I recently noticed some red noise patches to appear in the shadows and blacks of ALL my photos.

Below is a link of sample shot with the problem im referring to:

http://www.flickr.com/photos/ctpcheng/8068082339/in/photostream

Feel free to download the original or view it at the original size!

If you pixel peep at 1:1, you will notice the red noise i was referring to around his grey jacket and back pack. After noticing this problem, I've started seeing it EVERYWHERE on all my shots. I was shooting a birthday party in low light yesterday and the grey curtains were just plagued with this noise.

I sent a sample image to one of my professional friends who also uses the same camera and she said it might be a lens issue because she hasn't experience this before or has heard of this happening to anyone. Most of my shots have been done with the 35L lately so ill have to switch up my lens and run more tests...

So do i have a defective sensor? is this happening to everyone? suggestions?

Heres one of the first images i took with the camera:

http://www.flickr.com/photos/ctpcheng/7941965494/in/photostream/

If you look at the shadows near the top center part of the picture, you can see that red noise and lots of vertical banding...

Looking just at the first image, it looks pretty normal to me. (Have not had a chance to check out the second yet).

My suggestions:

1. Stop pixel peeping and enjoy that excellent camera.

2. Don't take static shots at 1/3200 second and ISO 400.

3.If some level of noise bothers you, use a little bit of noise reduction in your Lightroom 4. The 'Color' slider should do it for this image, but you should learn the full capabilities of NR in LR 4.

I do enjoy replies like that. So helpful.

My the door on my Mercedes benz squeeks when i shut it - Ignore it and enjoy the car!

Right... Its a $3500 camera and he is seeing a problem. For a camera, a bad sensor is pretty much the worst thing.

On the other hand, I dont see any banding/noise in the sizes I looked at. Some noise on a sensor is expected in the darkest shadows.

I definitely see the vertical banding. Did you do anything to try and pull any detail from the shadows, or is this truly an untouched original?

kodakeos: When I wrote my first post I just looked at the first image - that is all that he had posted. I did not see any unusual noise in that - did you? (By the time I got my post in, he had posted the second image. But I did not see this when I made my first post).

I did feel, based on just looking at this first image only, that it was pretty normal for a 5D Mk III (I have one, as I see you do also). I believe my suggestion to not take a shot such as this at 1/3200 second and ISO 400 was appropriate and helpful.

I also noted that this is his first and second post on FM. Based on my observation of his first image, I feel he is being overly critical (again based on my experience with a 5D Mk III). Not at all like you analogy of a squeak in the door of a Mercedes.

Your apologies for jumping all over my case with out knowing all the facts would be appreciated.

PS: Looking at the second image, I agree something looks strange there, and as Ben noted I wonder if the image was underexposed and pulled up in post-processing.

Welcome to FM, xopher (your first two posts).

xopher wrote: I sent a sample image to one of my professional friends who also uses the same camera

xopher wrote: If you look at the shadows near the top center part of the picture

I suspect that the camera is not defective. The way to check is to make some side by side shots with someone who has the same camera.

Calgary is a large enough city that somebody on FM in Calgary with a 5D3 must be willing to help out.

If you make shots with the same lens from the same position (two carefully aligned tripods or one tripod with the two cameras aligned by checking the corners of the frame in the viewfinder) with the same ISO, shutter speed, an aperture, then you'll be able to look for differences.

If your technique is really good (using a single tripod in the exact same spot), you'll be able to manually align the images in Photoshop and use the Difference layer treatment, exaggerated by a super S-shaped curve.

Shoot test shots at all ISOs. Use the same aperture for all the shots: f/8. Adjust the shutterspeed as you adjust the ISO. Be sure to have proper exposure (your two posted shots seem to be good in that regard). Use a cable release (remote release) or use the 10 second self-timer. Use mirror lockup. Wait at least 3 seconds after mirror up for vibrations to damp down.

Take it outdoors on a sunny day with the sun behind you or over one shoulder, between the hours of 10 am and 2 pm for the most repeatable consistent constant lighting.

The point of all the rigorous technique is to eliminate as many variables as possible.

No camera is perfect. It is a truism of testing that if you go looking for problems you will find them and then you have to determine how big the problem is and what is the cause -- usually it is experimental error.

I think if I had your camera, based on those images, I would be very happy with the camera. I think if you printed those images quite large (20 x 30 inches) you'd not see your 'problem' and you'd be quite happy with the camera. Pixel peeping at 100% is equivalent to even larger.

What is the largest you have ever printed a DSLR picture?

Can someone in Calgary help xopher do a comparison?

These are perfectly normal with 5d2 and 5d3. Shooting with canon sensor and fast lenses, those shadows noise are expected. Your below image has shown very clear what you described
http://farm9.staticflickr.com/8034/8014286718_ff49200cc6_o.jpg

be patient, canon will fix it some day...no need for comparison...just google for zillion post about these shadows issues.

Here is a black frame with a 0.3 sec exposure at ISO 400 that ought to show hot pixels. I can see banding if I pull it up about 3.5-4 stops, but no hot pixels standout.

If you do a similar test and see hot pixels, I'd say its not normal. This is my second 5D MK III and none had hot pixels.

scalesusa wrote:
If you do a similar test and see hot pixels, I'd say its not normal. This is my second 5D MK III and none had hot pixels.

Maybe he should also try the sensor clean reset similar to the 5d2? Put the lens cap on, put it in sensor clean mode and wait about a minute before turning it off. In a lot of cases where a 5d2 developed hot pixels, this would clear them out.

Would also check if "Highlight Tone Priority" and "Auto Lighting Optimizer" are on. If these are JPEGs out of the camera, check the picture styles, compression ratio, etc.

If you are going to go hunting for stuck pixels be aware the photoshop and Lightroom raw processors have hot pixel supression, so do not compare across imaging apps.

mttran wrote:
These are perfectly normal with 5d2 and 5d3. Shooting with canon sensor and fast lenses, those shadows noise are expected. Your below image has shown very clear what you described
http://farm9.staticflickr.com/8034/8014286718_ff49200cc6_o.jpg

be patient, canon will fix it some day...no need for comparison...just google for zillion post about these shadows issues.

+1

Put your flame suit on.

alundeb wrote:
mttran wrote:
These are perfectly normal with 5d2 and 5d3. Shooting with canon sensor and fast lenses, those shadows noise are expected. Your below image has shown very clear what you described
http://farm9.staticflickr.com/8034/8014286718_ff49200cc6_o.jpg

be patient, canon will fix it some day...no need for comparison...just google for zillion post about these shadows issues.

+1

Put your flame suit on.


Why, isn't it the canon facts.... ...the question i have: Is it the combination of system noise, DR and on board AWB process that causing these effects? the 14 bit system seems amplifying these effects more than older 12 bit one...that what i see from my 1ds2 and 5d2.

Threads like this one, we need heavy duty decoders to chimp in so we all can understand the logic behind these issues.

You know, it goes like this:

1) Someone sees unexpected noise / banding in the images.
2) He goes to a photo forum to ask about the problem.
3) People suggest that the sensor may be bad, and start comparing images.
4) More people chime in and say they never see this on their own sensor.
5) The thread starter says that he doesn't see it with his own sensor either in most images.
6) Some technical people suggest a testing method with a black frame to isolate the read noise signature.
7) Comparison among many sensors show that the read noise signature varies little between sensors. All have it.
8) Some try to find out what provokes it in images, but don't come to any conclusion.
9) Then there are more or less helpful suggestions:
10) You are not supposed to see it on a calibrated monitor or in prints.
11) The problem does not exist. It is not in my camera. It is invented by obsessive forum members.
12) Other Canon cameras are not any better. Get over it.
13) Get a Nikon or Sony Camera with an EXMOR sensor.

alundeb wrote:
You know, it goes like this:

1) Someone sees unexpected noise / banding in the images.
2) He goes to a photo forum to ask about the problem.
3) People suggest that the sensor may be bad, and start comparing images.
4) More people chime in and say they never see this on their own sensor.
5) The thread starter says that he doesn't see it with his own sensor either in most images.
6) Some technical people suggest a testing method with a black frame to isolate the read noise signature.
7) Comparison among many sensors show that the read noise signature varies little between sensors. All have it.
8) Some try to find out what provokes it in images, but don't come to any conclusion.
9) Then there are more or less helpful suggestions:
10) You are not supposed to see it on a calibrated monitor or in prints.
11) The problem does not exist. It is not in my camera. It is invented by obsessive forum members.
12) Other Canon cameras are not any better. Get over it.
13) Get a Nikon or Sony Camera with an EXMOR sensor.


On item 13, you mean..." Go get Nikon or Sony Camera ...., Bye"

Is there a on-line forum with pure technical and fun stuffs so we can relax and learn

In addition to "normal" noise, what I am seeing is 35L @ f/1.4 CA/fringing in the powdered sugar, which is pretty much to be expected.

Both shots are with the 35L @ f/1/4, so maybe it would be better to choose some other examples that eliminate the CA/fringing factor.

John_T wrote:
In addition to "normal" noise, what I am seeing is 35L @ f/1.4 CA/fringing in the powdered sugar, which is pretty much to be expected.

Both shots are with the 35L @ f/1/4, so maybe it would be better to choose some other examples that eliminate the CA/fringing factor.

We have seen so many 5d2 reported images past 4 years and CA was not the main part of these issues.

the only unusual noise i see here is in this thread

i think your camera is functioning properly.

That waffle shot must have the exposure pushed or there's something wrong with the camera.

goosemang wrote:
the only unusual noise i see here is in this thread

-1, most canon shooter know these 5d2/3 shadows issues at wide DR scenes. Here is couple from OP website @ iso 100:

http://farm9.staticflickr.com/8034/8014286718_ff49200cc6_o.jpg
http://farm9.staticflickr.com/8039/8014499257_9831218f38_o.jpg
http://farm9.staticflickr.com/8296/7941965494_4dedc809da_o.jpg