Why are my TIFF files becoming so large fro LR to PS?

Discussion in 'Digital Darkroom' started by patrick_regan|1, May 9, 2016.

  1. I'm trying to develop my digital workflow for processing my scanned film negatives. Right now, I have a bunch of 16-bit greyscale TIFF files that are between 200-400 MB (about 14k x 14k pixels). I do some batch processing (exposure, curves, etc.), then I'll open one as a smart object in PS to do dust and blemish removal, and any fine tuning.
    If I create just one new layer in the TIFF file, when I go to save this new file, I get an error message that I've exceeded the 4 GB file size for a TIFF! Can anyone help me figure out why my file size increased tenfold? Is the bloat happening when LR hands off the TIFF to PS (i.e. the newly created file), or is it happening when I save the file with the additional layer?
    I don't get this bloat when I have LR create PS files in psd format, just TIFF. I don't want to use psd, though, because I can't open the smart object in ACR within PS (basically, having LR hand off as a psd defeats the purpose of using LR).
    What other workflow aspects do I need to consider?
    Thanks,
    Patrick
    P.S. I'm using whatever is the latest Adobe CC flavor of LR and PS at the time of this posting.
     
  2. I do not use Lightroom but one thing I'd check first is whether the image remains a grayscale TIFF when you send it to PS, or whether it maybe is converted to RGB, which would increase the filesize of course. From there on, each layer takes that 1,2GB of memory (for three channels, if you have alpha channels for the selections, it becomes more again) per layer, so you'd reach 4GB a lot faster than you'd expect.
     
  3. digitaldog

    digitaldog Andrew Rodney

    You'll have to save as PSB to overcome the size limitation of both TIFF and PSD. Or reduce the layer overhead.
    Layers per se don't increase size, the pixels in them do. So if you had a duplicate layer vs. one that had a fraction of the pixels (a selection) or better, an adjustment layer, the size would be different, smaller.
     
    Uhooru likes this.
  4. Hi Patrick
    Here is what I believe what is going on.
    1) Lightroom will send it over in 16 bit RGB which triples the size right there.
    2) Since you are sending a Smart Object, you effectively have to double the size. 1 copy for the orignal image (which with the adjustments commands) and a second copy which is the rendered version which is used as the PS Layer
    3) When saving the TIFF file, a third flattened copy is created and inserted in the TIFF file as well. So #2 and #3 gives you another 3X factor by itself..

    So without trying anything even adding additional layers you all ready have a 9X factor growth. Getting a little more then that with Layers is not hard to imagine.

    Not sure how you don't get a large jump using a PSD either unless you changed the workflow to opening the file as "original" without LR adjustments.
    Hope this helps some
     
  5. digitaldog

    digitaldog Andrew Rodney

    Here is what I believe what is going on.
    1) Lightroom will send it over in 16 bit RGB which triples the size right there.
    2) Since you are sending a Smart Object, you effectively have to double the size. 1 copy for the orignal image (which with the adjustments commands) and a second copy which is the rendered version which is used as the PS Layer
    3) When saving the TIFF file, a third flattened copy is created and inserted in the TIFF file as well. So #2 and #3 gives you another 3X factor by itself..​
    Depends on the Export/Open As settings. Shouldn't that twice the size with high bit?
    A Smart Object increases the size based on what's embedded (that's all an SO is).
    The flattened part of the TIFF has to be embedded or LR will not handle it.
     
  6. I see no option in Lightroom to export a grayscale image. Furthermore all of the B&W options in LR are are done in RGB.
    I'm curious what you would scan at 14K x 14K. If I scan 6x6 film at 4000 ppi, I get 8500x8500 files (433 MB for a 16-bit TIFF) which are grain-sharp. Any more would bloat the file size without adding any useful information. A flatbed won't even do 4000 ppi. 2000 would be about the maximum usable resolution.
     
  7. digitaldog

    digitaldog Andrew Rodney

    Right, everything processed in LR is RGB based. New in LR6 is the ability to export CMYK, through Print to JPEG in Print module, which isn't useful here. Otherwise, the only color model out is RGB based.
     
  8. Here is what I believe what is going on.
    1) Lightroom will send it over in 16 bit RGB which triples the size right there.
    2) Since you are sending a Smart Object, you effectively have to double the size. 1 copy for the orignal image (which with the adjustments commands) and a second copy which is the rendered version which is used as the PS Layer
    3) When saving the TIFF file, a third flattened copy is created and inserted in the TIFF file as well. So #2 and #3 gives you another 3X factor by itself..​
    All these things seem plausible. Certainly Item 1 seems to be a big part of the problem. My first thought was "man, that's pretty dumb that it sends my greyscale TIFF as RGB," but one can in fact add color corrections to a greyscale image in Lightroom (split toning, for example). I guess LR isn't "smart" enough to know if color corrections have been applied?
    Edward, yes, I completely agree with you about my current scanning techniques. I was glossing over my need for such large files, and just trying to figure out what was going on with the files I happen to have at the moment. I don't need a file this large, and my scans at 6400 dpi on a V600 far exceed the true (resolving) resolution of about 1200-1600 dpi. I posted the question because "someday" I hope to get my hands on a large format camera, and scan large format negatives. I'm wondering how people do this with giant files like this. It would seem to me that the capabilities of LR to round-trip files to PS and back is limited to file size. I guess I need to plan for LR not having a role in my post-scan workflow?
     
  9. lwg

    lwg

    I avoid doing any adjustments in Lightroom to film scans. Then I choose to edit the original file when I open it in Photoshop (from LightRoom). That way LightRoom doesn't make any changes to the file.
     
  10. I avoid doing any adjustments in Lightroom to film scans. Then I choose to edit the original file when I open it in Photoshop (from LightRoom). That way LightRoom doesn't make any changes to the file.​
    L G, that's the workflow I was using previously with one modification: if I choose "open original" from the dialog, LR still creates a duplicate file (a "... -edit.tiff" file), so I add the extra step of opening the file in Finder/Explorer, doing my PS edits, then saving the file. LR periodically refreshes the directory where the file is stored, and will show the updated file in the grid view. So, I can still use LR as a catalog, but that's literally its only usable feature (for my current workflow, that is).
     
  11. Lightroom does not make permanent changes to any image file, RAW, TIFF or JPEG (et. al)). All edits are contained in an XML file correlated to the image file. Photoshop, on the other hand, permanently changes any file you save and close.
    At least 80% of the changes I need to make can be done in LR. If I want them to be permanent, I export the edited files to a TIFF or JPEG file with a different name or different directory. For example files for photo.net are reeduced to 700 pixels on the long edge, and saved as a JPEG in a sub-directory named PNET.
    Photoshop has very powerful editing tools, but the ones I use most often is the "Canvas Size" tool to add a border for printing, or for precision straightening and cropping. Each edition of Lightroom seems to have another tool formerly found only in Photoshop. If you want reversible edits in Photoshop, you must use adjustment layers, which are only possible with certain tools, and save the results as a 16-bit TIFF or PSD file. Color and exposure edits are much more intuitive in LR than in PS. Dust spotting is now as easy in LR as PS, and can be applied to any or all other photos taken at the same time.
     
  12. My solution is to open the TIFF file subsequently in PS and save it as JPEG. Then I import the JPEG into the LR album. I then either delete the Tiff file from the hard drive and LR or mark it as rejected and with 0 stars. As the time stamp is retained in all 3 file formats (CR2, TIFF, JPEG) the newly imported Jpeg file retains its proper place in the sequence in LR. The drawback is that it causes a backwards loop in my workflow. I save all chosen files for each album, with XMP changes, to a "... CHOSEN" directory. The Jpeg file is stored there for future export purposes. I must also copy it to the original pictures directory in order to import it into LR from the proper place. The jpeg file is smaller than CR2. That advantage is wiped out by the duplicate file. However it is still 9x smaller than the TIFF,
     
  13. Great! We'll post that back 3 years in a time machine to the OP.
     
  14. for sure. Here at P.net we have achieved time travel.o_O

    I'm thinking about ordering this enlarger kit. What do you think:D
    Spiratone-48-12-PP-Sunray.jpg
    1948 December, Spiratone ad
    If I hurry maybe I won't be too late. ;)
     
  15. But where's the USB port to transfer your 16 bit TIFF files into it? And you might not be able to get firmware updates any longer. :(
     

Share This Page

1111