Home · Register · Search · View Winners · Software · Hosting · Software · Join Upload & Sell

Moderated by: Fred Miranda
Username   Password

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1
       2       end
  

Archive 2010 · Problems with Uneven Depth of Field
  
 
stargazer78
Offline
• •
Upload & Sell: Off
p.1 #1 · p.1 #1 · Problems with Uneven Depth of Field


This photo was taken with a 24mm lens @ f/11 on a 5D2. RAW conversion was done through DPP, and sharpened in Photoshop (200% USM, 0.3 Radius, 0 Threshold).






Full Frame







100% Crop of Area A



Looking at the 100% Crop of the area labeled "A" (above). The leaves are reasonably sharp, but the background is very soft. Even though the area is close to the edge of the frame, I can't attribute the softness to the lens. A soft lens presumably would've rendered both the foreground and the background soft. Since the foreground is plenty sharp, I can only assume that the background softness is due to insufficient depth of field @f/11. But then I look at the area labeled "B" (below).






100% Crop of Area B



Here, we see the leaves from the same tree as before. The foreground leaves are still sharp. However, this time the background also appears sharp and in focus. What's even more puzzling is that the background here (near center) is actually much further away than the background from area "A" (near edge). Even with field curvature, I would've thought that the background in Area "A" would remain in focus.

I have many other samples that are similar to this, from different lenses. I often find my wide angle lenses can render the foreground detail very sharp across the frame. But sharpness in the background is very uneven. The center background is often in focus while the background near the edge look out of focus (even if they are physically closer than the center background). It's almost like the depth of field near the edges is significantly more shallow than the DoF at the center.

I don't know what to make of this. Anybody have a theory or explanation?






Nov 21, 2010 at 11:08 PM
skibum5
Online
• • • • • •
Upload & Sell: Off
p.1 #2 · p.1 #2 · Problems with Uneven Depth of Field


stargazer78 wrote:
I don't know what to make of this. Anybody have a theory or explanation?



field curvature, where the focal plane moves in closer and closer (or farther and farther) the closer to the edges you get, lots of wide angles have this and most lenses to at least a tiny extent

where you place the main focal plane in the center can have some affect and if placed in a certain spot can help things




Nov 21, 2010 at 11:30 PM
Richard Nye
Offline
• • • •
Upload & Sell: On
p.1 #3 · p.1 #3 · Problems with Uneven Depth of Field


I don't know about the DOF, but you must have been using a polarizer to make that sky so uneven.


Nov 21, 2010 at 11:39 PM
n0b0
Offline
• • • • •
Upload & Sell: Off
p.1 #4 · p.1 #4 · Problems with Uneven Depth of Field


Atmospheric interference?


Nov 21, 2010 at 11:43 PM
EB-1
Offline
• • • • • • •
Upload & Sell: Off
p.1 #5 · p.1 #5 · Problems with Uneven Depth of Field


Was that taken with a zoom lens?

EBH



Nov 21, 2010 at 11:51 PM
Specularist
Offline
• •
Upload & Sell: Off
p.1 #6 · p.1 #6 · Problems with Uneven Depth of Field


stargazer78 wrote:
Even with field curvature, I would've thought that the background in Area "A" would remain in focus.


Why, though? A curved field means the distance for best focus will vary across the frame, so it stands to reason that if the centre is focused at infinity the edges may not be.

I strongly suspect this is simple field curvature, but out of interest, who made the filter you used in this shot?



Nov 22, 2010 at 12:03 AM
stargazer78
Offline
• •
Upload & Sell: Off
p.1 #7 · p.1 #7 · Problems with Uneven Depth of Field


n0b0 wrote:
Atmospheric interference?


I don't think it is, since the blurry background in Area "A" is closer to the camera than the sharper background in Area "B"




EB-1 wrote:
Was that taken with a zoom lens?


Yes, it was taken with a 17-40L @ 24mm.






skibum5 wrote:
field curvature, where the focal plane moves in closer and closer (or farther and farther) the closer to the edges you get, lots of wide angles have this and most lenses to at least a tiny extent


That is one of my suspicions as well. But I never though field curvature would have such a noticeable arc. The 17-40L is Canon's cheapest and smallest "L" series zoom lens. Perhaps a strong field curvature is part of the design compromise. I wonder if the 16-35L would be an improvement in that regard...




Nov 22, 2010 at 12:16 AM
stargazer78
Offline
• •
Upload & Sell: Off
p.1 #8 · p.1 #8 · Problems with Uneven Depth of Field


Specularist wrote:
Why, though? A curved field means the distance for best focus will vary across the frame, so it stands to reason that if the centre is focused at infinity the edges may not be.


I did not focus at infinity. The focus point is actually at the little tree in the foreground. And so the field curvature would have to be very strong if the distant canyon in the center is in focus while the closer terrain at the left remains out of focus.




Specularist wrote:
I strongly suspect this is simple field curvature, but out of interest, who made the filter you used in this shot?


I was using the Hoya HD series CPL.



Nov 22, 2010 at 12:25 AM
EB-1
Offline
• • • • • • •
Upload & Sell: Off
p.1 #9 · p.1 #9 · Problems with Uneven Depth of Field


Zooms often suffer from field curvature though that one is a bit odd. EBH


Nov 22, 2010 at 12:42 AM
gdanmitchell
Offline
• • • • •
Upload & Sell: On
p.1 #10 · p.1 #10 · Problems with Uneven Depth of Field


A is nearer the side of the frame, so it could simply be normal softness closer to the edge of the frame - the plant also looks less sharp to me in example A.

Another possibility is that the lens is out of alignment and that it affects the far left side of the frame.

Dan



Nov 22, 2010 at 03:24 AM
 

Search in Used Dept. 



ChrisDM
Offline
• • • • •
Upload & Sell: On
p.1 #11 · p.1 #11 · Problems with Uneven Depth of Field


In the case you posted, the depth of field at 24mm/f11 should be something like everything from 5 feet to infinity should be in focus, so I would first double check your technique then consider your lens has a problem.... My 24-70 did something like that, where across the lefthand side of the plane of focus it would lose focus (even though it appeared in the plane of focus). Sent it in to Canon, they realigned elements, problem solved....


Nov 22, 2010 at 04:00 AM
Gochugogi
Offline
• • • • •
Upload & Sell: Off
p.1 #12 · p.1 #12 · Problems with Uneven Depth of Field


Heat also causes landscape backgrounds to appear out of focus.


Nov 22, 2010 at 04:39 AM
gdanmitchell
Offline
• • • • •
Upload & Sell: On
p.1 #13 · p.1 #13 · Problems with Uneven Depth of Field


Gochugogi wrote:
Heat also causes landscape backgrounds to appear out of focus.


True, but not generally in the way that this imager is OOF. You are more likely to see a "heat wave" effect that sort if "breaks up" the image. In addition, the effect is typically seen with very long focal lenses, not very wide lenses.

Dan



Nov 22, 2010 at 07:22 AM
RogerC11
Offline
• • • •
Upload & Sell: On
p.1 #14 · p.1 #14 · Problems with Uneven Depth of Field


What the heck is going on with the sky??


Nov 22, 2010 at 12:14 PM
jcolwell
Offline
• • • • • • •
Upload & Sell: On
p.1 #15 · p.1 #15 · Problems with Uneven Depth of Field


I think the background softness in A region is due to lens softness. The plants are much closer, and so the effect is much reduced. This result is very similar to tests I've done with the 17-40L and various ultra wide primes. The 17-40L is very sharp in the centre and to the nearer edges (i.e. within a radius of 12mm of the centre), but softens towards the far edges and in the corners. For example, the 17-40L is very sharp in the centre, even a bit more than the Tokina AT-X Pro 17/3.5, and the AT-X is noticeably sharper in the corners. For most purposes, both lenses provide acceptable results. Pixel peeping is not one of those purposes.

Here's some examples. The first photo shows the full image, with the red rectangle showing the location of the first comparison set. The following three comparison sets all show 100% crops of the AT-X image on top and the 17-40L on the bottom. Photos taken at f/8, 1/125 sec, ISO 100, tripod, etc. The first set is near the centre, the second set is near the left edge, and the third is the extreme top left corner.






















Nov 22, 2010 at 12:41 PM
The Image
Offline
• • • •
Upload & Sell: Off
p.1 #16 · p.1 #16 · Problems with Uneven Depth of Field


My 1st 16-35mm L II did the same thing, there were areas that clearly should have been in focus, but were not, You might want to send your 17-40 to Canon Factory Service Center to have the specs checked and possibly calibrated.


Nov 22, 2010 at 06:52 PM
gasrocks
Offline
• • •
Upload & Sell: Off
p.1 #17 · p.1 #17 · Problems with Uneven Depth of Field


Is it just me or is someone expecting perfection from a so-so zoom lens on a hot day with a CP?


Nov 22, 2010 at 07:10 PM
stargazer78
Offline
• •
Upload & Sell: Off
p.1 #18 · p.1 #18 · Problems with Uneven Depth of Field


gasrocks wrote:
Is it just me or is someone expecting perfection from a so-so zoom lens on a hot day with a CP?


I'm pretty sure it's just you. Especially since that particular photo was taken on a cold February morning. I'm no meteorologist, but I'm almost certain that a heat haze is unlikely under a 50-degree weather.




jcolwell wrote:
I think the background softness in A region is due to lens softness. The plants are much closer, and so the effect is much reduced.


I don't believe lens softness would explain the discrepancy between foreground and background sharpness. I think depth of field would be the reason for that, not lens sharpness. Shooting at f/16 instead of f/11 probably would've helped.



The Image wrote:
You might want to send your 17-40 to Canon Factory Service Center to have the specs checked and possibly calibrated.


I don't believe my lens is not suffering from any kind of decentered focus --- I've already provided examples proving otherwise. Foliage on the left side wouldn't have remained sharp if that was the case. One of the first things I do after buying a lens is a test for asymmetrical focus, and so I know my 17-40L doesn't suffer from it.



I'm convinced now that the first explanation offered in this thread (by skibum5) is also the most accurate. I always thought that field curvature would only be noticeable at larger apertures. I never considered it would still be visible at f/11 exposures. But today, I was reading the Photozone review of the EF 24-70mm f2.8... and sure enough, they were demonstrating the exact same problem I experienced with the 17-40L! Shot with settings (24mm f/10) nearly identical to the samples I provided above. His samples demonstrate the effects of field curvature at infinity focus, whereas my samples demonstrate field curvature at close focus (approx 2m). Excerpt from the Photozone Review of the EF 24-70mm f2.8:

PHOTOZONE wrote:
The scene below has been taken at 24mm f/10. Now at f/10 you'd normally just rely on the AF which suggests a near-infinity focus distance here. You'd waste some depth-of-field by this but normally everything would still be sharp at this setting (based on this kind of scene). The result can be seen BELOW the sample scene. The extreme corners are blurry - despite f/10 - whereas the lower center is perfectly sharp.









Notice that the sharpness of the bottom center and bottom corner are identical at one focus distance (2m) but not at a different focus distance (infinity). I think that a lot of times, people attribute poor corners to lens softness when in fact the real culprit is field curvature. I've shot my 17-40L countless times at f/11, and have seen how sharp the edges and corners can get. But then once in a while I get strangely soft results like the ones provided in the sample above. Now I know the problem is with field curvature, which is a design compromise (not a lens defect that can be fixed).





Nov 24, 2010 at 01:53 AM
mMontag
Offline
• • • •
Upload & Sell: Off
p.1 #19 · p.1 #19 · Problems with Uneven Depth of Field


I had that lens for a while - for the dollar spent it's actually a pretty good zoom landscape lens - but I've noticed the same effects you're seeing. Have to agree with the above assessments that the zone "A" is nearer the edge and 100% crops can be a harsh judgment tool.

For landscape - try focus bracketing in relation to your anticipated hyper focal distance focal point. Find your strongest HF setting - then focus bracket - best of three wins. I found f13 to be the better aperture setting for landscape on the one I had. If a lens has field curvature - it is what it is - you just need to find it's strongest focal point in relation to the scene.



Nov 24, 2010 at 03:09 AM
gdanmitchell
Offline
• • • • •
Upload & Sell: On
p.1 #20 · p.1 #20 · Problems with Uneven Depth of Field


There are lots of things that can cause corner softness, but in the second example a couple slots before this comment there is little to no specific evidence of field curvature being the issue. To show that you would need a sample in which the level of optimum focus stayed the same but the distance from the camera to the optimally focused subjects was different. The second example cannot show that, given the closely spaced comparison points and the fact that the degradation is more associated with position in the frame than anything else. Your original example does not demonstrate it either, at least not convincingly, since all elements in the defocused area are somewhat out of focus.

I have seen a number of examples purporting to demonstrate corner softness in which the primary issue was simply that subjects in the corner were not at the same distance from the camera as those in the center. Landscape examples are particularly prone to this issue.

I still think you are just dealing with normal differences in resolution in this sort of lens as you move from the center towards the frame edges. I would not rule out a decentering or other problem that makes focus uneven, though you could by comparing performance on opposite sides of the frame at equal distances from the center.

Finally, as I regard this as being within normal lens performance boundaries for a 24mm lens when viewed at 100%, I have to point out that this "problem" would not likely be visible in even rather large prints

Dan



Nov 24, 2010 at 01:53 PM
1
       2       end




FM Forums | Canon Forum | Join Upload & Sell

1
       2       end
    
 

You are not logged in. Login or Register

Username   Password    Retrive password