Nope. Only time I have ever seen banding is when I was 'punishing the pixels' in photoshop before I realised I should be working on 16bit TIFFS not 8bit. Since doing everything in 16bit I've never seen it again. I've never seen banding on a RAW file that I can recall.
Brody LeBlanc wrote:
Hmm, I've had the 5DII for a week now and I'm very upset to find out about this banding nonsense... All of the 12X18" prints I've been doing look so good. Oh well! Guess I'll have to sell it....
And this is the well documented proof? sometimes the internet makes me sad.
Seriously, at 1600 ISO I can not create any sort of noticeable banding. I even did that retarded 100% fill-light thing. Still no banding. Maybe I'm just distracted by the outstanding image quality I'm getting... Who knows?
You need to shoot JPEGs and push shadows and lower mids a stop or two and you'll see them. You have to work really hard to band with RAW converted to 16-bit TIFF or PSD.
are you saying your 5d2 doesn't have banding issues at low ISO? let say clean no banding at all?
I am talking about directly viewing a 5d2 raw file., not converted jpegs or pushed jpegs.
Please do understand that I shoot only from base ISO to 1600, from what I've seen even at 5d2's ISO 100, for me is not the IQ I am looking for.
The features you mentioned are all pluses... but in the end it's all the IQ. If you can guarantee me that 5d2 doesn't banding issues... I'll buy it -- this is the reason why I ask here, I need a new body.
I did not say that it is impossible to find banding in 5D2 images. In fact, I have written before that it is possible to find banding... in very rare cases, when you look for it at 100% magnification, typically in an underexposed monochromatic area of the frame of a very smooth gradient, more with green tones than anything else.
What I have also said is that I've shot some tens of thousands of frames on my 5D2 over about three years of use - including all sorts of subjects ranging from urban and natural landscapes to street and night photography. In a small number of images - most certainly smaller than 1% - if I look for banding I can find some slight evidence of it at 100% magnification, but it has yet to be affect a print of one of my photographs.
I did have one photograph in which I thought it was going to be a problem. Here is the image:
This scene posed some of the most difficult exposure challenges imaginable. The dynamic range between the bright highlights and the shadowed cliff face considerably exceeds that of current DSLRs. With digital capture, the reflected highlights of the sun on the water limit the bright end of the exposure scale since if they blow out the quality of the image will be awful. So I had to use a "darker" exposure than I might otherwise have preferred. However, the cliff face at the left was not only in deep shadow but was also a very low contrast subject. In post I wanted to use fill, adjust black point, and use a luminosity curve layer in this area to bring back just a bit of the very dark shadow detail. When I did so, and while viewing at 100%, I could see some banding in the adjusted areas that had been almost completely black before applying the curves adjustment. I worried that this might be visible under close inspection of the 16" x 24" print that I was planning to make so I compensated by adding a bit of very fine grain noise in that area to mask what I feared might be slightly visible noise banding.
However, before making the final print I thought to try a test print of that area of the frame without the added noise layer... and the banding that was visible at 100% magnification was not visible at all in the print. So I ended up deleting the layer that was to mask the banding.
The obsession with supposed perfection at 100% seems just plain strange to me. No image is perfect at 100%, and there are a number of ways besides noise that an image can reveal its "flaws" at such a size. But you will never see this stuff in a print, even a very big one - just as you won't see a ton of other stuff. And don't get me wrong - I'm most certainly not saying I don't care about print quality. I care about it deeply. What I am saying is that this concern is entirely misplaced. It is like worrying that you made a mistake when you added a teaspoon of salt to your soup and put in 3 grains too much salt!
By the way, wander on over to whatever other camera brand forum you want to investigate and you'll find that each brand has its own merry band of posters who will swear on a stack of bibles that Nikon or Sony or whoever also builds inadequate and imperfect cameras that have this, that, or the other terrible flaw.
But somehow we manage to struggle on and make photographs.
Finally, having seen too many of these threads, I can anticipate the counters to my position on this issue. "You are a fan boy." (I'm not.) "You don't push the camera with your postcard shots." (Yes, I do.) "Your standards are low." (They are not.) "Canon can do better." (No doubt.)
Regarding that last idea. Yes, of course, Canon and all manufacturers will continue to improve their cameras as they have done for decades. I'm all for that, and I welcome any improvements. However, photographic equipment never has been "perfect" and it never will be. But somehow, photographers working with much more primitive and troublesome equipment and processes than we have today managed to make astonishingly wonderful photographic images. The idea that you are kept from doing so by the shortcomings of the 5D2 (or whatever other camera you use) is silly.
Hmm, look like all my brand new 5x5d2 actually were 5d (1+1/2) - they have more/less shadow banding when DR is greater than 8/9 stops. I don't use filters like others. Other than couple internet reported low iso constraints, 5D2 does well. Man, this is old news
okay guys I will give 5d2 a chance... I will ask my two nieghbors to bring their 5d2s.
Will test them with bad and good lighting in a controlled environment.
My banding test verification will be simple: just view the RAW files, 100% in either DPP or in Windows CR2 veiwer. No push\pull mods., no crazy pp manipulations, just native RAW file expsoure.
Again thank you.... I just want to be sure that the $2,500 money I am investing is well worth the IQ I am getting from 5d2. I do not care about its features like, screen, 21MP, weather seals, video, etc.
My plan is the camera will be used only in-house [ studio ], it will be on a tripod all the time with a computer controlling it., CR2 data will be pass thru wireless lan. So again for me the added features are not really much of need.... I am targetting IQ.
dhphoto wrote:
BANDING only exists if you don't use the gear properly.
A silly statement. I used my gear quite "properly" and while i agree the 5D2 is a great camera, to claim banding exists only in improper applications is absurd.
Quite frankly i find threads like these a bit entertaining, and I'm thankful to be 100% brand-agnostic and therefore immune to blind loyalty. (that wasnt directed just to you dhphoto)
Banding is always there since Canon 1D (actually not that tragic on 1Ds). Simple advice is to not try lightening shadows too much. As for example 5DMK2 has some latitude in highlights, its possible to slightly "burn out" photo and then pull it back in PP and no shadow banding problem.
Or shoot HDRs.
Or live with simple fact that not everything is needed to be shot perfectly exposed and sometimes when shadows are black and highlights burned it doesnt hurt photo quality, if photo is good itself.
As far as colors go, 5Dc has better color accuracy than 5DMK2, thats all. Metamerism index says 84/80 for 5Dc and 80/76 for 5DMK2. If you WB correctly and maybe tweak colors a bit with proper tools, it shouldnt be so bad for 5DMK2. And Capture One as RAW developer can help too.
Ofc its not A900 or 1DsMK3, but then 1DsMK3 is quite a bit more expensive and A900 is out of production and probably without future, unless you want SLT/mirrorless system.
cputeq wrote:
A silly statement. I used my gear quite "properly" and while i agree the 5D2 is a great camera, to claim banding exists only in improper applications is absurd.
That's not what I said, don't deliberately misquote me. I said "BANDING only exists if you don't use the gear properly"
dhphoto wrote:
BANDING only exists if you don't use the gear properly.
This quote is definitely not in engineering cook book ...at least not mine.
An output of 5D2 camera is one of some billion combinations between aperture, focal length, light incident/reflection angles, shutter speed, temperature, weather, subject distance, optical lenses alignment, imaging sensor, af sensor, IC, power, media, lens filters, DR, iso amplification, sun & earth orbits, subjects reflection, MTF combo, your mood of the day .... etc...
Whoever said they knowing their camera inside out then they really don't know what they are talking about. Trust me, no one shoots the same way... just alots of bs to pretend low iso shadows banding was not happening with some 5d2....
All my 5x5d2 have it in the 100 iso shadows when DR hits 8/9 stops and wider.
Why 5d2 has DR low iso banding shadows constraint but some other canon model does not have it
Any camera should be free of banding in the first place...
Darn, Canon can't design a cam for a beginner like me
Mescalamba wrote:
As far as colors go, 5Dc has better color accuracy than 5DMK2, thats all. Metamerism index says 84/80 for 5Dc and 80/76 for 5DMK2. If you WB correctly and maybe tweak colors a bit with proper tools, it shouldnt be so bad for 5DMK2. And Capture One as RAW developer can help too.
Has anyone actually tried it with a calibrated camera? You know, the 24 color swatch Macbeth chart. When I shoot with two, three different cameras side by side and create custom profiles for each of them for the given scene, it's impossible to tell which camera is which, from the color perspective anyway. Color is one of the easiest things to fix this way.
To me this all sounds a bit like looking at two thermometers showing a slightly different value and without any further piece of information declaring one to be more accurate than the other.
Gochugogi wrote:
You need to shoot JPEGs and push shadows and lower mids a stop or two and you'll see them. You have to work really hard to band with RAW converted to 16-bit TIFF or PSD.
No offense, but if you're buying a 5D2 and shooting jpeg with it, and then pushing your exposure, you deserve to be shooting a rebel.
I would thinkt hat most people who use a 5D2 would know proper exposure, and know that if they are shooting jpeg...they are pretty well effed if they botch the exposure.
bushwacker wrote:
okay guys I will give 5d2 a chance... I will ask my two nieghbors to bring their 5d2s.
Will test them with bad and good lighting in a controlled environment.
My banding test verification will be simple: just view the RAW files, 100% in either DPP or in Windows CR2 veiwer. No push\pull mods., no crazy pp manipulations, just native RAW file expsoure.
Again thank you.... I just want to be sure that the $2,500 money I am investing is well worth the IQ I am getting from 5d2. I do not care about its features like, screen, 21MP, weather seals, video, etc.
My plan is the camera will be used only in-house [ studio ], it will be on a tripod all the time with a computer controlling it., CR2 data will be pass thru wireless lan. So again for me the added features are not really much of need.... I am targetting IQ.
I beleive there were perhaps some differences in bodies or firmware or something which resulted in differences in banding performance. On my copy of the 5d2, i could induce banding in LR3 using a fill light slider of around 35...shooting Raw in either iso 100 or 200 usually (landscapes)' and mostly exposing to the right.
I also believe large parts of banding or other noise complaints can be to how someone has calibrated their monitor luminance/brightness.
For instance, my home imac gives me lovely rich colors and deep blacks. I can email myself (work address) a perfectly fine jpeg processed, yet at work it will look much different - blacks are much less deep and unclipped at all and seem to be raised, showing lots of shadow noise that i never saw at home. My work uses cheap TN panels so i dont bother to try to correct, but i wonder if differences in monitor brightness and contrast could be contributing to perceived differences in noise performance to different viewers.
Noise should be measured by lab instrument than our global media (monitor/printer). Imagine, how bad the banding effects look like if we have 12bits/higher media platform
