False Color and Bleeding Example

Thanks for looking into this Alberto. Are the two ‘Capture’ Sharpening algorithms different? The images above were produced at default settings (though I did increase RT’s threshold later to 20 after noticing ART’s).

Even after turning off sharpening in both I still see more bleeding in ART than RT. Barely noticeable and obviously nitpicking but probably due to some differences in the defaults elsewhere (wb, chroma noise or local contrast?).

On the other hand it’s obvious that the False Color Suppression is doing its job well, at the cost of some chromaticity shifts (easy to see when moving the slider by looking at the red in the hand sign).

They are essentially the same algorithm, but applied at different stages in the pipeline. Anyway, I would not use “false color suppression” to fix the moire’ here, but I’d rather apply some local smoothing. Something like this (attached arp):


DSC_2161.NEF.arp (13.0 KB)

They are different. They share some parts, but also some parts are different, as Alberto already mentioned.

1 Like

@JackH Forgot to mention that I finally could download your raw file. Don’t know what caused the issue not being able to download it at first try ;-(

Yes, the smoothing does indeed get rid of the false color, nice one Alberto. Though the snow behind the fence now looks green :slight_smile:

The netting would probably cast a green shadow anyway…

Indeed, but I agree with @spidermonkey, it looks realistic to me…

Also, ping @LuisSanz, this might be a good benchmark for improving RCD

1 Like

The illuminant being the converter? :upside_down_face:

1 Like

Thanks Ingo. Makes sense since ART’s is faster, at least on Win 10 and my i7 (circa 2010).

Here is my moiré reduction technique at work (via GMIC).

1 Like

Sharpening in ART is faster than in RT when being in preview. In full processing (Queue/SaveAs) RT Capture Sharpening is faster (at least in my tests).

1 Like