Acceptable Sharpness@f/2.8

[james_mccormick]

<p>Hi Guys,<br>

To start with, I must confess, yes I am pixel peeping! I want to ask you all this question - I have recently purchased a 24-70/2.8L and have been mildly disappointed with the sharpness of this lens even compared to my 50/1.8!<br>

Do you think the attached image is acceptably sharp for this lens? It was shot indoors at around ISO400, and a shutter speed of around 1/500 at f/2.8. To be clear what I have done to it in ps: 0.5px sharpen on the eyes. Some exposure, highlights and shadows adjustment. <br>

Am I being unrealistic in what I expect from this lens?<br>

Thanks in advance,<br>

James</p><div></div>

[james_mccormick]

<p>I forgot to mention that I reduced the clarity a touch in camera raw to soften the skin a little.</p>

[rjtully]

<p>Looks like the eyes and Over-alls are in focus. DOF apparently put the rest out. I like PS unsharp at 150 1 4 for this picture but that is just me.</p>

[randallfarhy]

<p>As Richard notes, the eyes and over-alls are in focus, the eyes more so. I see nothing that would concern me about the lens being soft in this image as the softness appears to be a function of the shallow depth of field created by F2.8. proximity to subject and the longer end of the lens' range.<br>

A better test would be to shoot a wider focal length, at F5.6-F8 and look for even sharpness from edge to edge. A good subject for this kind of test would be a tree line, a building or such. Just be sure to keep the subject parallel to the focal plane.</p>

[john_a5]

<p>"even compared to my 50/1.8!"</p>

<p>That is an interesting statement, if you look at the DxO rating for that lens, it is higher than for the 50/1.2L--only the Zeiss Macro 50/2 has as high a rating among <a href="http://www.dxomark.com/index.php/Lens-with-Camera/Compare/Compare-lenses/%28lens1%29/197/%28lens2%29/187/%28onglet%29/0/%28brand%29/Canon/%28brand2%29/Canon">50mm lenses for Canon</a>---In fact, the only lens for Canon I saw with better resolution was the 85/1.2LII lens!</p>

<p>Anyway, this zoom lens is <a href="http://www.dxomark.com/index.php/Lens-with-Camera/Compare/Compare-lenses/%28lens1%29/165/%28lens2%29/187/%28onglet%29/0/%28brand%29/Canon/%28brand2%29/Canon">not as sharp</a>, but still a very good lens. At the size you posted, I agree with the others here, but f2.8 isn't the optimal f-stop for this lens. Digital cameras seem to show the diffraction much more than does film and stopping down is best if one can do it--given the effect you are after. That said, I do shoot a lot of environmental portraits in low light with this lens and find the results to be very good--even in very large prints as long as you stay away from the corners. You just need to bring the file together with the sharpening in your RAW processor a bit more than you might with the 50--if you use Photoshop, CS5 sharpening is a quantum leap over previous versions and probably equal to the best out there.</p>

[paddler4]

<p>Also, if you want to test the sharpness of a lens, reducing clarity is not a great idea--it will make the image look soft. I agree with others that the eyes look to be in focus nonetheless and that more appropriate test is in order.</p>

[sravan]

<p>Clarity reduces apparent sharpness. Try applying clarity adjustment only to selected areas using some brush / mask instead of applying it all over.<br>

To me it still seems good. The eyelashes are well defined and the weave on the jeans part of hte dress is perfectly visible. To me the whol epicture does have sufficient sharpness. Not sure what you are feeling is missing.<br>

BTW photos of young kids with perfect skin is a bad test for sharpness in my mind.</p>

[jkaufman]

<p>Another consideration is whether you used a tripod while shooting. The effects should be essentially negligible at 1/500 sec, but even negligible things become obvious when pixel-peeping.</p>

[PuppyDigs]

<blockquote>

<p>That is an interesting statement, if you look at the DxO rating for that lens, it is higher than for the 50/1.2L--only the Zeiss Macro 50/2 has as high a rating among <a rel="nofollow" href="http://www.dxomark.com/index.php/Lens-with-Camera/Compare/Compare-lenses/%28lens1%29/197/%28lens2%29/187/%28onglet%29/0/%28brand%29/Canon/%28brand2%29/Canon" target="_blank">50mm lenses for Canon</a>---In fact, the only lens for Canon I saw with better resolution was the 85/1.2LII lens!</p>

</blockquote>

<p>DxO musta tested a really good 50 1.8. The 3 I owned were terrible below F2.8 and suffered heavy barreling at 2 meters or less. Both my 50 2.5 and 50 1.2 were much better at F4 and faster.</p>

[mark_pierlot]

<p>No lens performs optimally at its maximum aperture. Some (such as the EF 85/1.2 L II) perform better wide open than others, but none is at its best.</p>

<p>My 24-70 is a lens I "agonized" over until I finally decided to get rid of it in favour of faster primes. My primary application for the zoom was informal portraiture of my kids indoors in available light (which accords with the image you posted). Since I found my primes to be sharper at wider apertures than the zoom, coupled with the fact that I didn't really need a zoom indoors anyway, I felt that the money that was tied up in my 24-70 would be better put towards a lens I'd get more use out of.</p>

<p>But if the long-awaited EF 24-70/2.8 L <strong><em>IS</em></strong> ever arrives, I will almost undoubtedly acquire a copy. In many situations, the gain in sharpness that IS provides trumps resolution.</p>

<p>By the way, none of this is intended as a criticism of your image, James, which I consider to be a very expressive portrait of your son. Ah, the innocent wonder of childhood!</p>

[dan_south]

<p>At f/2.8, don't expect sharpness away from the plane of focus. How sharp are the pixels of the objects on that plane?</p>

<p>Besides, it you take an extraordinary image, viewers will tend to overlook technical issues such as sharpness. ;-)</p>

[smooth_carrots]

<blockquote>

<p>Another consideration is whether you used a tripod while shooting. The effects should be essentially negligible at 1/500 sec, but even negligible things become obvious when pixel-peeping.</p>

 

</blockquote>

<p>I had to laugh at that statement. What's the point of using a tripod when photographing a child? Children are never still (except at 1/500 sec!)</p>

<p>As for the OP, no point in comparing a zoom lens wide open to a prime lens renowned for its sharpness. Your lens is fine.</p>

[g dan mitchell]

<blockquote>

<p><em>"I forgot to mention that I reduced the clarity a touch in camera raw to soften the skin a little."</em></p>

</blockquote>

<p>Without commenting on the overall sharpness of the image - there are many potential reasons for less than sharp results - I'd just like to point out that an adjustment to the overall photo to soften skin is probably not going to get your the results you want in many cases. </p>

<p>For example, it is typical that you want sharp eyes and eyebrows and other things, but you may want a bit softer skin. This requires some selective work in post - not a wholesale softening of the whole image.</p>

<p>Dan</p>

[robert gordon]

It's the expression that counts!

[Robin Smith]

<p>The image looks very processed to me with plastic skin. Whatever you did in processing it I would stop - it could conceivably be necessary for a model shoot but not for a young boy. It looks to me as though you missed focus - his dungarees or nose are in focus and his eyes are not - that is why it does not look sharp. Of course, it's only a small image so it's hard to tell.</p>

<p>Yes, the 50mm is sharper, but at 50mm the 24-70 is pretty good and I don't think anyone would say it is "not sharp".</p>

[james_mccormick]

<p>Thanks for all your comments. I don't tend to look too much into MTF charts as I find that they do not really represent a real world scenario, but that is just my opinion.<br>

@Mark Peierot, I fully agree with your statements fully. I shot most of my images for the two hours at f/2.8 merely to test the lens wide open. under normal circumstances I would have stopped to at least f/4.<br>

Another reason I ask is that I am currently moving into the area of paid shoots just to fund my hobby (at the moment as I am nowhere near skilled enough for any more than that). Alot of mums around here want environmental shots of their children playing rather that your typical uber-clean studio shots (although I must admit I do like those kinds). So my next question really is, would you have been happy to part with cash for shots like that? Would I be over charging if I charged £70 for a 2 hour shoot?<br>

@Robin Smith, whilst I appreciate you taking the time to respond, I must disagree with you slightly. No selective processing was applied to the little boys skin only a global drop in clarity in camera raw and certainly not enough to make his skin look plastic. Also, no contrast adjustments were made. I pulled the highlights and shadows a tiny bit just to spread the histogram and widen my dynamic range. It is interesting that the consensus is that the eyes are in focus some of whom seem to imply that they are the sharpest part of the image! I am not moaning at you here as you are entitled to your opinion and I certainly accept that :-)</p>

<p>@Robin Gordon - so true! But did I get that expression? ;-)</p>

<p>Again, thanks for all your comments, I think I got a pretty good copy of this lens so will stop pp'ing and just get shooting and improving my technique!<br>

James</p>

<p> </p>

[mark_pierlot]

<blockquote>

<p>The image looks very processed to me with plastic skin.</p>

</blockquote>

<p>To me, it looks like the skin of a young child. I don't see any evidence of doctoring. It is, perhaps, a bit over-lit.</p>

<p> </p>