Follow-Up: Which one to choose - Zeiss 21mm, TS-E 24mm, 16-35mm II?

[bastian_bauwens1]

<p>Hi all,</p>

<p>Thanks for all the valuable comments you provided me with. They really helped me think again about which lenses to keep. In order to make the decision, I went out today and took some shots with all the lenses. Back home I converted the images with DxO Optics Pro 6 (all with the same parameters) to TIFFs, imported them into Aperture, applied the same sharpening to all images and exported a 100% crop from the top-right corner as 100%-quality JPEG.</p>

<p>The lenses I used in the test were:<br>

1. Canon 16-35mm f/2.8 II L @ 24mm, f/11, ISO 100, 1/60s<br>

2. Canon 17-40mm f/4 L @ 24mm, f/11, ISO 100, 1/60s<br>

3. Canon 24-105mm f/4 IS L @ 24mm, f/11, ISO 100, 1/60s<br>

4. Zeiss Distagon 21mm f/2.8 @ 21mm, f/11, ISO 100, 1/60s<br>

5. Canon TS-E 24mm f/3.5 II L @ 24mm, f/11, no tilt/shift, ISO 100, 1/60s</p>

<p>Below are the 5 images - can you find out which image belongs to which lens? I'd like to know whether anyone can do it better than I could. If you want to participate, post something like 1-C, 2-A...</p>

<p>Cheers,<br>

Bastian</p>

<p>A:<br>

<img src="http://www.bastianbauwens.com/lenstest/A.jpg" alt="" width="844" height="562" /></p>

<p>B:<br>

<img src="http://www.bastianbauwens.com/lenstest/B.jpg" alt="" /></p>

<p>C:<br>

<img src="http://www.bastianbauwens.com/lenstest/C.jpg" alt="" /></p>

<p>D:<br>

<img src="http://www.bastianbauwens.com/lenstest/D.jpg" alt="" /></p>

<p>E:<br>

<img src="http://www.bastianbauwens.com/lenstest/E.jpg" alt="" /></p>

[john_a5]

<p>At 24mm F11, all should be sharp--if the camera was on a tripod--at 1/60th, however, I expect the worst performer is going to be your 24-105mm at that focal length. There is some residual chromatic aberration, particularly in A, sometimes DxO doesn't get it all with lenses it has modules for but that is more rare. I also noticed a little bit more in C and D than in B and E which are pretty clean. If DxO weren't involved, I would tend to think those with more were the zooms. But at this size, we can't make any hard comparisons to the quality of the images and some hand work, in DxO could probably remove the residual CA.</p>

<p>By the way, E seems a bit "muddier" in the bricks than the others.</p>

[alan_bryant1]

<p>My wild guess: 1-B, 2-A, 3-D, 4-E, 5-C</p>

[dan_south]

Did you focus all lenses manually? The ts lens cannot be auto focused, so your technique is a factor. Did you apply

distortion correction in DxO? I don't see the distortion that I would expect from some of the lenses. It looks as though you

used auto white balance, because the last image is more orange than the others.

[ilya_e]

<p>The best would be if you could use tripod, did only minimum post (no sharpening, no noise removal, etc) and did not crop anything. But photo.net won't allow you to post such large images. If you can host those images somewhere and post links that would be great. Then it would be easier to judge.<br /> Did you use LiveView or focusing screen with manual lenses? A is way too soft and C looks the sharpest to me and I like it the most out of all.</p>

[mark_pierlot]

<p>I agree with Alan that the last image (E) is from the Zeiss (4). It has significantly different colour rendition than the others, whose colour is consistent.</p>

<p>I also agree with John that E is less crisply deatailed, which would be surprising if it were taken with the Zeiss.</p>

<p>I'm not going to hazard to guess which lenses produced the other images.</p>

[dan_south]

<p>Another comment: it's very difficult to judge the sharpness of a lens when the prominent plane of your subject isn't parallel to the sensor. We have no idea what point you focused on, but there's no way that the whole wall is in focus when taken from this angle. At a minimum, the way these photos were composed makes comparison of corner sharpness impossible, because not one of the corners is actually in focus.</p>

<p>I'm not convinced that you're really looking for info about these lenses. First, you wanted to have a "poll," and now you're trying to play "stump the band" with a group of ill-conceived test shots. In the meantime, I posted links to "real world" photos taken with the Canon 24 mm TS-E II lens, but you never made a comment. Are you actually trying to evaluate the quality and suitability of these lenses, or is this all just a game to you?</p>

 

[Robin Smith]

<p>Well, I think that this shows that if you are shooting at f11 and using your processing workflow it doesn't really matter which lens you get in terms of sharpness, so why not get the cheapest (17-40mm)? As others say we are unable to judge distortion from these shots, which could be a factor - although perhaps you eliminated this in DX0. I also bet you would get as good, if not better, image than some of these using the humble 24 f2.8 prime too-and that is really cheap in comparison.</p>

[pbizarro]

<p>My experience, and opinion, reflect on what other posters have already mentioned (in this thread, and in your previous one). The lenses you are considering are all top quality, so in the end you will have to base your decision on elements that pertain to yourself, not us. This stuff is subjective!<br>

At this level, sharpness (or minute differences of) should not be a consideration. Even the L zooms, stopped down to f/11 or f/16, will be sharp enoug in the corners. From the lenses you mention, I have experience with the 17-40, the 16-35 MKII, and the 24-105.<br>

The 24-105 is a good lens for "travel shots", but in my opinion it has some limitations when shooting more considerate and "paused" photos from a tripod, for example. The barrel distortion is a nuisance, and it does not go away at f/11; I shoot a lot of seascapes, and I got fed up of having to correct the horizon line. Useful range, though, so if you think you like a 24-XX zoom range, consider the 24-70 L zoom as well, much better optically speaking.<br>

The two wide angle L zooms are very good, and you would be hard pressed to spot differences at f/11 or f/16. The 16-35 MKII takes 82mm filters, and as such, if you want to use Lee or Singh Ray graduated filters, you will need the "monster" type of Cokin holder (X-Pro or something like that?). The smaller holders will vignette in that lens. However, it is nice to have a bright and clear view provided by the f/2.8 lens in the viewfinder!<br>

Today, I am using the 24 L MKII, this lens is perfect for my purposes, and for the way I "see" landscapes. I like this focal length, to the point that I can previsualize the shot and decide where to place the tripod, even before setting up the camera. Experience and knowing what you want, breeds familiarity; know yourself, and you will know what you need!<br>

The 24 T/S MKII is very attractive as well. It will allow a great DOF at faster apertures, which in turn can, say, freeze a field of flowers blowing in the wind (allows a faster shuter speed). However, I also wanted my 24mm lens to be able to shoot nightscapes and stars, so the faster the aperture, the better. If you like 24mm, the 24 L MKII is simply the best there is in Canon land.</p>

[bastian_bauwens1]

<p>First, I truly apologize if anyone got the impression that I am not really interested in evaluating the quality of these lenses. This is not a game to me, but rather an important decision for which I don't have the adequate time to evaluate all aspects in practice by myself.</p>

<p>I think Robin Smith nailed it down - for now I will return the expensive primes and keep one of the zooms - probably the 17-40, as I'd like to use the Singh-Ray P filters with the lens. I can still upgrade to a better prime if I realize during my travel that I need a different lens. Paulo, I really appreciate the hint re. the filters.</p>

<p>I did apply distortion correction to these shots, by the way, as I'd do this normally, too. In the end, the software is part of the process, and if this gives one lens an advantage over another, I think that must be considered. I was also quite impressed by the results after the post processing, as the differences were a lot more significant before.</p>

<p>Also, @ Dan South: Sorry that I did not react to your photos with the TS-E, but I simply did not find the time in the last 2 days. I was lucky enough to catch an early flight yesterday to go out shooting with the lenses myself again. Nevertheless, the shots are nice, and the one that still makes me think I should return the Zeiss but maybe buy a TS-E in the near future is the one with the arch. Perspective correction seems really powerful in such situations. So, if you had an example of a landscape, I would appreciate this even more. I do understand the value of perspective correction for buildings, and I do understand it in principle for landscapes, but I'd like to see a real-world photo where this really made a difference with trees, mountains, or anything similar. Your help is truly appreciated - help me justify buying this lens, because otherwise I'll spend the money for a new bicycle. ;-) And regarding your comment about the wall being in focus - the hyperfocal distance was only 1.72m, and the wall was at least 3 meters away from me. So it would have been possible that the whole wall was in focus, right?</p>

<p>By the way, the solution is 1-D, 2-C, 3-A, 4-E, 5-B. I must say I am quite impressed that John A was able to see the difference in chromatic aberration in the shots. It seems others have a far better/more trained eye than I do.</p>

<p>I hope that I can soon finish the gear selection and then get ready for shooting. All the best and a good weekend to everyone!</p>

<p>Cheers,<br>

Bastian</p>

[dan_south]

Hi Bastian,

 

I promise to have some landscape examples up for you next week. Also, perspective correction (by applying rise with the

lens) was applied to all five of those earlier examples, not just the arch. You would definitely see keystoning in any of

these images if you had used a lens without shift/rise capability. The photo with the long wooden boat in front of the

building also shows the application of tilt for extended depth of field.

 

It's interesting that you singled out the arch photo, as I think it's the weakest of the five, but yes, the application of rise is

very evident in that example.

 

One more thing. Consider the Lee holder system. It's more robust and more easily customized, and it's not much bigger

than the P system. Lee has wopide angle adapter rings that place the holder closer to the lens than the regular rings. Plus

you can trim the holder down to two or even one slot with a screwdriver. Cokin P has to be sawed! These are features

that you'll doubtless find attractive when operating in the 17-22 mm range.

[john_a5]

<p>Bastian,<br /> As to chromatic aberration, the first image has a bunch left over on the silver window frame (Compare that to B that has none here or C, which has just a hint of the red). Otherwise, if you study the transitions in the downspout, you will see the differences between the images. I wasn't surprised that this first images was the 24-105mm as at their widest, zooms are usually at their worst in this regard.</p>

<p>I am also not surprised that E is the Zeiss, with the odd color balance as compared to the Canon lenses.</p>

[ray .]

"but rather an important decision for which I don't have the adequate time to evaluate all aspects in

practice by myself."

 

Use your time to photograph. Take pictures. Subject, light, composition, feel. In the end that's what it

should be about if photography is your motive, not relatively minor differences in optics.

[bastian_bauwens1]

@ Dan South: I had a look at the Lee filter system and it does look good. One question though - how about color

accuracy? I have had serious issues with some Cokin filters (whites became pink), and that's why I wanted to move to

Singh-Ray. Are the Lee filters as accurate with color as the Singh-Ray filters? If so, the fact that they are cheaper (at

least if you buy sets) would be a definite asset.

 

Thanks and cheers,

Bastian