Vignetting problem on EF 24-105/f4L IS

I posted test results for this lens on here about two weeks ago from the French photomag Chasseur d'Images. They said while the lens is exceptionally sharp at all apetures and focal lenghts, vignettage and distortion were majot issues, especially at the short end wide open. You got exactly what they found out on their high-tech test bench. Distortion is almost 1.5% barrel, what you'd expect from a kit lens and it shows up in your photo just like the dark corners show up. In this kind of situation, bump up the ISO and close down the lens by two or three stops each. Of course that won't help the distortion but at least the corners will be clean.

Hi Julius

Did you have the lens hood and any filters mounted? If so, is the vigenetting still as severe with them removed?

Thanks Joe for the advice. No I have not read that French report. And Cliff, I did not use any filters nor any lens hoods. I double checked my result by shooting at floor tiles but the result was just as disappointing.

However, when I mounted my EF16-35/f2.8L to shoot at 16mm, the vignetting and distortion are similar. So I guess this is a typical Canon problem at the wide end which has escaped my attention due to the fact that I have been shooting primarily on the 10D/20D APS-C sized reduced frame. The edges of my "L" lenses were rarely used. Hence I had never noticed this problem.

I have heard in the past that Canon's wide angle lenses leave much to be desired. Now I know why. I should perhaps heed my friend's advice of buying the Contax 21mm lens instead.

Hmm, that's bad on such an expensive lens...

I am attaching another shot of the same floor tiles for comparison.

Maybe Canon wishes to encourage people to use 1.6X and 1.3X bodies?

:-(

Happy shooting, Yakim.

I can hardly agree, Yakim. With the arrival of the EOS 5D, FF is Canon's advantage over all competition. As such I just cannot understand the vignetting problem I discovered by chance.

Since my couriosity has been aroused, I tried a couple of more EF lenses on my 5D. Sure enough the same vignetting of corners appeared on the EF50/f1.4 standard lens and the legendary EF135/f2L portrait lens.

My observation has been confirmed by a friend who owns the 1Ds2. Funny none of us noticed the vignetting problem on "L" lenses back in the dark ages of film. His explanation is that perhaps it has something to do with how digital sensors collect light. Thus the problem has become more pronounced after we switched to digital camera bodies.

Perhaps the great Bob Atkins has a more thorough scientific explanation.

and is this only the 5D sensor, or is it the same story with 1Ds and even tele lenses vignetting?

Most lenses vignette wide open. Perhaps it is just visible on 5D more beause people use more these new contrasty and saturated picture styles that amplify vignetting. Take 2 shots, one in "neutral" mode and another in "landscape" mode, and I'm sure "landscape" shows more vignetting. It is the same on film cameras with different films.

Also, your sample shots are underexposed, that also makes vignetting more visible.

You are bound to get light fall-off at 16mm, but at 135mm - I would be very concerned. There's something wrong here. I find it hard to believe that this normal for the 5D! I think I would have Canon ckeck out that 5D if it were mine.

I think I've read a review somewhere that says that this lens is an exceptionally good performing 35-70 f/4 - go outside those parameters, and you face compromises, especially at the wide end as you have found. I do think it would be interesting if someone managed to test the vignetting issue on FILM, because this would give some indication as to the extent that the problem is lens or sensor optics induced. I would not be at all surprised if some noticeable differences occurred.

Roison, all 4 shots that I have uploaded on this thread was shot using my new 5D. I don't have an 1Ds2. Only my friend has it. He is the one who actually alerted me that this vignetting problem is prevalent on FF DSLRs. He even sent me a shot of sky using his EF135/f2L on his 1Ds2. I became more alarmed and managed to duplicate the same problem shot using my 5D on my EF135/f2L.

Ahto, I will try your suggestions tomorrow when the sun comes out. My shots were taken using evaluative metering. I added +1 exposure compensation for the floor tiles and adjusted the levels on PhotoShop, taking away only that part of the histogram where there is no data. I did not want to blow out the highlights. My tests were not meant to make Canon gear look bad. Heck I ONLY own Canon SLR gear. I just wanted to share my discovery that even "L" lenses show excessive vignetting when shot using a 5D. Perhaps this problem existed previously in the film era but I have not noticed it. I am not about to retest them using film.

Cliff, assuming my 5D is not functioning properly (which I highly doubt), how do you explain my friend KY's vignetting problem shot using his 1Ds2 and EF135/f2L? You may view his test shot at http://www.ec-photo.com/341421-1/135+2L.jpg

Mark, I concur your statement. However, the issue has evolved to be more than just vignetting on the EF24-105/f4L IS lens. It is now on all EF lenses when shot through FF DSLR bodies. I agree someone should repeat these tests using film and compare the results thus far with digital.

wow. that is just totally unacceptable from a so called L lens. thanks for the heads up.

"I think I've read a review somewhere that says that this lens is an exceptionally good performing 35-70 f/4"

I think I saw a pile of comments like this on the Fred Miranda forum.

My personal opinion is that the 4.3x zoom was a stretch. . . .and it shows. For the cost of this lens. . .I would expect premium quality throughout the range. If you can't do that with a 4.3x zoom -> then make it a 2.9x zoom like the 24-70/2.8L. If I want prosumer performance. . .or a lens only usable in bright conditions at F8 . . .then I would have already bought a 28-135/IS.

I think most of us who were waiting for a 24-XX/4L expected top quality across the range, but were expecting a compromise in the form of a reduced zoom factor. (ie, a 24-70/4L).

I'm not sure the 24-105mm is to blame here as these effects were observed with a range of lenses

These shots are all underexposed, in ACR they would need at least a +1 stop of exposure, maybe +1.75. In that the vignetting is heavily exaggarated by the exposure. The floor tiles shot is probably close the the lens's closest ability to focus, at such close distances the distortion is also heavily accentuated.

I'm not making excuses for the lenses, the distortion isn't great on the top photo, but although I haven't got mine yet to test against the 24-70L, I wouldn't be surprised if the distortion is similar. The vignetting is annoying even if it is very easily fixed in ACR. I hope PTlens get some profiles for the 5D/24-105L soon and then I'll quit worrying about the lens and start hating how slow my computer runs while batching 300 files through PTLens!

This lens was never to be as good as the 24-70L in every way, I think that for it's main clientele (not landscape/portrait/architecture) it is a pretty good compromise through those focal lengths. For me it will be a stellar wedding lens and useful street/PJ lens.

Lest everyone starts to believe that all pictures made with this lens and camera combination at f/4 are unusable, have a look at different situations first - like the attached image.

Know the limitations of any machine you buy, before you buy.

I hope that anyone who's unhappy with the 24-105 actually takes it back to the store, not just think about it. That'll help drive the price down for fans like me. The word limitions starts with "L" too.

Another example...

here...

ALL - no exceptions - wide zooms and other primes too - have some degree of vignetting. Because your shots are greatly UNDEREXPOSED such problem is HUGELY (falsely) amplified. Get a CORRECT exposure and watch the vignetting almost disappear.

Also, on the floor samples, it seems the lens was *not* perpendicular to the floor as the amount of distortion is different on the top. Any wide lens would do that. If the lens had been perpendicular your feet or the tripod's would show (lest you reversed your column).

How would you then explain 135 mm lens vignetting?

It is interesting to note that even one of the sample images posted by Canon Japan shows significant vignetting issues (lens EF17-40mm f/4 L, at f/8)

http://web.canon.jp/Imaging/eos5d/html/eos5d_sample_4e.html

Right on Mike and Peter! Thanks for not blaming all observed vignetting effects on underexposure. Ben and Giampi, I have corrected those so called underexposures but am amazed to see vignetting exaggerated instead of "almost disappeared". Now kindly explain that phenomenon.

Exposure correction.

exposed correctly

exposed correctly again

First of all, I'm not an optics guy, but let's assume you have a small amount of center-edge light falloff present in any lens as a matter of physics. Take a picture of a entirely white screen. You will have a white center and very slightly darkened corners.

Now adjust levels in photoshop. There is no black object, so you pull the black point out until you can see the corners darken. You are exaggerating the contrast in the picture, so it develops really bad vignetting. You can probably generate this phenomenon on almost any lens given the right conditions...hence the problem appearing on the 50 and the 135. I'm pretty sure that in your pictures, you can increase or decrease the apparent vignetting by moving the black point.

I noticed both your pictures were fairly low contrast pictures of somewhat light objects. Perhaps if you repeat this experiment with different weather or different scenery the vignetting might be less apparent. I'm not saying the vignetting is not there, but digital concerns may compound optical problems.

wy

Since exposure is not your problem then it must be blamed on the lenses and the camera. Take them back to the store for a full refund. Canon wide angles an "L"s suck, especially their wide angles. Switch to Contax, Leica, Carl Zeiss, whatever.

Thanks Wayne. Your explanation makes sense. I shall reshoot these test pictures tomorrow in broad daylight.

Don't bash EOS here pal. Just stop the bloody lens down at stop or two and bump up ISO instead of shooting it wide open (as someone said before) where vignetting will show up at its worst. You're sounding like a troll dude! WTF?

Ben when you get the 5D can you test out your 24-70mmf2.8L on it and see if it does the same thing as the above shots do? If it does I think it is the FF sensor to blame and not the lenses. As Julius has tested it on a number of lenses, but I am interested in knowning about the 24-70mmf2.8 L as I might be getting it, But I am not sure yet as I wait for more reviews on the 24-105mmf4L IS. Thanks.

DK.

As a Canon Professional Services (CPS) member, I actually bought both the 5D and the EF24-105/f4L IS lens from Canon Hong Kong. While CPS membership gives me much faster service, refunds are NEVER customarily offered by any business establishment here.

I will try my best to get down to the root of this vignetting problem. But I am afraid that I am stuck with these two camera+lens. I hope this problem can be rectified. Otherwise it would be a big lesson for me after spending roughly US$4500.

And you know, I'm gonna spend that same $4500 and enjoy the results while realizing that even this great system has its limits. Cheers.

I shot slides for 15 years using Canon lenses. Every lens suffered from some light fall off around the corners when wide open. Wide zooms were the worse. Tele primes had the least. Most light fall off disappeared by stopping down 1 or 2 stops. Besides, slide mounts (or slight cropping for prints) covered most of the light fall off. It was mainly when I viewed unmounted chromes with lots of sky light fall off was evident.

Incidentally, I believe "light fall off" is the correct name for your problem. Vignetting occurs when something such as a filter or hood blocks the edges of the image.

Seems to me that some of the photo.net-members have been bitten by the so called "L-bug", and that the disease has affected their judgement:

"Most lenses vignette wide open."

"Your photos are underexposed."

"Don't bash EOS here pal. Just stop the bloody lens down a stop or two and bump up ISO instead of shooting it wide open."

Why would anybody pay 1250 $ for a lens that needs to be stopped down to f5,6 or f8 to avoid pitch-black corners? It just doesn't make sense.

Julius,

Just for fun, shoot a roll of ISO 50 or 100 slide film with this lens and tell us what the results were concerning light fall off in the corners. My post on the test results was done on here about the 16th of September, if you want to go back and look. I compared the 24-105 L to a number of other trans-standard zooms, mostly Canon.

Distortion and light falloff clean up nicely in Photoshop CS2. You could even make an action to do it for you.

DI

To understand how sensor optics can cause vignetting, it is useful to refer to some concepts in optics. The exit pupil of a lens is the apparent position of the aperture diaphragm when viewed from behind the lens. You can think of it as being the apparent point through which all principal rays pass as the image is inverted by the lens for projection onto the sensor. A principal ray is the central ray of the cone of light that is focussed by the lens onto each point of the image. The apex angle of the focussed cone of light will depend on the aperture in use, and will be wider for faster apertures.

The distance of the exit pupil from the sensor/film plane defines a triangle whose base is largest across the diagonal of the image. The angle between the central lens axis and a side that extends from the centre of the exit pupil to the corner of the image is also the off-perpendicular angle at which the principal ray strikes the sensor at the corner of the image (by opposite angles, for those who remember their Euclidean geometry). The microlens layer which lies over the Bayer colour filter array on top of the actual sensor silicon is typically only capable of accepting light to focus onto the sensor photodiodes provided it lies in a cone typically with no more than a 25-30 degree apex angle - i.e., within 12-15 degrees of vertical before light falloff begins to be significant. Film, which lacks the microlens layer, has no such limitations.

Of course, all wide angle lenses are susceptible to various forms of vignetting - there is an excellent discussion of these that also illustrates some of the concepts discussed above here:

http://www.vanwalree.com/optics/vignetting.html

When using a fast lens wide open with a digital sensor, the cone of focussed light has a much wider apex than when using a narrow aperture. More of the rays in that cone will lie outside the acceptance angle of the microlenses at the corners of the image. Again, film will only be subject to the normal causes of vignetting discussed by van Walree.

A comparison with the performance on film will reveal the extent to which vignetting is caused by features of the lens design or by additional limitations imposed by sensor optics. Of course, it is possible to correct for vignetting (and drawing/barrel distortion) in software.

As a footnote, the high level of barrel distortion at 24mm with the 24-105 is a consequence of trying to design the lens with the exit pupil further from the film plane to help with vignetting, coupled with the 4+ zoom ratio of the lens. Figure 2 in this van Walree article should give a basic understanding as to why this is so:

http://www.vanwalree.com/optics/distortion.html

My judgement also led me to post examples of the lens WIDE OPEN in different situations showing little to no light fall-off/vignetting. So if I'm smitten by the L disease the effects are rather benign...<twitching left eye> I think </twitching left eye>.

The combo in question...

Wayne,

The light fall off, that Julius describes, happens at 24mm. It is not possible to see at which focal lenght your pictures were taken.

The leaves were taken at 105mm. This new picture of the dog statue (by WS Lam) was taken at 24mm and f/4. Still no fall-off.

Dog again...

Oh, ignore the dog...that was using a 20D. Disregard.

Here's an FM forum thread from a month ago when this lens first came out. The following are a couple of in-depth reviews therein reporting vignetting issues with this lens of FF in certain situations from the get-go:

Mark1958:

I got this lens yesterday and took it to a brick old building and did some comparisons against the Canon 24-70L, 35/1.4L, canon 70-200L IS, canon 70-300DO. I also did some comparisons against the tamron 28-75 but have not looked at those. I have a sharp 24-70 that I got from Mike Hatam about 6 months ago. All testing done on with 1Dsmk2, tripod, IS OFF, leveled. SInce i was shooting all canon glass, I used large jpg mode with default settings. I concentrated on 24,35,50,70, and 105mm focal lengths and shot at f4,5.6,8 and 11. Here is a summary of my findings. If i get a chance I can try to organize some examples to support my claim.

The strength of the new canon zoom is between 35-70mm. At these focal lengths, this lens did pretty well even wide open. The contrast was slightly less good compared to the 24-70L but not so different that I was concerned.There may have been a tad more CA with the new zoom but not significant.

At 24mm, there was significant vignetting. It was worse at f4 but there all the way through f8 when i stoped looking. THe canon 24-70 was sharper at the edges and corners at 24mm. THere was a significant difference here. THe center part of the images were close and any differences were not concerning.

At 35-50mm, I became impressed that the the 24-105 was sharper than the canon 24-70 at f4-8 especially at the edges. Infact, the zoom was almost as sharp as my canon 35mm/1.4 prime. THe prime had better contrast and still a bit better.

At 70mm, the new canon was much better than the 24-70. I always felt the 24-70 was weak on the tele end anyway. The corners and edges were substantially better with the new canon. In fact, the 24-105 at 70mm was better than my 70-300DO and close to the 70-200L IS but again the 70-200 is still the best zoom I have used in the tele end.

At 105mm is where things fell apart again. The 70-300DO was much better especially at the edges. THe center was not as sharp but the differences here were not as significant. THe 70-200 was the best especially a the corners.

So once again there are a bunch of tradeoffs, f4vs2.8, weight, IS etc etc. This is not a perfect lens as most zooms are not. There are clearly weaknesses and strengths. I think this lens would be great on the 1.6x crop sensor cameras and perhaps even the 1.3x crop cameras. It is not outstanding on the full frame except in the 35 to 70mm range. So i am not that estactic about this lens but will use it a bit before I feel like I can make a final decision on its value in the lens stable. Mark

Mike Hatam:

Great job Mark - I agree pretty much with everything you've written.

I just spent 2 days using the 24-105L as a "walk-around" lens on my 20D and 1DsII. I also did some carefully controlled tripod tests against the 16-35L and 70-200IS.

Keep in mind, I've only tried this one copy of the 24-105, so I don't want to make any summary conclusions yet (there is likely sample-to-sample variation).

Here's my tentative early impressions...

The 24-105L suffers from quite a bit of vignetting, when shot at f4 on the 1DsII. The vignetting is mostly gone by f8, and doesn't show up at all on the 20D.

The IS is very effective, and very silent. I had to keep checking to make sure it was "on", because I could barely hear it.

The image quality is very nice on the 20D, and I'm very satisfied to use it as a "walk- around" lens on that body. It's better than the 17-85IS that I was using previously, and it's also better than the Tamron 28-75 I was using (focus is more accurate).

On the 1DsII, it's not sufficient for any critical work. The corners get a bit soft, and the vignetting is an issue. There's also significant distortion, especially at the wide end. Stopped down to f8, the lens is pretty sharp - roughly as sharp as the 17-40L. It beat my 16-35L @ 35mm, but lost badly to my 70-200IS at 105mm. I guess no big surprises there.

I'll be keeping this lens, and using it as a walk-around. But I won't be using it for any critical work on the 1DsII. I think the 24-70L is better suited for critical work on the 1DsII. Mike

I think I am with Henrick: Why would anybody pay 1250 $ for a lens that needs to be stopped down to f5,6 or f8 to avoid pitch-black corners? It just doesn't make sense.

I really would like to see a review from either Riechman or Castleman on this lens, preferably against a 24-70/2.8L. Given the cost of this lens. . .I certainly can wait!

What? Vignetting on a FF DSLR? I thought every EOS user knew that such heresy was just the false propaganda of Nikon users.

I'm not a Nikon user. We are taking about lenses, not religion.

Lenses and sensors.

Yes, lenses and censors. The 24-105mm, 16-35mm and 135mm will probably be excellent wide open on my 1,6 crop camera. Just to bad, I can't afford them.

We all know that wide angle lenses tend to have light fall off, especially zooms wide open. From what I have seen so far, it seems that the 24-105 wide open displays severe ligth fall off, as first reported by CD'Images, and then in other forums.

Whether or not this is acceptable to you, and the results you expect to obtain, given the high price tag, only you can decide. To me, the light fall off and distorsion at 24mm is very bad for the price, it puts this lens on par with the 24-85 USM. I was thinking of buying this lens for my EOS 1V, as a travel companion to my 70-300 DO, but now I am not so sure. I shoot mostly landscape and travel, where the zoom range and IS would come in handy. But I would need much better performance at f/4.

When travelling, the weather is not always sunny, so good performance wide open is a must. Even with IS, with ISO 100 slide film, f/4 is of the essence. For landscapes, the story is different, as shooting at f/11 from a tripod gets rid of the problems described here. But then again, the 24-85 USM is also good under those settings.

I think Canon made too much of a compromise with this lens. I know that light fall off and distorsion can be corrected in post-processing, but there is something to be said about getting it right in the camera in the first place. And please, this is not a Canon specific problem, so all those comments on "see I told you, Canon sucks" are really not necessary and unhelpful.

Where do you find the "see I told you, Canon sucks"-remarks?

Here are a lot of test shots with the 24-105mm:

http://www.dphotojournal.com/canon-ef-24-105mm-f4l-is-usm/

One of the test shots is taken with the 5D at 24mm f4, and it doesn't show any light fall off:

http://hosting.photo.net/photodb/image-display?photo_id=3759354&size=lg

So maybe Julius problem has to do with his camera, since it also happens with the 135mm?

Closing the italics. Check.

That Cliffs of Dover picture is interesting because it was a sunny day. We'll see what original poster here gets under similar conditions, as he intends to do.

His exposure corrected pictures are still not correctly exposed, hence the still over exaggarated vignetting. Correcting exposure in PS and exposing properly in the first place are two very different things. That church steeple shot is a prime example.

>>His exposure corrected pictures are still not correctly exposed<<

Of course! Changing levels in PS doesn't remove the issues caused by wrong exposures. And all his pictures look "foggy"...

The shot taken with the 135 shows a very visible pattern of horizontal lines. What are those?

Those are the posts that start rumors.

Even though ligh fall-off is a "normal" issue with most wides, especially ultra-wide zooms, before making a statement with the word *problem* in it someone should do due diligence and perform some serious tests.

The picture posted by Julius are grossly underexposed, they appear "foggy" and are in wide gamut RGB. All of which are signs of someone just starting out in digital photography and/or potography in general.

While we all were beginners at some point, to start a rumor about a problem is of no use to anyone.

As for the "correct exposures" posts they are not! Adjusting levels is not the same as having a correct exposure. A *correctED* exposure is not making a *correct* expsure. In additon, the edits were not good either, thery were still underexposed.

Here's a corrected version. But, be that as it may, the lens may indeed show light fall-off at certain focal lengths. No one would argue with that but, to use these pictures as irrefutable proof of a "problem" (and it wouldn't be a problem either) is not a reasonable thing to do.

Level adjusted and COLOR space changed.

I am not surprised that this lens vignettes when wide open - particularly when set to wide-angle. As annoying as it is, the very best lenses vignette when wide open and I have some of the latest lenses from Leica (both R and M), Zeiss for Hasselblad and Zeiss for Contax - not to mention a few L lenses too. They all vignette obviously when wide open. Even my example of the legendary 35mm Summicron ASPH vignettes significantly when wide open.

It may be distasteful but it is a fact of photographic life. f8 usually cures it in most cases.

All the best,

James

>>Seems to me that some of the photo.net-members have been bitten by the so called "L-bug", and that the disease has affected their judgement<<

Ah! This is not about L lenses, it's about the poor samples provided as "proof" of the "problem" described herein and called out, for all to see, in the post HEADING.

If such a normal issue as light fall-off was occurring to an unacceptable degree to ANY given lens by ANY manufacturer proper respect should be paid by the methodology involved in making such judgement BEFORE starting rumors on the NET.

As it is, there is already a post elsewhere on this site claiming the Canon 5D (the camera, not the lens) has "vignetting problems".

Know what I mean?

That's how so many posts on this site have been talking about the "notorious backfocus issue of the 70-200L" many other non-sensical rumors.

Because the purpose of this forum is, among other things, to help people make the RIGHT decision about their equipemnt, flooding the place with unsubstanciated rumors won't help anyone.

True, if someone decides to believe anything they read I suppose they are at fault. But, again, since this is a photo community it does no good to ANYONE to start rumors based on a few, grossly underexposed shots posted in the wrong color space.

And if the original poster is a member of CPS I thought he would know about ligh fall-off, color space and proper exposure. Finally, if the thought he had a defective sample CPS is excellent about taking care of those things.

So, given all of the above, what is the REAL purpose of posting this here?

All I know is that I'm having a blast with the 24-105/4L IS on a 1.6x body. No complaints whatsoever (yet).

It is very easy to close this discussion. Just post some pictures taken with the 24-105mm on a full frame camera at 24mm f4 without severe light fall-off. Or with any of the other mentioned lenses for that sake.

Unless James is right and all lenses have severe light fall-off on full frame cameras, and therefore have to be stopped down to f8. If that is the case, I'm not so envious at ff-camera-owners.

>>It is very easy to close this discussion<<

That's just the point. There is NO discussion to close, because there is no discussion to begin with.

FACT: Light fall off is a fact of life of ANY lens, including LF lenses.

FACT: the samples by the original poster are grossly underexposed and in the wrong colorspace.

FACT: the poster made a claim, both in the post heading and the post, which he didn't properly back with data (properly executed tests).

So, if he wishes to start another thread with properly conducted tests about *his* sample of the lens he should choose a more appropriate heading. Such as "Problem with MY 24-105L"

Paolo, it is indeed not a Canon-specific problem. Except that Canon is the only manufacturer denies there is any problem with using lenses designed for 35 mm film on FF sensors. And apparently the users are in denial too, saying that this can't be true and it's a lens problem or user problem etc. but no way could it be that this is precisely the sort of thing that optical engineers for many camera and lens makers have been warning about and why they haven't been excited about making FF sensors. The technology to do it without compromises like this doesn't exist. You can still get more of a frame than if you had been using 1.6x sensor so in that sense there is no problem - you still get something for the money, in fact a lot.

>>Canon is the only manufacturer denies there is any problem with using lenses designed for 35 mm film on FF sensors<<

Ok, you tell us the % of light fall-off increase of FF sensor vs film.

I have just been browsing Pbase for photos taken with the (full frame) 1Ds Mark II. And although it is possible to find pictures with light fall off, it doesn't seem to be a real issue:

http://www.pbase.com/cameras/canon/eos_1ds_mark_ii

Giampi: your corrected photo still shows vignetting. You have lightened the corners, but also the centre of the image, which is now blown out.

Ikka: I think these discussions of brand X against brand Y are pointless, and as you say, it is not the issue here. However, I think that your statement saying that other manufacturers are not making FF 35mm chips because of the optical constraints is not correct. They are not making them becasue they do not have the technology in-house to do it, and because they are expensive to make. If you ask any photographer what his/her preference would be, I have no doubt that the answer would be FF. Who asked for cropped sensors in the first place? As they become cheaper to make, it will all go back to normal.

This is not saying that cropped sensors cannot deliver the goods, far from it, but they are a transitional phase of the development.

>>Giampi: your corrected photo still shows vignetting. You have lightened the corners, but also the centre of the image, which is now blown out. <<

My corrected photo is/was NOT meant to discuss vignetting. It was ONLY meant to show the gross underexposure of the original and the edited versions posted by the author.

I have already talked about "light fall-off" at length and unless someone can come up with the figures I have asked it's a moot point.

What is the % increase of light fall-off between FF sensor VS film. Simple question. All those who seems to know can't post the answer.

I am a newbie in digital photography (or simply photography), your response enlightened me a lot.

I won a lottery in Las Vegas, and bought a few Canon dslrs, 1.6x to 1.3x to FF, and a few Canon lenses from 15 to 500. Interestingly, they all showed dark corners! :o

My dear Professor, incidentally I also got this lens and I decided to take a test shot:

http://www.ec-photo.com/352182-1/Exp+_0.jpg

Oh! Dark corners! I have removed the lens filter, and use Av mode at f4, and yet the camera chooses to expose like this. What should I do? Should I send it to Canon?

You ingenious remark inspired me! I dial in +1 exposure compenstion and shoot again:

http://www.ec-photo.com/352184-1/Exp+_1.jpg

I saw some red things disappear in the upper right hand corner in photoshop, is this normal? Oh, the dark corners are still there! Help!

Ok, never give up, dial in +2 compensation:

http://www.ec-photo.com/352186-1/Exp+_2.jpg

I saw 2/3 of red things disappear this time, is this normal Professor?

Well since there is a +3 compensation available, though I don't know what is it for, just try it:

http://www.ec-photo.com/352188-1/Exp+_3.jpg

Oh Professor I finally make the darke corners become very inconspicuous, do I get a pass in your photography class?

BTW, Winnie looks much better with all red things removed:

http://www.ec-photo.com/352190-1/Exp_3_100_crop.jpg

I know nothing about colour management, just export the RAW in stupid RGB and upload to the web, am I correct in doing this?

If there is a Nobel prize in photography, let me be the first to nominate you Professor Giampi.

Oh btw I am the KY guy Julius referred to who alerted him that many lenses vignett on FF dslr, am I eligible to be a Nikonian now?

KY Lee... you have no information about the lenses or any other conditions you used to take pictures. it kind of nullifies any possible use of your samples.

rest of the forum... I have read and mildy experienced that wide angle lenses show vignetting when wide open. that explains more than half the issues brought up in this thread. another issue that may be not commonly known and I have only read about once, is that the reason FF DSLR's are more prone to vignetting than film will be, is because when light hits a photosite at an angle, the light is more likely to be reflected off of the microlens as the angle gets steeper. really good light recording occurs when the light strikes the microlens/photosite at +-20 degrees from normal.

http://www.imx.nl/photosite/comments/c001.html

this article has more indepth stuff.

anyway. the point is wide angle lenses will cause vignetting in a FF dDSLR more than you may be used to with film. longer telephoto lenses should be less of an issue but maybe there is something else. there has to be an explanation for why your 135mm lens does this. most of the sample shots I have seen have been fine.

Byron,

I suppose we are talking about 24-105 vignett on FF dslr right?

"My dear Professor, incidentally I also got this lens and I decided to take a test shot:"

"I have removed the lens filter, and use Av mode at f4"

I am not posting to convince anyone, just terribly disapoointed with the atmosphere of supposingly enjoyable gear discussion I found here.

I feel sorry that I didn't mention the shots are from a 1Ds2 rather than a 5D, if this nullify anything, let it be.

If you really want to know more shooting info from the test, in order to facilitate fruitful discussion, just let me know. Otherwise, please ignore my post.

Enjoy your gear.

KY

mostly I would point out that you didn't mention the focal length. In what I can gather from reading briefly over the previous posts. most of the people agree that wide angle shots will vignette. and a few have pointed out that with that lens it is a problem in general.

So the consensus is that all the vignetting in this forum is a result of either the lens you mention or the wide angle. the ONLY variant from this argument is a previous post that shows vignetting with a 135mm f2 lens shot at f/ something or other. that is the only shot that shows vignetting on a lens that really probably shouldn't have any. which makes it the only shot that shows that the FF DSLR might be at fault. which is kind of what I can gather that people are debating here... is it the lens or the camera?

really I haven't read all the posts. there too many. but I am curious to know any faults that can be found with any camera. especially if I am interested in getting one one day. the lens (24-105) I could really care less about since I have no plans to get one.

Byron,

It is always good heading back to objective discussion.

My tests shots are all done at 24mm end.

For 135 2L, read the first comment in this page: (link)

I am nobody here, don't take my tests too seriously, if you never think it is a problem (like me actually who treated this only as a phenomenon of optics and digital sensors), forget about this issue, or just fix it with software.

Light fall off at edge is more promiment, but not limited to wide angle lenses:

http://www.ec-photo.com/348448-2/400.jpg http://www.ec-photo.com/348303-2/500.JPG http://www.ec-photo.com/348300-2/150_001.jpg http://www.ec-photo.com/348297-2/85.jpg hhttp://www.ec-photo.com/348291-2/70-200_70.jpg ttp://www.ec-photo.com/348294-2/70-200_200.jpg

All wide open, Av priority. Forget about the WB and other irrelevant technicalities, I know my skill sucks.

Someone did perform a scientific test and posted in dpreview before showing the % of light loss when compared to film, making wide lenses not as fast as it claims on digital, say 85 1.2 performs like a 1.4. I lost the link, and please ask someone in that forum.

KY

>>professor Giampi<<

If you are beginner using sarcasm without understanding what is being disussed here, or any of my posts, is not going to help your "argument".

The question I asked is very simple, and it still goes un-answered.

>>Someone did perform a scientific test and posted in dpreview before showing the % of light loss when compared to film<<

Good, where is it? I have never heard of it, I haven't read about it here, magazines or anywhere on the NET. Yet, if such test was conducted it would be of great importance and interest to many. How was the test done, do you know? What was the answer, you seem to know...let us know what those numbers are.

>>making wide lenses not as fast as it claims on digital, say 85 1.2 performs like a 1.4.<<

What are you talking about? Are you claiming that that an f/1.2 lens will perform like an f/1.4 on a digital body? Never heard of that either. In fact, metering done with my Sekonic hand-held meter translated EXACTLY the same on my 10D as they did on my Bronica ETRs, Rolleiflex and EOS3. How's that possible?

>>I lost the link, and please ask someone in that forum.<<

AH! My dog ate my homework, ask him...yep!

Again, ligh fall-off is normal. The claim made by some here is that the ***5D*** causes/shows greater than normal ligh fall-off.

Now, with each claim must come proof and it hasn't yet.

***YOUR*** esteemed Professor.

I too was a bit surprised at the amount of light falloff on the 24-105 lens at the wide end. I tested mine all the way to the minimum f/22 and could still see significant falloff even with the smallest aperture at 24mm! That was surprising to me. I had to increase to 35mm to get to a point where it was nearly eliminated. It has nothing to do with underexposure or color spaces. It's pretty typical to get light falloff at the corners at the wide end of a zoom lens with the aperture wide open when using a full frame sensor. I think there are a lot of people used to the 1.6 form factor that will not realize this and will be surprised because, well, they've never seen the "corners" before. The only thing I found surprising is that this lens is nearly unusable at 24mm with anything less than f/8 aperture and even then you'll need to retouch them to get rid of the falloff in some shots. Such is not the case with the 24-70 f/2.8 L lens. It too shows falloff in the corners at f/2.8 and 24mm but you only have to stop it down to about f/5 to completely eliminate it at 24mm. So my surprise here is not that it happens on the 24-105 but that it is nearly impossible to get rid of it at ANY aperture at 24mm. That, to me, is a bit extreme.

Mike

I seldom responded, in any internet forum, like the way I did. Anyone who read the whole thread should know what is going on here.

FYI, I finally retreive the source of info I quote:

".....The sensors with the widest acceptance angles (like the Kodak KAF sensors used in Oly E-1 and Leica Modul-R) are down just over a stop at 20 degrees. The Sony ICX413 (Nikon D100 and D70, Pentax *ist, Minolta 7D) and the unnamed Canon 6mp in 10D and 300D are down over 1.5 stops. And by 26 degrees (f1.0) everybodie's down 2 stops or more.

I actually took the response of one particular sensor, the KAF-4101CE from the Oly E-1, and integrated the angular response, to get the following T-stop calculations. These are the nearest common f stops culled from the table which goes from 1 degree to 26 degrees in 1 degree increments.

fstop tstp loss 1.00 1.32 -0.78 1.21 1.44 -0.52 1.41 1.60 -0.35 1.80 1.92 -0.19 2.08 2.18 -0.13

So, if you put an expensive f1.0 lens on your DSLR, it performs like a lens 3/4 stop slower would, on film or a sensor with no microlenses...."

Anyone who can prove this wrong, but using words is not as convincing as using data.

Sorry for the table format shift:

fstop tstp loss

1.00 1.32 -0.78

1.21 1.44 -0.52

1.41 1.60 -0.35

1.80 1.92 -0.19

2.08 2.18 -0.13

>>.It has nothing to do with underexposure or color spaces<<

No one said it did.

more fact reminders:

1) Light fall-off exists with all lenses to some extent or another (and it's not called vignetting though). Vignetting is used to describe a light loss due to phsysical obstruction. Light fall-off is due to lens characteristic.

2) There has been no authoritative study that has proven that FF digital sensors are any different than film, as it has been contented here, and cause "light fall-off" *in addition* to the lens's own.

While there have been *theories* about light loss due to angle, etc... no proof or direct comparison has been done that I know of or that can be taken as a universally accepted reference.

Another claim has been made here that f/1.2 equates to f/1.4 on a DSRL again, no proof of that either. And, it would be an entirely different subject as it would have to do with sensor CALIBRATION not light fall-off.

It's important to understand the issue of light fall-off has been around since the beginning of photography. But DSRL haven't so, it's only fair that claims be supported by proof. It doesn't do any good to anyone otherwise.

>>".....The sensors with the widest acceptance angles (like the Kodak KAF sensors used in Oly E-1 and Leica Modul-R) are down just over a stop at 20 degrees. The Sony ICX413 (Nikon D100 and D70, Pentax *ist, Minolta 7D) and the unnamed Canon 6mp in 10D and 300D are down over 1.5 stops. And by 26 degrees (f1.0) everybodie's down 2 stops or more.<<

You are changing the subject now. None of these cameras are FF.

Again, the contention has been made here and elsewhere that ***FF sensors*** add to light fall-off because of the image circle of lenses designed for 35mm film. So, where's the data? What % increase is it for FF vs film?

I have read the theories and optical bench studies/test about light angles. But, those have to be into perspective and real world use (i.e. with a FF camera and lens attached and do the same for film).

It's obvious that if you change the angle of a light capturing device you will change the amount of light being sampled. That's a given.

Again, the claim has been made here that FF *add* to light fall off and *more* so than film.

I just want to know *how* whoever made that claim arrived to that conclusion.

My 5D will here on Thursday so, I will take some pictures with and a 1v and see if the light fall-off increases between the two. It would be empirical but, better than nothing.

Just so it's clear: I am not out to disprove one or the other. I do NOT want rumors about ***FF sensors*** and light fall-off to get started here (or elsewhere) as they have already begun (see the thread about "5D vignetting?").

Those rumors do NOT help the photo community. Instead they cause a lot of confusion for people who are trying to get some honest information about a product or technology.

I have already ordered my FF camera so, it's not an issue for me EVEN if it turns out to increase light fall-off compared to film. As I have said, I am not out to prove one or the other. I just want real figures/samples of the claim made against *FF sensors*.

Bench tests of sensors are useful but, don't relate to real world. For example: which lens directs light at 26 degrees at the sensor plane? See what I mean...

One can say a set of tires is rated up to 300mph and another to 200 mph but, if you drive at a max speed of 80mph does it *REALLY* matter? I think not.

Since the 10D was mentioned above here's some real world samples taken with the 70-200 f/4L lens wide open:

again...where's the fall-off?

Those were un-cropped, unedited JPGs (reduced for web of course).

Now, about the f/stop table: now you are saying something different.

Every one knows there's a difference in f/stops between the center and the corner, that's light fall-off.

But, to say that a 1.2 lens on film is equivalent to a 1.4 on digital has nothing to do will fall-off. It would suggest that the sensor tested was calibrated such to give different readings.

Again, the shots I posted were all metered with a Sekonic and the reading translated perfectly on my 10D as it did on my film camera. There is absolutely no difference in metering ambient or flash between my film cameras and my DSRLs.

24-70 at 32mm...no where's the 2.5 f/stop light fall-off?

ok, a Fisheye..that will surely show light fall-off...ooops, it doesn't. (don't confuse the different coloration of the sky due to the sun position).

I guess the test bench about the 10D was....wrong? No, it probably wasn't. But, it has NO relevance whatsoever to REAL shooting conditions!

One with the 24-70L @ 24...again, no fall-off to speak of. (keep in mind the sun is to the left so the sky will be lighter on the left side).

another with a 50mm @ f/2.8...

>>and the unnamed Canon 6mp in 10D and 300D are down over 1.5 stops.<<

my pictures say otherwise.

>>And by 26 degrees (f1.0) everybodie's down 2 stops or more.<<

f/1.0 and 26 degrees may be cool at the bench but, where do they occur in real life? Which particular lens has f/1.0 and transmits light at 26 degrees on a 1.6 sensor at the sensor plane?

So, while light fall-off has always been and will be (at least until we find some radically new way of making lenses) something we live with (to some extent or another) the *claim* about ***FF*** sensor causing ***additional*** light fall-off (compared to film) under ***REAL*** world shooting situations and ***real*** lenses remains to be seen.

As I have said, my 5D will here tomorrow. I have a killer few weeks coming but, soon enough I'll be able to take some test shots along with a 1v and see what the difference (if any) between it and film is and whether it's really worth worrying about it or not.

Not to minimize the problem but I would stop down to make the best of this lens on FF. Would have improved this shot too.

Maybe you guys should take a time out and play some pong, agree to disagree or something, because now it's just going in circles:

'FACT: Light fall off is a fact of life of ANY lens, including LF lenses.'

...later, same poster...

'a Fisheye..that will surely show light fall-off...ooops, it doesn't.'

'One with the 24-70L @ 24...again, no fall-off to speak of.'

http://www.luminous-landscape.com/reviews/lenses/28-105.shtml

I'm not sure why the webpage file name is 28-105 but it's actually a comparison between the 24-70 f/2.8L and the 24-105 f/4L.

I suspect the url will change when Reichmann notices the screw-up.

>>because now it's just going in circles<<

If you can't undestand what's being said, sure. Otherwise it's a simple concept and it's not a matter of agreement or disagreement. It's a matter of FACTS.

Here's an easy recap:

Light fall-off is a fact of life. Even if/when you can't see it.

A rumor was started about the 5D without proof or data to support it.

Another post later indicated that the 10D would have anywhere between 1.5 to 2.0 stops of light fall-off hence, my samples.

The point is: in the lab one can place a sensor on a machine and project light onto it at various angles, that's fine. It will tell the theoretical limits of it. But, unless those number translate into real world use they mean next to nothing. Hence the car tire analogy.

Thus, to start rumors on the net without data to back them does not help the photo community. I hope we can all agree to that.

I hope this makes it clear.

As for the original poster he himself went from talking about his lens to faulting the FF sensor for the light fall-off problem in his samples.

Finally! A good review from a major website regarding the new lens!

It is good to read that optical quality at F4 holds up. Dissappointing to read about vignetting at 24mm at F8.

If I was rolling in dough. . I would buy this new lens. But alas. . .the concept of a "budget 4L" has now past.

Giampi,

What's teh point of posting all those images taken with a 1.6 crop sensor? Of course there's no fall-off. It's a cropped sensor.

The issue here is fall-off with a full frame sensor. What am I missing?

My full frame sensor should be here in a few hours. ;-)

so i invented those quotes, then? you guys have fun with this thread. ta.

I have not followed this thread for a few days and surprised to see such big reactions from Canon fans esp Prof Giampi. If this problem has been tested and proved then just send the 'shit' back to Canon. Canon produced lousy lens before .. but this time on an expensive "L" Zoom 24-105mm /F4. Light fall off on many zoom lens are common no matter what kind of sensor or body used. When we buy Canon compatible lens from "S" or "T" we knew the problems exisit from day one. Julius, thank you for your bravery of bringing this problem up to the forum and standing to so many attacks. I don't agree someone is spreading the rumors and this is just "consumer" and "user" feedback and if Canon is serious about development new zoom lens for FFDLSR, they should look into it seriously and find ways to fix them by software. cheers

yuck. glad i have D70's and D2x with a 17-55.

Thanks Lawrence for the support and encouragement. Recent test reports of this lens on several well known web sites have confirmed my original findings. Being the first bad news messenger, I was bound to be shot by die hard Canon loyalists. But I never expected such vicious personal attacks, twisted words, and verbal terrorism.

I now fully sympathize Michael Reichmann and the abuse he has been subject to. There are indeed net surfers who hides in studios day and night firing up thousands of irresponsible posts. Being photographers, perhaps their time would be better spent shooting more pictures rather than being embroiled in endless gear talk.

Eric, not every Canon user is a bully here. Had Nikon produced full frame DSLRs, I bet its lenses would show vignetting when fully open. With Canon gear, we at least have a choice of sensor size. I now await rigorous test reports on the 5D from impartial reviewers.

I cannot believe there is a thread this large over a "problem" that has been known since the earliest days of photography. Didn't any of you people ever shoot 35mm? Or is this a thread of green newbies with too much money to spend on, and too much time to worry about, expensive digital equipment?

Yes, wide angle and fast lenses have vignetting on "full frame" (i.e. on the format their image circle was designed to cover). This is not light rays striking microlenses at a shallow angle or any other "digital artifact", nor is it a "flaw" in the design of the 5D. It is the reality of designing a wide and/or very fast lens.

What's funny is that Reichmann did this test years ago with a 16-35L and a 1Ds and found *gasp* vignetting. Only it was the exact same vignetting he had found with the same lens on 35mm. He subsequently declared the "light ray angle" theory bunk and went out and made beautiful pictures with his 16-35L on his 1Ds just like he had done on 35mm. You know what? The vignetting didn't ruin the lens or the shots. It's rarely as pronounced in "real world" shooting as it is in test shots, especially test shots which are underexposed and which have their levels set in Photoshop to maximize the effect.

Julius Wong - there are a lot of photographers who would have little trouble making beautiful photographs with the equipment you've purchased (Reichmann being one who is already on record praising the 24-105L). But hey, if you have a problem with it, by all means, return it and get your money back.

"I cannot believe there is a thread this large over a "problem" that has been known since the earliest days of photography. Didn't any of you people ever shoot 35mm? Or is this a thread of green newbies with too much money to spend on, and too much time to worry about, expensive digital equipment?"

dude, you can not show me anything, film or digital, that Nikon has done that is anywhere near as pathetic as this. these images belong in the toy camera/holga forum. Have fun.

Hi Julius,

Great Portfolio! Do you mind if I use your 'Wind Power' image in one of my Wind power vs nuclear power essays for college?

I would really like too :)

Thanks

ps. Maybe you could try out another 24-105L and see if it's any better?

To Eric above,

The famous Nikkor 35mm f/1.4 lens shows appreciable light fall off and softness at f/1.4. I know, I used one with my FM3A.

Did anyone see such a strong light fall-off on EF-S lenses? For example, how is the combo 20D + 10-22 behaving?

I am betting that the Contax N digital camera and its lenses will not produce the kinds of fall off exhibited here from some of the Canon lenses. Why? Because (I'm given to understand) Contax opted for a thorough redesign of the lens mount and optics to go with their FF digital. It would be very interesting to see some comparisons, esp. with the wides.

Thanks Daniel for enlightening me on Reichmann's old findings. Yes, I am relatively new to photography and also rather naive. I guess I expected too much on a new "L" lens. I have since realized the design limitations of zoom lenses and came to accept this lens. Sorry to have caused a stir in this community. But I am glad that truth comes out at last after much discussions. Now can we close this thread?

Eric and Joshua, please, let's not morph this thread into yet another ugly "Canon vs Nikon (or Contax or any other) brand loyalty" thread.

Sonny Jet, you are welcomed to use my wind power photo for your college essay. BTW, thanks for the praise.

To everybody else: Happy Shooting! Hope to view more of your wonderful images on photo.net.

My understanding of vignetting is that the problem of reducing light from a large field onto a postage-stamp sized piece of film or digital sensor is considerable: the world is big and chips are small, and middle and large-format systems have a much easier task.

You can literally see this when moving the front standard on large format lenses up and down and how, as you move the cone of light towards its margins, distortion and light fall-off increase.

Second, lenses may be designed to reduce optical distortion and light fall-off at the cost of other qualities. I've worked with lenses in graphic arts darkrooms that created 1:1 negatives that could almost be matched up to their originals -- though the cost was a huge piece of glass and the need for massive lighting and long exposures. I simply would not expect a small format camera to produce the flat field and evenness asked for here, and I assume that lens designers for small cameras are not assuming that either.

To make that lovely image of the figures by the bay -- it is a gem! -- even better, I would use a larger format camera, and since the details in the front and off in the distance and exquisite colors of that light are such an important part of it, I'd put it on a tripod, too. I actually don't see the advantage of small format here -- a place the photographer could go back to and find all the essential elements in place with differences in light and passersby that might be seized with each visit to even greater effect.

It seems to me that small camera work wins up close and social, where you can pull it out in the middle of a conversation or walking down a street, and where the distortions of that fancy zoom are worth the cost of being able to select the focal length quickly.

Not long ago I fiddled with one of them while having coffee in a cafe, and which my companion put in my hands, and without leaving my chair I made a dozen images of the scene across from me that each had their own virtues and interest. I then did what I would normally do with my own 35mm rangefinder with a 35mm lens on it: I had to get up, walk forward and back, left and right, and tilt and turn -- a very different experience and very different kind of image -- not bettor or worse, but different. I left the experience thinking that I'd love to have such a digital camera and zoom: it would help me work in tight places (or when I am too old to move! Ha!)

Writing this leads me to think about how I was trained to respect the materials: to experiment and see what kinds of pictures this or that film, lens, format, etc., was happiest doing and then exploit the advantages and avoid the disadvantages. I think I have also learned to choose to follow my nose and shift to equipment that would do the trick. As detail became more and more important to me I shifted to middle format, but when I thought even more detail would be better and started working with large format I realized I was losing the mobility that I had gotten used to; middle format on a tripod you can pick up and carry in an instant, and your small cameras with zooms are, of course, even faster, lend themselves to even quicker snaps, and might even be more fun, too!

Enjoy!

Bruce Spear

ps: Now that I see photo.net inviting me to attach images I'll attach a snap of one of my American students being given instructed in the limitations of American foreign policy and where I think the small digital camera (Canon s50) wins: nobody noticed or even cared that I was snapping away, I had plenty of time to play with the composition, and the optical distortions and light fall-off seem part of the compostion. I've tried photographing such cafe scenes with middle-format, and as you can imagine, pulling out and firing a Hasselbad is more or less equivalent to firing a cannon in the middle of a church gathering -- an interesting idea, but not that here, where I was fascinated with the interaction and humor of the people, and especially my student: he was listening and learning without giving up too much of his ground, and I think he won a moment of friendship, too. B

Daniel, the well obstruction theory is not bunk, however, it may or may not contribute in a specific sensor and lens design combination. It is obvious that it would contribute disasterously with Leica rangefinder wide angles, and we know that Kodak applied lens-specific correction in the FF DSLRs due to this. How do we know that Canon isn't doing the same with the 5D? It should be easy enough to test. Namely, shoot an evenly illuminated flat and homogeneous surface with an intermediate or high ISO setting. Now, if you calculate the noise from the center of the frame and compare it with the edges, if there is more noise at the edges (pumped up by vignetting correction), you have a corrected image. If the noise is exactly the same, then you probably have the natural effects of the sensor and lens in the image.

I would not trust Michael Reichmann's comment about the vignetting being the same on film and digital. I don't even think he'd know how to measure the effects correctly. In order to compare images across media, you need a contrast reference (gray patch scale) which is then used to adjust the curves so that the gray patches match and then you can start comparing vignetting between scan and digital image.

First of all let me make myself's background clear.

I am a Canon user for film some 14 years ago, I used to use Canon EOS630 film back with the standard zoom EF28-70mm f3.5-4. I quit Canon system due to its bad performance for wide angle lens not due to dark angles but the bad distoration. I took mostly landscape so AUTOfocus is not critical for me and I have found Contax a more suitable system for myself. I purchased Contax 167MT and subsequently a few Contax Carl Zeiss Lens a 18mm f/4, 45mm f/2.8, 60mm f/2.8 macro, 200mm f/4 and no more till now. I am taking slide and film but was also interested in digital as this is gonna be the future (before DSLR I got G2 and Pentax Optio S). Since Contax stop production for all its product, I have no choice but to goto DSLR of other brands, I buy mounts for my Carl Zeiss lens and use it on Canon 20D, because I am used to Canon system in my photography history. I begin my Contax Film experience since 1999 until now.

Now if FF sensor causing the dark angles, I will say MOST LIKELY.

Forgive me for being such subjective...but let us forget which brand (Nikon, Canon, Olympus or whatsoever)....let us see the reason why film back is better than sensor.

Olympus website did say the answer (please forgive me if I extract anything wrong).

Extract from Olympus website:-

http://www.olympus-esystem.com/dea/products/e300/feature/index.html

Because the CCD image sensors used in digital cameras are flat, they are only able to accurately receive light that hits them straight on. For this reason, when a 35mm film camera lens is used with a digital camera, colour reproduction tends to be inaccurate and brightness insufficient at the image periphery, where light is apt to strike the CCD obliquely, making it impossible to realise the full potential of the CCD. Olympus's Zuiko Digital lenses, on the other hand, are designed exclusively for digital SLRs. By ensuring the same level of imaging quality at the periphery as in the center, these lenses are able to maximise the full potential of the E-300's 8-megapixel CCD. Also, since the Four Thirds System specifications result in a lens mount diameter that is much larger than the image circle, the combination of camera body and lens ensures that the light continues to travel in a straight line to the image sensor.

SO STRAIGHT LIGHT is the problem....with wide angle, straight lights is becoming more difficult.....I THINK.....

Any sensor no matter CMOS or CCD will have the same situation, if it is a 1.6x crop the problem won't appear, I think I don't need to explain, but FF will not escape with the problem unless there is new technology.

Here is a picture I took with my Contax Carl Zeiss Lens on Canon 20D body. I am only interested in Canon's sensor....unless I got a lot of $$$ I won't change to expensive Canon lens. But for doing fast job Canon lens is for sure. I am talking about the situation of myself (I only took landscape and slow objects). I do not mind using manual focus, even for cats. I only buy the sensor....which Contax cannot provide forever (I am not interested in digital N, those are Japanese design Carl Zeiss).

>>What's teh point of posting all those images taken with a 1.6 crop sensor? Of course there's no fall-off. It's a cropped sensor.<<

Did you read the thread?

A contention was made about the 10D vignetting. My pictures directly address (and disprove) the contention made.

One thing is to theorize, another is to prove it under normal shooting conditions (i.e. NOT in a lab).

"A contention was made about the 10D vignetting." Giampi, please DO NOT put words in my mouth. I have never alleged anything about the 10D. I don't know what is your problem, but you have a tendancy to mud sling people. Please leave me out of that. Frankly, I am getting very tired of you and your TWISTED arguments. Case closed!

Geez, guys, lighten up! All wide angles and zooms vignette and demonstrate curvilinear distortion wide open. That's a given. Yes, my 24-105 vignettes more than my 24-70 F2.8 on a 5D. Can't say I was that surprised, really.

Now get the PTLens freeware plugin for Photoshop (or the standalone app) and just fix both issues...

Simple.

DxO optics does something similar, so does Photoshop's standard lens correction module. But PTL seems to work well, plus it's simple and free.

You won't notice these problems on well captured real-world images most of the time but, if you do, there really are several straightforward fixes in post-processing.

So why not go out there and take some pictures...?

Hope this helps.

Jon Bower (www.apexphotos.com)

Hi guys! I've been delighted in reading your posts about EF 24-105 L IS.I recently bought such a lens for use on a 5D FF and I've been imemdiately struck by a relevant degree of underexposure at f 4. Any other conditions being equal, pictures appear darker at f 4, as compared to exactly the same shot got at 5.6 with doubling time exposure.This is not just vignetting, the full picture is affetcted until down to the center, although the effect is more relevant at the borders and corners. A beige wall at f8 becomes light brown at f 4, where an azure sky becomes blue. This is particularly appreciable at extreme focal lenghts and in subdued light, but still evident in several different light conditions. Such a peculiarity becomes much less evident - if at all -going from 5.6 to 8. Although I was not expecting prime lenses performances from a 4.3x zoom, I think this the major drawback of a lens with more than honest walk around features, far however from what one could expect from L. I've been wondering whether such features were typical of the lens project and meet production standards, or affected just my own sample. I applied to my country's Canon assistance, having no satisfactory answer on the phone. I then wrote to the customers care office in Italy and at Canon Europe, getting no answer at all! I felt rather alone with my problem and I'm happy so many persons have have been dealing with it. I think that 28-70 is by far too heavy and I will keep the combo until I get tired of it and the way Canon deals with me...

Our use for the Canon 5D + 24-105 is professional in respect of the university where I work. With experience going back to a brownie box we can determine the working range of equipment and adapt for any given shooting situation. However, the degree of 'light fall-off' and distortion of the Canon 24-105mm FF is not an acceptable limitation for our use, given the quantity of photographs we take and tight post processing margins. Our own tests show that the lens even with post processing is not useable with impunity below 35mm focal length at about f5.6. We concur with all the pictures shown on this board so far. Separate smaller range lens would provide the necessary quality and range and it is unfortunate that Canon and many reviewers imply that those qualities carry through to the 24-105 lens with little impairment.

Commenting on angle of light to photcells. Wide angle lens for reflex cameras are in fact retrofocus lens, which means that their back focal length appears to be closer to a standard 50mm, thus enabling them to fit normally in front of the mirror! Optical design magic does the rest to achieve the wide angle. Telephoto lens also use the same magic in the opposite way. This means that the angle of light rays subtended from the lens back element to the sensor is similar for most front focal lengths and is therefore NOT an issue affecting digital photocells.

You may have noticed that if you observe a stopped down iris from the front of a wide angle lens, it appears a slightly larger size and also 'follows you' round when viewed from the edges cf. the middle. This is clever lens design compensating for natural light cone diameter reduction towards the corners. The design question is 'how much will the market let us get away with?' My reply is that the professional market (as well as keen amateurs) want to know the real design spec before buying - Hate Hype, Feed Facts.

Fall-off v. exposure. The worst apparant case will be at the correct average exposure as this is the most linear and eye-sensitive part of the light-to-digital-to-light conversion process. Greater compression, hence apparant brightness flattening will occur away from the centre exposure range.

I have a 24-105L that had showed significant at 24mm f/4-f7 (almost gone at 28mm) on my 5D (where you would see it) but not as severe as the images posted above. I sent the lens to Canon for adjustment and now the vignetting is minimal. This lens was pretty awesome to begin with -- now it's superb.