2

I create artwork on Photoshop and occassionally print this artwork on physical products. That said, I save all of my artwork in PNG since is of better quality and has transparency.

Shen you save a PNG on Photoshop (not Save for Web as 300ppi needs to be retained), I am prompted with a small window that gives me the options for compression.

I am presented with "None/Fast" and "Smallest/Slow". When I take use "None/Fast", my file size is around 300Mb. However, when I use "Smallest/Slow", my file size drops down to about 3-5Mb. This is a tremendous file size reduction.

Afterwards I open both of the saved files (the 300Mb and 5Mb), and they are identical on screen and both have 300ppi retained.

What's the catch here? It seems to good to be true to have the file size compress to this tiny size without any visible quality loss (I'm zooming right in, no difference). Thank you.

  • No catch. Just what it says on the label. – joojaa Nov 12 '17 at 10:12
  • PNG was historically not intended to be for printing. While you may not run into any immediate problems as it can support color profiles and handles the typical necessities of printing fine, you may run into issues if you use it out of habit. How? You may find lack of support or conversion issues that rely on older iterations of PNG due to its legacy. – biscuitstack Nov 12 '17 at 15:43
1

The quality is the same. Both are lossless compressions. "Smallest/Slow" uses more computing power and takes a bit longer (thus "slow").

Not all images can be compressed the same amount. Large areas of the same color can be compressed easily, while images with a lot of noise or fine structures will be hard to reduce.

  • Also computers are quite much more powerful today than in the past so speed might have been a issue in the past. Or for example in batch processing, but today the bottleneck is disk so its rather pointless. – joojaa Nov 12 '17 at 10:13
0

PNG and ZIP files are based on same principle. The compression software tries to find repeating patterns. One copy is saved into a symbol table and the actual instances are replaced by a reference to that symbol table. Also the symbol names take space, so the most often used symbols have the shortest names.

The original content is possible to reconstruct without losing anything. Term "lossless compression" means just that.

One can specify the wanted compression depth. That is related with how many computing steps the algorithm uses to find and sort the repeating patterns.

Surely there exist images that do not squeeze especially much when saved as PNG - for example random noise. The limited number of possible brightness values still quarantees that allways some patterns can be found, but the overhead can be eat the benefits.

This explanation is a coarse shadow of the whole thing. Find more by searching for "PNG compression" and "Deflate compression".

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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