9:30 A
STORM
- Finish O/N run
- crashed hal when computer memory filled up. Need more external harddisks!
- Should do a test run on if we can save data on the fly from Hal over the network straight to Monet. Still probably can’t write to a drive that Monet is currently using for analysis.
Coding
- fix color contrast for 3D so that the sum of all layers in channel is not saturated
- fix colorbar
- incorporate MosaicViewer into STORMrenderBeta
- Fundamental rendering now active in STORMrender
Notes on color contrast during display color Z
- contrast appears higher since we compute a contrast correctly for all pixels at all heights in the given channel, but we apply that mapping in imadjust to each Z position independently mapping something [cmin cmax] to [0,1] so some pixels still get sent to max brightness at all Z, and these add up to a lot of white and a lot of brightness.
- Quick fix implemented: Now instead of mapping all these channels to
[0,1], we map only to[0,1/Zs]or[0,1/sqrt(Zs)]. - Though the more accurate way to do this is to compute for each z-slice in each channel, the fraction,
znormof total intensity which comes from that slice, and map to[0,znorm]. - New alternative approach: normalize by mean intensity of the z-section relative to all sections. Also make sure this number does not exceed 1 for any individual section
znorm = min(1,mean(thisz(:))./mean(allz(:))/(.5*Zs));
STORM analysis
- BX-C 2 color data was taken only in 2D. (problem with not doing 3D bead calibration first!)
- Fab7 A647 Pc-750 running DaoSTORM O/N.
Fly Work
- moved all emerging F1 test crosses to MAT alpha tubs to 29C. This way F1s emerge at 29C and start crosses at 29C.
- Flip these vials tomorrow morning
- collected Esc/esc, Mat-alpha tub, and Da/Gla x Esc[6]/CyO virgins