Wallet full of $100 bills

[Mark Elberson]

Man this thread is a hum-dinger! I can't believe there has been so much back and forth (of polar opposite ideas and beliefs) without the least bit of resentment or ill-will. That's a true testament to the members (ok, most of them) on this forum. Kudos to John and Jon for having a lively, entertaining and educational discussion for the rest of us to enjoy :-)


By the way, I agree with both of you! I think you've both figured it out but it appears that you're <span style="text-decoration: line-through;"]arguing about discussing two different things :-)

[Jon Ruyle]

You're saying that the 85mm f/1.8 'wins' because it's slower, and becauseit's a longer focal length (no cropping).

I didn't mean that the intrinsically slower lens wins. All I'm really saying is that stopping down tends to improve image quality, especially stopping down from a very fast aperture. Cropping tends to make image quality worse (do you diagree? If you crop a picture, don't flaws get magnified?)


Lets forget specific lenses for a moment. Lets say you have two lenses, A and B. You take a picture with lens A and crop it, throwing away the outer 60%. Then you take a picture with lens B, but stop down 1 1/3 stops beyond the aperture you used to take lens A. Now compare the pictures. Do you agree that lens B has been given a huge advantage in this test? Even if lens A is intrinsically better, lens B may well come out ahead in this comparison. In fact, lens B would have to pretty much suck compared to lens A to lose this contest. Right?


Well, B's advantage in this contest is exactly the advantage you give to all your lenses by shooting full frame.

The other factor is that a system has many parts and features. The faster frame rate and better AF tracking are significant differences between the 5DII and the 7D. If the 5DII with it's better IQ misses a shot because the subject was running toward you too fast for the AF to track it, or misses the perfectly-timed action shot of the burst because it's frame rate is half as fast as the 7D, better IQ doesn't help.

I agree absolutely.

If 'better' means narrower angle of view to avoid the need to crop away resolution, the crop body wins, especially from a cost perspective at the supertelephoto end.

Yes, if the crop body has more pixel density (which they usually do, and in particular it is no contest between the 7D and 5DII).


If you aren't using your whole APS-C sensor anyway, FF does not help you, and you want smaller pixels.

We're
comparing an 8mp APS-C to a 21mp full frame, which is unfair, I
admit.

Yeah - PZ's wide open center MTF for the 85mm f/1.8 is 1813 with the 8
MP 350D, and 2267 on the 15 MP 50D.

Fair enough, but if
you scale up the 8mp scores (by looking at lenses that were tested on
*both* 8mp and 15mp that had similar scores on 8mp), the nifty fifty
still comes out ahead. I still believe the nifty fifty on FF would be sharper than the 85 f/1.2 on APS-C, provided you compare the lenses fairly. (The nifty fifty has other disadvantages, obviously... worse bokeh, far worse build quality....)

But, we can agree to disagree, 'eh?

Agreed.

[Jon Ruyle]

I can't believe there has been so much back and forth (of polar opposite ideas and beliefs) without the least bit of resentment or ill-will.

It doesn't make sense to get mad at someone just because they have a different opinion or see things differently. Unless of course, you're married to that person []

By the way, I agree with both of you!

What? [:|] Okay, tell me what John is saying that you agree with. Maybe you can make me understand his point of view. A fresh perspective might help us both

I think you've both figured it out but it appears that you're <span style="text-decoration: line-through;"]arguing about discussing two different things :-)

Do you mean


Thing 1) Is it true that it make sense to compare APS-C vs FF for effective F and f or actual F and f? and


Thing 2) Is it true that by comparing effective f and F instead of actual, the IQ edge goes to the lens on the full frame camera?


If that is what you mean, then I think you are right. Two separate points that we seem to be mixing up and arguing at the same time. We should keep them separate.


Okay, I've done enough babbling for today. I'm going outside to take some pictures.

[Mark Elberson]

Okay, tell me what John is saying that you agree with

Agree with John:
I thinkit's unfair to compare the 7D with the 50 f/1.2 to the 5DII with the 85 f/1.8. If the the 85 f/1.8 is already sharper than the 50 f/1.2 on the same full frame sensor then of course it will be sharper than the 50 f/1.2 on a 1.6 FOVCF sensor.


Agree with Jon:
If you are buying a 50 f/1.2 specifically for the FoV and DoF on a 1.6 FOVCF camera then you would probably be better off with the 85 f/1.8 on a full frame camera. You and I share a similar philosophy in that I feel that wide-angle EF glass and exodicallyfast glass (50 f/1.2, 85 f/1.2, 200 f/2) are much better served on a full frame camera.


The fact is, even though the 50 f/1.2 may have a similar FoV and DoF on a 1.6 FOVCF camera as the 85 f/1.8, there are many other factors that contribute to the overall IQ of a lens systemmaking itessentially impossible to evaluate them empirically. Given the option, I'd rather shoot with the 85 f/1.8 on a 5DII but that doesn't make it better than the 50 f/1.2 on a 7D nor does it make either setup "right". I thought it was interesting to hear the thought process of two very well versed photographers describing their point of view.

[Jon Ruyle]

Agree with John:
I thinkit's unfair to compare the 7D with the 50 f/1.2 to the 5DII with the 85 f/1.8. If the the 85 f/1.8 is already sharper than the 50 f/1.2 on the same full frame sensor then of course it will be sharper than the 50 f/1.2 on a 1.6 FOVCF sensor.

Yeah, I admit that wasn't a good example because the 50 starts out less sharp than the 85.


But even if you start with a less sharp lens, it will become more sharp when you put it on FF.


How about comparing the 85 1.2 to the 135 f/2? Both are pretty sharp, right? Is that unfair?


According to photozone, "The Canon EF 85mm f/1.2 USM L II is one of the little wonders in the
Canon lens lineup - at least
on an APS-C DSLR. The center performance is nothing short of
breathtaking and the borders are only
slightly weaker."


I'm not trying to load the game with a weak lens on the crop side. But also according to photozone, MTF50's are as follows:


85 @ 1.2 on 15mp crop: 2400 center, 2075 border


135 @ 2 on ff: 3306 ff, 2769 border.


Not close.


(Oops, I said I was done babbling for the day. Shows you can't trust anything I say....)

[neuroanatomist]

How about comparing the 85 1.2 to the 135 f/2? Both are pretty sharp, right? Is that unfair?

Yep, still unfair. IMO, comparing different lenses and different bodies at the same time is just meaningless. Bryan's statement on his [url="http://www.the-digital-picture.com/Help/ISO-12233.aspx]About ISO 12233 Chart[/url] page ("Lenses should be critically compared to each other only with test samples from the same camera body as it is the combination that is tested,") and the PZ disclaimer near the top of every list of tested lenses ("Please note that the tests results are not comparable across the different systems,") both pretty clearly support that belief.


It makes sense if you think about it - the MTF number from PZ are in LW/PH. Different lenses have different numerators. Different sensors have different denominators. It still might work, with heavy use of correction factors, if the systems and their performance were linear - but optical systems are usually inherently nonlinear.


It's really not about 'loading the game' one way or the other - as you (at least, it may have been you - man, this is a lengthy thread and I'm losing track!) correctly pointed out, these are 'systems' = camera + lens. The only you can make valid comparisons is to hold one part of the system constant and vary the other. Sort of like Heisenberg's Uncertainty Principle, except that I really don't want to drag this thread down into the nether regions of quantum physics.


Slightly tongue-in-cheek, I think it's unfair to use the 135mm f/2L in any kind of sharpness comparison. You could put that lens on an old[url="http://en.wikipedia.org/wiki/Canon_EOS_D2000]Canon D2000[/url] and it would still be razor sharp.


[quote=Mark Elberson]I thought it was interesting to hear the thought process of two very well versed photographers describing their point of view.[/quote]


Wow, thanks Mark! I certainly aspire to be a well-versed photographer...but I'm not there yet. [:$]





So, after much debate and flying pixels...here are my thoughts:


[b]1.[/b] The 50mm f/1.2 is out. All of my pixel peeping has convinced me that I'm just not going to be happy with it. That doesn't mean it's a bad lens, but I have a hard time abiding unsharpness, and I'd be using it at the widest and therefore unsharpest (not a word, I know) part of it's range.


[b]2. [/b]The 5DII is out. FF is still in my future...but as someone else pointed out in this thread (or maybe some other thread, or many other threads...did I mention I'm losing track and a little sleep-deprived in the bargain?) - the best camera/lens/whatever is the one that's available to you now. The 7D is available to me now, and more importantly, I'm quite happy with it. So, I'll wait a while longer and refine my lens collection, and see what the 5DIII is like. I'm hoping for some of the 7D's AF and fps to make it into that upgrade. To my way of thinking, lenses are long term investments, bodies will be changed out over time (in my case, I lasted 5 months between T1i and 7D - October, 2009 to March, 2010...and I promised myself that I'd wait until 2011 for another body). The 5DII is a great camera, but it's a year older in its product life cycle than the 7D, and IMO a replacement is likely next year (as mentioned by Vincent Laforet in his recent webcast). Also, I can't see any way to use the 72mm filter left behind from selling the 200mm f/2.8L prime directly on a 5DII.


[b]3.[/b] The 35mm f/1.4L is out, for now. I still think it's due to be replaced soon. But even if not, my shooting tests over the last few days resulted in an important insight for me - I do like and will definitely use the 35mm focal length indoors, but pertty much everything I shot at that length was a 'family memory' type of shot, and none really would have benefitted from f/1.4. My EF-S 17-55mm f/2.8 with a bounced Speedlite consistently gave me the desired outcomes for those situations.


[b]4.[/b] The 50mm f/1.4 is out - for mostly the same reasons as the 35L. I could see using the 50mm length with a wide aperture (on 1.6x) when my 2 year-old is a little bigger. But by then, I'll most likely have a FF body for portraits and around-the-house shooting (keeping a crop body for wildlife) - in which case, 85mm will be ideal (and then I'll be looking for a 135mm f/2L for those tight shots - it's that whole mouse and cookie thing again, although maybe we've moved on to the moose and the muffin by now...).


[b]5.[/b] That leaves the 85mm f/1.2L II. I really love the 85mm focal length around the house, and could use the extra stop over the 85mm f/1.8. Also, looking at both PZ's numbers and Bryan's ISO 12233 charts (comparing the two 85mm lenses on the same cameras, of course, sorry for harping!), the 85mm f/1.2L II when stopped down to f/1.8 is sharper than the85mm f/1.8. So, I'm giving up a little in AF speed (not much, I think), getting a wider aperture for better low-light performance and thinner DoF (as good as it's going to get on my 7D, in that focal length range at least - and longer isn't feasible indoors), and not giving up anything on sharpness at the wide end of the 85mm f/1.8.


[b]6. [/b]Denise - here's where you say (again!), "I told you so..." []





Thanks again to everyone for your input, and thanks especially to you Jon for sticking with the discussion!


(note: I reserve the right to re-bump this thread when the next bout of <span style="font-size: small;"]<span style="color: red;"]Lens <span style="font-size: small;"]<span style="color: red;"]Lust or Body esire strikes...)

[Keith B]

5. That leaves the 85mm f/1.2L II. I really love the 85mm focal length around the house, and could use the extra stop over the 85mm f/1.8. Also, looking at both PZ's numbers and Bryan's ISO 12233 charts (comparing the two 85mm lenses on the same cameras, of course, sorry for harping!), the 85mm f/1.2L II when stopped down to f/1.8 is sharper than the85mm f/1.8. So, I'm giving up a little in AF speed (not much, I think), getting a wider aperture for better low-light performance and thinner DoF (as good as it's going to get on my 7D, in that focal length range at least - and longer isn't feasible indoors), and not giving up anything on sharpness at the wide end of the 85mm f/1.8.

85 1.2LII is melt in your mouth goodness!

[Jon Ruyle]

Bryan's statement on his [url="http://www.the-digital-picture.com/Help/ISO-12233.aspx]About ISO 12233 Chart[/url] page ("Lenses should be critically compared to each other only with test samples from the same camera body as it is the combination that is tested,") and the PZ disclaimer near the top of every list of tested lenses ("Please note that the tests results are not comparable across the different systems,") both pretty clearly support that belief. [/quote]


I believe they say that because 1) if your goal is to compare lenses (as Bryan's and PZ's reviews are) you shouldn't compare across different systems 2) you have to know what you're doing to compare across different systems and most people don't (think how much confusion the mere mention of DLA caused on Bryan's site). It's even possible that they themselves don't know (PZ anyway... I'm not one to doubt Bryan's knowledge)


[quote=neuroanatomist]That leaves the 85mm f/1.2L II.[/quote]


Okay, we can agree on that I'm all for the 85 f/1.2 II.


[quote=neuroanatomist]Thanks again to everyone for your input, and thanks especially to you Jon for sticking with the discussion![/quote]


Well, thanks for putting up with me []


[quote=neuroanatomist](note: I reserve the right to re-bump this thread when the next bout of <span style="font-size: small;"]<span style="color: red;"]Lens <span style="font-size: small;"]<span style="color: red;"]Lust or Body esire strikes...)

As do I, the next time I have a wallet full of $100 bills. (Right now my largest bill is a $2).

[neuroanatomist]

I believe they say that because 1) if your goal is to compare lenses (as Bryan's and PZ's reviews are) you shouldn't compare across different systems

True.

2) you have to know what you're doing to compare across different systems

Bryan actually specifically mentions comparing bodies, as well, and indicates that the same lens should be used to compare bodies (e.g. the 200mm f/2.8L or 200mm f/2L IS, since he's got shots with those lenses from a wide array of bodies).


But, when you talk about comparing systems (i.e. camera + lens), it's pretty straightforward to compare systems, I think. If you want to compare the 200mm f/2.8L on a 5DII with the 135mm f/2L on a 7D, you just need some test targets (resolution, color, three-dimensional for DoF, etc.) and you need a200mm f/2.8L, a 5DII, a 135mm f/2L, and a 7D. Simple, right? The complicated part comes when you try to compare those systems without the components that make up those systems, and instead start making approximations and assumptions, most of which are invalid.

I'm all for the 85 f/1.2 II.

Right. But, should I hang onto the 85mm f/1.8 too?!? [*-)]

[Fast Glass]

Sorry for bringing this up again, but I agree with Jon on this one.

Thing 1) Is it true that it make sense to compare APS-C vs FF for effective F and f or actual F and f? and

It makes sense to compare effective F and f becuase thats what you get as your end result when you take a picture. You have to compare with effective F and f, there's just no other way around it.

Thing 2) Is it true that by comparing effective f and F instead of actual, the IQ edge goes to the lens on the full frame camera?

Yep, especially if you compare ultra-wides. The differance is huge.


It's about taking the shot at this particulare focal length, which is the best system for it? And the obvious answer is FF, I think John agrees with this. But to say that 1.6 crop is not that bad when using the same lens is not comparing systems and istherefore irelavant. Because it took a certain focal lenth to make that certainshot, and you cannot makethat ceratain shot without that certain focal length.[] There is always an equivalentin FF so it's a never ending story.


If you do not persue this thread any longer I don't blame you! But I hope Jon weighs in on this![:P]


No worries,


John.