Now, it seems that my understanding was correct for LR5 but is this a bug with it not embedding correctly or has something changed I haven't picked up on?
Something is wrong here, I'm seeing inconsistent results where images that should have my custom profile don't after import to a different catalog.Thank you Andrew for checking. Following a suggestion I have received, I copied the LR6 dng which had Update DNG Preview and Metadata performed on my original pc, from the 2nd pc which showed the incorrect profile back to my original machine. I copied it into the catalog with no changes performed during the import and then checked the profile. It read the correct Dual Illuminated Profile? Odd,
I was disappointed that it didn't work with anything except LRit actually does work with any raw converter that accepts either icc/icm or dcp containers for camera profiles... more so - any kind of target does work for any kind of such raw converter ( excluding the cases of intentionally unusable targets - like for example a target with just one patch :D )
Sounds like and awful lot of effort to go down a dark Adobe road with no exits. (:-)Bob, the reason I choose to go down the DNg path is that in the past I have had profiles go missing. Whilst the info was stored in the xmp, the actual profile isn't.
I would just keep the files raw and adjust the ones that need to be adjusted. I assume the purpose of this is to correct the inaccuracies of the camera colour. I have a 5D2 and while it's a great camera colour accuracy is not it's long suit. For less money I actual bought a Hasselblad H3D and the colour is perfect straight out of the camera. (I profiled it the and there was minimal difference.) I didn't use the ColorCheckR Passport though, I was disappointed that it didn't work with anything except LR, so I bought the SpyderCheckr, which produces HSL adjustments.
I now use the Canon for general things but where I want accurate colours I use a camera with that.
in the past I have had profiles go missing.why don't you just backup the catalog where Adobe likes to keep user supplied DCP profiles, that's it... Adobe in its infinite wisdom does not allow ACR/LR to pickup profile(s) from the same catalog where the raw file(s) reside(s) - only from a few predefined catalogs, so just backup them on a regular basis, just like you do backup your raw files or LR catalog, that's it.
why don't you just backup the catalog where Adobe likes to keep user supplied DCP profiles, that's it... Adobe in its infinite wisdom does not allow ACR/LR to pickup profile(s) from the same catalog where the raw file(s) reside(s) - only from a few predefined catalogs, so just backup them on a regular basis, just like you do backup your raw files or LR catalog, that's it.The machine is backed up regularly as is the catalog daily. The missing profiles were from before I adopted LR several yrs ago. I occasionally send clients raw files and I want the profiles embedded. It saves mucking around with larger psd's or tiffs. The other reason I have gone to dng is the ability to validate them.
Bob, the reason I choose to go down the DNg path is that in the past I have had profiles go missing. Whilst the info was stored in the xmp, the actual profile isn't.Exactly and it's a very useful road to go down when you have multiple processing machines running LR and don't want to (or need to) have multiple copies of DNG profiles installed. The profiles should just be embedded into the DNG.
Exactly and it's a very useful road to go down when you have multiple processing machines running LR and don't want to (or need to) have multiple copies of DNG profiles installed. The profiles should just be embedded into the DNG.or a more simple solution is to allow ACR/LR to pickup profiles from the same location where the raw files are stored...
or a more simple solution is to allow ACR/LR to pickup profiles from the same location where the raw files are stored...No, it's not simpler and it's not how the DNG container was designed (we've found a bug) and it's not what the OP, I and other's want! Imagine rendered images without the ability to embed an ICC profile. Awful! The same is true for those of us that wish to use the DNG container as designed.
Something is wrong here, I'm seeing inconsistent results where images that should have my custom profile don't after import to a different catalog.Andrew, Thanks for checking. I don't see quite what you are seeing. If the catalog is on the machine #1 where the profiles live, they are picked up correctly. If the dng's are transferred to another machine where I know the dual illuminant profile is not installed, then it fails to show the correct embedded profile.
https://dl.dropboxusercontent.com/u/96709542/ConvertedOnImportIUpdatedPreviewsMG_2354.dng (https://dl.dropboxusercontent.com/u/96709542/ConvertedOnImportIUpdatedPreviewsMG_2354.dng)
Taking a quick look inside the DNG, the profile name in the file you uploaded shows as "20130317 Jeff and Pauline Venning 3". It's a single illuminant D55 profile.
Sandy
What is DNG, is it a raw file format, or a profile? I have raw files, and I have profiles.raw file formats allow camera profiles to be stored within raw files...
NoYes
, it's not simpler and it's not how the DNG container was designed
(we've found a bug) and it's not what the OP, I and other's want!
Imagine rendered images without the ability to embed an ICC profile.
Awful! The same is true for those of us that wish to use the DNG container as designed.
I have to ask you, do you even use a DNG workflow?
Do you actually realize the raw files and the DNG camera profiles don't reside in the same location and many of us want them to do so, by having them inside the DNG?
The facts are, some of us have multiple machines and don't want multiple DNG profiles separate from the images that require the profile or it gets substituted and that greatly and incorrectly affects it's image appearance.
The facts are, the DNG container should allow us to embed this important data directly into said container. It's simpler (for us at least), cuts down in possibly missing files and ensures the raw and profile stay together.
There's no logical argument other than LR6 should do what it was designed to do with respect to DNG camera profiles and embed them into the container.
Not sure why people shout like someone here knows how something was designed or has any input to it. Talk to Adobe if you think there is a design issue.Neither, it's a file format container that can contain raw or rendered data (you can save a JPEG inside a DNG), and it contains profile data among lots of other data we can store within the container called DNG. Not much different than TIFF!
What is DNG, is it a raw file format, or a profile?
YesNo! For you (assuming you even use DNG documents) yes, for many of us that work with DNG (do you?) that want the profiles embedded, by design, the answer is clearly yes we want the DNG profiles embedded into the container. That's what I want, that's what the OP wants.
congratulations with a bug if it is a bug, I and others want to be able to (_also_) use profiles stored with the same folders where raw files areIt is a bug. It's not proper operation consistently like the old days.
you, as usual, trying to obfuscate.As usual, you don't get what others who've used a process designed a specific way for years want. Again, do you even utilize a DNG workflow like the OP and I (and others) or are you just here to be typically annoying, wrong to argue that a feature that's been available for YEARS isn't what we and the designers of the product desire? Your writings as others here have found (and reported), is best ignored.
PS: I want to reiterate that by no means I wish you to lose the ability to store your profiles in DNG files... however the more simple solution is to (also) allow to store profiles per catalog (even for LR users keeping everything in LR catalogs to store them there)...What an idiotic idea. What catalog for ACR users? What about users who work with more than one converter that supports DNG camera profiles? You're flat earth DNG workflow theories are bested totally ignored.
You even realize that some people actually want to keep raw files and relevant profiles in the same place ? and that those people do not try to impose their ways on you, but as usual display your arrogance by claiming that only your way is the right one ? get yourself a mirror and take a look in it.The text illustrates you don't have a clue, you do NOT have to embed the DNG profile into the DNG container IF you don't want to. It's getting clear now you don't use nor care how others use a DNG workflow AS designed for those what want said design features.
You would agree then that this is a bug, correct?Still not sure what's going on where it sometimes works and sometimes doesn't but your file comes imported into my catalog as Adobe Standard but look at the dialog containing the list of DNG Profiles:
Still not sure what's going on where it sometimes works and sometimes doesn't but your file comes imported into my catalog as Adobe Standard but look at the dialog containing the list of DNG Profiles:Andrew, thanks for that. I am however unable to see the graphic for some reason.
(http://digitaldog.net/files/DNGProfilebug.tiff)
The bug appears to be the DNG profile selected by default?
Andrew, thanks for that. I am however unable to see the graphic for some reason.http://digitaldog.net/files/DNGProfilebug.tiff
No!
For you (assuming you even use DNG documents) yes, for many of us that work with DNG (do you?) that want the profiles embedded, by design, the answer is clearly yes we want the DNG profiles embedded into the container.
What catalog for ACR users?
What about users who work with more than one converter that supports DNG camera profiles?
You're flat earth DNG workflow theories are bested totally ignored.
you do NOT have to embed the DNG profile into the DNG container IF you don't want to.
It's getting clear now you don't use nor care how others use a DNG workflow AS designed for those what want said design features.
It is you, as usual, trying to obfuscate and further add zero text to aid the OP or those of us investigating a bug in Lightroom. Just go away would you please?
OK, spending useful time looking over the bug and ignoring AlterEgo who's lost, it seems this IS a LR6 bug. I imported your file into LR5 and updated it after selecting your custom profile. Opened in LR6, it shows the correct embedded profile by default: 20130317 Jeff and Pauline Venning 3.Andrew.... That is the INCORRECT profile..... I have sent you a pm with screen captures. The name of the correct profile should be "5Dmk2 Dual Illumant Profile". If you do not see the words within the quotes, you are not seeing the correct profile. On the 2nd machine using the same file, I see a profile called 20130424 Catalina! That is th only profile installed on the 2nd machine....!
Andrew.... That is the INCORRECT profile..... I have sent you a pm with screen captures.Got anything named 20130317 Jeff and Pauline Venning 3? Sandy see's it as do I in the list of profiles when we look at your DNG. How did you build it (what product)?
Got anything named 20130317 Jeff and Pauline Venning 3? Sandy see's it as do I in the list of profiles when we look at your DNG. How did you build it (what product)?Here is a list of all the profiles I have on my system. As you can see, the Jeff venning profile does exist. It however is not selected as I have defaults set for import. One of the is to automatically set the profile to "5Dmk2Dual Illuminant Profile"
On my usual machine using LR6.1, the profile is shown. When I take it onto the other machine running LR6.1 and which doesn't have the profile installed on it, it shows another profile instead, one called "20130424 Catalina" which is on machine #2.That set of a light bulb or memory of an issue reported in the past where LR selects a profile it can't find alpha numeric (not necessarily Adobe Standard). On that 2nd machine, what happens if all you do is install a copy of 20130317 Jeff and Pauline Venning 3 DNG profile?
I suspect that's what happening here is that what LR is seeing is the "20130317 Jeff and Pauline Venning 3" profile as an embedded profile (its in IFD 0, which is where the embedded profile lives), but not as the assigned profile. So it's defaulting to using Adobe Standard.Multiple DNG profiles can be stored but I don't know what the trick is to Assign one. The bug is an Assign issue.
Sandy, I have read what you have said. Lr IS seeing the "20130317 Jeff and Pauline Venning 3" profile. However, THAT is not the profile that was embedded. This is I think where the bug is. I am setting the profile as "5DMk2 Dual Illuminant Profile", I run Update Dng Previews and metadata which should embed that profile, NOT the venning profile!
I suspect that's what happening here is that what LR is seeing is the "20130317 Jeff and Pauline Venning 3" profile as an embedded profile (its in IFD 0, which is where the embedded profile lives), but not as the assigned profile. So it's defaulting to using Adobe Standard.
It would be useful to see the LR5 processed file that works, and compare to what's in that.
Sandy
Multiple DNG profiles can be stored but I don't know what the trick is to Assign one. The bug is an Assign issue.Andrew, In my reply to Sandy, I have attached a dropbox link so you can see before and after LR6 dngs. You say multiple dng profiles can be stored.... The correct profile though isn't being stored surely? else the "5DMK2 Dual Illuminant Profile" profile would be visible within the dng. Correct?
Ok, I might now be able to shed at least some light on this.
Firstly, looking at the LR5 DNG vs. the LR6 DNG:
1. They are the same in that they BOTH have the "20130317 Jeff and Pauline Venning 3" profile set in IFD 0. Why exactly this is occurring I'm not sure, but this behavior seems to be the same for both LR5 and LR6, and doesn't contribute to the issue.
2. The DNGs differ in two ways:
a) The LR5 DNG has the "5D Mk2 Dual Illuminant Profile" embedded as a second profile. This is not present in the LR6 profile.
b) The LR5 DNG also has adjustment data (exposure adjustments, etc) embedded into its XMP data. Critically, this data includes the following linecrs:CameraProfile="5D Mk2 Dual Illuminant Profile"
So what appears to be happening here is that when LR6 imports and converts to DNG with a profile, it does not create an XMP adjustment record. So when the DNG is written, both the profile and the profile assignment are missing.
However, if you adjust the image, and then write it out, and XMP adjustment record is created, and everything works as it should. I think that this is why the second test worked - "I then made some adjustments" being the giveaway.
In summary, I think that under LR6, import and assign profile is not enough to mark an image as "modified". If it's not marked as modified, then adjustment data isn't written, and you get the problem as described above.
It would be interesting to try to import with a preset that both assigned a profile and made e.g., a small exposure adjustment, and see if the problem persists.
I'd agree, BTW, that this is a bug.
Sandy
Sandy, I have read what you have said. Lr IS seeing the "20130317 Jeff and Pauline Venning 3" profile. However, THAT is not the profile that was embedded. This is I think where the bug is. I am setting the profile as "5DMk2 Dual Illuminant Profile", I run Update Dng Previews and metadata which should embed that profile, NOT the venning profile!Try using a different profile than 5DMk2 Dual Illuminant Profile. Indeed, the profile embedded into the document Sandy and I have is embedded with Jeff and Pauline Venning 3. You might want to move that profile too, then try the test again. Let's see what gets sucked into the DNG and (what I consider another bug), what profile gets selected on another machine.
This seems then that if no exposure adjustments are made to the file before running the Update Dng Previews and Metadata command, that the profile is not correctly embedded in the dng. If exposure adjustments are made, then the profile is embedded correctly.Yes but checking that any update to an adjustment, not just exposure, changes the testing. Also, I think there are at least two bugs here. One, the inability of the correct profile to 'stick' or be assigned unless you apply an edit and two, that moving to another machine doesn't call the correct "assigned" profile from the previous machine.
It would seem then that people should be able to replicate this result by trying it for themselves. Would you agree?
One, the inability of the correct profile to 'stick' or be assigned unless you apply an edit and two, that moving to another machine doesn't call the correct "assigned" profile from the previous machine.
Yes but checking that any update to an adjustment, not just exposure, changes the testing. Also, I think there are at least two bugs here. One, the inability of the correct profile to 'stick' or be assigned unless you apply an edit and two, that moving to another machine doesn't call the correct "assigned" profile from the previous machine.Thanks for emailing Eric Chan re this Andrew. I have tried to fill out a bug report for this but the system sends me in circles unfortunately.
I emailed Eric Chan at Adobe about issue #2 but haven't heard back.