Interesting lens issues

[Shekar Dattatri]

In preparation for a big shoot next week I've been putting my recently acquired XL-2 (Pal) to the test. I came across some interesting findings, which I thought I would share.

1) The first thing I noticed was that there is virtually no difference in magnification between my 16x manual Canon lens and the new 20X lens at full tele! (yes, this applies for 4:3 as well as 16:9). I find this weird, to say the least. I then compared both lenses against my old XL-1 16x auto lens and found that the magnification with this lens at full tele was less than the other two, but not by a huge degree. So what's the deal with this 20x business?!

2) Wanting to test the camera at really low light levels, I boosted gain to +18 (shooting 50i) and ran the test in a room with very low ambient lighting. For the test I shot the colourful cover of a book that was lying around and did both close up and wide shots. Locking the camera on my tripod and, without changing settings in between shots, I tried the test with three lenses: my three and half year old Auto 16x (which came with my XL-1), my manual 16x Canon lens (about 8 months old) and my brand new 20 X lens. The results were quite startling.
The worst image came from the 20X - cold and very grainy - followed closely by the 16x auto lens. The manual 16x was a revelation. The picture quality was substantially better. If you pressed me to quantify based on eye judgement, I would say there was at least a 100% improvement in picture quality. Noise was considerably reduced and the tone was warm and 'friendly'. Whereas I would consider the shots with the first two lenses in that situation as eminently unusable, the same shot captured with the Manual 16x was eminently usable. I'm glad I have the Manual 16x and have decided to make this the default lens on my camera. I think I will use the Auto 20x only in situations requiring image stabilisation. I've heard it said many times that there couldn't possibly be any difference in image quality between the Canon auto and manual lenses as it's all the same glass. Well, I've also heard it said that the proof of the pudding is in the eating! I've eaten, and the winner for me is the 16x manual on all counts. If the flourite in the new 20x does something great I don't know what it is.

Incidentally, in case you're wondering if I got a bad 20x to start with, I ran the test in good light as well and, interestingly there was no discernible difference between the auto 20x and the manual 16x. Both were excellent and there was no difference in the tone either.

[Rob Lohman]

First of all remember that the 20x is just the difference between
the full wide and full telephoto range of the lens. It doesn't say
anything on how far it will go.

However, take a look at the Watchdog's lens guide for the XL2:

http://www.dvinfo.net/canonxl2/articles/article04.php

You can see the lens is a bit wider but also a bit more telephoto.
So you should see some difference in magnification.

You can also actually see the difference in field of view for the
various lenses on an XL2 on the following page:

http://www.dvinfo.net/canonxl2/articles/article10.php

[A. J. deLange]

The 16X lens has a maximum focal length of 86.4mm; for the 20X it's 108mm. Images of the same object taken with the latter should, at the same distance, be 25% bigger (108/86.4)= 1.25. I took pictures of an electrical insulator on a power pole with both lenses. With the 16x the insulator was 131 pixels wide. With the 20X it was 164 pixels wide. The ratio is 1.25. All is as it should be.

There really isn't any reason the low light performances of the two lenses should differ (given that enough light is available for correct exposure at the maximum aperture of each). Here are two photos taken with, respectively, the 16X and 20X lenses under dim tungsten light with +18dB gain (60i, 2:3):
http://www.pbase.com/agamid/image/39973415
http://www.pbase.com/agamid/image/39973422

I don't see much difference that can't be explained by the slightly different exposures and certainly not a 100% improvement in the quality of the 16X over the 20X. This is, of course, a subjective judgement on my part. Do others see an appreciable difference?

Wrangler's Note: Edited to make the above urls clickable links. -gb

[Lauri Kettunen]

<<<-- Originally posted by A. J. deLange : ... Do others see an appreciable difference? -->>>

On my monitor there is hardly any visible difference, but if insisted, I would say the 20x image has slightly less noise. For instance, when I compare the second color (peach) from left on the lowest row, the 20x image contains slightly less noise, and furthermore, it's bit brighter than the 16x image. Summing up, in my eyes a matter of taste.

Shekar, my aim is not to say your observation is a false one. Instead, I believe there should be some reasonable explanation which is not obvious from your explanation. For example, something like, did you set the white balance separately to each lens?

[Shekar Dattatri]

Thanks for the feedback from all of you. Rob What you point out about the 20x is correct. I guess I just automatically (and wrongly) expected to get 4x extra at the tele end.

Regarding the low light performance test, no, I did not set white balance separately for each lens. Will try that and see if it makes a difference. I'll be happy if both lenses are the same as it will save me from lugging so many lenses on each shoot. I'm not too net savvy, but if my new results still show a considerable difference between the lenses as before, I'll try and post the shots.