Canon EOS 5DsR raw processing issue in DPP

Printable View

10-25-2015, 02:53 PM
MiroslavMaric

1 Attachment(s)

Canon EOS 5DsR raw processing issue in DPP

I processed the same raw with DPP v4.3.3.1 and ACR v9.1.1. Check below. I guess it is moire effect or some very similar art effect visible in DPP only? In DPP I tested both Neutral and Standard picture Styles. Got the same. ACR is free of this art effect. Any idea why? Raw original file for processing/demo purposes only is available here, so please check with your raw converter. I used 16 bit tif conversion for both.

Attachment 156466

Thanks for comments!

Cheers,
Miro
10-25-2015, 03:28 PM
Don Lacy

Hi Miro, The DPP4 conversion is showing what looks like moire not 100% sure I get the same thing when I open the file.
10-25-2015, 03:31 PM
MiroslavMaric

Hi Don, thanks for the confirmation. Is it possible that Adobe includes some moire remover in ACR? I have no experience with moire effect.

Cheers,
Miro
10-25-2015, 03:36 PM
Don Lacy

Quote:

Originally Posted by MiroslavMaric

Hi Don, thanks for the confirmation. Is it possible that Adobe includes some moire remover in ACR? I have no experience with moire effect.

Cheers,
Miro

I am not sure I do not get the artifacts when using ACR which is strange and a problem since the current ACR profiles for the 5DS and 5DSR cameras are pretty bad if you have to use ACR I would try to find custom profiles online or make your own.
10-25-2015, 03:54 PM
MiroslavMaric

Quote:

Originally Posted by Don Lacy

and a problem since the current ACR profiles for the 5DS and 5DSR cameras are pretty bad if you have to use ACR I would try to find custom profiles online or make your own.

Thanks, I heard about ACR colors/tone processing issues & 5DSR. I have not tested ACR. Until now all my 5DSR photos have been processed in DPP and for the first time I have noticed moire effect.

Cheers,
Miro
10-25-2015, 11:37 PM
arash_hazeghi

Hi Miro,

I am not sure what you are talking about. In the sample you posted both ACR and DPP show moire in the upper left corner of the frame in the close crops. In DPP you see maze artifacts while in ACR you see maze artifact plus false color from moire so it's worse.

The 5DsR doesn't have a low-pass filter on the sensor so this is normal.

best
10-26-2015, 12:25 AM
MiroslavMaric

Hi Arash,

I know 5DsR doesn't have low-pass filter and moire effects is expected on some plumage details. IMO in above sample ACR color artifacts are less visible than in DPP. The overall color tone difference is obvious too but I have no tool to check it. Based on details of plumage, to further process the raw of this photo I would use ACR. I will have to check and compare how much printed photo is affected.

Cheers,
Miro
10-26-2015, 08:25 AM
David Stephens

Focusing on the upper left corner, I see more moire in the ACR than in the DPP conversion. What happens when you try to remove it?
10-26-2015, 08:56 AM
Don Lacy

I find this interesting I now see the moire in the ACR conversion but I see more color fringing in the DPP conversion which looks worse to me. I think these difference has more to do with how we see colors and patterns as individuals when I look at this image I see a white and gold dress even though it is blue and black http://www.wired.com/2015/02/science...es-color-dress
10-26-2015, 10:39 AM
arash_hazeghi

Quote:

Originally Posted by MiroslavMaric

Hi Arash,

I know 5DsR doesn't have low-pass filter and moire effects is expected on some plumage details. IMO in above sample ACR color artifacts are less visible than in DPP. The overall color tone difference is obvious too but I have no tool to check it. Based on details of plumage, to further process the raw of this photo I would use ACR. I will have to check and compare how much printed photo is affected.

Cheers,
Miro

Hi Miro

I Have no idea what you are talking about. Moire is quite visible in ACR (it has nothing to do with tone or color profile).

best
10-26-2015, 11:11 AM
arash_hazeghi

1 Attachment(s)

Quote:

Originally Posted by Don Lacy

I find this interesting I now see the moire in the ACR conversion but I see more color fringing in the DPP conversion which looks worse to me. I think these difference has more to do with how we see colors and patterns as individuals when I look at this image I see a white and gold dress even though it is blue and black http://www.wired.com/2015/02/science...es-color-dress

Hi Don,
here I blew up the crop, you can see the moire as well as the false colors in ACR output (red, purple and yellow)
10-26-2015, 11:51 AM
MiroslavMaric

Quote:

Originally Posted by Don Lacy

I find this interesting I now see the moire in the ACR conversion but I see more color fringing in the DPP conversion which looks worse to me.

Hi Don, yes, I agree, color fringing looks worse to me too.

Quote:

Originally Posted by David Stephens

Focusing on the upper left corner, I see more moire in the ACR than in the DPP conversion. What happens when you try to remove it?

Hi David, I didn't try to edit upper left corner. IMO the main issue is the rest of the sample - let say the name of this artifacts is color fringing.

Quote:

Originally Posted by arash_hazeghi

Hi Miro

I Have no idea what you are talking about. Moire is quite visible in ACR (it has nothing to do with tone or color profile).

Hi Arash, in addition to artifacts I'm talking about different color cast of ACR and DPP. I have no tool to check the right color cast. It's related to the camera color profile.
IMO the moire in left upper corner is a minor issue of ACR and the main issue is the rest of the sample with stronger color fringing visible in DPP.

I would like to see the sample of some 3rd raw processor.

Cheers,
Miro
10-26-2015, 01:19 PM
Don Lacy

Hi Arash, Yes I see it now I was not concentrating on that area but the brown area below it. I know there are issues with the ACR conversion with the 5DS cameras when it comes to color profiles . So would you recommend the 5DS over the 5DSR for Avian photography or is this a rare issue to worry about.
10-26-2015, 01:20 PM
arash_hazeghi

1 Attachment(s)

Hi Miro,

I think you are confusing several different technical terms

A) moire, this is a function of the sensor: it is present in both outputs but worse in ACR.
b) color fringing, aka chromatic aberrations, this is a function of the lens: neither sample shows CA.
c) color profile, color cast or WB, the outputs are different. Generally DPP's colors are closer to the original scene as seen by the eye this is because DPP can read the white balance information embedded in the RAW file. ACR often has strange color cast that you need to remove by profiling but not one profile works for all images.

Besides it seems like you turned off sharpness in DPP, the crop is very soft as a result, while the ACR crop has some sharpening applied to it. If you want to extract more details set sharpness to 2-3 and reduce the NR, (it applies too much be default)

I applied my recommended settings to your CR2 file and got the result below, it looks pretty good, you can try to remove moire in PS but I doubt if it will show in a print.

Attachment 156496

hope this helps
10-26-2015, 01:23 PM
arash_hazeghi

Quote:

Originally Posted by Don Lacy

Hi Arash, Yes I see it now I was not concentrating on that area but the brown area below it. I know there are issues with the ACR conversion with the 5DS cameras when it comes to color profiles . So would you recommend the 5DS over the 5DSR for Avian photography or is this a rare issue to worry about.

Hey Don, I think moire is a very minor issue, I doubt if it will be visible in large prints, and certainly not in online presentations I personally like the R version better. The only caveat is if you shoot video with it, moire can ruin the video and has no fix in that case...

best
10-26-2015, 02:36 PM
MiroslavMaric

1 Attachment(s)

Quote:

Originally Posted by arash_hazeghi

Hi Miro,

I think you are confusing several different technical terms

Hi Arash,

Ref to your points:

a) Moire is related to filter&sensor combo. We are talking about same file and raw processing by software not about different filters&sensors. IMO ACR shows a bit more moire than DPP.

b) My main complaint is that DPP sample shows magenta cast (call it somehow) easy visible on some pixel areas, at least 30% of the sample area is affected. I have no idea why but it seem to me that you ignore this artifact or you don't see it on your monitor. This is not related to the lens or/and chromatic aberration and I have never told this.

c) Color profiling is another story. Till now I have processed almost all my 5DsR RAW files (around 500) with DPP and didn't check them in ACR. DPP is my master tool for 5DsR raw processing.

I prefer soft sharpness in picture style mode and offered original RAW file for download. The OP DPP sample was processed in Neutral style mode with Sharpness parameters, Strength 0, Fineness 1 and Threshold 1. Everyone could check it and test if moire or magenta cast artifact are affected by level of sharpness. DPP follows picture style mode settings and ACR ignores it. Below is DPP sample processed in 3 picture style modes and ACR sample from OP (default ACR sharpness can be changed in ACR for preview only). I don't see the issue mentioned in point b is resolved in any picture style mode.

Attachment 156506

Cheers,
Miro
10-26-2015, 06:33 PM
arash_hazeghi

Hi Miro,

I don't understand what you are trying to prove here. You seem to talk about different things at the same time, sorry I can't quite follow.

As for your last images, you are using the incorrect settings for DPP. As I have mentioned before, you should NOT use the "unsharp mask" setting or mess up with the picture style. Instead use the "sharpness" parameter with standard picture style as I showed above, if you have the DPP4 guide it explains how to do this. Carefully note that in DPP, sharpness is not unsharp mask, it has no fineness or thereshold parameters that you mention above and try to adjust. It only has a single adjustment value.

Good luck with your choice of RAW conversion.

best regards