Fred Miranda wrote:
What they wrote on the FAQ sounds right.
From what I posted before, I understood that when doing their resolution tests, they obtain readings at the sharpest focus on the center (to check the center resolution, then do the same for the corners and extreme corners) independently. Therefore their tests are field curvature invariant.
I agree that we should not make our choices based solely on graphs but it's a good starting point.
But this method of testing is not in fact invariant with respect to real-world results. Say you have two lenses A and B, where A has no field curvature and B has strong field curvature. If you test them according to TDP's methodology, which is a single shot at sharpest focus at the image center, then in the corner, A will most likely be sharper than B (assuming there are no other aberrations suffered by either lens). But if you test it according to Photozone's methodology, B *could* be sharper than A.
The problem with testing the center and periphery separately, then, is that doing so masks whether or not a lens is adversely affected by field curvature. Although I haven't actually bothered to look at the numbers or the test charts for these lenses, what has been claimed in this thread essentially suggests that the EF 35/2 is less well-corrected for field curvature than the EF 35/1.4L. When I want to compare the performance of two lenses, I absolutely do want to know whether it suffers from such an aberration. Compensating for it in this manner is highly misleading, even if Photozone's opinion is that it's not relevant to real-world usage.
lhdvries wrote:
Interesting thread-I just returned from returning my 35L to B&H after deciding the 35f/2.0 was all I needed.
No wonder things cost so much at camera stores. Based on posts like this and MANY, MANY other ones, people seem to use retail stores as research centers, buying and returning stuff for really no good reason.
troy12n wrote:
No wonder things cost so much at camera stores. Based on posts like this and MANY, MANY other ones, people seem to use retail stores as research centers, buying and returning stuff for really no good reason.
AGeoJO -- Did you correct the distortion of the 40mm in PP? It looks like you have some upward shift from your vantage point in the first shot. I would have expected some barrel distortion from this lens. Good looking shot, just trying to get full data. Thanks.
Gunzorro wrote:
AGeoJO -- Did you correct the distortion of the 40mm in PP? It looks like you have some upward shift from your vantage point in the first shot. I would have expected some barrel distortion from this lens. Good looking shot, just trying to get full data. Thanks.
Jim,
Thanks! No, no distortion correction was done on the shot.
wickerprints wrote:
But this method of testing is not in fact invariant with respect to real-world results. Say you have two lenses A and B, where A has no field curvature and B has strong field curvature. If you test them according to TDP's methodology, which is a single shot at sharpest focus at the image center, then in the corner, A will most likely be sharper than B (assuming there are no other aberrations suffered by either lens). But if you test it according to Photozone's methodology, B *could* be sharper than A.
The problem with testing the center and periphery separately, then, is that doing so masks whether or not a lens is adversely affected by field curvature. Although I haven't actually bothered to look at the numbers or the test charts for these lenses, what has been claimed in this thread essentially suggests that the EF 35/2 is less well-corrected for field curvature than the EF 35/1.4L. When I want to compare the performance of two lenses, I absolutely do want to know whether it suffers from such an aberration. Compensating for it in this manner is highly misleading, even if Photozone's opinion is that it's not relevant to real-world usage....Show more →
That shows that testing a lens is not an easy job. At least we know how they are testing them.
You can either dismiss field curvature or take it in consideration. It seems like Photozone's reviews dismisses field curvature since it believes it gives too many variants. Field curvature can be affected by distance as well making it very difficult to test. For example, some lenses have pronounced field curvature at infinity but not at normal distances. etc.
From the review tests along with "quick tests" on this thread, it seems like the 35mm f/2 have a more pronounced field curvature at closer range. When focusing on the edge of the frame (Like Photozone's and Photon tests), it gives sharper corners but when focusing on the center, (TDP and twistedlim tests) we get blurred edges (because of perhaps field curvature)
The easiest way to test if your wide angle has a pronounced field curvature is to use Live View.
Take a picture of a chart on the wall (or a brick wall). Align it perfectly and focus on the center of the target. Then check your image at 10x on the center and then on the edges. Then do the same but this time focus on the edge. If there is a considerable different between center and edges in both shots, your lens exhibits a noticeable field curvature at that distance.
Some photographers may erroneously think they have a bad lens when dismissing field curvature. IMO, the best way is to first know the limitation of your lens (Since there is nothing we can do about field curvature) and then know which settings should to be dialed to get the best picture from it.
Fred Miranda wrote:
Some photographers may erroneously think they have a bad lens when dismissing field curvature. IMO, the best way is to first know the limitation of your lens (Since there is nothing we can do about field curvature) and then know which settings should to be dialed to get the best picture from it.
This is, I think, part of the reason that many people report poor performance with ultra-wide zooms - along with their failure to take into consideration DOF issues when they compose images with very close foregrounds and the lower right/left corners.
Fred Miranda wrote:
You can either dismiss field curvature or take it in consideration. It seems like Photozone's reviews dismisses field curvature since it believes it gives too many variants. Field curvature can be affected by distance as well making it very difficult to test. For example, some lenses have pronounced field curvature at infinity but not at normal distances. etc.
From the review tests along with "quick tests" on this thread, it seems like the 35mm f/2 have a more pronounced field curvature at closer range. When focusing on the edge of the frame (Like Photozone's and Photon tests), it gives sharper corners but when focusing on the center, (TDP and twistedlim tests) we get blurred edges (because of perhaps field curvature) ...Show more →
I will add that the presence of field curvature per se is not an indicator of a lens' quality (or the lack of it). One just has to be aware of any lens' behaviour to best exploit it. Canon's 14L II, the subject of another ongoing thread, exhibits pronounced field curvature. Diglloyd has tested it with real subjects (results behind a paid subscription wall).
The Canon 14mm f/2.8 is a good example of how distance affects field curvature. If you focus on a foreground element, FC is almost imperceptible. However when focus at infinity it is more pronounced. By knowing this, you take take better advantage of the lens instead of discarding it as a bad copy...
We know that all lens designs have compromises. We just need to get to know our own gear based on our photography style. I don't believe we have so many bad copies out there. Some copies may have decentering and focus issues (backfocusing for example) but other than that there is not much to worry about.
BTW: If you want to test for decentering, shoot a blue sky or white wall wide open and then check the vignette of the photo. If it's more pronounced in one corner than the others your lens may need a tweak.
"Normal" people who look at photos don't do forensic evaluations and analysis of every inch and aspect of the image.
Looking at these examples side by side tells the tale. The pics from the f.1.4 vs the f/2 look the same.
With lower noise in extremely high ISO improving with every new camera release.......having f/1.4 lenses won't be all that important........unless you want to take pictures of black bats at midnight.
Jan 12, 2012 at 11:56 PM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
Fred Miranda wrote:
BTW, since we are discussing portability on this range, have anyone tried the Voigtlander Ultron 40mm f/2.0 SL-II for the Canon mount?
I own the Voigtlander Ultron 40 f/2 with Canon mount. and I use it both on my 1Ds3 & 1D4. It's a very good lens for it's tiny size. At this moment I'm using it when travelling in S.E.Asia. It's great for that purpose. And the image quality is nice. The price is also rather low for what you get. The build quality is better than the Canon 35 L. All metal except the very small rubber on MF ring. It has the Zeiss build quality.
It cost about $ 400 here in Bangkok. The price also include a hood and a close-up lens
Tom K. wrote:
"Normal" people who look at photos don't do forensic evaluations and analysis of every inch and aspect of the image.
Looking at these examples side by side tells the tale. The pics from the f.1.4 vs the f/2 look the same.
With lower noise in extremely high ISO improving with every new camera release.......having f/1.4 lenses won't be all that important........unless you want to take pictures of black bats at midnight.
You are right and that's why there's the 35L for those that need or want to take picture of black bats at midnight.
Jan 13, 2012 at 11:45 AM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
Tom K. wrote:
"Normal" people who look at photos don't do forensic evaluations and analysis of every inch and aspect of the image.
Looking at these examples side by side tells the tale. The pics from the f.1.4 vs the f/2 look the same.
With lower noise in extremely high ISO improving with every new camera release.......having f/1.4 lenses won't be all that important........unless you want to take pictures of black bats at midnight.
I don't buy fast lenses because I like shoot in the dark. The main reason for me is being able to blur the background. And also the bokeh. So lower noise in high ISO doesn't help
Fred Miranda wrote:
We know that all lens designs have compromises. We just need to get to know our own gear based on our photography style. I don't believe we have so many bad copies out there.
Hear, hear - and a merry double plus one! :-)
I know that this isn't news to quite a few folks and that others will probably never be able to accept it, but the majority of the huffing and puffing about lens performance is pretty darned irrelevant.
I'm not at all saying that lens performance is irrelevant. What I am saying is some of the following:
1. As Fred points out, no lens is perfect. I like to say that every lens has a "personality" - strengths and weaknesses that might make it more or less suited to your particular photographic needs.
2. Almost all of the purported lens flaws or factors that differentiate the performance of one lens from that of another turn out to be far less significant that people imagine them to be. While some lens might perform .05% better in some idealized measure of, say, resolution than some other lens... in virtually all cases this will make utterly no difference at all in your photographs.
3. Meanwhile, other things that seem more mundane can make a difference. For example, in the case of the 35mm primes, accept for the moment that that both lenses are actually excellent optical performers. With that notion in mind, there could be objective advantages in having a lens of smaller size and weight and lower cost. (I know that when I'm carting around a bag full of lenses and bodies and couple of tripods I have little or no interest in carrying bigger and larger lenses.)
It is so easy to get caught up in the Big And Very Expensive Lens contest. The choices seems safer - "it is the most expensive, it must be the best!" The gear itself can be so fascinating - "I love really well built stuff!" The "pro" label can cause us to take leave of our senses - "This is the gear the pros use!" And building on all three of those, who doesn't want to look like a Real Photographer when they are out in the field - "My lens is bigger and more expensive than your lens!" (Like the guys carrying around the 300mm f/2.8... on their local photo walk! :-)
All of this conspires to make it very difficult to resist the notion that the biggest and most expensive lens is the lens that you need. Sometimes it can be. Most often it isn't.
And, in the end, it is all about making photographs, isn't it?
Take care,
Dan
(And, yes, I do own some Big Expensive Lenses, too... ;-)
Jan 13, 2012 at 12:33 PM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
With lower noise in extremely high ISO improving with every new camera release.......having f/1.4 lenses won't be all that important........unless you want to take pictures of black bats at midnight.
1- Your reasoning works if your goal is to freeze motion (like a bat flying :-)): higher ISO will help you here to get faster shutter speed.
2- Your reasoning is not working your goal is to isolate your subject from the background: apperture is the key here.
