Conditions:
When watched in Develop (Preview too) Canon EOS 20D RAW (CR2) under default LR setup, all controls zeroed, in Camera calibration I can not find any system-wide preset and it says ACR 2.4.
Win XP SP2, Celeron 2.7, 1GB, office-grade 17'' profiled LCD
RESULTS:
I get poor ''develop'' color results:
1. red gets far too orange
2. on dark or in shade noise completely replaces fine details (eg. wool sweater) and details/noise controls do not help !
3. in deep shade dark brown-to-black posterisation is huge (especialy when denoise is used)
For the 2-seconds load time (when LR probably displays embeded jpeg until it makes it's own preview) the red color is fine.
Colors (and other too) are fine when RawShooter is used.
Pls help me make LR work because I love idea of LR workflow!
I DO NOT have PS to tune it...
regards
LR 1.0 EOS20D CR2 RAWs WRONG WROG COLORS
1) The Canon orange-reds problem is well-known. In camera calibration under reds try -6 and 14.
2) It has been stated here that NR and sharpening will be getting attention very soon.
3) See number 2.
LR 1.0 EOS20D CR2 RAWs WRONG WROG COLORS
Lee, I know you've mentioned this before, but I really must be missing something because implementing it yields results that are so way off. I've uploaded a 4th image with these settings -- notice that the yellowish-greenish sickly tint is still very much there!
Here are examples of pictures: http://public.fotki.com/eshwars/lr-issues/ All I did to get the JPG image was the following:
1. Opened the CR2 file in DPP, ''Convert and Save'' with Image Quality=10, Output Resolution=350 dpi. Output image is named: DPP_20070408_4038.JPG
2. Imported and opened the very same CR2 file in LR. Used ''Export Photos'' to JPG. Image quality=100, Color space=sRGB, Resolution=350dpi. Output image is named: LR_20070408_4038.jpg
3. I tried ''compensating'' for LR based on suggestion in another forum by modifying the Camera Calibration settings in LR. Hue=-40, Saturation=+10. I've tried other values as well, but it just doesn't look as good as DPP does by default. Output image is named: LR_20070408_4038_compensated.jpg
4. I implemented Lee Jay's suggestion of by modifying the Camera Calibration settings in LR. Hue=-6, Saturation=+14. Skin tone is still greenish+yellowish. Output image is named: LR_20070408_4038_-6+14.jpg
Unfortunately, I think the only viable conclusion here that LR just isn't meant to target Canon users -- it's just way too impractical to make CR2 images look pleasing :(
This problem is mentioned in so many forums (not just this). This is obviously a widespread problem Canon users are running into across many cameras.
BogdanH mentions in another post (on this forum) the trouble he had to go through by calibrating using color charts to improve his results (although he acknowledges that despite this, colors in LR are still off and not as pleasing as DPP). Are all Canon users seriously expected to do that?
I hope Adobe will step up and address this issue.
%26gt; 4. I implemented Lee Jay's suggestion of by modifying the Camera Calibration settings in LR. Hue=-6, Saturation=+14. Skin tone is still greenish+yellowish. Output image is named: LR_20070408_4038_-6+14.jpg
That was only the settings to fix the reds problem. I haven't messed with the greens and blues.
%26gt; Unfortunately, I think the only viable conclusion here that LR just isn't meant to target Canon users -- it's just way too impractical to make CR2 images look pleasing
Well, I don't use DPP because the images aren't very good and because the workflow absolutely stinks. Sometimes DPP gets good colors, sometimes it misses by a wide margin.
In your example, the LR result is much better but the WB is a bit too warm.
I've been looking and looking at the four versions, and don't know which is ''right.'' I suspect that version 3 is closest, but that due to natural mottling of a baby's skin, is less
%26lt;br /%26gt;''pretty'' than some of the others. Of course, this is on my monitor, a laptop which needs its profile updated...
%26lt;br /%26gt;
%26lt;br /%26gt;What inherent temperature were the first two developed with?
%26lt;br /%26gt;
%26lt;br /%26gt;
%26lt;span style=''color: rgb(102, 0, 204);''%26gt;%26lt;/span%26gt;
%26lt;font br='''' /%26gt;%26lt;/font%26gt; color=''#600000'' size=''2''%26amp;gt;~~ John McWilliams
%26lt;br /%26gt;
%26lt;br /%26gt;
%26lt;br /%26gt;
%26lt;br /%26gt;MacBookPro 2 Ghz Intel Core Duo, G-5 Dual 1.8; Canon DSLRs
First, thanks for suggestion.
Second, I feel this is a joke: RSP (or even RSE being 2-years old free s/w) can make Canon CR2 (quite standard, widely used) into decent TIFF/JPG, world leader's ''professional targeted'' app can not and sells it for 299$.
I thought it was my lack of knowledge or misinstallation/ misconfiguration of some kind- but if not...
Let's face it:
1. no color quality at all or even everage color
2. noise instead of details
3. slow
4. when keywords used then slow solow slow..go for cofee until an image refreshes....slow
5. it has lot of bugs (wrong miniature generation, sudden change of focus in library mode and other no-sign-of-quality-control effects)
Please convince me Adobe did not kill RSP just to not have to stand-up decent software.
''Please convince me Adobe did not kill RSP just to not have to stand-up decent software.''
Yeah, that's what Adobe did...kill RSP just so they could develop inferior software for you...that's what they spent the last year doing alright. Only problem with that theory is it ain't true...the vast majority of users DON'T have the problems you've outlined and they (Adobe) bought RSP to get Michael Jonnson-who _IS_ working on Camera Raw/Lightroom and therefore doesn't have time to continue developing a single platform product-hence Adobe ''killed'' RSP. Well, killed ain't exactly accurate...they just ceased to develop it-it ain't like they went out to your machine and made it blow up, right?
I shoot with Canon and I have zero problem getting excellent (and accurate) color out of Lightroom (and Camera Raw). So, the questions begs to be asked, why can I get good color while you and a few others seem unable to?
Since you haven't posted any examples, I can't tell you what you are doing wrong...but it's pretty obvious that there's something you are doing that is wrong. Perhaps it's white balance, perhaps it's exposure, perhaps it's the fact that _YOUR_ camera doesn't match up with the camera that was used to calibrate LR/CR in the first place.
As for Eshwar's posted example...it would be way easy to adjust the posted shot to match any of the rendered examples-or make it better. The key is to learn how to use the tone %26amp; color controls provided by Lightroom. While it may certainly be the case that Camera Raw/Lightroom fails to ''match'' the default color and tone out of the camera jpgs (and DPP), it's simply not true that with proper settings LR/CR is unable to match-pretty much exactly the results of DPP-should you fall in love with that flavor of rendering.
Lightroom can't think for you...you have to bring a little something to the table. All Lightroom can do is provide the tools for you to use (or not).
%26gt; Second, I feel this is a joke: RSP (or even RSE being 2-years old free s/w) can make Canon CR2 (quite standard, widely used) into decent TIFF/JPG, world leader's ''professional targeted'' app can not and sells it for 299$.
RSP had horrible color right out of the box. I fought with it for hours until I finally bought Color Engine. Unfortunately, ACR is not compatible with Magne Nilsen's color profiling approach so that isn't an option. I don't understand the difference between how most other applications handle profiling and how ACR does but, apparently, this would be hard to fix in the extreme from what I've read. Jeff probably knows all about this and could explain it.
There are issues with calibration and the default profiles in ACR and that's why there are tons of web pages on the subject of doing your own calibrations. For example, here's a script for getting the job done:
http://www.xs4all.nl/~tindeman/raw/acr-calibrator-l.html
Because of Eshwar's post, where he said:
''BogdanH mentions in another post (on this forum) the trouble he had to go through by calibrating using color charts to improve his results (although he acknowledges that despite this, colors in LR are still off and not as pleasing as DPP).''
...I just want to further explain my post from
http://www.adobeforums.com/cgi-bin/webx/.3bc39f40/4
In short, I was just made a suggestion there. It was ment as one of possible solutions. Thinking further, that's why those Presets in LR are ment for -for making fine predefined calibrations which serve as basis for further settings.
IMO, DPP gives very good results at (almost) default settings... as long photos are well exposed! But if photo is even a bit overexposed, DPP doesn't help much -because lacking of any ''recovery'' (and many other) settings. In such case(s), DPP makes all benefits of shooting raw, almost worthless.
LR is at v1.0, so it's hard to expect to be the best raw converter out there... and contrary: we shouldn't pretend LR is the best allready. I just hope we won't wait too long for update.
Greetings to all,
Bogdan
%26gt; LR is at v1.0, so it's hard to expect to be the best raw converter out there... and contrary: we shouldn't pretend LR is the best allready. I just hope we won't wait too long for update.
Be careful about this one. LR is in 1.0 but the RAW conversion engine is at 4.0. LR uses the ACR engine for the imaging.
Thank you Lee
Nice to here someone recognise this.
%26gt;Be careful about this one. LR is in 1.0 but the RAW conversion engine is at 4.0. LR uses the ACR engine for the imaging.
That's true, Lee, but I'm not sure much has changed in the way images are processed from version to version. My 20D still uses the 2.4 profile, and images processed by ACR 2.4 don't look noticeably different from ACR 4.0.
Yes, the profiles haven't changed, but the processing pipeline is drastically different.
%26gt;Yes, the profiles haven't changed, but the processing pipeline is drastically different.
Not arguing, just curious, but why do processed images not look any different? By that I mean with the default settings, which for the 20D have not changed. I agree that one can do a lot more with an image than before, but for the images that I have a problem with (orange colors mostly), I still cannot get quite right.
The differences are in the fact that it can process JPEGs now, and in the sharpening and noise reduction, plus the additional features like fill light, vibrance, and recovery. Color hasn't changed that I know of.
OK, I see what you're getting at. Yes, fill light, vibrance, and particularly recovery are absolutely great additions. I'm just hoping that Thomas can get around to revisiting his calibration of the 20D. :) BTW, sharpening and noise reduction have been in ACR for a long time, at least since 2.4.
Yeah, but Jeff says they're being improved soon.
Lee, thank you for pointing out the fact about ACR engine... I didn't think about that much, because I never used PS seriously.
Well, to be honest, I'm still at testing LR's database capabilities. I use Develope too (of course), but not at the level where I could say some final opinion.
Still, for my taste, default results from DPP are better than from LR (as long raw image is inside correct exposed range). On the other side, LR has some very needed features... once LR's overall speed increases, basic bugs dissapear and some of requested fetures appear, I believe I could completelly switch from DPP to LR.
Greetings,
Bogdan
%26gt; ''I shoot with Canon and I have zero problem getting excellent (and accurate) color out of Lightroom (and Camera Raw). So, the questions begs to be asked, why can I get good color while you and a few others seem unable to?''
That is a good question. And it ISN'T because we all don't know what we are doing. This is a serious problem that Adobe needs to address.
There is a repeated joke in some forums that Adobe must like orange. Many people who don't like orange as much as Adobe have attempted to calibrate ACR, often using scripts such as the Thomas Fors script, (that is what I have used), but also others. I have published such results a number of times before, but I'll concentrate on the red hue %26amp; red saturation values that so many people with so many camera models end up with after using this or other scripts:
D200: -24, 42
D2X: -20, 20
K100D: -27, 40
*istD: -24. 27
LX1: -20, 42
DS: -26, 35
LX1: -26, 19
A2: -26, 0
K10D: -13, 7
5D (daylight): -11, 4
5D (incandescent): -19, 14
It isn't simply that people have arrived at those figures ''somehow''. It is that, typically, they are preferred to the default ACR rendition. These are BETTER than the default. I use the above *istD %26amp; K10D figures, and I don't get adverse comments from judges in competitions and exhibitions. But before I used such figures, I had to spend a lot of time in Photoshop trying to get the colours right.
Look at them - what stands out? So many calibrations, and so many preferences, result in significant negative red hues, and significant postive red saturations. For many photographers, and many camera models, over a significant period of time. What is going on?
This isn't about using scripts to correct for in-model variations. The idea that those people all ''happen'' to have cameras that differ from whatever Adobe tested in such a way that they need a negative red hue and a positive red saturation is just too silly. There is something more sytematic here.
What this suggests is that the Adobe standard, which is to have transformation matrices between CIE colour space and the camera's native colour space coordinates, systematically yields different results from what the scripts are trying to achieve. The scripts are trying to make the RGB values after conversion match the values that a GretagMacbeth ColorChecker ''should'' give. And for many people, the latter is preferred to what Adobe try to achieve.
The fact that some people may have no problem with the standard Adobe calibration doesn't make the problem go away. It never will go away, until it is solved. Instead, it offers the opportunity of asking WHY is there this difference, and can we learn the underlying cause from that?
Without suggesting that this is the cause, I'll identify a hypothetical cause. Suppose that it was revealed that the experts who were satisfied with the standard results all used ProPhoto in Photoshop, and the rest of us use Adobe RGB instead. (I use Adobe RGB). The answer wouldn't be for everyone to use ProPhoto - it would be to satisfy both groups.
Adobe mustn't try to claim that so many customers are wrong. Those of us who use these scripts are pro-ACR, (and perhaps pro-DNG), and are trying to correct a significant and frustrating problem in our raw converter of choice. According to various forums, some of those who are less pro-ACR give up and try other raw converters. It is becoming ''well-known'' that Adobe get reds/oranges wrong, and make grass too yellow. That needs to stop.
Interesting that you mention Adobe RGB because Red/Orange is one of the colors all these cameras can capture well out of the gamut of Adobe RGB. Yes, I do use ProPhoto RGB (unless I'm going to the web). There are a lot of colors that cameras can capture that Adobe RGB can't contain and since the internal working color space of Camera Raw/Lightroom is PP RGB chromaticities in linear gamma it might be that some users are getting less than optimal red rendering...
On the other hand, Bruce Fraser and I tested 3 Canon Digital Rebels...two were similar serial numbers (similar lots) and one was considerably different (made at least 9 months before) and not only did Bruce find that all the cameras varied in ISO sensitivity ranging from about 80-125 when set at ISO 100 but the optimal calibration settings changed considerably as well. Which does indeed lead one to think that camera to camera variation is a worse problem than most photographers think (or that the camera makers would like to acknowledge).
But that is a considerably different situation from what was being described where the poster's skin color was off. That was simply incorrect adjustment not a fundamental failure of Camera Raw's tool set.
Thomas Knoll would have to explain the inner workings of how he profiles the cameras...but I do know that he doesn't have the time to sample lots of individual cameras/model...and he can't control Nikon %26amp; Canon Q%26amp;E and unit variation...which was why he put the calibrate tab in there in the first place. I do know that Thomas and the Camera Raw team are ''working on color'', but I can't really say anything else on that subject.
The trick is to identify the differences between ''incorrect settings'' and a fundamental flaw in the rendering concept. What exactly is ''technically correct'' vs ''matching the camera jpg''? I really don't camera about matching the camera jpg nor the color rendering of the camera company's software...
Lee Jay -
I shoot with a 20D as well and of course I have noticed this noise problem in the shadows as well when working in LR. I really just recently switched over to RAW mode - having shot Jpegs with this and my previous cameras up until last year.
I really don't want to go ''backwards'' and revert back to JPEG.
Has anyone come up with recommended settings for this camera and certain ISO combinations for Noise Reduction, Sharpening that you know of?
I shoot ISO 100 when I possibly can - but b/c I shoot a fair amount of sports - that is not often possible - so a great number of my shots are ISO 400 and higher. So I probably need to isolate those images and apply a fix for this across the board - at least to get to a decent starting point.
I can't really get good conversions in LR 1.0 with regards to noise. So I don't use it for RAW conversions above ISO 200 yet, and use RSP instead. I'm hopeful about the promised improvements in the next release. I'm hoping to put RSP to bed at that time but time will tell.
Jeff,
Here's the issue. The red area just behind the propeller on the fuselage (where it says 'Yak 54') is Ferrari red. There's no hint of orange in it in real life. It's so darned red it almost hurts.
The LR image is with the calibration settings at zero. The RSP image is with Magne's color engine. The RSP+CE rendering of that red color is pretty close to accurate. The LR rendering is way, way too orange. This is from a RAW on the Canon 5D, LR does the exact same thing to images from my 20D of the same airplane.
http://photos.imageevent.com/sipphoto/samplepictures/Canon%20Orange%20Reds%2043. jpg
So...what's it look like of you set Red Sat to -15 (or -20) and Red Sat to +10? Better?
Adjusting the Calibrate function is pretty much the same as running Magne's profile in RSP....it's adjusts the ''default rendering''.
RSP and Phase One have BOTH suffered from ''default renderings issues'' (depending on the cameras) which were solved (or improved) by using different ''profiles''. Which is pretty much the same thing as adjusting LR/CR defaults...
If I set red hue to -6 and red sat to +14, it comes very close. In fact, I now have a preset for that. This is very similar to what others have posted about lots of cameras.
To me, it's not just about a choice of default rendering it's about accuracy - how close is the rendered image to the actual object in the real world? Default renderings should seek to be as color accurate as possible and then the photographer can launch from there and do whatever the heck they want to including completely inverting the colors! But I think the goal should always be to start out accurate. That's why many people buy the GretagMacbeth color checker and attempt to create their own profiles - to get accurate color to start from.
Hi again!
First, I'm gratefull for all your comments and suggestions- I appreciate them very much. I thought PRO s/w was not about having Pro to use it to produce any decent results- However it looks not that easy and I wonder:
1. If you one should create manually presets (depending on light, ISO) and apply to every shot manually ... why that profile is not built into LR ? Maybe ISO and light conditions could be read off the CR2 ?
2. Profiling- I've got IT8 target, I could profile my camera (although in the light of massive evidence that most people get same problem with many Camera makes and models I less and less believe it is it)- where can I (in the ''Pro'' LR) choose which custom profile to apply to particular shoot (if it can not read ISO and light conditions form the raw file) ?
3. Would single point calibration (hue+sat) be sufficient for the whoole picture or I should develop single picture multple time and ''collage'' it in PSP I have (for the non-linearity) ?
4. Does Canon adjust s/w for each camera unit individually?
That would explain why it's internal jpg'er does not suffer from implied quality fluctuation in production lots.
best regards
1) RSP had the ability to apply NR and sharpening by camera and ISO. I *think* ACR 4 beta has something like this but I'm not sure. If so, LR is sure to get it sometime in the future.
2) Generally, despite the problems I've shown here, I tend to get pretty good color out of LR without messing with it. However, there are these cases with deep reds that seem to cause problems. It would be nice to have better accuracy right out of the box, but it's not too bad even as it is on most shots.
3) More likely, a single calibration (camera calibration panel) for a camera will be sufficient to get close. Some people use two (daylight and tungsten).
4) No. The variations were are seeing between ACR and reality are likely much larger than the camera-to-camera variations. Jeff is almost certainly right that there are camera-to-camera variations, but the goal of the profiling should be to hit somewhere near the center of the distribution. Right now, it seems to be outside the tail of that distribution for certain colors.
Despite what I've said above and the fact that I tend to be pretty hard on LR and ACR, this color issue is not a reason for me not to use the product. NR and sharpening are and they seem to be getting attention.
The color responce of sensors would need calibration based on the white balance of the light-most sensors have a metameric failure when shot under tungsten since there is so little blue light (and the blue is the most dense separation filter). So, it makes sense to do a calibration under both standard ''daylight'' and a separate one under ''tungsten''.
ISO doesn't seem to factor in nor does less strong white balance variations like say from D50 to D65.
But the spectral responce is camera specific so each camera would need to be tested and calibrated.
As for finding a perfect ''canned calibration'' for a wide range of cameras, no, not likely without the camera makers help...the sensors ARE made in lots (punched out of sheets or disks of silicon) but neither Nikon nor Canon (nor the other makers) are likely to telegraph that their cameras aren't ''perfect''...ya know?
To me, it simply doesn't seem to be a big deal to adjust the color rendering calibration based upon the camera and save it as a preset...then apply it upon import into Lightroom. You really only have to do it once for each base white balance per camera.
To read about the manual process, see:
Out of Gamut: Calibrating Camera Raw in Photoshop CS. It show Camera Raw version 2.x hosted by Photoshop CS, but the controls (and process) are the same.
%26gt; To me, it simply doesn't seem to be a big deal to adjust the color rendering calibration based upon the camera and save it as a preset...then apply it upon import into Lightroom. You really only have to do it once for each base white balance per camera.
Well, if you want to do it right, you have to have a calibration target, you have to shoot it properly, and you have to either adjust by hand (difficult in LR, a little easier in CS) or use a script a user has developed. None of this is what the average user of LR should be expected to do, at least if Adobe wants to sell zillions of copies to the general public. Some pros and enthusiasts may undertake this effort, but the vast majority will not.
It would certainly be better if the canned profiles tended to hit the center of the statistical spread of a particular camera type instead of being outside the spread, especially in easy conditions like the one I showed - direct sunlight. Most sensors are well-balanced for such lighting conditions. Tungsten is much more difficult as you said because it lacks blue light.
Anecdotal evidence suggests this is a consistent problem with ACR across a wide range of cameras, not a fluctuation away from the mean for a particular camera in a series.
Jeff
I see your point here. I loved RSP, and even though it had the profile for camera 5D, it still really struggled to get decent skin tones. I think partly bc of what we are talking about here, the individual differences between cameras within the same model, but also bc RSP Hue, Sat controls did so in a global manner.
So, i will get the color checker card.
Qu: is as simple as taking 2 images with my 5D (during the day, away from direct sunlight), with daylight and tungston white balance settings, and then following Bruce Frazers excellent explaination? Or do i need to more anal about the images i take.
You have to take the images under the different types of light, not with different white balance settings.
Manually adjusting the WB for that condition using the camera presets?
No. Adjusting during the RAW conversion using the gray patches on the card. The instructions for the procedure describe this process.
I appreciate the fact that as Jeff states there is camera to camera variation, and I have callibrated my 20D both using Bruce Fraser's manual method and Thomas Fors' script. I still have problems with certain types of images that perhaps is not related to the general question of Canon reds, but I'd like to see what some of you think.
I take lots of photos of insects on flowers, and living in California many are California Poppies. Shooting raw+jpeg I'd see an obvious difference between the two. In this example, the first image is the jpeg. The flower is clearly orange, but there are differences in shading and hue.
http://gattograsso.com/temp/poppie-jpeg.jpg
And here is the raw using Lightroom defaults. It is almost uniformly a single shade of orange with all of the subtle shading gone. Notice the difference in the histogram compared with the jpeg.
http://gattograsso.com/temp/poppie-raw.jpg
Callibration adjustements will do nothing to fix this. DPP, however, will render the raw almost identical to the jpeg. Any thoughts on this?
That's a whole other issue - contrast. I gave up on DPP after only a short time, but from what I remember it likes to choose very high contrast on some images. Images like this one with very uniform colors may do well with that approach.
You can get an almost arbitrarily high contrast in LR if you want to.
Fred, check the histogram...you are clipping the red channel of the raw file and that's why you are loosing textural detail.
Canons have a history of red channel clipping due, I think (and so did Bruce Fraser), to the relatively wide band tri-color separation filters Canon seems to favor and the tendancy for Canon sensors to flood from IR. We thought Canon's red separation allowed too much IR light to come in..it does tend to reduce noise because of increased signal to noise, but saturated magentas, reds and yellows (all using large amounts of red) really tend towards flooding and near clipping.
HSL and Curves would prolly help out a lot...but Canon is deffinately doing some red channel compression when making on board jpgs...
Thanks Lee and Jeff. I figured that it must be a different issue. I'll keep playing with things and see what I can do.
%26gt;Canon is deffinately doing some red channel compression when making on board jpgs
They must be doing the same thing in DPP, because histogram from a processed raw is very similar to jpeg. I have to underexpose quite a bit to keep reds from clipping in this kind of image. Thanks again, I'm learning a lot.
Thanks Lee for explanations.
Could anyone share his Canon 20D [daylight?] profile/preset (v. 2.0.3)?
I'd check how it (applied to my pics) perform- could prove just fine...
(and pls state if it can be used also for other purposes than just this test)
regards
And ditto for the Canon 30D please - thanks
Well, I spent a couple of hours trying to get the raw image to match the jpeg without success. It turned out that it was only very slightly overexposed in the reds, and +2 recovery fixed that. I found that I could get detail back by pulling back brightness and a little contrast, but I could not get colors anywhere close to correct with HSL or tone.
So I decided to see what other raw converters would do with this image. First, since I had them, I opened it in Preview and GraphicConverter. Preview (which I presume would be what Aperture would do) was pretty good, not quite as good as DPP, but much closer than Lightroom. GraphicConverter was not good at all, like Lightroom with brightness way down.
Next I downloaded trials of Photo Mechanic and Capture One LE. Photo Mechanic was very close to DPP, but Capture One was better even than DPP.
So I have to conclude that this type of image just can't be handled correctly by ACR at this time. Four out of five other converters got it mostly right, so I don't think that there's anything fundamentally wrong with the image. I guess I'll just have to resort to using DPP for these and hope that ACR changes for the better in the future. If anyone has any insight as to how I could go about adjusting it in Lightroom I'd be very appreciative.
Let us download the RAW.
How do I set that up, Lee? Do I just put a link to it, or is there some other mechanism to make it download?
Not Lee, as you'll see, but yes, a link to the file on a website would be fine. I was going to suggest the same, and would like to take a look.
%26lt;br /%26gt;
%26lt;br /%26gt;
%26lt;span style=''color: rgb(102, 0, 204);''%26gt;%26lt;/span%26gt;
%26lt;font br='''' /%26gt;%26lt;/font%26gt; color=''#600000'' size=''2''%26amp;gt;~~ John McWilliams
%26lt;br /%26gt;
%26lt;br /%26gt;
%26lt;br /%26gt;
%26lt;br /%26gt;MacBookPro 2 Ghz Intel Core Duo, G-5 Dual 1.8; Canon DSLRs
''So I have to conclude that this type of image just can't be handled correctly by ACR at this time.''
No, you should conclude _YOU_ don't know how to adjust for this sort of image...each raw convertor will have their own rendering and their own special set of tools to adjust the rendering. Coming up with a series of corrections to ''match'' the camera jpg shouldn't be that hard if you understand how the tools work. Once you arrive at this setting then you can save off a preset so you can arrive at optimal settings easier in the future. As Lee Jay said, post the original raw somewhere so others can take a whack at it...
Double posted, sorry.
Here's a link to the raw file:
http://gattograsso.com/temp/_MG_4703c.CR2
Thanks for your interest.
If anyone would like a side-by-side comparison, this is the out-of-camera jpeg that accompanies the raw file:
http://gattograsso.com/temp/_MG_4703j.JPG
I learned that option-click (on a Mac, anyway) will download the files.
(I also shoot with a 20D)
I must admit to being fairly frustrated with skin tones in BOTH RSP and Lr. I thought it was just part of my being both red/green %26amp; blue brown ''Color Blind'' (Color deficient is a better term, I clearly see many shades of these colors tho some tend to merge together). I've been able to fake it by juicing specific tones using TAT on each pic but still feel a lot of my skin tones seem to edge towards blacker and orangy/browner than what real life looked like to me. You can flip through my flickr stream at www.flickr.com/photos/kadath to see what I mean but here are a few specific disappointments skin color wise:
http://www.flickr.com/photos/kadath/460458424/
http://www.flickr.com/photos/kadath/460416993/
Again a lot of this might be limitations with flash or simply my own ineptitude setting White Balance, but when everything else looks groovy to me except skin tones, I gotta wonder...
Both look fine to me on my Laptop, skintone wise.
Don
Don Ricklin, MacBook 1.83Ghz Duo 2 Core running 10.4.9 %26amp; Win XP, Pentax *ist D
http://donricklin.blogspot.com/
The flower jpeg looks blown (even though it isn't) and the grass looks a little blue. So instead of trying to duplicate the JPEG, I tried to create an image I liked better using LR 1.0.
http://photos.imageevent.com/sipphoto/samplepictures/_MG_4703c.jpg
No comments:
Post a Comment