Hi folks. Just got Gigapixel and it is capable o f some amazing results but sometimes I think the effect is too much even on minimum settings. So how about adding a mix slider so you could mix in some of the original image? Better still, a masking brush where you could reduce the effect on parts of the image.
1 Like
I like the general idea of your suggestion but would suggest taking it a little further, even.
the various AI models Gigapixel uses tend to produce varying results for varying areas in the image, often one model producing better results in some part of the image than others but not so good results overall.
It would be extremely useful if there was a (layered?) interface, where one can freely select the different areas of the image and define where each of the available AI model is to be applied for best overall results.
If such a feature is too much to ask, the maybe at least a ‘transparent crop’ or a ‘cropped layer’ function, which allows to select only part of the image for processing, inserting a transparent background into the exported image/layer for all the ‘cropped away’ parts, so that the processed image area is then still in the exact place of the original after being upscaled, so it can be used as a layer for blending upscaled areas of the image together in other software.
This would at least allow for quick and easy copying/blending those (cropped) upscaled results into another upscaled copy of the same image where a different and more suitable AI model is used.
If a AI model selector isn’t something that can be realized then at least such a transparent crop would still be quite useful to save time when manually blending together image sections upscaled with various AI models for best overall final results.
These are wonderful suggestions. I’ll be sure to show this to our development team for future implementation.
