Daan B wrote:
On the sRAW: this is basically a re-sampling/re-sizing too, right? So, why would it not make the effect less visible (as a constrained re-size in PS would do too)?
More or less, sRAW is a demosaiced RAW, with discarded rows and columns.
No re-sample or re-size.
I won't argue with the experts who say it's real, but there could be other factors that make it worse, such as your monitor's gamut limit, a 6-bit monitor that doesn't handle gradients well, the monitor calibration, the particular software used to process the image, whether you're viewing in a web browser without color management, or viewing a print with a bad printer profile. Unlike shadow banding, the highlight banding seems easier fix in processing.
I don't see sky banding from my 7D. What I sometimes see are subtle chages in sky tone that are there in the sky. These take different forms, vertical and horizontal. I think they are due to temperature variations and different amounts of faint cloud. The problem is that we often don't see these but they become apparent when we examine our images. After seeing the effects I now look for them in the sky. I think I see them, or is it my imagination? They are faint.
Not making an issue...been reading this thread and contemplating sending my 7D in for some "tweaking". My main issue, as others claim is highlight noise. Not sure if I'll call it banding; but I call it noise. Short and sweet. I feel this shot is properly exposed. Please correct me if I'm wrong. Love to learn. First photo is 100% crop of the RAW image right out of the box. All I did was down ressed to 100 resolution and applied Smart Sharpen amount 36, radius .3 and converted to sRGB. Second photo is to demonstrate the original, full size shot and camera settings, ISO etc. All exif information is intact for questions there.
I know sky shots are not the easiest for IQ. So, to keep it simple, two questions: is this to be expected as normal with this type of shot or is it a bit more noisy than your experience?
ziffl, the issue exists in camera and has been acknowledged by Canon. I fail to see what LR 2.7 vs 3.3 has to do with this. I never had this problem with my 40D (or my 10D for that matter).
Bobster2, I have a high end monitor and it gets calibrated at least once a month using Greytag Macbeth Eye-One.
I can't comment on how easy the problem might be to fix in pp, but the point is when one spends $1800 on a camera, one shouldn't have to fix every picture for this problem. It would be kind of like buying a brand new $40K car that pulls badly to the right and saying the solution is just to steer to the left all the time. True, but you shouldn't have to.
Just to clarify one thing. I have a persisting problem with noise above 800 ISO. I had a problem with quite noticeable vertical banding in highlights, but that has been fixed. Two very different and separate problems.
Why are the brightness and contrast settings set > 0? If you set that after import, then could we see the image as it was imported with no settings applied?
TeamSpeed wrote:
Why are the brightness and contrast settings set > 0? If you set that after import, then could we see the image as it was imported with no settings applied?
If your question is in referance to my post, these are the default settings for ACR 4.0, that I'm aware.
EverLearning wrote:
ziffl, the issue exists in camera and has been acknowledged by Canon. I fail to see what LR 2.7 vs 3.3 has to do with this. I never had this problem with my 40D (or my 10D for that matter).
Bobster2, I have a high end monitor and it gets calibrated at least once a month using Greytag Macbeth Eye-One.
I can't comment on how easy the problem might be to fix in pp, but the point is when one spends $1800 on a camera, one shouldn't have to fix every picture for this problem. It would be kind of like buying a brand new $40K car that pulls badly to the right and saying the solution is just to steer to the left all the time. True, but you shouldn't have to.
Just to clarify one thing. I have a persisting problem with noise above 800 ISO. I had a problem with quite noticeable vertical banding in highlights, but that has been fixed. Two very different and separate problems....Show more →
Ziffl is trying to be helpful, because it's a pretty well accepted fact that as it pertains to noise, the LR3.x versions made a significant and measurable difference when processing 7D RAW files vs. LR2.x.
I'm not saying this because I'm a LR user, but just because I've read too many posts all saying the same thing, that folks with 7D's and L2.x were generally pissed about noise performance when converting RAWs, but after LR3.x, opinions almost universally changed.
I don't have any threads handy, but you could try a search. There were some direct comparisons posted.
So, while it may or may not be your issue, I think trying the demo version of LR3.x could be worthwhile for you. You could experiment w/o having to drop any money.
mptnest wrote:
Not making an issue...been reading this thread and contemplating sending my 7D in for some "tweaking". My main issue, as others claim is highlight noise. Not sure if I'll call it banding; but I call it noise. Short and sweet. I feel this shot is properly exposed. Please correct me if I'm wrong. Love to learn. First photo is 100% crop of the RAW image right out of the box. All I did was down ressed to 100 resolution and applied Smart Sharpen amount 36, radius .3 and converted to sRGB. Second photo is to demonstrate the original, full size shot and camera settings, ISO etc. All exif information is intact for questions there.
I know sky shots are not the easiest for IQ. So, to keep it simple, two questions: is this to be expected as normal with this type of shot or is it a bit more noisy than your experience? ...Show more →
Empty Nester,
First off, let me say I'm likely not going to fix anything, because I'm not an expert, but here are some thoughts.
I think noise is excessive in that shot, yes, but a couple things.
It's not in critically sharp focus so that's not helping. It basically prevents any real post-processing from being done to fix things up a bit, and my guess is most of the awesome shots you see by all the great shooters on this forum are post-processed with some localized noise reduction, or in PetKal's case, shot with an 8MP APS-H sensor with nice big pixels Now I understand that we're talking about the noise in the sky, so you're not worried about getting the sky in focus, but if the bird was in focus, you could then do something about the noise in the sky.
The magnification on a 7D image is more extreme than any before it, so I've accepted that if I want to pixel peep and get what I consider critically sharp, I need to nail exposure, use really really sharp glass, nail focus, and then possibly apply noise-reduction, which I almost never do right now. I'm too lazy, but I'm ok with my laziness...
A couple of questions:
What lens were you shooting with? I see 300 @ f5.6, so I'm just guessing a 300 f4 IS?
What do you use to convert RAW's? Edit: I see you use ACR4.0. I'm not very familiar with RAW converters as I just use DPP. Along with being lazy, I'm also cheap. If you convert your RAW's with ACR4.0 and then also try the very latest DPP version, do things appear any better?
dwweiche, Ziffl may very well be trying to be helpful, but the different issues being discussed are being confused or blended. There is a banding issue and there is a ISO noise issue and more recently introduced into the conversation is the superiority of LR3 over LR2 for removing noise.
From what I have read, I have no reason to not believe that LR3 is better at reducing ISO noise than LR2, but neither version's noise reduction is going to clean up actual banding. With regards to the noise, it is good that there are options for cleaning up noise (I have Noise Ninja and am very happy with it) but the 7D was supposed to be a solid improvement in noise management and yet my 40D matches it. Even the best noise reduction program will cause some degree of image softening. So the best way of dealing with noise in an image is to not have it there in the first place. When a camera offers 12800 ISO, one would certainly expect quite clean images at 1600 ISO. I don't think it is unreasonable to expect this or a completely clean image at 800 ISO.
If I have come across as 'testy', I apologize. But I have spent a lot of money on gear, and I agonized over the buy/don't buy decision with the 7D. Ultimately, I bought it and I would call it one of the two or three worst photography purchases I have made in nearly 30 years. Again, please remember that THE main reason I bought it was for its supposed improved ISO performance and that I have/had a number of important trips that would benefit greatly from reduced noise at higher ISOs. For this reason, it is a complete bust. Adding to my frustration is the significant time I spent testing camera bodies, three separate times, and documenting my findings for Canon. Then the camera coming back twice with new problems. And as mentioned, I am not very pleased with the local retailer who had (to date) been the beneficiary of almost all my purchases.
BTW, and on another frustrating note, my FTP site isn't working. I spent well over an hour tonight trying to get it to work, but with no luck. I have the picture with banding, but no way to post it. It will be Thursday at best before I can try to fix it again. So unless somebody can suggest another way for me to post the banding shot, it's going to have to wait. Sorry.
Take a picture of the sky with the lens out of focus which should make the banding easier to see.
Or for that matter, take a picture indoors of a piece of paper, out of focus, exposure adjusted so the histogram is to the right, but not too far to the right that it comes close to clipping.
digitalbug30d wrote:
100 percent crops...geez you could go up to 300 percent and see Aliens....
since I wouldnt prin 100 percent crops..meh
My print engine uses 360 PPI for native resolution, and that easily makes for printing 100-150% crops, even when the whole image is printed. And banding is still visible in prints, but not as easliy as on a monitor. If you are not interested in that level of image quality, that is up to you.
I have taken a look at your crops, Daan, and do see what you are seeing the crops.
So, for gits and shiggles, I downloaded the RAW file expecting to see the same thing, and I can unequivocally say that this condition does NOT show up for me in:
1) DPP
2) LR 3.3
3) PS CS 5 (ACR 6.4, I think)
4) Capture One 5
... on my weekly-calibrated 27" monitor.
Perhaps this is something going on with your conversion software?
Geez Lance, talk about "gits and shiggles", your statement takes the cake! Based on the quote of mine you are commenting on, you have no way of knowing if I read one article on LR3 or if I am the world's most knowledgeable non-owner of LR3. Plus, despite me explaining it a multitude of times now, you still don't get it. One, highlight banding has nothing to do with noise management. Canon has already admitted they missed adjusting one of two dials on the sensor the first time it was for service. They made the adjustment and the banding went away. Two, a 7D owner should not have to use any form of noise management at 800 ISO, nor apply it heavily at 1600 ISO. Even the best noise management software will result in some degree of loss of sharpness. Why even offer 12800 if 3200 isn't usable? Three, Canon's correction of my banding and noise should not have thrown my WB way out of wack. And yes this too can be fixed with LR, but why should I have to adjust every picture? The WB was fine before their most recent adjustment.
This is a discussion of problems with the 7D. It is not a 'religious' war by Canon haters (I was thrilled with my 40D by the way) nor is it supposed to be an infomercial for LR3. And I truly am happy for those of you who are happy with your copy of the 7D, but it doesn't change the fact that some of us got bad copies.
Daan B wrote:
The thing that is holding me back right now, is the highlight banding I have seen in a number of 7D files.
This sort of stuff is drives many of us crazy. Some individual poster announces, in tones that suggest that everyone already knows this, that he sees is some critical, terrible, awful, deal-killing problem with some camera that makes it unsuitable for real photographic work.
Rather than stating that such a thing exists - which is arguable at the minimum - it might be a bit more appropriate to post something along the lines of "I've heard rumors of banding in highlights on the 7D and I think I may see it in this example. Does anyone else see it? Any explanations?"
This sometimes makes we wonder about the motives of people who post these things.
Now I understand that we're talking about the noise in the sky, so you're not worried about getting the sky in focus, but if the bird was in focus, you could then do something about the noise in the sky.