Shatulsky Vlad

The image you are attempting to quantize does not contain a 32 bit ARGB palette.

1

https://github.com/drewnoakes/nquant/blob/be8bc2e78ea412853629e8dbf98e38bed2853f72/nQuant.Core/ImageBuffer.cs#L21-L23 There is actually a workaround for this case. In my case converting the image to Format32bppPArgb and then using Quantize on it - decreased original image size by ~3...