129 replies to this topic

[Alex_Mustard]

Thanks for taking the time to do all these Stephen. Really appreciated by many.

I have to say I am shocked by how badly the 17-35mm performs behind the Seacam super dome, even with a dioptre. Especially as the 12-24mm performs so well behind the same dome/dioptre combo.

Just as a comparison here are 100% crops of edge sharpness from my tests of the 17-35mm at 17mm with a tiny 4" Subal dome. Note this is a different dome to the tests I posted above

And there is no doubt that the corner sharpness is much better with the 4" than you are getting with the super dome. Now I know that the Seacam Superdome is good and much better than a Subal 4" dome - so something is going wrong here. Maybe your 17-35mm is mis-aligned lens or there is a slip in focus in the testing (I know how much work doing these tests is - it is easy for the focus to lock on the background just as you press the shutter). I can't believe it is the Seacam dome.

Anyway here are my tests. Both way open at F5.6 and both at 17mm behind the tiny Subal 4" dome (no CA correction or sharpening applied). They are 100% crops from top right.

This one +4 dioptre:


This one no dioptre:


Alex

[loftus]

Thanks guys for all these tests; one thing that is so striking is the number of variables, and that it really behooves anyone who cares about this to do their own testing for their own setups. I'll be getting in the pool as soon as our Florida freeze passes.

[Paul Kay]

Alex

I'd agree, many thanks to Stephen - the effort is much appreciated, and also that there is something amiss with the 17~35 somewhere. That said, its a lens I used to own in film days but never got results that really satisfied me - even using the mid-port from Subal (the same as the one you are using).

[kaarlin]

Stephen and Alex, thanks a lot for your test shots.

I didn't use a tripod and exactly the same corner on each shot, sorry for that. For me I wanted to know what this lense would do in the most extreme conditions: so, very close to the subject, widest aperture and the lens used at 14mm. I think this are the most extreme conditions I would ever use a lens, so the results can only be better than what I saw.

I think Alex showed us that the 17-35mm with a +4 diopter (and the right dome) is the best option for shooting on reefs!

Edited by kaarlin, 05 February 2009 - 06:40 AM.

[Paul Kay]

Looking again at Alex's comparison shots above, I'm surprised at the difference between diopter and none diopter shots but this probably indicates that the 17~35 has relatively poor close focus performance - its not close focus corrected. It has a closest focus distance of 0.9 feet which means that its on the edge of its focus capabilities without a diopter, so its going to be pretty poor tested behind a small dome without a diopter. If this is the case then it may well respond better to a stronger diopter in a larger port too as this will bring its focusing range closer allowing it to operate as though it is focusing further away on a closer subject (when the lens is set at infinity where it performs well, with a stronger diopter it should be focusing on a virtual image closer than 3 x the radius of the dome from the dome - which will limit its furthest focus but may improve its performance). Worth a try!

[StephenFrink]

Thanks for taking the time to do all these Stephen. Really appreciated by many.

I have to say I am shocked by how badly the 17-35mm performs behind the Seacam super dome, even with a dioptre. Especially as the 12-24mm performs so well behind the same dome/dioptre combo.

Just as a comparison here are 100% crops of edge sharpness from my tests of the 17-35mm at 17mm with a tiny 4" Subal dome. Note this is a different dome to the tests I posted above

And there is no doubt that the corner sharpness is much better with the 4" than you are getting with the super dome. Now I know that the Seacam Superdome is good and much better than a Subal 4" dome - so something is going wrong here. Maybe your 17-35mm is mis-aligned lens or there is a slip in focus in the testing (I know how much work doing these tests is - it is easy for the focus to lock on the background just as you press the shutter). I can't believe it is the Seacam dome.

Anyway here are my tests. Both way open at F5.6 and both at 17mm behind the tiny Subal 4" dome (no CA correction or sharpening applied). They are 100% crops from top right.

Alex

Alex - I didn't buy that the results could be so bad either. I know 17-35 is better than the tests suggest, so clearly the port extension we assumed was right, is not. After all, until full frame, the 1.5 crop forgave many sins. Plus, I shoot Canon, so have never obsessed over the 17-35 Nikkor. But, now that we need to support that lens with full frame Nikons, the test has to be more scientific.

Today I tested 17-35 with PVL 25, 30, 35, 40, 45, 50, and 55; and with PVL 40 (the one that looked best to my eye in the field) with +1, +2, and +4 B+W diopters. I'll analyze and post those results after we wrap our class this afternoon.

Stephen

Edited by StephenFrink, 05 February 2009 - 11:18 AM.

[wolfeeldiver]

I think many in their quest for finding the optimal combination for a dome port and lens configuration may be expecting more out of the configuration than is possible. We must not forget the fundamentals that are work here: i.e. trying to take a lens designed for air use, and designed to shoot optimally a flat surface at some some arms length or longer distance away... adding a diopter, putting that air lens behind a dome port and a water interface is a whole different ball game. Now it must focus on a much closer and much much more curved virtual image, something not originally intended to do.

I'm sure there is some commercially availble combination of ports, spacers, diopters that will get it close.. but I imagine there is only "so much" that can be expected from a land lens not being used as originally intended.

Pie in the sky dreaming here: Perhaps what the market is craving for, is a special use diopter. That is designed to focus close up on a curved surface. A surface that mirrors the virtual image created by a dome port. Hmm...

Edited by wolfeeldiver, 05 February 2009 - 12:38 PM.

[StephenFrink]

17-35 More:

Here's the best I could do with Seacam Superdome, PVL45, full shot, 100% center, and 100% lower left corner. Almost ALL of these tests I've been doing are sharp in the center, btw. I've only been testing for ultimate corner resolution.

No diopter, this is the best of PVL 25, 30, 35, 40, 45, 50, and 55:

Edited by StephenFrink, 05 February 2009 - 01:24 PM.

[StephenFrink]

Here are the diopter tests, showing full view (which I had to back up for, because diopters reduce wide angle coverage, the greater the diopter the greater the angle of view reduction), the shot taken from same distance as no diopter picture (note how many bars of color on color chart are lost as a result of adding diopter), and the 100% corner crop.

Here is B+W +1 diopter:

[StephenFrink]

+2 diopter:

[StephenFrink]

+4







Again, all are sharp in center, whatever port extension, whatever diopter.

I've avoided any editorial comment, preferring only to show the results, obtained as empirically as possible, and invite one to make own conclusions. All I'll add here is that if I was shooting full frame Nikon and looking for a wide angle zoom, I would consider the 14-24 very favorably, at least based on results with Seacam superdome and PVL 50.

[loftus]

I would consider the 14-24 very favorably, at least based on results with Seacam superdome and PVL 50.

Steve,
Would it be fair to say, that if one does not have a superdome, the 17-35 with +4 would be your next choice? (On FX)

[james]

Steve,

How do we contribute beer money for you and more importantly your assistants! They look redder and redder in each photo :-)

But seriously, thanks for performing these tests and posting the results.

Cheers
James

[StephenFrink]

Steve,
Would it be fair to say, that if one does not have a superdome, the 17-35 with +4 would be your next choice? (On FX)

Jeff - Depends. The +4 costs significant wide angle view to achieve sharper corners. Actually, I'd probably use it without a diopter and go to a higher ISO and smaller aperture to enhance corner performance. The new cameras do so well at higher ISO, that's the direction I'd likely go first.

The 17-35mm with +4 is not wide enough, for me, for the things I'd normally shoot with a lens like that. Plus, the pincushion distortion and chromatic aberration is exacerbated with +4 so I'd probably find another way to shoot without it. But, that's just me because the wide angle view is so important.

Having said all that, maybe the smaller wide port works better than the superdome. Clearly, it does wonderfully for Alex. Decisions, decisions.

James - Thanks for the kind words. The timing worked well since we had all the cameras, lenses, housings, and port extensions here for my class in Bonaire. Plus, I was curious and this was the only way I knew to do at least semi-controlled testing. My class seemed to enjoy the exercise as well, so it was fun. Semi-fun, anyway.

Edited by StephenFrink, 05 February 2009 - 03:45 PM.

[photovan]

Echoing the praise from loftus great work to you and all your helpers.

Hey Admin... given how this thread developed to look at a few combinations, is it worth shifting/renaming stephen's tests into another more generalised wide angle tesing thread?

[divegypsy]

Clearly Stephen Frink is NOT maintaining a constant distance from and a constant angle with respect to his subjects in his 17-35mm lens tests. This can be clearly seen because the target slate for even the "center" lens area varies in size from frame to frame between the "no diopter" view and the views with the various diopter strengths. The position of the woman holding the "center sharpness" slate is inconsistent to the building structure behind her. And how close the slate is to the edge of the frame varies.

All these factors, I believe are contributing significantly his conclusions, which I believe may not be correct. Particularly the conclusion that the various diopters are resulting in significant variations in the angle of view of the lens at 17mm.

To get a really valid comparison of the lens with the different diopters, EVERYTHING else needs to be consistent. It would be better to simply hang the slates along the edge of the pool, not have them held by people who are clearly changing position from picture to picture. And when taking the pictures, make sure that you keep the center of the image really centered on the "center sharpness" slate and the far edge of the "corner sharpness" slate at the very same position in the corner of the frame.

Fred

[Alex_Mustard]

Really fascinating tests. I'll repeat my thanks again!

Thanks for taking the time to do all these Stephen. Really appreciated by many.

Having conducted similar tests I know how much work these are - in and out of the pool, taking off ports and lenses etc. Then preparing all the crops etc.

Anyway I have learnt a lot from your tests. Two things in particular are fascinating and I had not appreciated their significance before. First how sensitive the lens performance is to dome positioning (port extension length). With the 14-24mm if goes from zero, to hero, to zero within 10mm or port differences. Of course I knew it was important to get the right port extension, but I had never realised such small differences have such large impact. Fascinating stuff - and this could go a long way to explaining the problems people have been having getting this lens to work well underwater, particularly as not all housings offer such a variety of extension rings as Seacam.

PLV 55


PLV 50


The other factor I had not realised was so significant was the loss of coverage, both in air and underwater, caused by using dioptres (as a percentage it is a very significant difference). It was a good idea to shoot splits so both can be seen simultaneously.

No dioptre:


Same place with +4:


I think that this was a failing in my tests. I did not use a tripod to fix the camera position, using just a weight on the floor of the pool to line myself up. I didn't notice any change in coverage - but I could have easily "corrected" for this just by rocking backwards and forwards on the spot to keep the same corner points in the frame.

As your tests show - when the frame size is normalised - the dioptres improve corner sharpness with both lenses that can taken them: 12-24 and 17-35mm, but that normalising is hiding a considerable cost loss of frame coverage caused by the dioptres (seen on the 17-35mm examples above).

If dioptres cause such a reduction (as your examples above clearly show) I can certainly see much motivation for not using them. Their positive effect on corner sharpness seems irrelevant in comparison. It is embarrassing that I missed this factor in my own tests. Since the effect can be seen on land I am tempted to unpack my cameras (off to Indonesia for 4 weeks) to see it for myself.

Thanks for the educational postings - maybe I need to come on one of your workshops!

Alex

p.s. Darren, I have updated the subtitle of the thread, and yesterday I posted a news item on the front page promoting this discussion.

[Alex_Mustard]

Just to show how significant the reduced coverage caused by the dioptre was I have overlaid (inverted) the non dioptre shot. I think this shows Stephen's point "the greater the diopter the greater the angle of view reduction - note how many bars of color on color chart are lost as a result of adding diopter" very graphically.



Now I think that there was a little camera movement between these shots - as I had to skew the image slightly, but there is no denying that the dioptre cuts coverage by a large amount.

Alex

[Paul Kay]

So would it be fair to say that long standing statements about corner correction being 'acceptable' with a concentric dome port hold true - that 90~100 degrees is still the effective limit and is not substantially improved by using better, more modern camera lenses? I should have read up and quoted from "The Photographic Lens" by Ray again: "...the focal length of the combined lenses [diopter & camera lens] is less than the prime [camera] lens value...". I haven't found the formula yet but will try to do so.

[cpix]

Very interesting thread - and thanks for all the work (when I think how long it takes me to change a lens in my own housing.......).
Anyway, further to Alex's post - is the quality of the edge with the +4 diopter similar to the quality of the same part of the image without the diopter ? In otherwords, is the fact that the +4 diopter has a zoom in effect the major contributing factor as to why the corner sharpness is better - it just avoided the corners ?