is it possible that nobody noticed it? I used to use LR for managing all my photos, as i only shoot RAW. recently i've changed the camera though to Canon's 50D. I was looking forward to it's 7.1Mpx sRAW mode dor casual photos and full resolution for bigger projects. Unfortunately it lasted very shortly. Is it possible that nobody noticed strong magenta cast on bigger (7.1) sRAW files? Acctually if you shoot three photos - RAW, sRAW1 and sRAW2 each of them will look somehow different. Canon has already corrected this mistake in their DPP (it looked exactly the same there) but i still want to use LR for all the worklflow and things. Now i'm forced to use huge RAWs (at least 20MB each) instead of handy sRAWs.
Please do something with it,
ps. i'm writing this as the issue still hasn't changed even though ACR 5.4 RC was released :/
i'm not saying that it's broken - it's simply isn't working properly on 50D'd sRAWs - that's all. and i know it because i downloaded RC version this morning and the situation didn't change. i'm using Lightroom because LR and ACR are more or less the same thing as far as working on RAWs is concerned. My point is: if Canon could change the way sRAWs are processed in their DPP why Adobe can't? because of that i'm put in a little bit unconvenient situation where i have to decied whether i should buy huge CF cards - one photo is over 20MB in RAW - or use other software, which I find worse. I decided to write about it as the new update for ACR is coming out and probably the same thing will happen with LR - as it always did in past.
i'd love to get some news straight from adobe, as this issue seems to be 100% software, whether they're planing to do something with it or not.
Can you provide an example sequence of 3 images of the same scene:
- full raw
- sraw mode 1
- sraw mode 2
(We need the raw & sraw files, not jpegs/tiffs.) You can upload to a site
like yousendit.com. You can also email me at email@example.com if you
don't want to share that link publicly.
Please respond to Eric's request for sample files.
You say you use LR, and you downloaded Camera Raw 5.4RC. Just to be absolutely clear, are you aware that downloading Camera Raw has absolutely no effect on results when using LR? LR contains its own internal version of Camera Raw, and does use the plug-in. What version of LR are you using?
effect is clearly visible especially on human skin. original RAW looks exactly like jpeg straight from the camera, excluding NR and such things. file 0254 is 7.1Mpx sRAW
also one guy started discussion on other forum about this issue over here: http://photography-on-the.net/forum/showthread.php?t=692384
thanks for the concern,
and the files:
Thanks for the files. This is helpful. BTW, which firmware version of the
50D do you have?
effect is clearly visible especially on human skin. original RAW looks
exactly like jpeg straight from the camera, excluding NR and such things.
file 0254 is 7.1Mpx sRAW
also one guy started discussion on other forum about this issue over here:
thanks for the concern,
Perhaps a bit more detail:
It's proving to be a real nuisance on a wedding I shot recently. sRAWs files (and at least a couple RAW files) show a severe purple color cast in deep shadow regions or underexposed shots when those regions are recovered in Lightroom. And it's not confined to 800ISO & up, one shot had it at 200ISO :-(. The result is that the white balance of the picture gets badly compromised (the one picture has both a purple color cast, as well as a green).
I'm using the Camera Raw profiles (beta 2), but I'm twiddling my thumbs what to do about this. Digital Photo Professional develops the same pictures perfectly.
Any suggestions how to solve this would be most welcome. The firmware i'm using is 1.06.
Here's two sRAW1 shots for reference:
ps. you're probably aware of it, but here's a couple other links to those with similar problems:
Which version of Camera Raw are you using?
It's proving to be a real nuisance on a wedding I shot recently. sRAWs files
(and at least a couple RAW files) show a severe purple color cast in deep
shadow regions or underexposed shots when those regions are recovered. And
it's not confined to 800ISO & up, one shot had it at 200ISO :-(. The result
is that the white balance of the picture gets badly compromised (the one
picture has both a purple color cast, as well as a green).
I'm using the Camera Raw profiles (beta 2), but I'm twiddling my thumbs what
to do about this. Digital Photo Professional develops the same pictures
Any suggestions how to solve this would be most welcome. The firmware i'm
using is 1.06.
Here's two sRAW1 shots for reference:
You should see an improvement in the next dot update of LR, coming soon.
I'm on Lightroom 2.3, and the splash screen indicates that its Camera Raw 5.3
(I've tried to update lightroom, but it says it's currently up-to-date.
It will be released when it's ready. Keep checking this forumd and/or the Lightroom forum.
If you're lucky, the Adobe Auto Updater will notify you. If it does, buy a lottery ticket, you're on an extremely lucky roll.
"Soon" is all I can say. It will be announced here in the forums when it's
out, as well as the usual photo sites (e.g., dpreview).
I'm glad to hear it--I'll look forward to its release.
Any idea when to expect it--Is there a way to get notified when it's ready?
OK, just a brief edit to say that there still seems to be a bit of magenta cast when using the shadow recovery tool highter than about 60. I know that's pretty high, but it would be nice to have it perfect :-)
Apart from that, and for normal uses, it's a huge improvement.
Can you provide example sRAW mode 2 and full raw .CR2 files? After doing a
lot of testing with these since firmware 1.05, and all the way thru 1.07, I
have not seen any magenta cast discrepancies on the unit that I have.
Sure. It's fairly repeatable.
I took them at 800 ISO, as it becomes more evident at higher ISO's & just shot off three at our couch (RAW, sRAW1 & sRAW2)
It's the sRAW2 one that still has the magenta cast in the shadows.
Camera firmware is 1.07
Let me know if you need anything else. Here's the download for the zip file:
Sorry just noticed you only asked for RAW mode 2 & RAW.
Here's a couple more scenes @ISO400 in both modes. The wierd thing is Lightroom also displays them as though there's an exposure difference (DPP does not & they were shot with a flash set to manual). Plus there's the magenta cast in the shadows...
I also have the problem described here.
I have a Canon EOS 5D Mk2.
Photos taken at ISO200 and the smaller RAW format exhibit this issue.
All photos that were taken this year and processed with ACR 6.4 and 6.4.1 have exhibited this issue. I don't think this issue used to happen with the original version of ACR that came with Photoshop CS5.
I updated my camera firmware 2 years ago, but that is too long ago to have caused this problem as I was getting good results from it immediately after the firmware update.
In ACR, changing the Camera profile to any setting other than "Adobe Standard" helped. Also the slider below it named shadows helped when I moved it towards the green (left) side.
However it was difficult to get a satisfactory looking color balance overall, and there were still noisy magenta artefacts in the underexposed shadows.
Installed Canon Digital Photo Professional (188.8.131.52) on my PC today and compared the images. Wow! DPP ones look stunning on the default settings compared to the ACR processed ones. No color cast at all.
So I will return to using Canon DPP and then use PTLens to correct distortion, and manually correct chromatic aberration. Less automation but better quality results.
If this issue is addressed I'm happy to help test ACR.
Here are sample RAW files (CR2 and XMP) exhibiting the issue described:
With default parameters I see a slight difference between the embedded JPEG and the Camera Raw conversion, bug if anything the Camera Raw conversion seems a bit better balanced, with the embedded JPEG a bit too cyan... Is this the magnitude of difference that you're talking about? If so, it seems just a quick white balance away from pretty good color... I do see some issues in the deep shadows.
Please post a small version of the image with the color you like as converted through DPP.
Rich, I noticed in your files that white balance was set to daylight. Is the WB in your camera set to daylight? In camera raw, try using auto or as shot and let me know what you see. (leave camera profile on adobe standard). I also noticed you have firmware 2.0.7. You might try and update your firmware to 2.0.9 (the latest) and see if this issue persists. Let me know!
I previously spent 6 or 7 hours experimenting with different color balance settings in ACR, and processes also. I believe I have tried all the possible combinations and none make an appreciable difference to the result. My original post shows the measures that I found a midly effective in ACR (and it wasn't colour balance).
On the camera I have shot hundreds of photos as exposure bracketed panoramas on manual settings, ensuring that the white balance, ISO and picture mode are the same between all photos in the panoramas. Each panorama conisists of 80-90 photos made up of triple exposure bracketed shots.
For the panorama stitching process I have to keep development settings consistent between all photos in the panorama. Therfore I have saved a RAW "recipie" which I apply to all photos in a panorama set to ensure consistency. This recipie produced good results in 2010.
The darkest photos in these sets consistently show the issue described, even though the settings are the same for all 80-90 photos. And the same issues appear consistently in 5 panoramas I have shot in this manner.
As previously mentioned the white balance isn't quite right in the "good" photo (middle exposure in the bracket), but this is because this shared recipie tries to ahieve a compromise between the issue in underexposed areas vs. overall quality.
So, I don't intend to retry white balance settings and color profile adjustments again unless there is something new to test.
Also, I can upgrade the camera firmware, but this will prevent me from doing futher tests with version 2.0.7.
Finally I developed all affected photos in Canon Digital Photo Professional software and the issue was not present.
Most of my photos including the two examples I submitted were shot with Highlight Tone Priority enabled.
Research on this feature revealed that it modifies the raw files, compressing some of the highlights at the expense of more noise at the shadow end of the spectrum.
I've turned this setting off now.
It is redundant for HDR photography.
All the photos exhibiting the distinct magenta cast issue were shot in this mode.
Since the information is not visible in the EXIF data (that I can see), I can't tell which ones weren't shot in this mode to do a comparison.
It also explains why Canon Digital Photo Professional copes with the CR2 files OK.
Europe, Middle East and Africa