Too narrow an aperture = bad?

Superb! Fine quality testing Tim :)

I would guess again that this is sharpest between settings, like f/5.6-8? It's certainly moved up a bit. Also, the fall-off at f/22 and f/32 doesn't look quite so steep.

Edit: if these were taken from the same position as the 7D, you could take the 5D2 images, crop them down to 1.6x format, and have 8.2mp vs 18mp from the 7D... ;)
 
Last edited:
HoppyUK said:
Superb! Fine quality testing Tim :)

I would guess again that this is sharpest between settings, like f/5.6-8? It's certainly moved up a bit. Also, the fall-off at f/22 and f/32 doesn't look quite so steep.

Edit: if these were taken from the same position as the 7D, you could take the 5D2 images, crop them down to 1.6x format, and have 8.2mp vs 18mp from the 7D... ;)
Click to expand...

I thought about making a like for like comparison by cropping and resizing to match, although to be honest I'm not sure what it will prove. These are very diificult things to make meaningful comparisons of. I'm sure over the ages countless comparisons have been made between all manner of cameras by uprezzing and downrezzing files and at the end of the day what does it really prove? One or other file, or perhaps both, is going to end up *******ised if you want to view on screen.

The 5D2 will make better pictures if you use the whole sensor. If you crop it down to the size of the 7D you will potentially see more detail from the 7D, unless you crush the file so small that nobody can see the difference, or diffraction, noise and other miscreants render the advantage moot. Unless you are going to print, large, IMHO it's just like organising deckchairs on the Titanic as she sinks - a waste of time.

There are so many variables in any case. How do you shoot to make the most valid like for like comparisons? Should I not have used a 160mm lens on the 5D2? Should we be comparing equal apertures or apertures adjusted for equivalent DOF on the different formats, given the change in focal length? If we change those things then shouldn't we also change ISO in order to maintain shutter speed? Remind me what it is we are supposed to be comapring? (Rhetorical)

The bottom line is that as you stop down more and more diffraction will harm your IQ. What impact that has will be governed by your chosen f/stop and your final enlargement factor. At small size, such as for web, you can probably stop down an awfully long way. For giant prints (or computer displays) viewed at close distance, or for pixel peeping you probably cannot.
futureal33 said:
Was that performed with the 17-55 IS Tim?
Click to expand...
No. The 100/2.8L IS Macro was used on both cameras.
 
Have to say I agree Tim, and well put. But what's a rhetorical question? :D

The point I was driving at was how pixel density lowers the point at which diffraction impacts, and while I accept the theory, I don't accept that it has any meaningful significance in practise. If there was anything there to worry about, even though you were maybe not specifically testing for it, something would have shown up as not quite right.

The numbers (theoretical) can be quite alarming, yet we constantly see ever higher pixel densities and with them higher sharpness potential - proved by actual photographs - yet the theory might lead you to believe there comes a point when it works the other way. If that's no so, someone should tell Canon about it.

Thanks for your time with this. Excellent work, interesting stuff. Cheers :) What is evidently clear is that high f/numbers are not good for ultimate sharpness :eek:
 
HoppyUK said:
Have to say I agree Tim, and well put. But what's a rhetorical question? :D

The point I was driving at was how pixel density lowers the point at which diffraction impacts, and while I accept the theory, I don't accept that it has any meaningful significance in practise.
Click to expand...
Well in practice it might make a difference if you were to fall into the trap of stopping down for additional sharpness.

Suppose you are a wildlife shooter, strapped for focal length and using a 100-400 zoom lens with a 1.4X TC. You have the option of a 5D2 or 7D with which to shoot. You know the 7D will put more pixels on your subject. You also know/believe that the 100-400 sharpens up a bit when stopped down to maybe f/8 or f/9. If you stop down then with the TC you'll be at f/11 or f/13.

Now, which setting will give you the best results and the most usable detail....?

7D at 560mm and f/8;
7D at 560mm and f/11:
5D2 at 560mm and f/8:
5D2 at 560mm and f/11;

But what about....
7D at 400mm and f/5.6;
7D at 400mm and f/8;
5D2 at 400mm and f/5.6;
5D2 at 400mm and f/8;

How about if you need 800 ISO for adequate hand held shutter speed at 400mm and f/5.6, 1600 ISO for 400mm and f/8, 2500 ISO (Oh dear) for 560mm and f/8 and 5000 (Oh dear again) ISO for 560mm and f/11?

Which will give the best result? Answer.....

Probably the 7D with 500/4 with a 1.4X TC at 700mm, f/7.1, 200 ISO on a tripod. ;)
 
Last edited:
Haha! Nice work Tim :)

It's a compromise at the end of the day and with choices like that I suspect that ultimately the most important factors will be rather less to do with the lens and more about high shutter speeds and a big tripod :D
 
Andy in response to your question on why don't manufacturers optimise their lenses at the widest ap - I would guess it's because people are not generally going to want the DOF to be that shallow in most cases (even if they could optimise it wide open).

Fast lenses have the advantage of a shallow DOF when it's needed but how often is that in reality.
 
You must log in or register to reply here.
Back
Top