Re: [darktable-user] Scanning and EXIF information
On dimanche 15 novembre 2020 21:52:02 CET Andrew Greig wrote: > Thanks Stephane, > > I have taken your warning on messing with RAW files on board. I can > produce a text file with my information and store it with the RAW files. > Is it possible to change the data exported to a jpg or png to reflect > the settings for the images? And leave the RAW files untouched? You could add the scanning info as a caption in darktable, it won't be stored in the raw file, but darktable will keep it in the database (and sidecar), and export it to the jpg/png. Remco darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
Re: [darktable-user] Scanning and EXIF information
Thanks Stephane, I have taken your warning on messing with RAW files on board. I can produce a text file with my information and store it with the RAW files. Is it possible to change the data exported to a jpg or png to reflect the settings for the images? And leave the RAW files untouched? Gratefully Andrew Greig On 16/11/20 2:02 am, Stéphane Gourichon wrote: Hi Greig, You basically want to replace some EXIF information present in the raw file. A number of people (including me) would strongly avoid ever modifying a raw file. These are full of dragons. First I would make a backup of the raw files for fear or any change triggering some wrong checksum somewhere and making other software (or the camera itself) refuse to do anything with it, claiming corruption. If any software claims to be able to perform changes, I would open a binary comparison of before/after to check what is the actual change. And keep the original raw files in a backup, indefinitely. The situation is somehow similar to a more common one : camera clock was off and you need right times in EXIF after the facts. The typical case being: at some events (wedding, etc) several cameras were used, clock mismatch as always by minutes hours or days, making a mess of timeline order. Fortunately the case of adjusting time is somehow covered by darktable, which copies EXIF DateTimeOriginal to the XMP sidecar, as XML attribute of rdf:Description XML element like this : exif:DateTimeOriginal="2020:11:11 09:54:44" . Changing this line in the XMP while darktable is not running was enough for me to get correct datetime in Darktable and exported EXIF data, yay! I checked the source code (current git master), and alas this is specific to DateTimeOriginal, darktable current code does not consider overriding EXIF metadata from the XMP file, other than DateTimeOriginal. Would your need happen to me, I would either just ignore the problem and live with it, or modify darktable source code to save that data to XMP and read it back, then offer a Pull Request to the darktable team. It may be possible to manually edit EXIF values in the database, but I'm not sure this can even work, and furthermore the current design of the database is fragile with respect to the locations of raw files. More concretely, whenever the raw file is moved, or happen to be stored on an external drive that is mounted at a different place, you would probably lose your customization. With the XMP sidecar, darktable would load it again in those cases. Addendum: to handle comfortably the "several cameras with random time offsets take pictures of same events over a time range" case, I wrote a year ago a script to automatically adjust batches of pictures. It saved the day several times for me and I am considering publishing it as open-source. Anyone who reads this (even years after) please consider telling me if you're interested. -- Stéphane Le 15/11/2020 à 02.55, Andrew Greig a écrit : Hi All, I shot and developed a roll of B+W film, and I will scan it using my Canon EOS 5D iv with the EF 100mm f2.8 Macro lens. As I was shooting the roll of film I maintained a notebook where I wrote down the settings, what is the best way to substitute the info in the resulting .CR2 files with the info from my notes? I am really looking forward to loading these into Darktable/Negadoctor. Thanks Andrew Greig darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org -- darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
Re: [darktable-user] Scanning and EXIF information
Another way to handle the scans is import them into darktable, then update the EXIF data in the database. There is a lua script to adjust or set time for an image or set of images called, appropriately enough, adjust_time. It's located at https://github.com/wpferguson/extra-dt-lua-scripts. On Sun, Nov 15, 2020 at 10:10 AM Stéphane Gourichon < stephane_darkta...@gourichon.org> wrote: > Hi Greig, > > You basically want to replace some EXIF information present in the raw > file. > > A number of people (including me) would strongly avoid ever modifying a > raw file. These are full of dragons. First I would make a backup of the > raw files for fear or any change triggering some wrong checksum > somewhere and making other software (or the camera itself) refuse to do > anything with it, claiming corruption. If any software claims to be able > to perform changes, I would open a binary comparison of before/after to > check what is the actual change. And keep the original raw files in a > backup, indefinitely. > > > The situation is somehow similar to a more common one : camera clock was > off and you need right times in EXIF after the facts. The typical case > being: at some events (wedding, etc) several cameras were used, clock > mismatch as always by minutes hours or days, making a mess of timeline > order. Fortunately the case of adjusting time is somehow covered by > darktable, which copies EXIF DateTimeOriginal to the XMP sidecar, as XML > attribute of rdf:Description XML element like this : > exif:DateTimeOriginal="2020:11:11 09:54:44" . Changing this line in the > XMP while darktable is not running was enough for me to get correct > datetime in Darktable and exported EXIF data, yay! > > I checked the source code (current git master), and alas this is > specific to DateTimeOriginal, darktable current code does not consider > overriding EXIF metadata from the XMP file, other than DateTimeOriginal. > > Would your need happen to me, I would either just ignore the problem and > live with it, or modify darktable source code to save that data to XMP > and read it back, then offer a Pull Request to the darktable team. > > It may be possible to manually edit EXIF values in the database, but I'm > not sure this can even work, and furthermore the current design of the > database is fragile with respect to the locations of raw files. More > concretely, whenever the raw file is moved, or happen to be stored on an > external drive that is mounted at a different place, you would probably > lose your customization. With the XMP sidecar, darktable would load it > again in those cases. > > > Addendum: to handle comfortably the "several cameras with random time > offsets take pictures of same events over a time range" case, I wrote a > year ago a script to automatically adjust batches of pictures. It saved > the day several times for me and I am considering publishing it as > open-source. Anyone who reads this (even years after) please consider > telling me if you're interested. > > > -- Stéphane > > > > Le 15/11/2020 à 02.55, Andrew Greig a écrit : > > Hi All, > > > > I shot and developed a roll of B+W film, and I will scan it using my > > Canon EOS 5D iv with the EF 100mm f2.8 Macro lens. As I was shooting > > the roll of film I maintained a notebook where I wrote down the > > settings, what is the best way to substitute the info in the resulting > > .CR2 files with the info from my notes? I am really looking forward to > > loading these into Darktable/Negadoctor. > > > > Thanks > > Andrew Greig > > > > > > > > darktable user mailing list > > to unsubscribe send a mail to > > darktable-user+unsubscr...@lists.darktable.org > > > -- > Stéphane Gourichon > > > > darktable user mailing list > to unsubscribe send a mail to > darktable-user+unsubscr...@lists.darktable.org > > darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
Re: [darktable-user] Scanning and EXIF information
Hi Greig, You basically want to replace some EXIF information present in the raw file. A number of people (including me) would strongly avoid ever modifying a raw file. These are full of dragons. First I would make a backup of the raw files for fear or any change triggering some wrong checksum somewhere and making other software (or the camera itself) refuse to do anything with it, claiming corruption. If any software claims to be able to perform changes, I would open a binary comparison of before/after to check what is the actual change. And keep the original raw files in a backup, indefinitely. The situation is somehow similar to a more common one : camera clock was off and you need right times in EXIF after the facts. The typical case being: at some events (wedding, etc) several cameras were used, clock mismatch as always by minutes hours or days, making a mess of timeline order. Fortunately the case of adjusting time is somehow covered by darktable, which copies EXIF DateTimeOriginal to the XMP sidecar, as XML attribute of rdf:Description XML element like this : exif:DateTimeOriginal="2020:11:11 09:54:44" . Changing this line in the XMP while darktable is not running was enough for me to get correct datetime in Darktable and exported EXIF data, yay! I checked the source code (current git master), and alas this is specific to DateTimeOriginal, darktable current code does not consider overriding EXIF metadata from the XMP file, other than DateTimeOriginal. Would your need happen to me, I would either just ignore the problem and live with it, or modify darktable source code to save that data to XMP and read it back, then offer a Pull Request to the darktable team. It may be possible to manually edit EXIF values in the database, but I'm not sure this can even work, and furthermore the current design of the database is fragile with respect to the locations of raw files. More concretely, whenever the raw file is moved, or happen to be stored on an external drive that is mounted at a different place, you would probably lose your customization. With the XMP sidecar, darktable would load it again in those cases. Addendum: to handle comfortably the "several cameras with random time offsets take pictures of same events over a time range" case, I wrote a year ago a script to automatically adjust batches of pictures. It saved the day several times for me and I am considering publishing it as open-source. Anyone who reads this (even years after) please consider telling me if you're interested. -- Stéphane Le 15/11/2020 à 02.55, Andrew Greig a écrit : Hi All, I shot and developed a roll of B+W film, and I will scan it using my Canon EOS 5D iv with the EF 100mm f2.8 Macro lens. As I was shooting the roll of film I maintained a notebook where I wrote down the settings, what is the best way to substitute the info in the resulting .CR2 files with the info from my notes? I am really looking forward to loading these into Darktable/Negadoctor. Thanks Andrew Greig darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org -- Stéphane Gourichon darktable user mailing list to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
