1DSIII red noise in ISO 800

[dwdallam]

Here is a full crop of a section of background and the coat of a subject.

ISO 800.
Exposure +.86 in LR2.
Fill light +60.
Some other minor adjustments.
Noise reduction OFF.

 Is this normal for shadow areas when pushed hard in this way? I mean the fill light is pushing hard, not the exposure. The reason I'm asking is because I remember some talking about red noise with their MKIIIs and it was something bad in camera they needed to have serviced.

[Josh-H]

Here is a full crop of a section of background and the coat of a subject.

ISO 800.
Exposure +.86 in LR2.
Fill light +60.
Some other minor adjustments.
Noise reduction OFF.

 Is this normal for shadow areas when pushed hard in this way? I mean the fill light is pushing hard, not the exposure. The reason I'm asking is because I remember some talking about red noise with their MKIIIs and it was something bad in camera they needed to have serviced.

Why did u turn NR off in LR?

At ISO800 with my 1DSMK3 in LR with LR default noise reduction I have no such color noise anywhere at all. At ISO 800 I normally apply a setting of 25 Luminance NR [leaving color at default 25] to just smooth out the slight luminance graininess on screen at 100%. I dont have noise like that in any of my files at ISO800 with LR default NR.

If I turn NR off in LR then there will be some color noise at ISO800 - usually worse in the shadows as your example shows although that is a particularly bad example]. This is also the case on my 5D as well as my colleagues D3X.

LR's default of 25 for color seems to do a very nice job of cleaning it up - Leave it turned on IMO.

Edit - I vaguely remember a post by Jeff Schewe about LR default Noise reduction - but cant remember the details. The gist of it 'from memory' is that LR's default of 25 color NR is basically a setting of zero. I may need to be corrected on this as I cant remember the details of the post.

Edit - I went back as I thought it may have been in the eLR 2.0 tutorial that Michael and Jeff did but it wasnt - so not sure where I read / heard it. Hopefully someone more knowledgeable chimes in.

[dwdallam]

Why did u turn NR off in LR?

At ISO800 with my 1DSMK3 in LR with LR default noise reduction I have no such color noise anywhere at all. At ISO 800 I normally apply a setting of 25 Luminance NR [leaving color at default 25] to just smooth out the slight luminance graininess on screen at 100%. I dont have noise like that in any of my files at ISO800 with LR default NR.

If I turn NR off in LR then there will be some color noise at ISO800 - usually worse in the shadows as your example shows although that is a particularly bad example]. This is also the case on my 5D as well as my colleagues D3X.

LR's default of 25 for color seems to do a very nice job of cleaning it up - Leave it turned on IMO.

Edit - I vaguely remember a post by Jeff Schewe about LR default Noise reduction - but cant remember the details. The gist of it 'from memory' is that LR's default of 25 color NR is basically a setting of zero. I may need to be corrected on this as I cant remember the details of the post.

Edit - I went back as I thought it may have been in the eLR 2.0 tutorial that Michael and Jeff did but it wasnt - so not sure where I read / heard it. Hopefully someone more knowledgeable chimes in.

Well I usually do either leave it at default or bump it up for specific shots. Even if I max it, it does not clear these red flecks. I just wanted to make them obvious, although they are obvious at any noise reduction. Here is the same crop with noise reduction in LR at 30/30.

[Panopeeper]

The only relevant aspect is the actual exposure of that part (I don't mean the shutter and aperture but the amount of light collected there). If those pixels are almost "black", then nothing on earth will help you to *avoid* the noise (of course you can reduce it, but that's a different issue).

If you upload the raw file, I take a look at it and perhaps make an experiment. Btw, what is the background? SOmething textured or smooth?

[dwdallam]

The only relevant aspect is the actual exposure of that part (I don't mean the shutter and aperture but the amount of light collected there). If those pixels are almost "black", then nothing on earth will help you to *avoid* the noise (of course you can reduce it, but that's a different issue).

If you upload the raw file, I take a look at it and perhaps make an experiment. Btw, what is the background? SOmething textured or smooth?

It's a black muslin. The file is under exposed in the shadows of course. I just wanted to know if the red flecks were normal under shadow conditions. Here is a link to the RAW file:

http://dwdallam.com/shared/

Look for the CR2 file in the list.

Thanks pano.

[Panopeeper]

Questions:

1. I don't see the red noise. What settings have you used? The best would be to upload the XMP file.

2. The texture of the cloth (muslin) is exactly like the row-column pattern of the basis noise. However, it is not excessively noisy.

3. Have you noticed the greenish blotches in the ACR rendering? They are observable not only on the background, but on the dark blue jacket as well. They vanish with chroma NR from the jacket, but not from the background.

4. Side issue: what are the pink and white/bluish sparkles on the jacket, close to the shoulder and the pocket? They are not hot pixels.

[dwdallam]

Questions:

1. I don't see the red noise. What settings have you used? The best would be to upload the XMP file.

2. The texture of the cloth (muslin) is exactly like the row-column pattern of the basis noise. However, it is not excessively noisy.

3. Have you noticed the greenish blotches in the ACR rendering? They are observable not only on the background, but on the dark blue jacket as well. They vanish with chroma NR from the jacket, but not from the background.

4. Side issue: what are the pink and white/bluish sparkles on the jacket, close to the shoulder and the pocket? They are not hot pixels.

4. I saw those too and I think they are just link on his jacket. The xmp file is in the same location. Maybe there is a problem with the green?

3. I thought his jacket was black?

I'm wondering, so are you saying that this looks like a correctly operating 1DSIII? If so, I'm happy with that.

The green doesn't show up in print at all, nor did I notice it.

[Panopeeper]

4. I saw those too and I think they are just link on his jacket. The xmp file is in the same location. Maybe there is a problem with the green?

3. I thought his jacket was black?

I'm wondering, so are you saying that this looks like a correctly operating 1DSIII? If so, I'm happy with that.

The green doesn't show up in print at all, nor did I notice it.

The file with the extention is not an XMP file. Why do you edit and save an XMP file with Internet Explorer? That file is useless as XMP.

I tried to achieve the effect like on the first crop you posted; it needs lots of fill light plus pushing, or lots of pushing. It's your decision, but if you push the almost total black, then you need to increase Blacks to suppress the noise - or turn on NR.

All in all, it is very normal. The stoff in that area (to the right of the man's shoulder) is in the 11th stop; lower it is even darker; that's not in the range of the camera.

I found it much more interesting, that there is posterisation on the top of the man's shoulder (if you don't push it so much that the dark blue jacket becomes medium blue).

[dwdallam]

The file with the extention is not an XMP file. Why do you edit and save an XMP file with Internet Explorer? That file is useless as XMP.

I tried to achieve the effect like on the first crop you posted; it needs lots of fill light plus pushing, or lots of pushing. It's your decision, but if you push the almost total black, then you need to increase Blacks to suppress the noise - or turn on NR.

All in all, it is very normal. The stoff in that area (to the right of the man's shoulder) is in the 11th stop; lower it is even darker; that's not in the range of the camera.

I found it much more interesting, that there is posterisation on the top of the man's shoulder (if you don't push it so much that the dark blue jacket becomes medium blue).

OK thanks for the analysis. The xmp file is an xmp but my ftp program probably "auto" transferred it. I should have manually changed it to binary. What's the deal with the posterization? I've seen that in come B&W conversions I've recently done with some hard pushing into the blacks and highlights at the transition points between the contrast. Is this a possible camera defect?

[james_elliot]

I took the CR2 image and converted it to JPEG with Dxo. Exposure+0.85, Luminance+5. No NR.
I can't see any unpleasant artifact of any kind.
However I think that DxO is superior to ACR (I made some tests here).

You can download your converted picture there:
http://www.photo-lovers.org/portfolio/priv...rs/dwdallam.jpg
If you want me to try other parameters with DxO, let me know.

Friendly

What's the deal with the posterization? I've seen that in come B&W conversions I've recently done with some hard pushing into the blacks and highlights at the transition points between the contrast. Is this a possible camera defect?

[Panopeeper]

I took the CR2 image and converted it to JPEG with Dxo. Exposure+0.85, Luminance+5. No NR.
I can't see any unpleasant artifact of any kind.

You should have taken a look at the crop dwdallam attached in the first post. He made a very different rendering, without cutting off the noise by black level. Apparently his goal was to see the noise, while your goal was to hide it.

The "posterization" is present in the DxO rendering as well, but now I think it is simply the changing illumination on the jacket.

I think that DxO is superior to ACR

That is possible, but the above sample is no proof of that.

[Panopeeper]

What's the deal with the posterization? I've seen that in come B&W conversions I've recently done with some hard pushing into the blacks and highlights at the transition points between the contrast. Is this a possible camera defect?

I am not sure that this is really posterization; the jacket is slightly folding, its patches are in different angle to the light source. I see those shades in all three raw channels; this indicates, that it is not posterization. Perhaps you have better examples.

[james_elliot]

You should have taken a look at the crop dwdallam attached in the first post.

I did.

He made a very different rendering, without cutting off the noise by black level. Apparently his goal was to see the noise, while your goal was to hide it.

The initial settings for ACR given by dwdallam were Exposure +0.86, Fill Light+60, and that's why I set Exposure+0.86 and luminance +5 in DxO, and let the rest OFF.
I didn't set any black level of any kind and didn't try to hide any noise.
Be more specific and tell me what DxO settings in your opinion match dwdallam's ACR settings and I will be happy to try.

[Panopeeper]

The initial settings for ACR given by dwdallam were Exposure +0.86, Fill Light+60

1. Those settings are not reflecting the rendering the crop dwdallam posted; I tried a conversion, the result is different. The settings contain luminance and color noise reduction 30, while dwdallam wrote "noise reduction OFF", so I turned off the NR, but still, the result is not the same as the posted JPEG.

2. ACR applies +0.4 EV exposure adjustment to the 5D2 images; this is not visible, you have to apply -0.40 EV in order to cancel it. This means, that the exposure adjustment would be 1.26.

3. I have no idea how one could transform these settings in DxO parameters.

The point is still, that dwdallam did not try to optimize the outcome; he showed the noise in its almost full glory, although it could have been even worse by setting Blacks to 0.

I have no reason to doubt your assertion, that DxO creates a better output from 5D2 (or from Canon) raw files than ACR. All I said is, that this situation is not suitable to prove that. If I wanted to compare the converters, I would define some outcomes: how bright certain parts should be, how much can be annihilated by black point; then making the best conversion with both converters, once with and once w/o noise reduction. However, I personally am not interested now in the outcome; my interest was in the raw data (how noisy it is in reality).

[james_elliot]

I have no reason to doubt your assertion, that DxO creates a better output from 5D2 (or from Canon) raw files than ACR. All I said is, that this situation is not suitable to prove that. If I wanted to compare the converters, I would define some outcomes: how bright certain parts should be, how much can be annihilated by black point; then making the best conversion with both converters, once with and once w/o noise reduction. However, I personally am not interested now in the outcome; my interest was in the raw data (how noisy it is in reality).

I just checked again and black point is set to 0. I also used 1.2 as exposure compensation and the amount of noise is almost not altered in the output. Noise reduction is really set to 0, both for luminance and for chroma noise.

I think however that you are underestimating one point. Measuring the noise in the RAW channels (something I also do here) doesn't say everything about the noise after demosaicing.
Many algorithms (vector median was one of the first) do not use anymore the value of the pixel itself in the bayer grid (I mean for example that if we are rebuilding the (R,G,B ) values for a green pixel of the Bayer array whose value is v, we don't get G=v for that pixel after demosaicing).
As noise is a stochastic process, this can have for example an averaging effect on noise, and thus reduce the most important and visually unpleasant values.
Noise in the output depends of course on the noise in the RAW channels, but also on the demosaicing algorithm, even if no specific noise reduction is applied.

PS: Btw, I have no special interest in DxO, I even don't like some of its features such as, for example,  their perspective correction algorithm or their specific noise reduction algorithm.

[dwdallam]

I took the CR2 image and converted it to JPEG with Dxo. Exposure+0.85, Luminance+5. No NR.
I can't see any unpleasant artifact of any kind.
However I think that DxO is superior to ACR (I made some tests here).

You can download your converted picture there:
http://www.photo-lovers.org/portfolio/priv...rs/dwdallam.jpg
If you want me to try other parameters with DxO, let me know.

Friendly

James thanks. No worries as long as the camera is functioning properly I'm not too anal about converters.

[dwdallam]

Pano--thanks for your work here. I'm very satisfied that the camera shows normal operation. You're a great asset to this forum. Thanks again.