Tokina 10-17 vs Full Frame digital

This might also prove to be of interest:

http://www.robgalbraith.com/bins/content_p...cid=7-8739-8970

I need to get one of those balsa wood 1DsmkIII's - that would sure lighten my load and make for a much more maneuverable underwater setup.

Cheers

James

Lastly, I have done side by side tests, same subject, same lens, FF and smaller bodies - my opinion stands - and the only differential is the sensor and its size!!! I WANT a smaller camera as my 1DSs are TOO heavy but I simply don't care for the alternatives except the 1.3 crop sensors (a Leica M8 sounds lovely but not for underwater I think!) which leaves little option!

You said this before but you later qualified your statement to include using a FF lens and aperture that has no analog on the cropped sensor end. Because of that, your claim that "the only differential is the sensor and its size!!!" is patently false. You are not comparing like setups even if you are doing actual side-by-side comparison shots (which you've offered no evidence of).

I don't see why you don't consider a 5D if your 1Ds is too heavy. I ruled out the 1Ds2 in my latest search due to its weight. I considered the 5D but the fact is that the D2x is a better overall camera, offers matching IQ at base ISO, and offers lenses I prefer. I'm not at all convinced that the 5D wouldn't be a better choice, however. You do realize that the M8 only provides 8 bit RAW? I think the M8 is a joke.

I don't mind that you prefer the results your get with your rig that includes a full frame sensor, many people do, but what keeps this going is your continual insistence that there's a magical trait of FF that not only doesn't exist but cannot exist. If all that mattered in digital photography were the sensor we'd all be using Canon. That's not all that matters, though, and there are good reasons why some make other choices. Lenses matter a lot, they're the actual reason you like FF, they're the reason I chose Nikon, and they're the subject of Alex's original post. Camera bodies are surprisingly similar in performance (at least for now and for underwater), it's the lens choices that truly differentiate cropped versus full frame. Cropped is closer to its performance ceiling than FF is so eventually the argument will be moot. The high end will eventually be FF.

As for Canon claiming their lenses were digital-ready 10 years in advance, you're free to believe their propaganda if you like but the proof is in the performance of actual lenses on actual cameras. I see no evidence that Canon lenses are any more digital-appropriate than Nikon ones. If Canon knew so long in advance, why are they continuing to have such problems with wide angle? For having such great foresight, you'd expect their execution to be better.

Frankly, I would expect all manufacturers to know well in advance that digital would come and eventually replace film. I'd also expect them to know that there would be some design considerations. I would NOT expect them to know how to properly engineer digital lenses until they gained experience with real digital bodies. Canon could not possibly do that 10 years in advance.

Here's my lens comparison:

Lens Type Nikon Canon

180 Fisheye 10.5 15

Fisheye zoom 10-17 No

Cropped fisheye 16 No

18mm No 17

20mm No 20

24mm No 24

WA Zoom 12-24 16-35, 17-40

Mid-Range Zoom 17-55 24-70

Fish/Macro No 50

100mm 60 100

150-200mm Macro 105 180

2:1 Macro 200 No

Macro Zoom 70-180 No

It doesn't format very well. I have assigned colors to the "winner" lenses but it didn't copy from my spreadsheet very well.

I hope this is useful. I've used most of these lenses on cropped and FF cameras.

Cheers

James

Craig

Alex posted about the zoom-fish AND FF! As I've said before when reality and theory don't coincide its rarely theory that has got things wrong. I KNOW what I can see and from your last post it is quite clear that you are not prepared to accept anything which doesn't fit with what you believe to be the situation. I have absolutely no prejudice about Canon vs. Nikon (or Leica - you clearly haven't used one or you'd appreciate their character) - all have their place and afficiandos - but I do stand by what I have described. You may not like it, but there you go. I've been in this game long enough (too long some may say) to disbelieve what my eyes tell me - once I used to carry out mtf tests on photographic lenses for a living, and I did once test two systems (not Nikon vs. Canon though!), technically via a tested mtf route and by actually taking 'identical' images. On paper their should have been little discernable difference between the images from either, but of course subtle variations do have a visible effect. So I never discount the fact that things may not be quite as simple as they seem.

My tests by the way were not on a fast lens and were well stopped down, and disappointed me as I'd hoped to use a lighter camera for my travel photography - the 5D does indeed remain an option (as does that sad M8 with its miniscule ultralight M series lenses and modified sensor lenses).

I know I digress here, but your mention of the M8 size prompted it. I just bought my daughter her first camera, a D40 with the 18-55 kit lens. What a cool small DSLR for easy travel for those with Nikon glass. (Maybe the D40x for those wanting the megapixels)

James - please, please add in the 60 EFS + 12mm extension tube MkII - its my 'standard' macro lens for FF in less the than ideal conditions that often prevail here (excellent weekend but Saturday was rough in the RIB and vis hovered from 3~5m) and its very, very good.

Loftus - you are right - we've never had it so good, D40X, 400D - superb little cameras along with the rather more expensive M8! I have to say that an M8 appeals but I doubt that it will last as long as the very heavily battered old Leica M4-2 I have downstairs (30 years I think) - it looks as rough as can be, but winds one oh so smoothly, and the shutter still sounds (very quietly that is) as good as ever, but I digress too.

When reality and theory don't coincide, it's almost certainly the person's understanding of reality and theory that are mistaken. Any theory that's disproven by reality isn't a theory at all. I don't know what about this discussion even qualifies as theory; digital tonality is simple mathematics.

Yes, Alex started a discussion of lenses being the big difference between DX and FF. I agree and, as far as I can tell, you do too. I'm objecting to your claim that FF offers better tonality. I assert that any differences you see are due entirely to lens effects.

I KNOW what I can see and from your last post it is quite clear that you are not prepared to accept anything which doesn't fit with what you believe to be the situation.

Not without any shred of evidence I'm not, but I'm open to you correcting my misunderstanding of what I "believe to be the situation". Otherwise, I'd like either some explanation for how your claims could possibly be so or some examples of how it actually is so in practice.

As for Leica, any modern camera that prides itself on image quality yet only offers 8 bits of raw is a joke. I refuse to subscribe to a Reichman-esque photography religion as I know how digital works. If you want to take an M8 underwater as a viable alternative to your 1Ds with its "superior tonality", more power to you. I can't wait to see those sunbursts. I assume you shoot RAW?

On paper their should have been little discernable difference between the images from either, but of course subtle variations do have a visible effect. So I never discount the fact that things may not be quite as simple as they seem.

That's because MTF numbers are a simple summary of a complex issue and don't tell anywhere near a complete story. Lens testing does not teach you how light works nor does it tell you anything about digital.

My tests by the way were not on a fast lens and were well stopped down, and disappointed me as I'd hoped to use a lighter camera for my travel photography - the 5D does indeed remain an option (as does that sad M8 with its miniscule ultralight M series lenses and modified sensor lenses).

I would love to see comparison images of identical scenes shot with identical perspective, matching DOF, and identical lighting and exposure. I'd also like to see the equipment setups and exposure information. I'm eager to see this tonality superiority you are talking about.

If you can't find suitable tonality because a 1.5 crop can't provide it, I don't see by your reasoning how an M8 with its 1.33 crop, inferior resolution, and massively truncated 8-bit data would possibly do better. I'm curious to know why you think it can considering 27mm is a whole lot more like 24mm than 36mm. Of course, we all know that sensor width has nothing to do with tonality, right?

The point of this discussion for me is to get you to realize that the differences you see come from your preference of lens and shooting style, not the sensor itself. These are lens differences that are inherent, as a practical matter, between formats which was exactly Alex's point. I have no objection to your preference for the specific image characteristics you claim to see just as I have no objection to your preference for rectalinear lenses over fisheyes. The best tonality achievable today in a full frame sensor is not as good as what is achievable with an available cropped sensor camera (the 1DMk3).

James, a couple things on the lenses:

DX has an equivalent for the 24mm, it's the Tokina 17mm which you generously introduced me to ;-)

DX has an equivalent for the 50mm, it's the Nikon 35mm f/2 (it's just not labelled "micro" even though working distance is acceptable). In reality, the midrange zooms take the place of the 50mm for that purpose and, if you use them with a flat port, some of those can offer acceptable magnification without diopters. I consider the 17-55/dome a good midrange zoom for sharks, but the 24-85/flat port good for fish portraits. It's what I used for bigger frogfish. Haven't looked into the best choice today.

I think this comparison summarizes things though. For macro, there is more power and more flexibility with DX. For fisheye, DX has more choices. For other wide angle, FF has more primes. IQ is a more complicated issue.

"The point of this discussion for me is to get you to realize that the differences you see come from your preference of lens and shooting style, not the sensor itself."

You can't! My comparison was on the same lens same exposure, etc., etc. Bottom line is that there was smoother tonal transition on the FF. Remember that all sensors ARE different and hence will inevitably produce differences.

Photography is primarily a practical operation with a straightforward result - an image which anyone can look at and see whether they like or not. Bottom line for me is that there is a difference in what I look at and somewhere there is a reason. I'm sure that your technical explanations have sound reasoning behind them but I do still see a difference - a smoother degree of tonal transition. I suspect that we are arguing over minutiae and its just going nowhere. Perhaps you might be prepared to accept that there are differences in sensors (after all it would be surprising if differently sized sensors of varying design age, manufacturers and so on were all identical - wouldn't it) and that this might have some bearing on things?

I preseume that images shot on the D2X in its normal and crop mode would exhibit idential images if viewed at 100% on screen but this won't be true of different sensors and what I see is that FF sensors (certainly my 1DS images) exhibit tonality which appears smoother to me. This does, as you say, suit my shooting style better, and is probably what I look for and notice. I just don't think we're going to agree on this one!

You can't! My comparison was on the same lens same exposure, etc., etc. Bottom line is that there was smoother tonal transition on the FF. Remember that all sensors ARE different and hence will inevitably produce differences.

It doesn't surprise me that you say this since you have not once attempted to engage in meaningful dialog. All you do is make the same claim, that you see better tonality, yet you never offer a single image that demonstrates this or explanation for how this might be so. I would think that a photo pro who is earning money with his images, who has studied this in depth and who has done direct comparisons might have some images to share.

If you are comparing cropped versus full frame using the identical lens, are we to assume that both cameras are Canon? Which models? Why are you so definitive with your claims yet so unwilling to provide information to back them up? What proof do you have that the camera you tested is representative of all cropped sensor cameras when it comes to tonality? Since your preferred body is a 1Ds, might we assume your cropped sample is similarly old and your opinion years out of date?

I presume that images shot on the D2X in its normal and crop mode would exhibit idential images if viewed at 100% on screen but this won't be true of different sensors and what I see is that FF sensors (certainly my 1DS images) exhibit tonality which appears smoother to me.

What you see, apparently, is better tonality out of your 1Ds than out of some camera that you haven't named. Hardly the same as claiming that full frame offers better tonality than cropped and markedly different than the other tangent you went on regarding OOF areas using fast aperture wide angle shooting.

You said that full frame sensors offer better tonality than cropped frame sensors. Now you say that the differences have to do with something other than sensor size, or am I misinterpreting? Are you suggesting that the D2x crop mode has identical IQ to DX because the sensor is still DX, or perhaps is it that only FF enjoys the benefit of improved tonality? Have you ever compared the D2x to your 1Ds? If not, why did you make such a blanket statement?

Tonality characteristics of a sensor have utterly nothing to do with the size of that sensor and I'm glad that you accept that now. It may be that your tested cropped sensor camera did not offer tonality that compared to your 1Ds but it is a fact that the D200 matches the tonality of the 5D or your 1Ds. There is no way the D200 or the 5D could improve upon their tonality without more ADC resolution. I said this before and you dismissed it with some quip about "reality and theory" yet your reality had nothing to do with theory. I'd suggest you take a look again at the performance of today's cropped sensor cameras.

Anyone can test my assertions for themselves - just try shooting a few similar frames on FF and cropped - many shops will allow a few trial frames to be taken - and they can see for themselves.

I've tried to explain how images form under a vast array of factors which may affect the final image, but here is a question for you to bring things entirely back to the 'sensor'. Take a very similar PICTURE taken on the same film - one on 5" x 4" and one on 35mm - which will exhibit smoother tonality? This is, in effect, using an identical but much larger sensor. Or perhaps Ansel Adams et al really were wasting their time? I suspect that we are discussing two entirely different things - pixels versus the image - and it is the final image which is important, the interplay of numerous influences on the image simply allows us to create the type of image we like. We could go on forever.... but I reckon other readers must be getting bored by all this by now. I spent a whole afternoon discussing lens design with a fried (who designs lenses oddly enough) and discussion fora like these are not really the place for such long winded, in-depth discussions.

.... but I reckon other readers must be getting bored by all this by now..////........... and discussion fora like these are not really the place for such long winded, in-depth discussions.

To be honest, your refusal to offer proof to your comment is now making me question how much conviction you have in it's validity. I would like to see if you or Craig are correct, not because I have a dog in the fight, but because seeing something pointed out that I may not have recognized by myself could be meaningful to my photography or the selection of my next camera body .

I am not trying to stoke the fire

Paul, you could just say that you don't have any such examples. It's not our burden to prove your assertions. There's no way I could disprove your assertions by doing your tests for you and I'm sure you know that. Any counterexample I'd offer you'd dismiss as user error or my lack of discriminating taste.

I don't know why you are trying again to divert the argument toward film. This is digital; there are very specific technical reasons with digital that prove you are mistaken, so chasing an argument about why large format film is perceived differently or making an appeal to authority with Ansel Adams doesn't improve your case. This isn't a topic that requires God to answer.

Yes, when viewing distances and sizes mean that differing ppi numbers are used in different formats, then tonality will differ. That's what's happening with your film examples, but viewing full frame at 300 ppi and cropped at 200 ppi is a BS comparison. You aren't getting better tonality, you are becoming less demanding of tonality. A sensor cannot possess good tonality unless each pixel does, so if we are "discussing two entirely different things - pixels versus the image" then you are wrong. Tonality characteristics of a sensor ARE about each pixel. Who cares about small prints when we are talking about full frame dSLRs? You sound like Reichman here. Are you going to call me a "pixel peeper" now?

Yes, we could go on forever with me chasing your diversions around in an endless circle. You've already admitted that sensor size has no bearing on tonality, otherwise D2x full and cropped modes would be different, but you can't bring yourself to admit that you were wrong even though that was your fundamental assertion. What you have is a preference for your Canon 1Ds that leads to a blind belief that your Canon product simply performs better in every conceivable way than competing products and it's hard for you to accept that other formats and other manufacturers may have caught up to your 1Ds. Your camera is still a fine camera but its performance is exceeded by other Canon FF models, the D2x and possibly the D200/1D3 as well. This is just how it is with digital.

Now, if your latest point is, in fact, what you've meant all along you could have simply said that you prefer using higher resolution in your output because it "improves tonality" and full frame offers you more resolution to work with. I would understand that point, but I would also understand why you might be hesitant to say it. After all, its well known that full frame doesn't offer dramatic improvements in resolution (currently) and there are DX cameras that outresolve (and out-"megapixel") your 1Ds. Canon full frame sensors are great; there's no need to make up mythical advantages where they don't exist.

I'll remind you, and others, what you said that started this. From here:

Just to jump in and add my two pennorth -jumping to a 5D has both upsides and downs. (I actually have 1DS and 20D in Seacam - its a long story!).

 

Up is image 'quality' and full frame format which is 'different' to 1.6x crop and 'different' to Nikon's 1.5x crop in much the same way that medium format film was/is different to 35mm. This said, unless you appreciate the differences then it might not be such a big deal.

You not only said there where "qualitative" differences between sensors due to their crop but that those differences were quite dramatic (equating them to the differences between MF and 35mm film). When challenged, you said those differences were in "tonality". Then you said the tonality differences were dependant on how you shot and that FF allowed the use of lenses not available on DX. Now you assert that the differences are not in the pixel but the image, likening it to the difference again between MF and 35mm film, but that cannot be because current DX cameras have nearly as much "film area" (pixel count) as FF sensors do. If its not in the pixels themselves and its not in the pixel count, it doesn't exist.

Your claim is pure prejudice for larger format sizes and nothing more. Meanwhile, buyers concern themselves with the actual performance of real camera bodies and lenses they can buy today. Full frame lovers just can't understand why anyone would buy anything other than Canon because they live in a fantasy world where their cameras are better because they are full frame. Yes, there are real reasons why people freely choose between DX sensors and FF and those reasons change with each generation of product. It is not about the sensor, it is about the system.

Iggy, I agree with you. I have no desire to argue for arguing's sake or to prove anyone wrong. I don't like myth and legend propagated though. I think it's worthwhile for people to learn and understand the true differences between cameras so that they can make informed decisions based on their desires and needs. It's not helpful to provide bad information.

I'd like Paul to list those characteristics of FF that are, indeed, unique so that we can understand them. I want them to be real and substantiated though. It's a fact that FF sensors offer capabilities with existing lenses that DX doesn't but the opposite is also true. How that applies to underwater shooting is an interesting discussion and one that Alex originally put forth in this thread. Paul's point about the 24mm f/1.4 lens is a good one and I'd like him to show more about why he thinks it's a compelling choice for underwater shooting. Those shooting the Tokina 10-17, after all, are doing just that.

I have no dislike for FF, quite the opposite. I have no dislike for Canon either; I owned and liked my Canon gear before it was stolen. Choosing between them is difficult enough without dealing with the propaganda surrounding each company's products.

Still going! I can list some more tangible differences. I have been shooting a lot of topside lately, so they may not all be relevant...

To get the identical shot on a DX camera vs a FF camera can be difficult.

Where you can use a 300 f2.8 on a FF body, you need a 200 f2 on a DX camera. This will give you the same perspective, DOF control, etc.

To shoot fish w/ a 50mm f2.8 lens on FF you'd need approx a 35mm f2 lens to get the same perspective, DOF control, etc

Where you can use a 16-35 f2.8 lens for wideangle, you'd need a 12-24 f2 to get the same perspective, DOF control, etc. - a lens that doesn't exist.

If you underexpose a shot with a DX sensor camera at ISO 400 you have at least one less stop to work with to bring the shot back if you aren't tolerant of noise. This one is easy to show examples although I don't have any though since I don't shoot DX.

For commercial photography, shooting at ISO1600 is often critical. I have shot shows at ISO1600 and f2.8 and still had underexposed shots from the 1DmkII that could be brought up and were perfectly acceptable. If I'd been using DX I'd have had to run a noise suppression program - adding an additional step to my workflow.

Anyhow, this is sort of a general post based on my experience. As most of us user our underwater cameras as general workhorses (shooting topside and below) this is still important stuff.

Craig's comments are pretty narrowly focused where there is literally no difference in performance. Shooting underwater macro at ISO100 you can get the same results with just about any system.

Cheers

James

Craig's comments are pretty narrowly focused where there is literally no difference in performance. Shooting underwater macro at ISO100 you can get the same results with just about any system.

Where you can use a 300 f2.8 on a FF body, you need a 200 f2 on a DX camera. This will give you the same perspective, DOF control, etc.

 

To shoot fish w/ a 50mm f2.8 lens on FF you'd need approx a 35mm f2 lens to get the same perspective, DOF control, etc

 

Where you can use a 16-35 f2.8 lens for wideangle, you'd need a 12-24 f2 to get the same perspective, DOF control, etc. - a lens that doesn't exist.

It is eye opening that Canon, who has long espoused the "advanatges" of FF, now referrs to the mkIII as their "new flagship", "the most advanced EOS Digital SLR ever produced" and "the new standard". Not sports camera, but dslr. That's not to say that cropped is better than FF, but rather to say that depending upon application either can be preferred over the other. I like the D2x underwater quite a bit, but for topside, from what I have read to date, I think I'd prefer the MkIII. I think resolution disparities between 10 and 16 mp, is a red herring given the other shortcomings digital capture needs to corral first.

I guess it was inevitable that the provocative opening to this thread would develop into a general FF vs DX discussion.

For me FF has some clear advantages regarding high ISO noise performance and low DOF at wide aperture. High ISO noise control is crucial for photojournalists covering news and sport etc. And low DOF is v important for portrait and fashion photographers etc.

But I agree with Craig that both features have limited benefits for underwater photography. The majority of UW photos are base or low ISO - where DX and FF give comparable results. A sports or news photographer earns their money from capturing a moment - and ultimate image quality is less important (annoyingly the only pro sports shooter I know well shoots Nikon, although he accepts he is an acception). Underwater our images are used in a different ways - and the UW photographer earns from images used large in advertising, books, prints and magazines. Generally the stuff that sells is glossy and colourful and base ISO (however good a camera's high ISO performance is it is not a good as its base ISO) and high image quality. Under these criteria DX is cheaper than FF and with lenses like the 10-17mm has more flexibility for comparable image quality.

For me there are two flaws in my argument. First most people don't do UW photography to earn money. And second many underwater photographers want to use their expensive camera on land too. Have a look at Paul's website to see some of the awesome land photos he has taken with his UW camera (I wish I had that ability). It is the land photography strengths where the FF option makes more sense.

To get back to the Tok - this is undoubtedly and important lens. Fisheyes are easier to design and certainly easier to get better image quality out of then wide rectilinear zooms. And this is not a DX vs FF issue. While their remains much criticism of Canon's wide zooms, the Nikon 12-24m DX is not a perfect lens. Underwater wide rectilinear zooms are very critically effected by poor dome port setup - where fisheyes are not, further enhancing their suitability.

I have seen my Tok 10-17mm images on my computer now. And the image quality is spot on as long as you use it with a big fisheye dome and small extension ring.

So my feelings on the debate is that for the types of images that we generally take UW, FF has little advantage over DX. And to answer the initial question, a lens as powerful as the Tok adds a big + to the DX solution. Although will I be able to resist a FF D3 later in the year? I better start selling some of those DX photos as it ain't gonna be cheap.

Alex

p.s. To try and wrestle the discussion in a new direction with another provocative question? Would Canon underwater photographers benefit from a prosumer-spec DX sensor body to make use of the Tok 10-17mm? A 5D level camera, but more like the D200 in philosophy (fast, DX sensor)?

p.p.s. It is great to have Craig back!

Always the calming voice of reason, Alex.

Let's see some 10-17 pics!

Actually I was trying to be deliberately provocative in some of that. I think a lot of good info has come out in this discussion as a result of an initially opinionated statement!

Alex

p.s. So as not to drag things off topic - you can see wide angle 10-17mm shots of mine here: http://wetpixel.com/forums/index.php?showtopic=19316