No CA removal in DPP?

sparkie

Leading Member
Messages
783
Solutions
1
Reaction score
33
Location
NY, US
Can't seem to find any CA adjustment sliders on DPP. And I don't want to use CS ACR as the colour and shadow detail RAW conversions are as refined as DPPs (IMO)

TIA
 
And I don't
want to use CS ACR as the colour and shadow detail RAW conversions
are as refined as DPPs (IMO)
Are you talking about CS (ACR 2.4) or CS2 (ACR 3.1)? I find CS2 gives me the best RAW conversions ever. But unfortunately it needs to be calibrated, as the default settings are way off. To calibrate, you need a Gretag/MacBeth color checker and this script:
http://fors.net/scripts/ACR-Calibrator/

This older tutorial may also be worth your while:
http://www.creativepro.com/story/feature/21351-1.html

Regards
Stefan

--

»We've experienced the fact that the perceptions of an expert surpass the precision of measuring instruments.« Nakabayashi-san of Canon, Inc.
 
DPP also does it automatically if you select the medium or high
quality setting, but it doesn't seem to be as successful as EVU.
--
http://edwardkaraa.shutterpoint.com
http://pbase.com/edwardkaraa
...in RAW mode in DPP? -- I process in RAW at the highest tiff
output setting. I wasn't wasre that DPP or EVU for that matter did
CA automatically. are you sure about this?
Yes very sure about it. Just try to convert a file containing CA with EVU, DPP and ACR (without CA correction) and you will see the difference. I believe C1 does it as well.

--
http://edwardkaraa.shutterpoint.com
http://pbase.com/edwardkaraa
 
When using the system described in the links, how sensitive did you find it to the lighting condition that the ColorChecker was shot in. I know they (ColorChecker's) are supposed to be pretty light condition flexible, but was more than one "profile" needed?
 
Yes very sure about it. Just try to convert a file containing CA
with EVU, DPP and ACR (without CA correction) and you will see the
difference. I believe C1 does it as well.
There is absolutely no lateral CA correction in EVU like there is
in ACR. Absolutely positively not.

Jason
EVU has automatic CA detection and correction. There are no sliders like in ACR, and it works quite well in some situations and it doesn't in others. God I have to be so patient on these forums.
--
http://edwardkaraa.shutterpoint.com
http://pbase.com/edwardkaraa
 
The script does not create a profile, which is OK since ACR still doesn't support custom profiles. It calibrates the color defaults, that's all. I used studio strobes to photograph the color checker. For me, those defaults are not etched in stone once calibrated - I "reserve the right" to tweak them as I go, depending on many factors of an image, including light source (but after neutralizing white balance of course). This visual tweaking is highly subjective and I readily make use of it.

Regards
Stefan

--

»We've experienced the fact that the perceptions of an expert surpass the precision of measuring instruments.« Nakabayashi-san of Canon, Inc.
 
...before you start snapping at each others' heels, please make sure you're talking about the same "CA" first. Could it be that one person means "Chromatic Aberration" and the other person thinks CA stands for "Color Aliasing"?

I believe there is no way to remove chromatic aberration whith any rate of success automatically, without user input, while automatic detection works pretty well on color aliasing. That's what's happening in most of today's RAW converters.

Regards
Stefan

--

»We've experienced the fact that the perceptions of an expert surpass the precision of measuring instruments.« Nakabayashi-san of Canon, Inc.
 
...before you start snapping at each others' heels, please make
sure you're talking about the same "CA" first. Could it be that one
person means "Chromatic Aberration" and the other person thinks CA
stands for "Color Aliasing"?
I believe there is no way to remove chromatic aberration whith any
rate of success automatically, without user input, while automatic
detection works pretty well on color aliasing. That's what's
happening in most of today's RAW converters.

Regards
Stefan

--
»We've experienced the fact that the perceptions of an expert
surpass the precision of measuring instruments.« Nakabayashi-san of
Canon, Inc.
 
No I am talking about CA as in Chromatic Aberrations. Manual controls allow you to remove them completely. But if you do a simple comparison with the same file processed in EVU and DPP, you will find that EVU succeeds in removing a large part of it in some cases, and very little in some other cases. But EVU results are always better than DPP in this respect.
...before you start snapping at each others' heels, please make
sure you're talking about the same "CA" first. Could it be that one
person means "Chromatic Aberration" and the other person thinks CA
stands for "Color Aliasing"?
I believe there is no way to remove chromatic aberration whith any
rate of success automatically, without user input, while automatic
detection works pretty well on color aliasing. That's what's
happening in most of today's RAW converters.

Regards
Stefan

--
»We've experienced the fact that the perceptions of an expert
surpass the precision of measuring instruments.« Nakabayashi-san of
Canon, Inc.
--
http://edwardkaraa.shutterpoint.com
http://pbase.com/edwardkaraa
 
huh? my original post was about Chromatic aberation correction. we are talking about the same thing.

I dont use EVU, as I think DPP is a better RAW converter. I personally do not think DPP does automatic Chromatic aberation removal, if it does it is just so minimal I dont notice it. As when i open up the file in CS the CA is very evident in the image.

Canon should have included manual sliders for CA removal just like CS ACR. I hope they will in the next release of DPP.
...before you start snapping at each others' heels, please make
sure you're talking about the same "CA" first. Could it be that one
person means "Chromatic Aberration" and the other person thinks CA
stands for "Color Aliasing"?
I believe there is no way to remove chromatic aberration whith any
rate of success automatically, without user input, while automatic
detection works pretty well on color aliasing. That's what's
happening in most of today's RAW converters.

Regards
Stefan

--
»We've experienced the fact that the perceptions of an expert
surpass the precision of measuring instruments.« Nakabayashi-san of
Canon, Inc.
--
http://edwardkaraa.shutterpoint.com
http://pbase.com/edwardkaraa
 
Yes, unfortunately the automatic removal in DPP is very minimal but you can verify it by processing the same file twice, once with high or standard quality setting, and another with the speed setting. And, yes, the least Canon can do is include the manual sliders in a so called professional software, instead of the useless stamp tool which doesn't work. However, do not underestimate EVU. It has much better color rendition and sharpness control than DPP, and whatever you can do with the RGB curve, you can do it better with the tone curve in EVU.
huh? my original post was about Chromatic aberation correction. we
are talking about the same thing.

I dont use EVU, as I think DPP is a better RAW converter. I
personally do not think DPP does automatic Chromatic aberation
removal, if it does it is just so minimal I dont notice it. As when
i open up the file in CS the CA is very evident in the image.

Canon should have included manual sliders for CA removal just like
CS ACR. I hope they will in the next release of DPP.

--
http://edwardkaraa.shutterpoint.com
http://pbase.com/edwardkaraa
 
EVU has automatic CA detection and correction. There are no sliders
like in ACR, and it works quite well in some situations and it
doesn't in others. God I have to be so patient on these forums.
You are wrong. If you're going be longitudinal type CA showing up more in DPP than EVU, simply look to the excessively noise reduced and blurred output of EVU. Of course, we're talking about transverse or lateral CA which is the ONLY CA ACR and ACR simply resolves these details that EVU does not. If you go to the DXO page you will see they also do not claim to remove longitudinal CA as it is basically impossible. All you can do is fun filters to remove some of it, not actually correct it since sensors do not have 3D information to do so.

If you don't believe me just email Westfall of Canon and you'll get your answer. DPP is the professional solution from Canon, if they ever had CA correction it will get it first.

Jason
 
do not claim to remove longitudinal CA
as it is basically impossible.
That would be my line of thinking, too. Perhaps not impossible, but extremely difficult for an automation process, which would risk to mistake other artefacts for CA (e. g. sensor blooming) and therefore do a lot of "automatic damage" trying to remove it.

The fact that CA can come in all colors of the rainbow, depending on the lens used, and is often asymmetrical, adds to the difficulty.

Regards
Stefan

--

»We've experienced the fact that the perceptions of an expert surpass the precision of measuring instruments.« Nakabayashi-san of Canon, Inc.
 
That would be my line of thinking, too. Perhaps not impossible, but
extremely difficult for an automation process, which would risk to
mistake other artefacts for CA (e. g. sensor blooming) and
therefore do a lot of "automatic damage" trying to remove it.
It really depends on your definition of "correct." A good place to look is the DXO page:

http://www.dxo.com/en/photo/support/form_pb_img_quality.php

Note they have more info most about the lenses than most, yet they can only correct lateral CA but not the other forms.

This is because lateral CA is a very linear things. The red that was supposed to land on this pixel lands on this pixel. Just move the color info over and it is truly corrected.

Longitudinal CA on the other hand means the red (for example) is focused on an entirely different plane than the sensor is on. Now the question is can any program fix an out of focus image? Well deconvolution can do some neat things, but we can all imagine how such a selective deconvolution could get VERY messy quickly. And when you're done deconvolving an out of focus image (at least to today's standards) anybody with a critical eye wouldn't say the image is truly corrected, just patched up. Instead of correcting longitudinal CA most actions just do some selective reduction or filtering, not true correction which would actually yield a much sharper image than the original.

Back to lateral CA correction like ACR. It would be absolutely ludicrous for Canon to have an automatic CA correction utility and a huge database of lenses and not brag about it. This feature otherwise costs a fortune and would have required a huge amount of work to implement.

Jason
 
I just have to disagree with your findings concerning CA correction.

The difference I see between EVU and DPP is in the finest details which is the main reason I use DPP instead of EVU. Where EVU will tend to blur the boundaries between high contrast areas, which might explain why the CA is reduced, DPP will sometimes enahance it. I've seen this with blooming and other color artifacts.

As far as sharpening is concerned, how exactly does EVU give you "much better" sharpness control over DPP. Other than actually seeing the sharpening on screen (alone with those dreaded halos) I find DPP's sharpening preferable by a long shot.

Ted
huh? my original post was about Chromatic aberation correction. we
are talking about the same thing.

I dont use EVU, as I think DPP is a better RAW converter. I
personally do not think DPP does automatic Chromatic aberation
removal, if it does it is just so minimal I dont notice it. As when
i open up the file in CS the CA is very evident in the image.

Canon should have included manual sliders for CA removal just like
CS ACR. I hope they will in the next release of DPP.

--
http://edwardkaraa.shutterpoint.com
http://pbase.com/edwardkaraa
 

Keyboard shortcuts

Back
Top