Why does MarkerFaceColor in Figure look like crinkled aluminum foil rather than a single color?

Attached is a Figure with 40-50 data points showing the value of D at specific values of x and y. MarkerSize and MarkerFaceColor are controlled by two LUT's so that small values of D give small spots colored blue and large values of D have large spots colored red. Until recently, the code behaved as expected, each marker having one color. Now the markers have weird radial streaks ("crinkles"). Any suggestions on where the crinkles in MarkerFaceColor might be coming from? The colormap is 'jet', but changing the colormap did not solve the problem. Thank you for your help.

[george. holzwarth]

Here is a script which exhibits the problem on the original computer running R2018a under Win7.

% TEST_MarkerFaceColor_V2
clear all;
close all;
x = linspace(0,3,4)
y = x

figure (1)
plot(x(1),y(1),'Marker','o','MarkerSize',20,'MarkerFaceColor','r','MarkerEdgeColor','k'); hold on;
plot(x(2),y(2),'Marker','o','MarkerSize',20,'MarkerFaceColor','g','MarkerEdgeColor','k'); hold on;
plot(x(3),y(3),'Marker','o','MarkerSize',20,'MarkerFaceColor','b','MarkerEdgeColor','k'); hold on;
plot(x(4),y(4),'Marker','o','MarkerSize',20,'MarkerFaceColor','k','MarkerEdgeColor','k'); hold on;
pause (3)

The markers are incorrectly colored on this computer. Markers show pie-shaped zones of different intensity, as in the fig files included with the original query.

However, the marker color is correct when this script is run on 2 other computers (R2017b, Win10), and (R2018a Win7). Changing the renderer did not solve the problem on the "guilty" computer. Any suggestions for possible origins of the problem? Is something set incorrectly in Matlab? Thank you. GH

[Image Analyst]

I think they look find in the PDF, don't they? Is that the same as what's in the .fig. I didn't open the .fig file - too many steps required to see it -- post a PNG image if it's different than the PDF.

This is what I see with Windows 10 and R2018a:

[Greg]

Both (.fig and .pdf) look fine here. Graphics driver, openGL, or resolution setting problems?

[george. holzwarth]

Thanks, Image Analyst and Greg, for your suggestions. Greg's suggestion to look at openGL issues was on target. I found a Matlab memo "Resolving Low-Level Graphics Issues" to be helpful on details on how to change from the original setting (hardware-accelerated OpenGL) to Basic hardware-accelerated OpenGL. Upgrading graphics driver was not needed. Perhaps R2018a reset this from its previous setting in R2017a on this computer. GH

Answer 1

[george. holzwarth]

Problem was solved by changing the OpenGL Implementation from "Hardware-accelerated OpenGL" to "Basic hardware-accelerated OpenGL". Glen suggested looking into this topic. See Matlab website memo "Resolving Low-Level Graphics Issues" for details on how to make the switch. GH