Spring 2024 - Wavefront, new ISETBio integration

This version has been modified to serve as a repository that can be called from other ISET repositories, including ISETBio (dev branch) and ISET3d. In the past, we could not have ISETCam and ISETBio both on the path, as they had many duplicate functions. Those are now gone.

We also separated out the validation scripts into a different repository (isetvalidate). We use the ieValidate() from that repository to validate any code changes.

There is one major computational change. We use wavefronts (specified as Zernike polynomials) to specify the shift-invariant optical image. In the past, we specified the optics by an OTF directly, which meant pre-specifying the frequency samples. This led to problems with very high dynamic range scenes we are currently modeling in ISETAuto. These problems acrose because we had to interpolate the OTF from the specified frequencies, and this approximation had small errors that were apparent for HDR (e.g., flare) calculations.

We now specify the optics as a wavefront aberration using Zernike polynomial. In oiCompute, we calculate the PSF and OTF on the fly, using the spatial sampling that is matched to the scene. This eliminates the artifacts for the high dynamic range scenes. There are essentially no differences from the prior calculations (as checked in ieValidate).

To use the prior method, if you must, you can set the name of the oi.optics to 'opticsotf'.

The new code also includes split pixel simulations and various other new functions that we will describe on the wiki page. When we update the wiki page (this summer?).

Final GUIDE version - Also tested with earlier versions of Matlab

In September 2020 we completed the migration from the original graphical user interface tool (GUIDE) to the new appdesigner interface. This was a substantial re-write, and the new releases going forward may not be consistent with older versions of Matlab.

This release is the final snapshot of the ISETCam repository before we pushed the appdesigner framework into the master branch. No further development will take place using the older GUIDE code.

Matlab 2018 support

Up to this point, we could run on Matlab 2018. We may start using the appdesigner tools going forward, which doesn't seem very compatible past 2018. We are making this easy to find by the tag.