|
If you convert to a lossy format (i.e. jpg), you will lose a tiny amount of quality from the compression, but for a single save-load-save cycle, the artifacts probably won't be visible. Your ideal transformation sequence is indeed to have a 16-bit psd as a starting point, transform, then export, but you may not be able to tell the difference if you went to 8-bit jpg first, as the 8-bit/16-bit difference will only affect the precision of calculating the color of downsample pixels that would end up in the same position as two pixels from the high resolution image (which is exceedingly unlikely to produce any posterization lines under the worst of circumstances), and as I mentioned, one extra generation of jpg artifacts probably won't be noticeable either. |
