Why red color looses details most in jpeg compression

Follow

Abhishek

• May 26, 2020 08:07
• Edited

I have an ecommerce website dealing with flowers and cakes, we get our photoshoot done inhouse and also the post shoot editing of images. While we upload them on website all images go through same image compression algorithm. But if we compare all images, then we find product images having red color in them loosing out most of its details only in the areas with red color.

I have attached an image below for your reference, here left hand side image is at 80% quality and right side have no compression in place. If you compare all these images, then you will find images have red color areas, they loose details most, at least visually.

I have read about jpeg compression but it says that color doesn't make any difference, but it is not the case here. What can I do to fix these issues?

 

0

• 

  Stephen Doyle

  • May 28, 2020 11:46

  Hi Abhishek,
  I can see you contacted our support team about this also, so I'll follow up on the request that you created.
  We can share the recommended solutions here also when we've assisted with the support request
  Thanks,
  Stephen

  0

  Comment actions Permalink
• 

  Benjamin Sutton

  • October 05, 2022 15:44

  Stephen - Wondering if you can share those details with us as well? Trying to find some resources on the compression of images uploaded. Thanks!

  0

  Comment actions Permalink
• 

  Stephen Doyle

  • October 05, 2022 16:04

  Hi Benjamiin,

  Apologies; this wasn't followed up on here because the original poster didn't reply to give us the specific images they were asking about so we could check the details and see if there was any issue.

  I did send them some general info, which I've pasted below in case that's helpful (and I've updated the original links I provided to be links to documentation that was added after May 2020 when I replied to this originally)
  
  If you have any issues with specific images in your account, I recommend creating a support request for my team directly and we can check it for you.

  Regards,
  Stephen

  To give some general answers, you may see better results using our automatic quality feature (with q_auto:good or q_auto:best) than with a fixed quality level, partially due to chroma subsampling, and also because our algorithm will analyse the output and check for visual changes in that output.

  There's a description of how this works here: https://cloudinary.com/documentation/image_optimization#automatic_quality_selection_q_auto

  
  You may also be seeing that the color profile from the original image has changed, and if so, you can force a specific profile via the 'cs' parameter: https://cloudinary.com/documentation/image_transformation_reference#cs_color_space

  Thirdly, if the original is JPEG also, you're compressing an already-compressed image and that will lose some information compared to the first time compression is introduced. If so, we recommend that you upload the original file in a lossless format like TIF, PNG, PSD, etc, and then deliver it in JPEG, WEBP, [edit: AVIF etc,] preferably via our automatic format selection: https://cloudinary.com/documentation/image_optimization#how_to_optimize_image_format

  0

  Comment actions Permalink

Post is closed for comments.