Archiving to Lossy DNG vs HQ JPG

Status
Not open for further replies.

Andrew NZ

New Member
Premium Classic Member
Joined
May 1, 2021
Messages
16
Location
New Zealand
Lightroom Experience
Intermediate
Lightroom Version
Classic
Lightroom Version Number
LR 5.6
Operating System
  1. Windows 8
I am a keen and serious amateur photographer with a particular interest in nature, landscape, wildlife especially birds, and also candid shots of friends. For the past 12 years I have archived images as full-res DNG. Recently, partly reflecting the size of DNG from my newly acquired Sony A6600 (previously NEX 5N), I have been considering alternatives.

My workflow has been import files to LR as DNG (recently via DNG converter since LR 5 can't recognize the a6600 raw files) then do a multi stage series of delete rejects / develop selected runs. At a minimum develop means working with the histogram to get overall exposure balance where I want it. In most cases it means getting colors how I want them too and for the majority of images it means taking them right to the point of being ready for web upload and ready for a final print oriented develop tweak. I almost never print my images however I do like to leave them in a state where this is possible and the quality has not been ruined. I aim not to need to significantly re-develop my images in future.

I realize that many photographers would consider archiving anything less than fuil-res raw files to be a contradiction however I am not convinced this is necessarily true for my requirements.

I initially thought converting full-res DNG to Lossy DNG was the answer. This was based on a mis-conception as to what a lossy "Convert to DNG" would do. I was hoping it would look at the post developed pixels (rather than undeveloped raw original pixels) and apply lossy DNG compression to that. On further reading I now believe it applies Lossy compression based on original pixels in the raw image and then passes those back to LR along with the history of development actions.

I am not all that comfortable with this option since in situations where I have, for example, a dark image which requires lots of boosting of deep shadows, the lossy compression might well seriously reduce the quality of the aggressively boosted dark pixels. Similar problems could arise with overly light images or images with a particular color close to fully saturated.

This led me to considering using exported post development HQ (90%) jpg for archiving. Based on my workflow assumption that I do not re-work images again some time in the future, it seems to me this might actually do a better job of preserving the pixels that matter while managing to achieve a decent amount of compression. Of course I lose the history of development stages but again that does not necessarily matter if I have done a reasonable job of the initial development

I know there is another alternative which is more HDD space but for various reasons I would prefer not to go that path.

Thanks for your patience. I hope I have managed to make the question / concern comprehensible. I would be interested in your insights and opinions.

Cheers Andrew
 
I'm not as interested in archiving finished edit versions as backing up the raw file of rated pictures.
  • I simply work from RAW and do not convert to DNG. Never saw a benefit. Why do you do it?
  • I backup my files and LrC files to an external drive.
  • I guess my 'archive' is a separate backup of any photo with a star rating (see below). These are copied to a 2nd external drive. This external drive is backup up to a cloud account. I export XMP files with my settings in them. I also backup an associated TIFF or DNG files from external editors/internal HDR/Pano processing but only the RAW and XMP are backed up to the cloud along with a directory listing of all associated files.
There is a question of whether or not the Adobe proprietary DNG format has had enough of an update. For this reason, you may want to consider archiving to TIFF.

Not sure your avoidance of more HDD, which is cheap, and would solve a number of your concerns.

For archiving protection, you may want to look at cloud storage.

P.S. - On backing up by star rating. I wrote a Python script that looks for the star rating in XMP, TIFF and DNG files. For those found, a wildcard backup is created following the file naming convention I use.
 
Welcome to the forum. This question come up often and the concerns that many will raise, as did Paul above, is that what you are doing is akin to printing an image and then throwing away the negative. Are you planning on stopping your LR subscription soon? Then I could see printing out copies of your processed images, but I would probably use a 16-bit Tiff format in a reasonably large, appropriate color space. But as storage is getting cheaper and cheaper, I concur with Paul about keeping the original file (raw or DNG) and then exporting the Tiff derivatives so you have them in the event that you do leave LR.

Good luck,

--Ken
 
I initially thought converting full-res DNG to Lossy DNG...I am not all that comfortable with this option since in situations where I have, for example, a dark image which requires lots of boosting of deep shadows, the lossy compression might well seriously reduce the quality of the aggressively boosted dark pixels. Similar problems could arise

My impression of Lossy DNG Is that it’s more of an option for archiving secondary photos, not the few valuable keepers. For example, someone might regularly come back from a shoot with 1000 images but only 20-30 are significant; instead of making the permanent decision to throw out the 900 “OK but not great” images they could be converted as Lossy DNG to save a lot of space while keeping them. If one was wanted later, Lossy DNG would provide a reasonable amount of raw editing flexibility up to a modest enlargement size.

But if you have an image where preserving maximum shadow detail would be critical to the value of the image, it sounds like that should not be converted to lossy DNG.

I was running low on storage space and considered some alternatives, but in the end they would take time to implement regularly, and taking the cost of that time into account, I decided that the cheapest/simplest way to deal with it was to leave the files as they are and buy three 8TB hard drives: One for primary storage and two more for mirrored backups.
 
Status
Not open for further replies.
Back
Top