How JPEG Works: The Art and Science of Imperfection

[ma3mool]

At its core, JPEG is a lossy image compression technique. What you may not realize is that achieving this required a fascinating blend of human perception, elegant signal processing, and clever engineering prowess.

Let's begin with a visual challenge: Can you believe the colors of the tiles marked A and B are exactly the same?

Read the full article here.

How JPEG Works: The Art and Science of Imperfection

 

[Gars]

I need to read that a several more times
what about WebP - I see enormous compression with alfa channel with virtually no loss of quality

ty for the great article

 

[TabTree IT]

This is a great article.Thanks for sharing.It was very useful and informative.

 

[ivankelley]

At a high level, JPEG is a lossy image compression technique. What you may not realize is that achieving this required a fascinating blend of human perception, elegant signal processing, and clever engineering prowess.

Let's begin with a visual challenge: What are the colors of the tiles marked as A and B in the image?

Read the full article here.

https://www.techspot.com/article/2792-jpeg-explained/

Thanks for sharing

 

[MarkHughes]

Great article, More like this please

 

[Revolution 11]

Great article, More like this please

Yes, more please. These articles about basic tech are the best on the site.

 

[nicodf]

"Effectively, we began this step with three channels (Y, Cb, and Cr), and ended up keeping the information intact in one full channel (the brightness component, Y) and got rid of half the information in the other two channels (Cb and Cr). By going from 3 channels to 1 + ½ + ½ = 2, we are now effectively at 2 channels, or 66% of our original image!"

I think this is wrong. In a 4:2:0 subsampling (replacing the Cb and Cr values of 2x2 blocks with a single one), the compression level should be 50%, not 66%. Instead of storing 4 Y + 4 Cb + 4 Cr = 12 values, you instead store 4 Y + 1 Cb + 1 Cr = 6 values (50% compression).

For the 4x4 case, instead of storing 16 Y + 16 Cb + 16 Cr = 48 values, we can store 16 Y + 1 Cb + 1 Cr = 18 values, or a 33% compression level.

 

[RudyBob]

That's a lot to know and I appreciate your time and expertise.

 

[p51d007]

Nice article! I "keep" all the photos I take in raw, but share the ones I want in jpeg for obvious reasons. Obviously the raw files will print better, for those I actually print, which these days,
isn't very many like it was in the old days.

 

[Julio Franco]

"Effectively, we began this step with three channels (Y, Cb, and Cr), and ended up keeping the information intact in one full channel (the brightness component, Y) and got rid of half the information in the other two channels (Cb and Cr). By going from 3 channels to 1 + ½ + ½ = 2, we are now effectively at 2 channels, or 66% of our original image!"

I think this is wrong. In a 4:2:0 subsampling (replacing the Cb and Cr values of 2x2 blocks with a single one), the compression level should be 50%, not 66%. Instead of storing 4 Y + 4 Cb + 4 Cr = 12 values, you instead store 4 Y + 1 Cb + 1 Cr = 6 values (50% compression).

For the 4x4 case, instead of storing 16 Y + 16 Cb + 16 Cr = 48 values, we can store 16 Y + 1 Cb + 1 Cr = 18 values, or a 33% compression level.

Thank you for the correction, noted and updated.

 

[Random Commenter]

I need to read that a several more times
what about WebP - I see enormous compression with alfa channel with virtually no loss of quality

ty for the great article

I believe WebP may be built upon JPG, or at least upon very similar principles. I found that by adding the jpg suffix, old mspaint was able to open the image.