colormap | Molflow+

Body

What is the color map in terms of a matrix for Texture Scaling/Gadient, that you use in molflow+?

I assume it is a nx3 matrix, rgb, but I want to have the exact one, so that I can generate identical plots in matlab by myself in some cases.

Thanks,
Chao

Submitted by Marton Ady 4 years ago

Submitted by Pascal Rene Baehr 4 years ago

I wonder what the general stance on the rainbow colormap is.

The trend goes towards more pleasant color schemes like Matlab's "parula" or various schemes according to Moreland (see Colorbrewer).

Rainbow has quite the history though, so I suppose a lot of people using Molflow are used to it.

For reference (from R's pal package):

Various schemes

Submitted by Marton Ady 4 years ago

I'm not sure to understand the question above.

Rainbow colormap represents the visible light spectrum, from 700 to 400nm, as the colors appear in a rainbow.

Screen Shot 2019-06-26 at 3.46.02 PM.png

Molflow/Synrad color scale:

Screen Shot 2019-06-26 at 3.44.21 PM.png

Submitted by Pascal Rene Baehr 4 years ago

There were a lot of discussions in the past about suitable colorschemes for scientific visualisations, with the publication from Moreland (see paper) being one of the most well known ones that give a good overview of the subject and why good color schemes matter (besides just being more pleasant for the eyes).

Quick summary: Rainbow colormap does it's job, that's why it usually is a standard choice, but it has some problems (e.g. if you account for color-deficient vision).

So my question would be, if different colormaps would be something people are interested in at all?

EDIT: Another good read regarding that matter from the ROOT team (see here)

Submitted by Marton Ady 4 years ago

Thank you for the clarification and the paper!

Just so you understand, I interpreted the "what the stance on rainbow colormap is" question as if you were asking what the definition of "rainbow" is, hence my previous answer.

Reading the paper, one remark is that the 5% of users with impaired color vision might put up with the grayscale scale already in Molflow, but it is indeed worthwhile to experiment with the cold-hot color scale, for example. So far only one user (from NASA!) mentioned that it would be nice to have control over the color scale, to match it to other software.

My suggestion:

It's a very quick change (in GLGradient.cpp and in the Facet::BuildTexture) to switch to cold-hot scale.
One can also check what "standard" scientific software use (i.e. "jet" scale in Matlab)
One has to see whether the cold-hot is visible on Molflow's blue and Synrad's red background!
If these tests pass, the "color scale" checkbox can be replaced by a combobox offering new options, not forgetting to update the config file and other references

It can be a nice programming task, I suggest to check with Roberto!

Cheers, Marton

CERN Accelerating science