photoshop performance thread

>>2720689
Is this just a ruse to crash people's computers/photoshop?

>>2720690
if your computer crashes doing this, you might need an upgrade

>>2720689
>20k x 20k
This will make it crash for most people here. And no, I don't have money for a better PC.

>>2720703
A good PC isn't a must have, but if you were to spend say 6 hours per day drawing, every day and know you'd really want to get into art, a good PC should be the first thing you'd seek for

>>2720706
That's true, but a 20k resolution is really unnecessary high for most uses. Sure it's nice, but it's by no means a must.

>>2720689
>Create 20k x 20k (Pixels), 300 dpi canvas
Why is the PPI at all relevant? 1ppi or >9000 won't change the size of the file, 20k pixels are still 20k pixels no matter how many inches you cram them in.

>>2720711
Empty layers and flipping the canvas multiple times doesn't make a lot of sense either.

>>2720708
The 20k resolution in this case is for the sake of testing performance, which you cannot test with a 1k resolution.

Also, most big production work tend to be around 15k, but end up resized to 5k.
>>2720711
Photoshop still processes DPI even tho you cannot see them.
>>2720718
Same thing as above, photoshop still processes empty layers even tho they are empty. An empty 20k x 20k layer is harder to process than a gradient filled 2k x 2k layer.

File: 44925.jpg (30KB, 460x460px) Image search: [Google]

30KB, 460x460px

>>2720711
I used to have a potato computer and going from 72 to 1 dpi almost doubled my performance on a 2000 x 1500 canvas.

if i did everything right... it took 8 seconds

>Photoshop still processes DPI even tho you cannot see them.
Processes it how? All PPI changes it the physical size of your print. It is literally just a tag in your file, it doesn't affect file size or how much memory it uses. 1ppi or 20000 ppi will perform the same if the amount of pixels is the same.
>An empty 20k x 20k layer is harder to process than a gradient filled 2k x 2k layer.
that's like saying a full bottle of water is harder to lift than an empty bottle.

>>2720729
>that's like saying a full bottle of water is harder to lift than an empty bottle.
I typed this backwards like a retard but you get the picture. Just test this yourself and look at the numbers. Empty layers don't add anything to your file.

>>2720729
just try for yourself man

how to economize computer's work with photoshop?

I usually work with +60 layers, some of them masks, with transparency or different layer modes

it only feel swow with some kind of brushes or when rendering a jpg/png save file

>>2720736
>I usually work with +60 layers, some of them masks, with transparency or different layer modes
Most work doesn't require this. Try to merge layers when you can. If you are doing some complex client work that requires this then upgrade your computer.

>>2720735
What would you like me to try? I get exactly the same file size, same scratch size and same amount of memory allocated whether the file has 1 ppi or 29999,999. I can add as many empty layers as I'd like and that doesn't change either, while adding a black to white gradient across one of these layers adds 1.12G to my file size, over 3GB to scratch size, and another 1,89GB of RAM allocated, according to task manager. As for the test in the OP, no matter the amount of empty layers, PPI, or how many times I flip the canvas, the amount of time it takes isn't affected.
Flipping the canvas several times after adding the noise layer, the shortest it's taken is a little under 2 seconds, and the most is about 5.

>>2720736
smart objects

>>2720753
Isn't that like a nested file though? It would still slow down the computer.

>>2720689
>Mfw still using Windows 98
>Mfw I have to work in quadrants with my art

Tried this with a I7 4790k with a slight overclock, 5th flip took 1.4 seconds.

>>2722028
I know jack shit about computers but shouldn't the amount of RAM determine how well you do in this exercise instead?

>>2720706
I'd try a sketchbook instead.