to perform better in many aspects (especially build & image pixels) ...but if it doesn't also improve dramatically in some other very important fields - such as sound and low light - then I will think twice about upgrading.

Well, the simplest way I can express it is that smaller pixels cannot perform better than larger pixels, all other things being equal (such as CCD size). Again, this is a limitation of *format* not camera and it's one area where HDV suffers relative to DV. While there's no physical way for a smaller bucket to improve dramatically over a larger bucket in terms of how much water it can carry, it is amazing how the actual hit in low light performance is dramatically reduced. John said the difference is "just barely noticeable," and that's thanks primarily to the H1's significantly improved DSP.

XL2: ea. 1/3rd-inch CCD has 480,000 effective pixels.
XL H1: ea. 1/3rd-inch CCD has 1,560,000 effective pixels.

The H1 chips have more than three times as many pixels as the XL2 chips of roughly the same size. Therefore the pixels on the H1 are about three times smaller than those of the XL2. So you'd expect the low light performance of the H1 to be about three times worse than the XL2, but it isn't, instead it's "just barely noticeable" because the H1 has a much better, much improved processor relative to the XL2.

Regarding audio performance and 4-channel 12-bit PCM vs. 4-channel 16-bit 192kbps, I think that's the subject for a case study. To me it's just amazing that we can have four channels to begin with.

[Tony Davies-Patrick]

Are you now using, or own the H1 Chris?

[Chris Hurd]

Nope. I do not own nor am I currently using the H1. I'm fully trained on it, and I have a fair amount of experience with it, but it's just a little too expensive for me. I'm holding out for an HDV replacement of the GL2, if there ever is one this year.

[Kevin Shaw]

Someone posted a link recently to a German web site which compared current low-cost HD cameras, and showed frame grabs indicating the XL-H1 is significantly better than others (e.g. Sony) in dim lighting. Combine that with the comment here from someone who's compared the XL-H1 to the XL1 and concluded they're similar in low light, and it sounds like the XL-H1 has promise in this regard. Based on my own experience with the Sony FX1 versus the Canon GL1/GL2, I'd say the Sony is about one F stop less sensitive with less image noise, so if you gained that stop of sensitivity back on the XL-H1 and had a clean HD image that would be worth talking about.

[Jon Bickford]

the issue with smaller pixels is very much like film grain, an IS0 of 100 will have a much higher resolution but it won't work in low light as compared to an ISO 800 film because if say 1 foot candle hits one pixel in the XL2 it has to be distributed across 3 or more pixels in the H1

to clarify my opinion on the subject:

of the 5 gigs i've shot 30 far with the H1 only one of them was done in HD as my clients needed something they could cut with so i'm pretty accustomed to working with it in SD and I can definetely say that is is a pretty big improvement in this area over the XL2, and i loved the XL2. the low light performance is very very close but the extra control over color, gamma, white balance etc. The workings of the new lens which has a much much better stabilization that doesn't do the rubber band snap when starting a pan, the focus distance readouts and the arbitrary focal length measurement of 1-99 which after awhile becomes very useful to know, "so and so's face looks best at a 23". That and the superior view finder all conspire to make this a much nicer standard def camera to work with.

In standard def the audio isn't compressed so it is the same quality of recording, one marked inprovement is that the headphones pre-amp is much stronger than the XL2 Which was nowhere near loud and clear enough for a boom operator to properly monitor audio on set.

I have been very very happy with my purchase thus far and have found very few limitations that you wouldn't expect on a camera with a 4 digit price tag.

-Jon

[Chris Hurd]

Thanks for that excellent reply John. So you're from Pedro. I think I said this once before already, but if you run into Watt, say hello for me.

[Greg Boston]

Originally Posted by Tony Davies-Patrick

Greg - I find the low light performance of the XL1 quite poor in natural light(as far as maintaining quality without grain), and although the XL1s is slightly better, I still find it cannot cope very well in low light without giving a lot of noise. If you just keep the gain to -3 or 0 it helps keep noise at bay, but then you just get no detail at all in the shadows when working in low light levels when the subject is moving a lot. I was hoping the H1 would vastly improve matters, but what I've seen and heard so far has not given me hope...although the H1 would at least possibly provide a 'cleaner' image at low light levels (anybody know?).

For a lot of my work I am shooting around water, so more often need a polarizer over the lens to cut down on harsh reflections, and will use an f/stop of around f/5.6-f/8 to maintain enough depth of field with some subjects. So I'm already placing my camera at a disadvantage of extra stops by adding the filter, even before light levels begin to drop. Sometimes I need to film fish in the wild from above the water and need that polarizer to provide enough details in the subject. Pointing a lens downwards eats light, and with the added filter it eats more, and without using some gain you'll just be filming dark shadows instead of subjects with colour and detail.

If the XL-H1 is not going to improve on the XL1s & XL2, then it is another negative point for the H1. I'd really like to see some footage of the H1 & Xl2 side by side used outdoors in low light levels without lights.

Jon's mention that the H1 is slightly worse than the XL2 has made me hold back my thoughts about buying the H1 just yet. With the added points that the H1 is also slightly worse than the Xl1s/Xl2 with audio, then the positives about the H1 are beginning to looking far less rosy.

With all due respect Tony, I think the XL-2 is 'light years' (pardon the pun) away from the XL1/s in terms of low light capability. The simple reason...better noise processing algorithms in the XL-2 allow you to bring in 6db of gain without seeing much grain if any. If you do, there's a noise reduction control that will wipe the noise right out of the picture at the expense of some ghosting if there is fast movement in frame. But you couldn't even do that with the earlier cameras. The biggest thing that convinced me to buy the XL-2 was the final shot of the night footage from Vienna, Austria. It was at +18db, yes, +18db and I couldn't believe my eyes. I still love the picture from my venerable Xl-1, but the Xl-2 gives me the benefit of 6 yrs. advancement in low light handling. And it shows.

-gb-

[Tony Davies-Patrick]

Greg - I agree with most of what you have said. I know that the XL1s improved on the XL1, and that the XL2 improved on the XL1s. My main question was would the XL-H1 improve on the XL2 in low light enough for me to shell out the extra cash. I have not yet used the XL-H1, so was trying to gain answers from those that have used one in low light outdoor work without extra lighting.
So far the answer seems to be that the H1 does not improve drastically over the XL2 in low light levels (at least not enough for my kind of work). I was tempted to sell all my gear for the H1, but am still on the fence on this one. I'm not interested in the new AF 20X lens (I prefer MF lenses), and have so far found no need for an image stabiliser for any off the shoulder work, or even when working from a boat in choppy waters. It will probably be wiser for me to stick with using two bodies - the XL1s & XL2 (plus all the extra camera and editing equipment that I've bought for them) rather than moving to the H1; and maybe wait until Canon adds a good wide angle to the H1. Hopefully they will later offer a package of either the body only or H1 body + wide angle lens (as long as the the new lens improves on the earlier AF 3X lens).