Double precision modelViewMatrix

Is there a way to use double-precision modelview matrix in the shader other than using a custom uniform?

My application often uses FoV below 0.05 deg which creates a problem and I guess it is related to the single-precision matrix handed to the shader.

Double precision floating point is not supported in WebGL shaders in general. Single precision (32bit) is the maximum precision.

1 Like

Thanks for the information! I didn’t expect that at all.

Maybe the only solution is removing unnecessary waste of digits.