PhotoShelter’s Improved Image Processing

PhotoShelter’s Improved Image Processing

Your images should be seen online in their best possible format, and processed in a way that’s up to par with the industry’s best practices. For those reasons, we’re excited to announce our improved image processing, optimized for both quality and performance.

Moving forward, all new images uploaded to your PhotoShelter account will be processed this way, resulting in the highest image fidelity. The next step is to process all 100+ million images that live in all PhotoShelter members’ accounts over the next few months.

Some of the benefits of this new image processing include:

  • JPEG quality: We use the highest JPEG quality possible while still keeping image size manageable, so your images’ fine details and colors are preserved.
  • Gamma correction: We also preserve fine details, micro contrast, and high contrast areas when resizing your images so that they have the correct brightness. Specifically, we use sRGB gama curve (2.2). Geek out on Wikipedia over gamma correction and how it affects our perception of color.
  • Color management: Our entire image pipeline is still sRGB, meaning that any image uploaded will be tagged as sRGB – this is the de facto standard for viewing and compatibility on the web.

Overall, you should see an improved sharpness and quality of your images as displayed on your PhotoShelter website.

You might not notice the difference off hand, but check out these before and after examples. Read the notes below for more information and then mouse over each image to show the improvements.

Note the improved sharpening of fine details, the removal of artifacts around the trees, and overall better highlight separation.

[himage][/himage]

 

Note the more pleasing rendition of the out of focus areas and skin texture, more accurate representation of fine detail, and overall better highlight separation.

[himage][/himage]

 

Note the much improved highlight separation and sharpening, and the lack of artifacts in low contrast areas like the center artists’s face.

[himage][/himage]

 

Note the improved sharpening of the building, better highlight separation in the foliage, and removal of artifacts in areas where the mountain meets the sky.

[himage][/himage]

 

Note the improved sharpening and micro contrast in the artist’s glasses

[himage][/himage]

 

Note the improved sharpening and highlight separation and a more pleasing rendition of the out of focus areas.

[himage][/himage]

 

For more general tips on color management and your digital workflow, check out this video with photographer Martin Bailey. Also feel free to contact us with any questions at support@photoshelter.com.

Next Post:
Previous Post:
This article was written by
There are 23 comments for this article
    • Lauren Margolis Author at 6:11 pm

      @Daryl You aren’t doing anything wrong necessarily! Unless you’re seeing issues with your images online or in printing. We’ve improved our systems to yield the best possible results, but of course you should also follow best color management practices before you upload your images to the web.

  1. Kevin at 7:20 pm

    How can I “note” anything when there isn’t an images to compare it too? Shouldn’t you have posted the version prior to the update and the new and improved?

  2. kevin at 7:25 pm

    Why do you think the compressed more narrow sRGB is what I want? Converting my wide gamut files to sRGB always changes what I created.

    I don’t “Geek Out” on wikipedia since I’ve spent more time than I care to correcting inaccurate data on Wikipedia pages created about my wife without my wife’s permission that are so factually incorrect.

  3. Kevin at 11:37 am

    Lauren,

    Thank you for the extra work as opposed to you, the blog creator, making the example for those looking at your blog. IF the goal is to show those who are considering PhotoShelter how you’ve improved and what they gain then the examples fail at that.

    • Lauren Margolis Author at 1:17 pm

      @Kevin I really apologize for the miscommunication – the examples above do indeed showcase the before/after, just hover your mouse over the image and read the note above to learn where you can specifically see improvement. Re: your earlier comment, I thought you wanted to see examples on your own site, and we use sRGB because it is the industry standard for displaying images on the web and the most widely accepted color space for displaying imagery on the web.

  4. Oscar Bjarnason at 12:29 pm

    Hi

    Does this mean that we should use sRGB and not AdobeRGB when uploading images?
    Or does this just refer to when your system generates jpg’s from tiff when clients don’t download the original file?

    Oscar.

    • Lauren Margolis Author at 1:35 pm

      @Oscar You can upload with any color profile and we’ll convert it to sRGB, preserving your original profile. Downloads will have the image’s original profile attached and be completely unchanged.

  5. Patrick Downs (@PatDownsPhotos) at 1:42 pm

    re “Color management: Our entire image pipeline is still sRGB, meaning that any image uploaded will be tagged as sRGB – this is the de facto standard for viewing and compatibility on the web.”

    Q: If I profile my images as Adobe RGB in Lightroom or in camera and then export them that way, they are converted to sRBG it seems. BUT, will the images be that too, in the sense that if someone downloads the original file from my PS gallery, it will be sRBG? Or will it be AdobeRGB as set in camera? Thanks! And thanks for the constant improvement s to PS.

  6. Rich Wagner at 3:31 pm

    Wow, the blog post is very confusing, as it sure sounded like the color space of the original WAS being converted to sRGB (and it better not be!). That would be infuriating. I was panicked until I read the responses here.

    And as far as “geeking out” on Wikipedia about gamma curves, the author better “geek out” on color management before posting on the topic. The gamma curve for sRGB is NOT 2.2, as it has a linear component at the “foot” end of the curve. That’s old news. I would hope that Photoshelter is using a standard sRGB profile for the conversion and not a hacked profile with an approximated 2.2 gamma curve.

    See:
    http://www.sjbrown.co.uk/2004/05/14/gamma-correct-rendering/

    And from the OP’s beloved Wikipedia:

    http://en.wikipedia.org/wiki/SRGB
    “Unlike most other RGB color spaces, the sRGB gamma cannot be expressed as a single numerical value. The overall gamma is approximately 2.2, consisting of a linear (gamma 1.0) section near black, and a non-linear section elsewhere involving a 2.4 exponent and a gamma (slope of log output versus log input) changing from 1.0 through about 2.3.”

  7. Rich Wagner at 4:01 pm

    I checked the images in this blog post, and the sRGB profile is being embedded in displayed images, rather than leaving them untagged, which is fantastic.

    Based on the embedded profile, it looks like Photoshelter is using the LCMS color management module (CMM) for color transformations, which is also fantastic. Marti Maria, the author of Little CMS, definitely knows the correct transfer function for sRGB is not gamma 2.2, so the color management should be spot-on. (You can verify this by searching this pdf for “sRGB” – http://www.littlecms.com/LittleCMS2.4%20API.pdf) LCMS is a well-respected CMM.

    http://www.littlecms.com/

    Having display images converted to and tagged with sRGB is great, as color-managed browsers (like Safari) should be able to display accurate color regardless of the color space of the original uploaded to Photoshelter (Adobe RGB, ProPhoto, etc.), or when the browser is used on a wide-gamut monitor. Now, to get all current browsers to use color management….

    –Rich

  8. Kevin at 4:40 pm

    Lauren,

    What I see doing the roll over as suggested is the new and better looses or compresses the blacks and look darker. So if less detail in the shadows in order to darken the highlights is your methodology I’d suggest someone go back to school. See above posts for why you’ve erred, again at PS. I don’t manage histograms and spend my time giving shadow detail and highlight detail to have PS remove it.

    GEEZ, don’t they teach color theory anymore or does the new generation just not care.

    Google chrome is now recognizing ICC profiles.

  9. Tom Parkes at 5:42 am

    The after images in the portrait examples now look over-sharpened to my eye, particularly the hair. If the new processing system is more accurately representing the original then that’s fine. Otherwise, it demonstrates the downside to a sledgehammer approach to sharpening. Overall, the difference is probably too small to worry over.

  10. Rich Wagner at 4:51 pm

    @Tom Parkes

    I agree – many of the images now look over-sharpened, but not a huge deal. Tagged display images is more important IMO.

  11. Mike Padua at 6:56 pm

    Forgive my ignorance: what is “highlight separation”?

    As a new user, I appreciate the improvements, the images look fine–though I honestly doubt I would have been able to tell without the A/B comparisons.

    • Lauren Margolis Author at 10:10 am

      @Mike this simply means that the highlights stand out more – so the brighter areas are brighter, mainly due to sharpening.

  12. paul at 6:37 pm

    On my side I really happy sRGB is used, it’s actually IMO a much better profile for web and surprisingly print. FYI Getty use also sRGB as their standard.

  13. Derek at 1:31 am

    Re: this “improved image processing,” why give us something we don’t need or want? I find this news very distressing. It assumes, as photographers, we don’t know what we are doing and need this type of paternalistic oversight. I spend I alot of time post-processing my web-destined images, which includes converting them to sRGB, outputting to highest level jpeg quality, sharpening them to the exact desired degree, adjusting gamma and color also to the exact desired degree, all on high-end, color calibrated hardware. And as such, I expect the integrity of my choices to be respected and maintained by my hosting provider, not altered post upload.

    Frankly the examples of your image processing improvement shown on the above linked page, do not seem like improvements to me at all, especially when it comes to the sharpening effect. The “after” images look over-sharpened and pixilated.

    At best, one would expect no alterations be made to the uploaded photos. At worst, if you feel you must cater to those unfamiliar with how to maximize the quality of their images themselves, you would at least provide those who do know how and choose to do it themselves the option of turning off this awful, annoying feature.

    This and PS’s limitations on nesting galleries, but especially this oversharpening (which makes my carefully prepped images look lousy), truly makes me think I need to go elsewhere for a hosting solution. Something I am considering seriously, especially if you tell me this can’t be switched off.

    • Lauren Margolis Author at 11:58 am

      @Derek We preserve your original color profile, so nothing is changed in the upload process. Also, PhotoShelter has supported infinite gallery nesting since we released the new Image Browser last spring. Please let us know if you have any questions about either image processing or gallery nesting – support@photoshelter.com.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>