EOS 5D Woes

[James_Symington]

Dear Les and BigWill,

I too got a 2nd 24-105mm after that initial recall and the corner sharpness improved slightly on the 2nd copy - fortunately.

I just did a quick comparison of my 5D and 1Ds II shooting against a blank white wall at home with a few of my lenses. The lenses used were:

24mm f1.4 L
35mm f1.4 L
85mm f1.8
20mm f1.8 (Sigma)
24-105mm f4 L

24mm f1.4 L
Vignetting pronounced at f1.4 and identical on both cameras, negligible at f2.8, absent at f4

35mm f1.4 L
Vignetting pronounced at f1.4 and identical on both cameras, negligible at f2.8, absent at f4

85mm f1.8
Slight vignetting at f1.8, negligible at f2.8 and absent at f4 identically on both cameras

24-105mm f4 L
Pronounced and identical vignetting at f4, negligible at f8 (the widest aperture I usually use for landscape purposes with this lens and successfully I might add) - and totally absent at f11

20mm f1.8 Sigma
Medium vignetting that improves up until f4 when it disappears - in identical fashion on both cameras

On the basis of this rough but revealing test - admittedly just with my sample of one 5D and one 1Ds II - I can conclude that this alleged 5D vignetting problem occurs in equal, identical measure to the 1Ds II. Duff 1Ds II? Maybe. Or maybe I just have one of the few 5Ds in the country that works.

More likely this 5D vignetting problem is really just a 24-105mm problem - and one that can be overcome simply either digitally or by stopping down.

As a landscaper shooting at small apertures on a tripod I don't care most of the time. But I have just come back from a month long trip around the Pacific and Chile where aside from MF cameras I only had the 5D with the 24-105mm and hand-held at f8 the results were sparkling with no vignetting that you would notice. It is the greatest all-round zoom lens I have ever owned even with its known limitations.

Draw your own conclusions from this test if you like - I have at least satisfied myself if nobody else.

Goodnight all,

James

http://www.jamessymington.co.uk

[Scphoto]

Draw your own conclusions from this test if you like - I have at least satisfied myself if nobody else.

Sounds fine to me, like you i rarely shoot at F4 anyway so its not an issue.

[jchrisc]

I just did a quick comparison of my 5D and 1Ds II shooting against a blank white wall at home with a few of my lenses.

The only lens I have that compares directly with any of yours is the 80mm f1.8

How does yours look compared with this (wide open)

[Lounge Lizard]

Thanks for your input James. I published some pictures here and I plan to publish more shortly of a 5D vs film shoot-out with the same lenses.

I frequently shoot at f2.8 or f4 - it's why I invested in L glass. What I am trying to establish is whether the 5D has a design flaw and all cameras are similarly affected or whether Canon has a QC issue. I am happy to persevere trying to get hold of a good body but I don't want to waste my time if such a beast does not exist.

I am not aware of any full-frame images that have been published on the web showing negligible or nil vignetting wide open on the 24-105 or 17-40 lenses. My results so far show vignetting even at f8 and vignetting on the 70-200 lens as well as a 50mm prime. There are claims made by some that there isn't a vignetting issue and there is a slight chance that these users have a good camera (if so, I want one just like that!) but, as yet, I have not seen any proof.

[BigWill]

I wouldn't dream of telling anyone how to think about a camera. That is up to each individual to make up their own mind about. What I (and I believe David and Les) are trying to achieve is that people are at least made aware of the facts about the camera and it's limitations so that they can make an informed choice. As a potential purchaser of this camera (or it's type) myself someday, I have a vested intrest in knowing its capabilities or lack thereof. I don't see why the fact that I don't currently own the camera should preclude me from speaking about it here as I believe my opinion is as valid as anyone else's around here and I find it odd that you have singled me out as the one who should desist from expressing an opinion as there are others in this thread who, like myself, don't own the camera but who haven't been chastised for expressing an opinion.

If you wish me to be silent on this matter then I suggest you exercise your moderator powers and remove my comments from this thread. You have the capability to do so if you choose. I leave the matter in your hands.

BigWill

[radim]

David, I'll try today to go back to Jessops on Strand and take few shots on my CF card if they still have same body I've tried this week.

Radim

[Lounge Lizard]

Thanks. The lens must be wide open in Av mode and choose a scene sufficiently plain and uniform so you stand a chance of seeing it. Take two identical shots - one at max apertue and one at f8 and then, using the thumbwheel, it's easy to toggle between them and see if the corners darken or lighten depending on which way you view them.

Better still, take the images home and post them here.

[alanS]

Sorry to go off subject just a little and defend EF-S lenses.

These lenses are specifically designed to get the best out of the digital cameras they fit. Whilst I accept that they can not be fitted to non compatible cameras I believe that Canon show an increasing commitment to EF-S and that if at some point in the future any purchaser wishes to buy other than an EF-S compatible camera they will find a ready market for any EF-S lens that they wish to sell.

I just find it a little off putting that some on these forums advise others to steer clear of EF-S lenses because of compatibility issues. I'd counter that by pointing out that at least EF-S lenses are specifically designed for the cameras they fit. Anyway, I rest my case.

[Lounge Lizard]

I quite agree Alan but mostly we point this out in case members are considering building up a lens collection and then upgrading to a full-frame body later on. In fact, I have an EF-S 10-22mm and it is excellent but I bought it knowing that I'd have to sell it if I went full-frame.

On the other side of the coin, it does suggest that Canon sees issues with full-frame such that it continues to put effort into developing lenses for APS-C size sensors.

[Lounge Lizard]

...when a major company get away with shirking their responsibilities to their customers by denying the existance of what has now become a well documented problem!

Well, that's part of the problem. Is it really 'well-documented'? We have some members here showing problems of various degrees and we have other claiming that the 5D does not vignette.

FTR, I only ever starting buying L lenses when I had my D30 so I never really examined how the lenses behave on film when used wide-open (all will be revealed shortly). Equally, this is my first full-frame dSLR. I am putting together the evidence and hope fully I'll have something conclusive to put up online shortly rather than deal with peoples' subjective opinion.

Also, as an aside, I decided to expand my testing to a three-way battle - 5D vs 20D vs 600 (film).

[James_Symington]

Hi David,

Good morning. What I tried to establish in my little test last night comparing my two cameras is that there is no difference between the 1Ds II and the 5D in their behaviour. The results I was getting look essentially the same as your posted photos and for lenses like the 24 and 35mm Ls vignetting is all but gone at f2.8 and on the 24-105mm it is all but gone at f8. What I don't quite get is why this translates to the 5D having any kind of problem at all? The proof of the pudding would be repeating the test on slide film as you will do - but I am sure it will be the same.

Anecdotally, the Leica 35mm f2 Summicron-M vignettes every bit as badly wide open as the 24-105mm does. I was so surprised that such a highly corrected uber lens behaved like this that I started a thread on photo.net ages ago because I thought something was wrong with mine. Wrong! Everybody's is like that and it is just a fact of lens design that this vignetting has to be incorporated. In the latest edition of Leica Photographie there is a comparison of their 4 current 50mm lenses and the f2 Summicron also has pronounced vignetting when wide open and they say so in the text too. That is supposedly the reference 50mm lens in 35mm photography.

Full frame digital cameras of the quality and resolution of the 5D and the 1Ds II are merciless in exposing the characteristics of the lenses used on them and this what I believe is the source of this issue. Because the 5D is the first 'affordable' full frame DSLR plenty of people are finding problems with their lenses. People may not like the vignetting characteristics of the 24-105mm but that is a whole different discussion. But I don't think the 5D has a problem at all. Just my view though!

All the best,

James

[James_Symington]

Hi Chris,

It looks the same as yours on both of my cameras!

Thanks,

James

[Lounge Lizard]

If mine looked like that, I'd be a lot happier but experience so far suggests it's much worse than that.

[chrisgatcum]

Hi folks

I guess it's time that I throw myself into the fray, not least because of the criticisms posted about my 5D test and AP's tests in general.

To start with I want to assure everyone that the editorial department does not pander to advertisers as has been suggested. I have never owned Canon kit so personally I have no attachment to the brand whatsoever (even if I did, my job is to report the facts as they are). While Canon may advertise their products and provide competition prizes, so do other manufacturers and this is between the company and AP's advertising / marketing departments. Editorial is a seperate entity.
Before I hear cries of 'he doth protest too much' (I may not be psychic, but I know how some people think) I will leave it at this - we (AP's testing staff) are impartial and we know we are impartial. I appreciate you can only take my word for that.

As far as my test of the 5D goes I really don't know where to start. The simple fact is that I can only test the camera I am given with a selection of lenses and can only analyse the results it gives. In this case I have several DVDs of images taken under wide-ranging conditions and given the early rumours of vignetting this was something I specifically set out to explore. In this instance - with this particular camera - I did not see any evidence of vignetting. So, the ONLY conclusion I could reach is that vignetting is not an issue. I could not report that vignetting might be a problem, as that would be speculative.

It has, however, become apparent since I conducted that test that some people are experiencing vignetting, and in some instances it is terrible. In his recent 5DvsD200 test, for example, Damien Demolder found the 5D did vignette slightly, although it should be stressed that this was with a different camera body and different lens - perhaps manufacturing tolerances have a part to play?

I hope this goes some way to explaining things, and I assure you that we at AP hold you as the reader in far greater esteem than the camera makers, and it is our endeavour to make sure the advice we give is correct.
If you have a 'duff' 5D then I would suggest the answer is to contact Canon. Don't accept 'no' for an answer and if you aren't happy about the service and / or product, keep going. I know it's not easy and it's definitely frustrating, but you do have your rights as a UK consumer.

Thanks for listening / reading

Chris Gatcum

[TerryS]

James_Symington wrote:
On the basis of this rough but revealing test - admittedly just with my sample of one 5D and one 1Ds II - I can conclude that this alleged 5D vignetting problem occurs in equal, identical measure to the 1Ds II. Duff 1Ds II? Maybe. Or maybe I just have one of the few 5Ds in the country that works.

Many thanks for your tests James - they are precisely what is needed in order to draw reliable conclusions about this issue. So it appears that there is no vignetting problem that is specific to the 5D - in this respect it performs identically with Canon's other current full frame body, the 1DsMkII (albeit that this conclusion is based on a rather small sample size!). The next step is to determine whether there is a difference in vignetting between Canon's full frame sensors and 35mm film, for which we eagerly await David's tests. Although really it would be better to compare James's 5D against a film SLR (using the same lenses), since otherwise we won't know whether any differences are genuine or are because David has a 'bad' 5D.

I think it is worth emphasising that your results show that "pronounced" vignetting does occur in some (identical) circumstances for both the 5D and 1DsMkII, since in this discussion (and previous discussions here and elsewhere) there have been claims that in these circumstances these bodies show no vignetting. Not only have such claims muddied the water as far as ascertaining whether or not the 5D has a specific problem, but I always thought that they were a little incredible and, indeed, unrealistic, since optical vignetting is an inherent property of any lens and is unavoidable. The question is not 'Does the 5D/24-105mm show vignetting at 24mm?' (the answer to which is inevitably yes), but rather 'Does the 5D/24-105mm show additional vignetting compared to that shown by the 1DsMkII or a 35mm film SLR?' and 'Does the 24-105mm show excessive (and unacceptable) vignetting at 24mm with all full-frame bodies?'.

James_Symington wrote:
... I only had the 5D with the 24-105mm and hand-held at f8 the results were sparkling with no vignetting that you would notice. It is the greatest all-round zoom lens I have ever owned even with its known limitations.

Although I am using the 24-105mm on a cropped sensor body (the 30D), I totally agree with you James. With its overall optical quality (vignetting aside), focal length range, image stabilisation, handling (including nice size & weight) and build quality, I fully concur with your comments. In this context I also agree with what you said in your first posting in this thread - optical vignetting is nothing new and exists to some degree with every camera (both film and digital) and lens combination. Consequently, in the case of the 5D there is a danger of blowing this issue out of all proportion, although in my opinion the 24-105mm does appear to vignette excessively (unacceptably?) at 24mm, as was concluded in an earlier article in the British Journal of Photography. Although as you have pointed out vignetting is easily corrected in software, such as the excellent and free PTLens available as standalone application or Photoshop plugin, see: PTLens Plugin and PTLens & Canon EOS 5D.

David Steel wrote:
On the other side of the coin, it does suggest that Canon sees issues with full-frame such that it continues to put effort into developing lenses for APS-C size sensors.

I think that this is more a recognition by Canon that the manufacturing costs of full-frame sensors will always be high and this cost factor will put them out of the reach of most beginners and many amateurs. Other than cost, I don't think Canon sees any "issues" with full-frame.

Terry

[Lounge Lizard]

Many thanks for your input Chris. I myself have never had any doubts about AP's impartiality though I might suspect that a few manufacturers ensure that they only send out 'top-notch' examples for review.

Though I am still testing and formulating my conclusions, I am suspecting that there is considerable variability in the cameras being shipped. It's worrying because the manufacturers' publicity and test reviews set users' expectations and there will always be the concern amongst buyers whether their particular camera is sub-standard or not.

I am now on body #2 and the vignetting performance appears to be similar to the first one (my side-by-side comparison with the same lenses on a film body is still being assembled) and this may not be too much of a surprise since there isn't a great deal of difference in the serial numbers suggesting that they are from the same batch.

For my part, I simply want a good body that works well with my lenses when used wide-open (that's what I bought them for in the first place) but the dilemma is whether there is an inherent design flaw meaning that I should forget this particular model and use something else, or whether there are good examples out there and I should keep striving to find one. I'd much rather be taking photographs for my pleasure than be endless testing to prove to my dealer or the manufacturere that a fault exists.

I feel that manufacturers should be more approachable and open about such issues and not leave consumers to do all the leg work and effectively QA the product for them - this is not the right approach. After buying Canon cameras over the years (Canon EXEE, T70, EOS100, EOS50E, EOS3, D30, 10D, 20D and now 5D) my brand loyalty has weakened somewhat. Thankfully, my local branch of Jessops is as keen to find an answer as I am.

[chrisgatcum]

David
Although I cannot respond in an 'official' capacity I would largely concur with your suspicions regarding 'variability' in cameras - a friend and I almost came to blows after I recommended the 5D and the one she got exhibited severe vignetting. She's on her third body and it's an entirely different camera (ie no vignetting...)
I sense a discussion with the 'big C' is imminent...

[Lounge Lizard]

Wow! Thanks Chris - it gives me hope.

[SteveEM]

Hi all,

Here is my two penneth..I am neither a user of Canon, Nikon, or Olympus, so have no axe to grind.

It seems to me that almost all the major camera makers rushed their DSLR's to market a little fast, to gain market share and before proper thought was given to the 'light' needs of digital sensors. We know they have 'depth' unlike film and so require light to hit them at 90 degrees. But most of the SLR makers simply popped digital sensors into conventional SLR bodies and used them with conventional lenses...and here the problems start. It is just that the full frame sensors highlight the issues more and are effectively a compromised design..as the bad vignetting issues show.

It seems to me that only Olympus with their E digital system designed PROPERLY from the ground up, with properly designed digital lenses have the issues 'sorted'. Of course they have suffered sales-wise because they came to the party later, but I think they have the answers..

Cheers Steve.

[TerryS]

Although I cannot respond in an 'official' capacity I would largely concur with your suspicions regarding 'variability' in cameras - a friend and I almost came to blows after I recommended the 5D and the one she got exhibited severe vignetting. She's on her third body and it's an entirely different camera (ie no vignetting...)
I sense a discussion with the 'big C' is imminent...

Thanks for the response Chris. So (off the record ) is it your opinion that the vignetting that is being reported with the 5D body and 24-105mm lens (at 24mm) is more likely to be due to variability of the 5D body rather than variability of the 24-105mm lens?

This is important because if an individual buys a 5D/24-105 kit and they don't have access to another full frame body then he or she has no means of determining whether the 'fault' lies with the body or the lens. Also, in my case, I am using the 24-105mm lens on a cropped-sensor 30D body (and have long since sold my Canon 35mm film SLRs) and so have no means of testing the lens for vignetting issues.

Terry.