Converting I;16B PNG to RGB ruins the image

[clairerity]

Hello! I have a image grayscale image (spectrogram) that looks fine on 32bit (F), but when i try to convert it to RGB, most data is loss, that means the image is almost fully black (majority of the white/gray parts turned to black!), would just like to know how to resolve this. Thank you very much in advance!

[radarhere]

Would you mind attaching a copy of your image?

[clairerity]

Hello @radarhere ! thank you very much for your reply :)

This is the original image in F mode (32 bit grayscale aka 1 channel)

when i try to convert it using .convert('RGB'), it becomes this

Almost all parts are lost :(

[radarhere]

From looking at the image, I conclude that it's not actually in F mode, but I;16B.

Unfortunately, Pillow does not have very good support for I;16 modes at the moment. Also, when I try to convert to RGB, my image becomes too bright, rather than too dark.

If you are looking for an immediate solution, I did

from PIL import Image
im = Image.open('input.png')
im.point(lambda i: i * (1 / 255)).convert("RGB").save("output.png")

but since yours is too dark, I would suggest

from PIL import Image
im = Image.open('input.png')
im.point(lambda i: i * 255).convert("RGB").save("output.png")

[clairerity]

Hello! thank you very much for this! although there was still some info loss when converting, it significantly improved using the code you provided! Thank you again :)