If the camera calibration profile was different, that would explain it.
The Auto Tone settings are calculated after the camera profile is applied both on import and inside the Develop module. Are you changing the Camera Calibration profile after the image is imported from your default setting?
(Cross-Post...funny the time stamp is way different? Anyhooo....great minds think alike!)
Hi, no I didn't change anything at all. I conducted the test as a controlled experiment as far as possible.
I just checked and confirmed the Calibration Profile is set to "Adobe Standard" for all the images.
One thing I didn't mention, because it shouldn't matter, is that I applied the Auto Tone in the develop module to a virtual copy of the original image. It is the VC parameters I have shown above (for example 14BD3913). I then imported the same image from my CF card with the Auto Tone Preset in the import (for example 14BD3913-2 above).
I think editing a VC should be the same as editing the original in terms of the results, but who knows!
I dunno Brian, it's a mystery to me. I've heard of other situations were users have had mysterious differences in auto-toning results (e.g. second try does something slightly different than first try..) some of which to the best of my knowledge have never been explained nor fixed..
PS - I also would not expect it to matter whether virtual copy or real, although I've never checked.
The reading differences in the table you posted are fairly small (±2 units max.) and probably not visible in Compare view. I see no difference (±0) in the files I tested between Import Auto Tone and Virtual Copy Auto Tone. As far as I can tell the only setting that affects Auto Tone is the Camera Profile selected in the Camera Calibration panel.
I suggest posting an original Import Auto Toned CR2 file with the issue, such as the above file. Export it as a DNG and post it on download site such as Dropbox. Please also post a table with the Auto Tone settings as above for the original file and virtual copy. We can then check and investigate further......other wise it's all speculation.
..other wise it's all speculation.
I'm having a sorta-memory of something else that matters (besides camera profile), like maybe white balance (?), but I may be dreaming..
I agree, the differences are small. My interest is in fully understanding the process. I don't expect the import to change my raw file in any way. If I set a preset in the import I would expect the preset to be recorded in the LR Cat, I wouldn't expect my raw file to be modified by the import. However, if I export DNG files to share with you, I would expect any changes made in LR to be applied ahead of the export, just like they are when I export a JPG. Wouldn't the best thing be to share the original CR2 files and the catalog? So, with that in mind I've exported as a LR Catalog with the original raw files. You can then select the files and see the history of exactly what was done in the develop module. I've also included a PDF that describes the step I took in my experiment. You should be able to copy from here LR Shared Cat
Thanks for your interest.
Rob's reply concerning White Balance in his last post appears to be the cause. The CR2 images I checked shot with a 5D MKII showed the same settings with Auto Tone regardless of the WB settings. So I picked out a picture that had a large neutral area (i.e.snow), similar to the white sofa in your 14BD3913.CR2. Using As Shot WB (5275, 16) versus Auto WB (6350, 10) produces a Whites settings of +36 and +37 respectively. All other settings were the same. What you are seeing is caused by the White Balance settings and is perfectly normal.
'Auto WB' is heavily influenced by the subject matter color for both in-camera Auto WB setting or LR Auto WB setting. For this reason I only use the in-camera Auto WB setting in LR with my Canon Speedlites and NOT the LR Auto WB. The color temperature varies slightly with flash duration and the Speedlite will record that to the in-camera Auto White Balance reading. This gives me a better "starting point" for the WB setting in LR. For everything else I use WB preset settings that have been calculated for "typical" lighting conditions (Sunny, Cloudy, Tugsten) using a ColorChecker Passport. You may still need to fine-tune WB for specific lighting conditions such as late day and Winter i.e. (low sun), which can be Sync'd across same-scene shots.
Hi, thanks for the thoughtful reply. If the WB was different in the 2 files I could understand that Auto Tone would do different things because the starting points would be different. However, in this case we are talking about the SAME raw file, imported from the camera's CF card twice. One is imported with no preset, then the Auto Tone preset applied, the other is imported with the Auto Tone preset set in the import module. Both files have the SAME WB setting (As Shot) and the same Temp and Tint, as you would expect. The only differences (in LR) are in the tonal values.
By the way, the image was shot using a fixed WB setting, not AWB. Auto WB was not applied in LR either, so I don't see how WB can be a factor when it is constant in both image files. Obviously something is happening that I don't understand, but I can't see that it is WB.
I suspect there is a little "undocumented feature" more commonly known as a "bug" in the LR software, it's just not significant enough to attract a fix.
Thanks for your thoughts,
I suspect there is a little "undocumented feature"..
I suspect you're right.
I received two email replies (BriPhoto & Rob Cole) to my above post, which are now missing here. I only mentioned Auto WB because of this statement in your PDF posted here Shared Cat:
"Apply auto WB and Auto tone to an unmodified photo and note the changes.
Apply auto WB and tone to another photo with very different exposure and tone and note changes to confirm that changes are different and do depend on the content of the photo being operated on (assumption)."
You are now saying:
"However, in this case we are talking about the SAME raw file, imported from the camera's CF card twice. One is imported with no preset, then the Auto Tone preset applied, the other is imported with the Auto Tone preset set in the import module. Both files have the SAME WB setting (As Shot) and the same Temp and Tint, as you would expect. The only differences (in LR) are in the tonal values."
I repeated Import of the two copies of your 14BD3913 CR2 file (14BD3913.CR2, 14BD3913-2.CR2) following the above procedure. The Tone values are identical for both using LR5.7 on my Windows 7 SP1 system. I cannot duplicate your issue as outlined above.
Is this the Auto Tone preset you are using on Import?
Two other things you can try:
1) Reset your LR Preferences file:
2) If that doesn't correct the issue I would try uninstalling LR5.7.1, delete the LR Preferences file, and then reinstall LR5.7.1 using a fresh download.
Hi, the PDF was describing a test I was doing to investigate the behavior of Auto WB as well as Auto Tone. All I have been discussing here is the Auto Tone issue. Sorry about the confusion and thanks again for looking into this with me.
Yes, I can confirm that I used the same presets as in your last post screenshots.
I decided this afternoon to repeat the test as you had done, using just the 2 files we have been discussing. I created a NEW LR Cat and repeated the test, exactly as before and I got the same results as you! The adjustments are identical! So now I have 3 catalogs. My original one, that has been in use for some years, the one I "Exported as Catalog" and shared, and now this new one. I have double checked the History on each image file to confirm the step performed were identical. So now I have 2 catalogs that show the same steps on the same files, one has the inconsistency we have been discussing here and the other doesn't. So I'm thinking it is something to do with preferences as suggested, or at least my Catalog file. Before deleting my preferences I've Removed the images from my original catalog and repeated the test - this time I didn't see the issue! So, the issue has mysteriously "gone away". This is probably the least satisfying result. I think it is a software bug, but I can't reproduce it now.
Thanks for working through this with me. I wish we had a definative result though.
Before deleting my preferences I've Removed the images from my original catalog and repeated the test - this time I didn't see the issue! So, the issue has mysteriously "gone away". This is probably the least satisfying result. I think it is a software bug, but I can't reproduce it now.
I hope you have up to date XMP files for all of the CR2 files you "removed." Do you have 'Automatically write changes in XMP checked in the Catalog Settings preferences?
So you're saying creating a new catalog and importing the same image files does not exhibit the issue?
This could have been caused by corruption of the LR catalog file. How often do you backup the LR catalog and do you have the below options for Test integrity and Optimize checked? How many pictures did you have imported into the LR catalog before removing them?
You may want to run system memory and hard drive integrity tests to make sure they are OK.
Yes, thanks, I do have the auto update on my XMP files and I do test the integrity of my catalog (and optimize) on every backup and I get the backup prompt on every shutdown. I always backup if I have done any work at all because I have had corruptions in the past, usually when traveling and using USB drives. I am not concerned about remove these images because I shot them specifically for this experiment, besides, I also always keep a copy of every imported image.
My catalog has 39,939 images in it.
Yes, creating a new catalog and importing again (or even deleting from then re-importing into the original (big) catalog) does not exhibit the issue. I can not reproduce the issue, though I could restore a backup of the catalog which documents the issue in the History panel.