DESI FITS IMAGES:
Each FITS file have four extensions following the CCD datasheet identification: 

Extension [1] correspond to Amplifier E
Extension [2] correspond to Amplifier F
Extension [3] correspond to Amplifier G
Extension [4] correspond to Amplifier H

Each extension has the follow size:
Horizontal= 50 prescan + 1024 pixels + 50 overscan
Vertical= 1024 pixels 
Note: If you open the fits with ds9 -> Open Multiple Extension Cubes, the overscan region should be located at the right side of the image. 
 
There are 2 sets with different electronics configuration (Pepe will comment about it):

SET 1:
Folder 20180326_140009.tgz for a PTC analysis
Folder 20180326_161040.tgz for a BIAS analysis (or master bias for PTC) 

SET 2:
Folder 20180326_175759.tgz for a PTC analysis
Folder 20180326_180538.tgz for a BIAS analysis (or master bias for PTC)


PTC Test :
There are 32 images, captured in a sequence of 2 flat fields and 2 dark field, from 0 sec up to 14 secs with increments of 2 sec.
 
i.e:
PTC_000.fits -> Flat @ 0 sec
PTC_001.fits -> Flat @ 0 sec
PTC_002.fits -> Dark @ 0 sec
PTC_003.fits -> Dark @ 0 sec
PTC_004.fits -> Flat @ 2 sec
PTC_005.fits -> Flat @ 2 sec
PTC_006.fits -> Dark @ 2 sec
PTC_007.fits -> Dark @ 2 sec
...
PTC_0031.fits -> Dark @ 14sec

BIAS Test :
There are 10 bias images, zero secs of exposure time.
Known Bugs:

1. Seems to be that the CCD dynamic range do not fit very well on the ADC range, Pepe is aware and looks for a low-noise solution. As far as I know (TBC), the SET 2 of images has a related configuration.

2. There are some images with 1 row less than others, Otger and David are aware and look for a solution.
3. Fits headers are still on debugging mode, then its use is not recommended yet.