Hi Jolyon, Hi Cedric,
Is it possible to get negative values when converting an image from cone catch to RNL Chromaticity? I can\\\’t understand what negative values would imply and if these would interfere with my results (e.g. would this change the delta-S?).
Yes, values will often be negative in RNL chromaticity space. The origin is the achromatic (grey/black/white) point. For example, the human cones are listed in order from longwave to shortwave (LW, MW, SW), and creating RNL chromaticity images for human cone catch images will mean the X channel is red-green (all positive numbers being red, all negative numbers being green I think). The Y channel is then red-green versus blue (i.e. blue-yellow), so again positive and negative refer to blue or yellow colours.
The lovely thing about this space is that you can use Euclidean distances to work out the delta-S (JNDs) between any values.
Hi Lais, I just realised the latest version with desaturation function wasn’t uploaded! I’ve just added it… version 2.1.0. Due to covid I haven’t had time to fully document the new release. Just run the “convert to cone-catch” function and the option will be there. This will also highlight any issues with the image. When dealing with dark images/backgrounds remember that using lower-reflectance standards is a good idea.
Thanks for uploading the new version! I can imagine how crazy things have been during this weird time – same here.
I now have the desaturation function but when I try to convert the image to cone catch having the function selected, I get the message “Unrecognized command: ‘Average Stack'”. I have tried to use different models to see if the problem was in the model I was using but I still get this message. Do you have an idea of why this is happening?
Oh! Sorry! I did not say anything about the standard. I am using the 38%, but I might use a lower-reflectance one. I will try this now. Thanks again!
Ahh crap – sorry – I forgot to add this new function I made for that script. I’ve added it now (re-download the latest version and it should work fine). Let me know how it goes.
Cool! This makes totally sense! Thanks Jolyon!
One more question, is the “desaturation function” already available? Some of my images were taken under dim light and I think this explains why I’m getting some odd results. I would like to try this function before excluding these images from my data.