Alley in Pisa, Italy

by Lars Kotthoff

submit your photo


Hall of Fame
View past winners from this year

Please participate in Meta
and help us grow.

Take the 2-minute tour ×
Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It's 100% free, no registration required.

I'm using Lightroom for photo workflow. At the end of it, I would like to have the best JPEG I can get, basically two options to choose from: - export from Lightroom to JPEG, using the 100% Lightroom quality - export from Lightroom to TIFF and converting to JPEG with ImageMagick

I know that neither 100% is really 100%, nevertheless, I'd like to choose the better options, which one is closer to the original file. Have someone measured the difference between the two? I've seen a thread about how to compare the difference, but never seen reasonable measurements.

Update: I'm shooting in RAW, using DNG, exporting different sizes. The question is: what tool produces better quality JPEG: Lightroom or ImageMagick?

share|improve this question

migrated from stackoverflow.com Jul 29 '11 at 12:37

This question came from our site for professional and enthusiast programmers.

3  
Note the first part of the comment to the question you link to - it is very important to understand that a computable difference (esp. small PSNR values) cannot tell the difference in image quality, as there is no available model of the perception and mental interpretation of an image. –  ysap Jul 29 '11 at 14:27
1  
Note that 100% quality in lightroom often results in 5 to 10 times file size increase with no visible change in image quality. –  Billy ONeal Jul 30 '11 at 2:01
add comment

4 Answers 4

I have both, so I ran a quick test for you.

Firstly, here's the original DNG. It's a snap of mine, converted from Nikon NEF by Ligtroom, no editing.

I then exported it with Ligtroom, same size, no sharpening, 100% JPEG quality, and no edits. Lightroom Version

(You can download it from here for pixel-peeping)

Then I converted the DNG through ImageMagick. Again, no edits, 100% quality:

ImageMagick Version

(Here's the downloadable location)

You be the judge which gives the better export. However, note the following:

  • The Lightroom version has been modified. Not only is the exposure different, but it's very slightly warped relative to the ImageMagick version. I think there might be some lens correction going on. If there is, I can't find it; all my "Develop" settings are at default.
  • The DNG is 8.6MB. Ligtroom's JPEG is 7.3MB: not a very significant compression, but that's what you get at 100% quality. Imagemagick gave me a 14.2MB file.

Here's my take-away from this experiment: It doesn't matter.

  • You're going to want to edit your photos. Your image editing will probably affect your results more than your choice of JPEG conversion tool ever will.
  • You'll want to increase your JPEG compression by reducing the quality setting. Otherwise, you'll get files with barely any compression at all; at that rate, why not go TIFF or PNG? By increasing the compression, you're going to lose quality, and it will probably be more significant than the mere choice of tool.
    • This is especially significant with ImageMagick giving the wacky larger results than the DNG; it blows that option out of the water.
  • How big of a deal are the quality differences anyway? You'll have to really pixel peep at full size to see any changes. If you get obvious visual differences (like my exposure shift), the differences due to conversion will be insignificant.
share|improve this answer
1  
Craig, I appreciate your work, however I would like to point out a few notes: 1: DNG to JPEG is not the same as DNG to TIFF to JPG. I explicitly written that conversion way previously, because I know the internals of DNG and how it is exported through Lightroom to JPG. Imagemage TIFF to JPG vs. Ligthroom direct JPG is comparable. 2: I want to store a high-quality JPG as most prints require JPG. That's it. I want the best JPG. Of course I do compress bu size and by quality, but I need a good baseline. And no, I have no rooms for tiffs or dngs in large (we are talking about TBs of data). –  István Jul 30 '11 at 7:19
add comment
up vote 3 down vote accepted

I've created analytical measures of a single photo. My conclusion is: you can get better quality with Lightroom->TIFF->ImageMagick->JPEG, but the file size is much larger and the difference does not seem to worth it. As for myself, I'll use direct Lightroom->JPEG.

Please note, that this isn't the same comparison that as Craig has done. He had compared the DNG's two rendering settings that have no connection with the JPEG conversion and compression. For further reading, this pdf describes same of the DNG internals that one shall be aware of. You should check the chapter "The Raw processor" starting at page 15.

share|improve this answer
add comment

if you are shooting in jpg then that's the most you can get, don't mess reconverting it. Otherwise store them in dng not in jpg, then suit your jpg compression for particual uses, like for web, Big, Normal, Small jpgs.

The algorithm is important but the source is more important. So in dng you can be sure you'll get the best of any jpg export.

share|improve this answer
    
Eric, I'm shooting in raw, using dng, exporting to tiff if required, etc. The queestion is really: what tool produces better jpeg: lightroom or imagemagick? –  István Jul 29 '11 at 5:25
add comment

Assuming that the parameters on both of the files are the same (i.e. highest quality) and you aren't up-scaling or down-scaling the the photographs, the resulting output should be effectively exactly for both problems. Now that is not to say that if you did a diff on the files they would be the same, but from the standpoint of the image quality and eye appeal you shouldn't be able to tell the difference.

That said though, if you are up-scaling or down-scaling the photographs and doing other manipulations to then you will tend to see differences because programs tend to use slightly different (or significantly different if they have a patent!) algorithms for the scaling and other operations that you do to a photograph. However, at this point I would argue that you should go with the software that you find works best for your workflow and who's output has the best eye appeal for what you are doing. If you need precise reproduction of something you are going to have to limit any manipulations you do in which case we return to the point about the JPEG output being effectively the same.

share|improve this answer
    
I think I will do a my measurements and experiments and I will publish the results, hoping that it will make my point clear. I hope that it can be done on this weekend. –  István Jul 30 '11 at 7:25
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.