Hands-on Canon GL2 Report, Part One
by Barry Goyette, August 2002
I’ve had about 3 weeks to play with the new Canon GL2, and I thought this was a good time to offer my impressions about the camera, and also to give you the details of a comparison test I made utilizing the GL2, the XL1S and the Sony PD150.
I’m an owner of a GL1 and XL1S, so I know these cameras intimately. The PD150 was borrowed from a friend of a friend, and I have to say that in the few hours I spent working with it, it is a formidable camera indeed, so I’m glad that somebody here at dvinfo suggested that I include it in the trials. Because of the limited time I had with the PD150, it is only being used here in direct image quality comparisons. Occasionally, I will include a comment about the GL1 as referenced to an improved feature on the GL2, but otherwise I will limit my comments to the GL2, XL1S and the PD150. Except where noted, all cameras were set to factory default, and exposures were made in manual mode. In most cases the white balance used was the default daylight setting on each camera. Three image quality comparisons were made.
High Contrast Line Resolution
While photographing test charts can provide some useful comparative information, my experience with Dv is that the compression and 720×480 file size quite often become a limiting factor that makes any conclusion based on a LOT of subjective evaluation. Utilizing a 1956 EIA resolution chart, the GL2 maybe showed a slight improvement in horizontal resolution over the XL1S, but maybe is the important word. Using, another chart of my construction utilizing line groupings varying in 5% increments, the XL1S, GL2 and PD150 all resolved the exact same set of lines. Other information on the chart showed that the XL1S was less sharp overall, with the GL2 and PD150 running neck and neck–higher contrast things look sharper on the PD150, lower contrast areas were sharper on the GL2. The GL2 image also suggests a slightly lower overall contrast than the other cameras. The pd150 and GL2 exhibited slightly more noise than the XL1S in the neutral areas of the test chart. Overall, the color balance of the GL2 was slightly more neutral than the pd150. The XL1S was marked by a strong red shift, especially in the highlights. One interesting thing of note is that the PD150 rendered the chart slightly taller than the Canon cameras. (I’ll check to see which camera is more accurate, but my hunch is that it is the Sony).
For the reasons stated above, I am suspect of any information generated from a typical line resolution regime, at least as it is relevant to the DV format. In my opinion, the proof is in the puddin’…..unless you photograph black and white lines for a living. In this test I photographed a small still life that contained items of a relatively neutral color range, and moderate amounts detail. In this test again, the XL1S stood apart from the other camera’s with it’s softer “almost out of focus” look and reddish color shift. The GL2 appeared slightly sharper than the PD150, although this appeared to be as a result of higher electronic “sharpening”. Actual resolution of the two cameras appears to be about the same.
The GL2 excelled above the other two cameras in its ability to render highlight detail. In the still life test, the cameras were matched by monitoring the zebra stripes in a particular area of the scene, the GL2 showed significantly more range in that highlight even when its exposure was increased by nearly a full stop. I believe this is the most significant improvement in the camera, when compared to the gl1 and XL1S, and is responsible for giving the GL2 more film like look than it’s predecessors. The PD150 rates a close second in this respect. Still images for the above comparisons can be found at http://homepage.mac.com/barrygoyette/PhotoAlbum5.html
I photographed my friend Brian Lawler doing a mock Apple “Switch” testimonial using the three cameras simultaneously. The PD150 produced what I thought was the most pleasing skintone, with the GL2 slightly less warm. Again, the GL2 image looked somewhat “sharpened”, but otherwise was a very close match for the PD150. The XL1S picture again showed excess red and softness. The cameras were matched with a zebra pattern on brian’s silver hair. Here the XL1S showed a glaring difference in the highlight area, one that is so great that I plan to redo the test to make sure. You can see the film at http://homepage.mac.com/barrygoyette/iMovieTheater3.html
About 1/3 of the way through you will see a clip that is glaringly different… that is the XL1S (the sorenson 3 codec has magnified the differences and problems significantly, so take what you see and divide by three). My apologies for the sound…
The sensitivity of the GL2 became something of a question mark as I put it through a variety of tests. With all cameras set to 1/60 of a second and 0 gain, The GL2 required an aperture ranging from .7 to 1.7 stops larger than the XL1s, depending on the lighting situation, to reach an equavalent zebra pattern in the highlights. (the difference with the PD150 was even greater). What confuses me is the inconsistency of this variance, but my guess is that is has something to do with the GL2’s lower contrast, and it’s improved highlight rendering that allows the exposure to vary somewhat (in comparison with the other cameras) depending on the overall contrast of the scene.
This lowered sensitivity is not necessarily a bad thing. In the still life test with the PD150, the GL2 f-stop was 3.2 versus the sony’s at 6.8. In this situation the GL2 with it’s smaller chip appeared sharper, and had lower (and thus more controllable) depth of field than the larger chip PD150.
One of the first things I noticed when I hooked up the GL2 to a monitor in my studio, was that the GL2 showed quite a bit of noise when I pointed at moderately lit, or darker areas of my studio. As I worked with the camera more over the coming weeks, I came to realize a couple of things…
First, that this noise wasn’t necessarily more than what I have experienced with my other Canon cameras. It was just sharper (and a bit finer), and that by adjusting the sharpness setting on the GL2 downward a notch or two this noise could be minimized or eliminated. In brightly lit situations, the GL2’s noise level is almost invisible, and on par with the XL1S and the pd150.
Second, when I registered this first impression, the camera settings I used were AV exposure mode and f/4 to f/5.6. Due to the lower sensitivity of the camera, I was forcing it to increase the gain quicker than the XL1S or GL1 would. In low light situations, I would recommend using the Manual exposure mode, or at a minimum, avoiding the AV mode to minimize this problem.
Please direct questions to the DV Info Net Community Forums.