Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

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

FM Forums | Canon Forum | Join Upload & Sell

1       2       3              6      
7
       8       9       10       end
  

Archive 2013 · Canon EF 24-70mm f/4L IS Resolution Tests!

  
 
RogerC11
Offline
• • • •
Upload & Sell: On
p.7 #1 · p.7 #1 · Canon EF 24-70mm f/4L IS Resolution Tests!


RobDickinson wrote:
I've only had a 25-105 a few weeks (loaner) and havnt used it much but I did use it on a sunset the other day and I got huge amounts of flare (not processed not published yet).

is this typical? Does the 24-70f4L have the same issue or not?

The flare issue is a known problem with the earlier copies of this lens. Brian from TDP has documented it on his review of the lens. The shape of the rear flange was changed in later copies to help resolve the issue. If you scroll down about halfway down the page you will see what he has to say about it.
http://the-digital-picture.com/Reviews/Canon-EF-24-105mm-f-4-L-IS-USM-Lens-Review.aspx



Jan 06, 2013 at 09:43 PM
RobDickinson
Offline
• • • •
Upload & Sell: Off
p.7 #2 · p.7 #2 · Canon EF 24-70mm f/4L IS Resolution Tests!


Hmm this is something else and a new lens, may just have been down to the shot. Will see if I can sort it out.


Jan 06, 2013 at 09:58 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.7 #3 · p.7 #3 · Canon EF 24-70mm f/4L IS Resolution Tests!


snapsy wrote:
That makes sense. I'd be interested to hear out the 24-70 II compares to the 24-105 at infinity focus once you've had some time with the 24-70.


I'm not Dan but I'ved used both and 24-70 II handily beats the 24-105 I tried for that sort of shooting, wide end, FF, edge to edge, distant focus, f/8 (or for that matter f/2.8, 70mm, distant object center frame). It also doesn't get the nasty PF behind branches and such against white clouds in such scenarios.



(I do find it interesting that he got the 24-70 II after all his railing about how nobody ever could possibly tell the 24-105L or old 24-70 2.8 results from anything else. He bought it just because who want two of the same lens You spend $2300 vs $800 just to not have to of the same Maybe he got it for 2.8 alone but then why not $1300-1500 for the old version or the tamron (which would give him 2.8 AND IS and probably do every last single thing better than the 24-105)?)


Edited on Jan 06, 2013 at 10:21 PM · View previous versions



Jan 06, 2013 at 10:15 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.7 #4 · p.7 #4 · Canon EF 24-70mm f/4L IS Resolution Tests!


gdanmitchell wrote:
I threw out that $200 number, without thinking. It actually is not accurate. My least expensive lens is actually my old 35mm f/2 which seems to be available for roughly $300. I have a soft spot for this lens and really enjoy using it when I want to lighten the equipment or when I want to optimize IQ at 35mm. I use it for several things - mostly street photography but also for occasional landscape photographs when I work with primes. Here is an example:

In a few other cases it is part of a small set that also includes
...Show more

I do notice that you happen to mention perhaps three of the sharpest low cost lenses in existence though and all with special capabilities .




Jan 06, 2013 at 10:17 PM
RCicala
Offline
• • • •
Upload & Sell: Off
p.7 #5 · p.7 #5 · Canon EF 24-70mm f/4L IS Resolution Tests!


gdanmitchell wrote:
It turns out that this concern is vastly over-rated, and that barrel distortion and other similar sorts of lens distortions can be corrected in post with typically no visible degradation of the image.
Dan


Well, this got me to wondering, so I decided to test it: http://www.lensrentals.com/blog/2013/01/you-can-correct-it-in-post-but

I guess it depends on what the definition of 'no visible' means. Resolution dropped to about 85% to 90% of original after correction, but that probably wouldn't be noticeable on a small print and certainly not an online jpg. But certainly in a large print the difference would be noticeable.



Jan 07, 2013 at 12:59 PM
gdanmitchell
Offline
• • • • • • •
Upload & Sell: Off
p.7 #6 · p.7 #6 · Canon EF 24-70mm f/4L IS Resolution Tests!


RCicala wrote:
Well, this got me to wondering, so I decided to test it: http://www.lensrentals.com/blog/2013/01/you-can-correct-it-in-post-but

I guess it depends on what the definition of 'no visible' means. Resolution dropped to about 85% to 90% of original after correction, but that probably wouldn't be noticeable on a small print and certainly not an online jpg. But certainly in a large print the difference would be noticeable.


Did you have a chance to see my examples of the difference in crops from images that would be about five feet wide? There are links in my earlier post.

I regularly print pretty large from files that I have adjusted this way, and I'm completely confident that any deterioration of the image from typical adjustments to things like vignetting, barrel/pincushion (and related) distortion, and CA are not visible in very large prints.

As I wrote before, based on the logic of the claims that there is a degradation I used to assume that these sorts of corrections would create a problem. But testing (e.g. printing) has convinced me that this isn't the case in any but the most extreme adjustments.

I would not make the silly claim that there is "no degradation" of the image, but I am confident that this degradation is so tiny as to escape notice in prints, even with rather careful inspection. In many cases you cannot eve see it if you look for it in 100% crops, and in the cases where you can find it there, a) you have to look really closely and compare side-by-side examples, and b) it still won't be visible in prints, much less on the web.

Take care,

Dan



Jan 07, 2013 at 03:10 PM
RCicala
Offline
• • • •
Upload & Sell: Off
p.7 #7 · p.7 #7 · Canon EF 24-70mm f/4L IS Resolution Tests!


Not really disagreeing with your points about images, Dan. It's only a 15% difference in resolution even at MTF50 and a dozen things other than laboratory resolution tests are more important to any image. I'm like you, I tend to love my inexpensive primes (the 40 f/2.8 lives on my camera, at least at the moment).

I got interested because I see circular logic applied in discussing a number of lenses (not particularly this thread, but all over) -- saying X and Y have equal resolution and I'll fix the one with distortion in post. That's perfectly fine, and perfectly correct to do. But in that case, the person shouldn't be making an argument that the resolution is equal (or better, or whatever), because it's changed.

Of course, what I still have to do is check to see if correcting 1% distortion has a lesser effect than correcting 4%. It may be the change is similar no matter the % distortion corrected, in which case we're back to square 1.



Jan 07, 2013 at 04:15 PM
Fred Miranda
Offline
Admin
Upload & Sell: On
p.7 #8 · p.7 #8 · Canon EF 24-70mm f/4L IS Resolution Tests!


RCicala wrote:
Well, this got me to wondering, so I decided to test it: http://www.lensrentals.com/blog/2013/01/you-can-correct-it-in-post-but

I guess it depends on what the definition of 'no visible' means. Resolution dropped to about 85% to 90% of original after correction, but that probably wouldn't be noticeable on a small print and certainly not an online jpg. But certainly in a large print the difference would be noticeable.


Many photographers, including myself, have seen the softness introduced by high barrel correction in post. I am glad you run some tests and wrote this article Roger.
You put your money where your mouth is.
Thanks,
Fred



Jan 07, 2013 at 04:16 PM
jcolwell
Offline
• • • • • • •
Upload & Sell: On
p.7 #9 · p.7 #9 · Canon EF 24-70mm f/4L IS Resolution Tests!


Thanks Roger.


Jan 07, 2013 at 04:26 PM
RobDickinson
Offline
• • • •
Upload & Sell: Off
p.7 #10 · p.7 #10 · Canon EF 24-70mm f/4L IS Resolution Tests!


Many of us on here are on a quest for as much 'image quality' overall as we can get. Often it costs a fair bit, in money time and effort.

Nice to see a good empirical test of distortion correcting on sharpness/detail, its about what I expected but at least we know now the numbers.



Jan 07, 2013 at 04:31 PM
gfiksel
Offline
• • • •
Upload & Sell: Off
p.7 #11 · p.7 #11 · Canon EF 24-70mm f/4L IS Resolution Tests!


Roger, excellent points and tests. Come to really appreciate all the work that you've done. Learned a lot!


Jan 07, 2013 at 04:33 PM
RCicala
Offline
• • • •
Upload & Sell: Off
p.7 #12 · p.7 #12 · Canon EF 24-70mm f/4L IS Resolution Tests!


I was honestly a bit surprised: I thought the corners would lose, but didn't expect similar loss in the center. That's what got me wondering if any amount of correction is going to cause similar problems.

I don't correct distortion all that often, I tend to shoot with lenses that don't have too much. What's eating my lunch right now is the 132,413 images that I corrected a tilted horizon in . . . Maybe I should be turning on that artificial horizon feature a bit more often.



Jan 07, 2013 at 04:36 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.7 #13 · p.7 #13 · Canon EF 24-70mm f/4L IS Resolution Tests!


RCicala wrote:
Not really disagreeing with your points about images, Dan. It's only a 15% difference in resolution even at MTF50 and a dozen things other than laboratory resolution tests are more important to any image. I'm like you, I tend to love my inexpensive primes (the 40 f/2.8 lives on my camera, at least at the moment).

I got interested because I see circular logic applied in discussing a number of lenses (not particularly this thread, but all over) -- saying X and Y have equal resolution and I'll fix the one with distortion in post. That's perfectly fine, and perfectly correct
...Show more

it has to depend upon degree of correction



Jan 07, 2013 at 04:41 PM
RobDickinson
Offline
• • • •
Upload & Sell: Off
p.7 #14 · p.7 #14 · Canon EF 24-70mm f/4L IS Resolution Tests!


RCicala wrote:
I was honestly a bit surprised: I thought the corners would lose, but didn't expect similar loss in the center. That's what got me wondering if any amount of correction is going to cause similar problems.

I don't correct distortion all that often, I tend to shoot with lenses that don't have too much. What's eating my lunch right now is the 132,413 images that I corrected a tilted horizon in . . . Maybe I should be turning on that artificial horizon feature a bit more often.



I wonder of DPP with DLO would work better?

As for horizons, I'm known to always get them wrong so use the built in level especially in portrait mode if available. Get it wrong a degree or two in landscape isnt too bad but in portrait you end up loosing a lot of frame.

BTW what effective FOV is the 24-105 post correction? 26-27mm?



Jan 07, 2013 at 04:47 PM
Fred Miranda
Offline
Admin
Upload & Sell: On
p.7 #15 · p.7 #15 · Canon EF 24-70mm f/4L IS Resolution Tests!


RCicala wrote:
I was honestly a bit surprised: I thought the corners would lose, but didn't expect similar loss in the center.

I don't correct distortion all that often, I tend to shoot with lenses that don't have too much. What's eating my lunch right now is the 132,413 images that I corrected a tilted horizon in . . . Maybe I should be turning on that artificial horizon feature a bit more often.


I would guess, the heavier that correction the bigger the drop in resolution...I'm also surprised to see that the MTF50 res. loss is pretty much even throughout the frame.

That's what got me wondering if any amount of correction is going to cause similar problems.

Do you want me to change your username to "LensGuruGod1" so you don't have to test this out?
Fred



Jan 07, 2013 at 04:51 PM
alundeb
Online
• • • • •
Upload & Sell: Off
p.7 #16 · p.7 #16 · Canon EF 24-70mm f/4L IS Resolution Tests!


Fred Miranda wrote:
I would guess, the heavier that correction the bigger the drop in resolution...I'm also surprised to see that the MTF50 res. loss is pretty much even throughout the frame.
Do you want me to change your username to "LensGuruGod1" so you don't have to test this out?
Fred


As I said some days ago, the main source of the loss in resolution from geometric correction, is the resampling of all pixels that takes place where some correction is applied. The different amount of stretching (~0% - ~5%), that one would think would give a proportional loss of resolution, is smaller than the 10-15 % Roger measured. Try to interpolate an image just 1% bicubic and you will see what I mean.

And I agree with the circular reasoning argument. Because of that, I have learned to appreciate lenses that don't need CA correction more than lenses with the highest resolution numbers. The loss of quality with CA correction is more severe than geometrical corrections.



Jan 07, 2013 at 04:59 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.7 #17 · p.7 #17 · Canon EF 24-70mm f/4L IS Resolution Tests!


Fred Miranda wrote:
I would guess, the heavier that correction the bigger the drop in resolution...I'm also surprised to see that the MTF50 res. loss is pretty much even throughout the frame.

Fred


maybe the algorithm tries to even it out a bit across the frame and save as much edges as possible

for barrel the center kind of gets bulbous and they probably try to compress that in upon itself while stretching the outer parts out a bit

i'd think it would vary upon the lens and exactly what sorts of distortion it produces and it might be less even for some



Jan 07, 2013 at 05:02 PM
Schlotkins
Offline
• • • •
Upload & Sell: Off
p.7 #18 · p.7 #18 · Canon EF 24-70mm f/4L IS Resolution Tests!


Well, I think this lens is pretty interesting. I have a 24-70 2.8 MK I and am looking to upgrade it. I mostly take that lens for travel and shoot stopped down so this new 24-70 looks pretty darn good, minus price of course. Resolution is approaching the new 24-70, lower distortion, than any other 24-70, 4 stop IS and a Macro mode that would work in a pinch. Plus it's 35% lighter and quite a bit smaller.

Chris



Jan 07, 2013 at 05:06 PM
skibum5
Offline
• • • • • •
Upload & Sell: Off
p.7 #19 · p.7 #19 · Canon EF 24-70mm f/4L IS Resolution Tests!


alundeb wrote:
As I said some days ago, the main source of the loss in resolution from geometric correction, is the resampling of all pixels that takes place where some correction is applied. The different amount of stretching (~0% - ~5%), that one would think would give a proportional loss of resolution, is smaller than the 10-15 % Roger measured. Try to interpolate an image just 1% bicubic and you will see what I mean.


good point, especially since we are dealing with fractions of pixels here


And I agree with the circular reasoning argument.


although some people are willing to live with certain degrees of distortion left uncorrected, so it's more complex than just universally account for full geometric correction


Because of that, I have learned to appreciate lenses that don't need CA correction more than lenses with the highest resolution numbers. The loss of quality with CA correction is more severe than geometrical corrections.


perhaps it has become a bit too much of CA doesn't matter at all these days and this should be kept in mind (you can't go overboard though and have to mind what parts of frame you are comparing)



Edited on Jan 07, 2013 at 06:39 PM · View previous versions



Jan 07, 2013 at 05:07 PM
macrobild
Offline
• •
Upload & Sell: Off
p.7 #20 · p.7 #20 · Canon EF 24-70mm f/4L IS Resolution Tests!



The unevenness in 70mm is too large ( in my opinion) and will the lens be like the old 24-70/2,8 , a lottery to get one good lens?
I tested 4 ex of the 24-70/2,8 before I got one I could accept .

Thank you RCicala for taking the time and to report your findings.



Jan 07, 2013 at 05:34 PM
1       2       3              6      
7
       8       9       10       end




FM Forums | Canon Forum | Join Upload & Sell

1       2       3              6      
7
       8       9       10       end
    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.