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

  

Archive 2012 · L lens copy variation oddities (very odd stuff)
  
 
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.1 #1 · p.1 #1 · L lens copy variation oddities (very odd stuff)


So I was trying multiple copies of some lenses and here are some odd results in one case:

At 70mm, lens copy 1 is sharper than copy 2 dead center frame, especially at f/2.8 wide open. Copy 1 also hits peak more quickly, since it has f/4 sharper than f/6.3 but copy 2 has f/6.3 sharper than f/4 (copy 1 has f/6.3 a bare trace sharper than f/6.3 on copy 2 so it's not that copy 2 is sharper at f/6.3 than copy 1). OK not SO weird, copy variation can exist, even in L lenses. If I focus on the edges copy 1 is also sharper, for the most part (at f/6.3, on one side only, copy 2, is a TRACE sharper than copy 1). Again, OK, so copy variation exists.

But here is the weird stuff, at 24mm I looked into how the rest of the scene looked too and not just as the focusing point.

At 24mm, copy 1 seems to be just a trace sharper than copy 2 in the center, both focused on object in the center. The object was a tree about mid-way deep into the scene, perhaps (best of a few live view focusing attempts was used). Now oddly, if I pan around the frame, other trees at various depths appear sharper with copy 2 than on copy 1, especially at f/2.8 where it is a VERY noticeable difference, at f/4 it is fairly noticeable, at f/8 a little bit noticeable. So copy 2 is sharper at the edges and corners right?

But wait, if I go and focus on an object at the edge or in a corner then I get sharper results on that object on the boundary with copy 1 than I do with copy 2! So now it seems like copy 1 is better at the edges and corners!

So it's like the two lenses are somehow placing the DOF differently or have different FC or something That might not be weird if it was two different lens designs but it is merely two different copies of the SAME lens!

What is going on

Copy 1 is basically as sharp or sharper when focused anywhere in the frame it seems at that object that is focused on, at 70mm center frame at f/2.8 it even beats copy 2 at f/4 (by f/6.3 they are very close). And the lower left corner at 24mm f/4 it is definitely sharper than copy 2 and a little bit so upper left corner 24mm f/4 and yet if I focus both on some tree mid-way into the scene in the center then copy 2 is almost as sharp on the tree that is focused on and the other trees, especially closer to edges, and super especially upper left quadrant, are a bit sharper with copy 2 than copy 1 at f/8 and very much so at f/2.8!

It's like copy 1 is a sharpness monster, whatever is exactly focused on is MAD sharp (well maybe not quite for far edges at 70mm, this lens model is OK there but nothing amazing) and yet has some sort of different and maybe odder field curvature (or maybe it has the normal FC) and yet can, for some scenarios at least, give real world results noticeably less sharp over the entire frame as a whole, depending upon where focus depth was place and where objects are in the scene. My vague impression is it might hold for more scenes than not, but I'm not 100% sure, especially if one gets into trying to find the ideal focusing placement for any given scene.

An interesting test would be to find the ideal depth to focus at for a given scene for each lens and then compare the two results and see if copy 1 doesn't manage to do better then than copy 2. I also wonder if on avg half scenes would fit copy 1 more easily and half copy 2 or if, in real world, the way the DOF works would make one fit natural scenes better much more often and if that would matter more than one copy being less sharp overall.

It seems to be easily to quickly get landscapes with extreme bottom corners, especially bottom left, 100% crisp with copy 1 but easier to get more distant stuff and especially the upper left and upper part to fit DOF with copy 2 perhaps

I will post samples later, but the differences in where DOF seems to be placed appear to be significant. I mean really easily noticeably different parts of the scene are in focus with one vs the other when both are focused on the same object. I know that can vary a little and sometimes a lot between different designs, but maybe it's actually more of a copy to copy thing than even a lens design thing

It makes it really hard to even figure out which lens copy is better in a practical sense. Maybe it would depend scene to scene. If I just focused on each part of a test chart with each and looked at the spot focused on, I'd pick copy 1 for sure and yet a quick scan over some real world scenes, focus placed on some central object almost makes copy 2 appear to be better, maybe even noticeably so once getting wider than f/8, at least on the wider to mid-range range. Then again on one shot copy 1 is crisp to depths on the near landscape corners and copy 2 is not.



Oct 09, 2012 at 07:51 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.1 #2 · p.1 #2 · L lens copy variation oddities (very odd stuff)


any comments as to what the deal is?



Oct 09, 2012 at 08:02 PM
RCicala
Offline
• • • •
Upload & Sell: On
p.1 #3 · p.1 #3 · L lens copy variation oddities (very odd stuff)


I may be oversimplifying (or just plain confused) but am I understanding correctly that Copy 1 had higher resolution on target, but copy 2 has a greater depth of field? I know there are more specifics but is that most of the difference?




Oct 09, 2012 at 10:00 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.1 #4 · p.1 #4 · L lens copy variation oddities (very odd stuff)


RCicala wrote:
I may be oversimplifying (or just plain confused) but am I understanding correctly that Copy 1 had higher resolution on target, but copy 2 has a greater depth of field? I know there are more specifics but is that most of the difference?



at 24mm:

copy 1 hits more areas of the frame with higher resolution when focusing exactly on the spot in the scene to be examined, i.e. want to examine top left, then focus on a branch up there, want to look at right side, focus on something at far right edge, etc. (At 24mm copy 1 was a bit sharper upper left, more noticeably sharper lower left, a trace sharper center, *less sharp lower right*, a trace sharper top right, a tiny bit sharper top, basically at least a little sharper everywhere other than the lower right.)

copy 2, in that scene, when both were focused on a tree dead centered in the frame and also about mid-way deep into the scene's DOF, showed that tree just a trace less crisp, but showed all the stuff in the corners or even remotely near any edges sharper (despite when strictly focusing on most edges and corners it tested a trace to somewhat more noticeably less sharp); at f/2.8 the difference was night and day copy 2 showed stuff all over anywhere remotely near the edges much sharper, f/4 the same bit less of a difference, f/8 the same but quite a small difference and maybe not best at every single spot now

So, especially when not stopped down to f/8 copy 2 seemed to give an overall sharper corner to corner image, but if I had a scene where I cared about one particular spot and not getting tons within a limited DOF then copy 1 was generally sharper so it's like the two copies were placing the DOF differently or something.




Oct 09, 2012 at 10:10 PM
abqnmusa
Offline
• • • •
Upload & Sell: Off
p.1 #5 · p.1 #5 · L lens copy variation oddities (very odd stuff)


feature of zooms ?


Oct 09, 2012 at 10:12 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.1 #6 · p.1 #6 · L lens copy variation oddities (very odd stuff)


RCicala wrote:
I may be oversimplifying (or just plain confused) but am I understanding correctly that Copy 1 had higher resolution on target, but copy 2 has a greater depth of field? I know there are more specifics but is that most of the difference?




at 70mm, i only looked at places exactly where the two were focused and copy 1 was better left edge, center and right edge (didn't check along top or bottom or corners)

at 70mm in the center:
for copy 1, f/4 was sharper than f/6.3 was sharper than f/2.8
for copy 2, f/6.3 was sharper than f/4 was sharper than f/2.8

copy 1 was noticeably sharper at f/2.8 in the center, probably sharper than copy 2 had reached even by f/4
they were close at f/6.3 but copy 1 was still just barely sharper



Oct 09, 2012 at 10:12 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.1 #7 · p.1 #7 · L lens copy variation oddities (very odd stuff)


RCicala wrote:
I may be oversimplifying (or just plain confused) but am I understanding correctly that Copy 1 had higher resolution on target, but copy 2 has a greater depth of field? I know there are more specifics but is that most of the difference?



I just did a 24mm test, one at f/4 and one at f/8, I focused at all sorts of spots and different depths into the scene and then tried to pick the sample from each copy that showed the best sharpness over the largest area of the frame.

The scene was very complex with parts near and parts far and all sorts of layered depths all over the frame.

In this particular case, the best overall compromise at f4 was better from copy 2 it was: bit better top left and left edge and bottom and bottom right and right edge and top right and top and deep center and close center and mid-right center but the same mid-left center and worse at the lower left than best overall sample shot from copy 2.

So trying to fit this complex scene into f/4, copy 2 seemed to give a sharper overall result, definitely noticeably so, despite largely testing less sharp at any given spot.

I'm sure there would be other scenes with DOF and subjects set differently where the results might flip. I will try a few more scene types. Some types might be shot more often than others.

At f/8 it became much closer between the two copy 2 still won slightly more areas of the frame but it was often close or tied, copy 1 did do better along the left side though, copy 2 did better lower right corner and a bit so at the right edge, the rest were copy 2 barely, barely winning or tied.

My impression was there were more spots within the scene that copy 2 could be focused that would lead to a larger part of the scene being within nice DOF.


Copy 2 sounds sort of better going by that and yet at 24mm one might often do landscapes and be at f/8 where copy can be made to perform pretty closely (although with more care needed and if you need to do some lower light hand-held stuff and try to cram a scene into f/3.5-f/5.6 it seems like copy 2 would make it a lot easier and produce better results, at least for scenes similar to this one) and at 70mm copy 1 is better, certainly f/2.8-f/4 anywhere near the center and 70mm might be used a lot for people and such where you might prefer ideal wide open or near wide open performance, for landscapes at f/8 and 70mm perhaps they become close.



Edited on Oct 09, 2012 at 10:39 PM · View previous versions



Oct 09, 2012 at 10:24 PM
John_T
Offline
• • • •
Upload & Sell: On
p.1 #8 · p.1 #8 · L lens copy variation oddities (very odd stuff)


With all due respect, I'm getting a headache. How about trying a bit of a picture is worth 1,000 words?

Just red circle the areas/differences you are referring too.



Oct 09, 2012 at 10:27 PM
Wahoowa
Offline
• • •
Upload & Sell: On
p.1 #9 · p.1 #9 · L lens copy variation oddities (very odd stuff)


I assume you do microadjustment for both lenses?


Oct 09, 2012 at 10:35 PM
 

Search in Used Dept. 



skibum5
Offline
• • • • • •
Upload & Sell: Off
p.1 #10 · p.1 #10 · L lens copy variation oddities (very odd stuff)


I wonder if the TDP focuses center frame only and then measures everything and the edges far where they may and what I'm saying doesn't fit into what he said that some copies that are sharp wide open at 70mm seem softer at wide angle edges and vice-versa....

In the end even the copy worse at 70mm is still solid center frame though, not a match for the 70-200 2.8 IS II but a match or near match and even a win compared to the various other 70-200/70-300 so it is still hardly bad.

And even the one with the seemingly trickier DOF appears to be able to made to come reasonably close to a 24mm prime (perhaps not if you got below f/4 or even f/5.6 though for many scenes other than ones that fit it perfectly). By f/8 it actually seems able to produce a slightly better overall result with care taken. At f/4, with this scene at least, it is not able to be as sharp edge to edge, the other copy clearly wins.

If there is a copy that has the spot by spot measured sharpness, especially at 70mm, and the easier DOF placement of copy 2 that would be a truly monster lens.

I will try a 24 1.4 II tomorrow and see how it behaves. And a few more scene types to see if the results hold for most scenes or not (that copy 2 is easier to get a larger portion of a scene in top crispness at the same time) also trying some flat field scenes in addition to the complex ones.

It is pretty shocking how when choosing certain particular spots to focus on within a scene that the two copies can render other parts of the scene so noticeably differently seemingly placing the DOF in very different locations. I figured that wouldn't vary much copy to copy and only lens design to lens design, at least unless something appeared to be horribly wrong with one copy which doesn't seem to be the case.





Oct 09, 2012 at 10:35 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.1 #11 · p.1 #11 · L lens copy variation oddities (very odd stuff)


Wahoowa wrote:
I assume you do microadjustment for both lenses?


doesn't matter, not using AF



Oct 09, 2012 at 10:36 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.1 #12 · p.1 #12 · L lens copy variation oddities (very odd stuff)


John_T wrote:
With all due respect, I'm getting a headache. How about trying a bit of a picture is worth 1,000 words?

Just red circle the areas/differences you are referring too.


I definitely hear you. Images would make it all clear much more quickly.

I will it takes some time to better edit, cut crop, upload, circle, post the images though. I will put them up some time over the next few days, perhaps a few more than that if colors start popping locally and I need to focus more on real shooting again and I don't get it done tonight.



Oct 09, 2012 at 10:37 PM
John_T
Offline
• • • •
Upload & Sell: On
p.1 #13 · p.1 #13 · L lens copy variation oddities (very odd stuff)


A slight lens element misalignment might cause what you are seeing. Maybe a mfg fault or it was dropped. Perhaps you need another copy or two to check against.


Oct 09, 2012 at 10:44 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.1 #14 · p.1 #14 · L lens copy variation oddities (very odd stuff)


John_T wrote:
A slight lens element misalignment might cause what you are seeing. Maybe a mfg fault or it was dropped. Perhaps you need another copy or two to check against.


I wonder which is the more 100% aligned, the one sharper at 70mm f/2.8 or the one that seemingly makes it easier to fit a complex scene into a more wide open aperture....

Anyway here is what I mean:
24mm f/2.8 focused on a tree in the center of the frame at about half-way into the scene's DOF perhaps.

200% crops so avoid jpg compression messing with detail and to simply make it all easy to verify focus

Copy 1 the tree focused on:
http://farm9.staticflickr.com/8169/8072365901_81ceecc4f5_o.jpg
Copy 2 the tree focused on:
http://farm9.staticflickr.com/8030/8072366285_713a1a5617_o.jpg

(so you can see both frames were focused on the trunk, actually picked best of half a dozen focusing attempts on the tree trunk, focused 10xs liveview; copy 2 is just a trace better for micro-contrast)

So focus was placed at the exact same spot for both shots and yet looking at a 100% corner crop form both you see:

Copy 1:
http://farm9.staticflickr.com/8316/8072366167_b7700753c4_o.jpg

Copy 2:
http://farm9.staticflickr.com/8042/8072360518_dfb515b995_o.jpg

So despite copy 2 being a hair crisper in the center copy 1 has much better corner performance....

or does it.... because if I specifically focus at something in the corner instead, 200% crops:

Copy 1:
http://farm9.staticflickr.com/8318/8072361140_6f593920c4_o.jpg
Copy 2:
http://farm9.staticflickr.com/8309/8072366619_76b4c9cf6f_o.jpg

and now copy 1 seems to have worse corner performance.

So it seems like the two copies must have different field curvature or something and must be placing things other than the subject that is focused on at different places within the DOF?

So you can have one copy looking better in most corners if you focus directly at the object in the corner and yet have worse corners if you focus somewhere else in the scene so it seems, as I said, that copy to copy variation can alter the field curvature characteristics.

It will be interesting to see which field curvature the 24 1.4 II tracks, more like copy 1 or more like copy 2.

Anyway enough of this nonsense for now, back to editing proper photos and other doing other things.



Oct 09, 2012 at 11:25 PM
RCicala
Offline
• • • •
Upload & Sell: On
p.1 #15 · p.1 #15 · L lens copy variation oddities (very odd stuff)


OK, now I get it. I think Copy 2 is slightly decentered, probably a slight tilt adjustment needed. If you focus on the corner with that lens, it then puts the rest of the lens out of focus a bit.

If I'm following everything accurately, which I'm very possibly not.

You wouldn't happen to have a Seimen's star chart handy (only a geek like me would even ask)? That might demonstrate the decentered lens for you. I'll bet you'd see a slight flare towards 2 o'clock or 7 o'clock.

http://www.lensrentals.com/blog/2012/05/testing-for-a-decentered-lens-an-old-technique-gets-a-makeover



Oct 09, 2012 at 11:41 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.1 #16 · p.1 #16 · L lens copy variation oddities (very odd stuff)


RCicala wrote:
OK, now I get it. I think Copy 2 is slightly decentered, probably a slight tilt adjustment needed. If you focus on the corner with that lens, it then puts the rest of the lens out of focus a bit.

If I'm following everything accurately, which I'm very possibly not.

You wouldn't happen to have a Seimen's star chart handy (only a geek like me would even ask)? That might demonstrate the decentered lens for you. I'll bet you'd see a slight flare towards 2 o'clock or 7 o'clock.

http://www.lensrentals.com/blog/2012/05/testing-for-a-decentered-lens-an-old-technique-gets-a-makeover


Not sure if I have that chart, maybe one of my charts has that as part of it.

In the above test I had actually focused both in the center and then noticed that trees and stuff away from the main target seemed to fall into different parts of the DOF between the two copies, in this case, the branch in the upper right fell more out of the DOF with copy 2 than with copy 1 (it's not that copy 2 is not capable of being sharp there, since focused up there it does as well as the other copy certainly, so it seems that either copy 2 is warping it out of the DOF that it should be within or copy 1 is warping it into the DOF when it should be out of it).

It's seems perhaps like the left side of copy 2 has the DOF closer to the front, especially the upper left quadrant, and on the right side it has the DOF farther back relative to the other copy (the scenes are so complex that I'm not 100% sure which one has it correct.... if either ). I need to check it out more.

Someone else here has been posting that they think they might have DOF closer on the right side (opposite of my copy 2).

Whatever the case, the two do seem to place the DOF differently within a scene.

If it was tilted, I wonder if they'd muck it up or if fixing the tilt would do something and make the superior center f/2.8 performance it has go away or, assuming copy 1 had the DOF arrangement correct, that copy 1 would be turned into an uber copy?


Edited on Oct 10, 2012 at 01:27 AM · View previous versions



Oct 10, 2012 at 01:17 AM
snapsy
Online
• • • • •
Upload & Sell: On
p.1 #17 · p.1 #17 · L lens copy variation oddities (very odd stuff)


I would verify that the lens apertures are calibrated. If not then one lens will have more DOF than the other. Did you notice any exposure differences?


Oct 10, 2012 at 01:26 AM
bbasiaga
Offline
• • •
Upload & Sell: Off
p.1 #18 · p.1 #18 · L lens copy variation oddities (very odd stuff)


Was the wind rustling the trees in one more than the others?

-Brian



Oct 10, 2012 at 02:43 AM





FM Forums | Canon Forum | Join Upload & Sell

    
 

You are not logged in. Login or Register

Username   Password    Reset password