What are people's (experimentally tested if possible) views on sharpening images depending on whether they are to be viewed on retina or non-retina devices. Assuming resolution in the file is not an issue, which might lead to the higher resolution retina image lacking sharpness simply because there is not enough spatial information available, my thinking would be that the image to be viewed on non-retina devices would need more output sharpening for two reasons:

  1. The image is being downsized (more) and therefore loses sharpness due to the interpolation.
  2. The image is then essentially being magnified on the non-retina display, which could be considered analagous to either viewing a print up close or printing it on a medium where the ink diffuses a sizeable distance.

Do people agree with this assessment and if so what steps do you take to achieve the extra sharpening for the lower resolution image. E.g. before resizing? After? Or Both?


2 Answers 2


Ideally, output sharpening is always dependent on the target medium. Optimal quality needs an image which was resized and sharpened for the intended viewing conditions.

A high-res display needs a larger image than a low-res display, and a screen needs differnt sharpening than a print, all of which should, eg., be handled automatically by the Lightroom export dialog if you set size and device correctly. (I.e. i wouldn't set sharpening to high or low for different devices but instead trust LR's logic to apply the correct amount based on the image dimensions.)

If this exercise is worthwhile is a different question, your typical audience probably can't tell the difference anyway, but for optimal quality, you need to serve different images depending on the device.

  • 1
    \$\begingroup\$ But lightroom doesn't know what the display size of the image is. What is the "correct amount"? When standard is chosen? Low? High? What factors would affect whether low, standard, or high output sharpening for screen should be applied, if not the display size or (analogously) how close a print is intended to be viewed. Sharpening specific to the image/taste is applied before output sharpening is applied, so it can't be a matter of "whichever you feel looks best". \$\endgroup\$ Jul 29, 2015 at 21:21

Let's get this straight - you're looking to double the amount of sharpening work in your workflow to provide differently (and manually tailored, because otherwise you're wasting your time) content for a small population who...

  • are mac users
  • possess the type of display you're targeting
  • have that display calibrated properly
  • were aware that they were receiving a theoretically superior product
  • are sitting close enough to their display
  • could tell the difference between the two sources if they weren't told which was which
  • would actually return enough benefit to you to make that investment worthwhile

Perhaps I'm missing something here but I can't see any cost/benefit analysis making this a worthwhile addition to your workflow.

  • \$\begingroup\$ "...double the amount of sharpening workflow": Double the amount of italic output italic sharpening workflow. Which would be an automated process done in batch, and would constitute only a fraction of my overall sharpening workflow let alone my total workflow. "manually tailored" When is output sharpening manually tailored, other than depending on resizing amount and output medium (which is the point here)? \$\endgroup\$ Jul 26, 2015 at 15:16
  • 2
    \$\begingroup\$ "small population who are mac users": I was using the brand name in a generic sense. Almost all phones and tablets on sale today have hi-resolution displays, and desktop products are now following this trend. See link \$\endgroup\$ Jul 26, 2015 at 15:19
  • 2
    \$\begingroup\$ "could tell the difference between the two sources if they weren't told which was which": You mean the one with normal output sharpening applied as compared to the one with extra output sharpening? That was my question, and what I was seeking opinion on. \$\endgroup\$ Jul 26, 2015 at 15:29
  • 3
    \$\begingroup\$ Please avoid extended conversations in the comments section of a question or answer. That is exactly why we have a chat room. More info: meta.stackexchange.com/questions/95937/… \$\endgroup\$
    – dpollitt
    Jul 26, 2015 at 18:32
  • 1
    \$\begingroup\$ I want to echo the chat sentiment, comments are not intended for discussion. I also don't think this answers the question really, which probably explains the volume of down voting... \$\endgroup\$
    – Joanne C
    Jul 26, 2015 at 21:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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