Image Capture in TIFF format?

[jkmorse]

As many of you may be aware, I and others have had issues with FITs incompatibilities when migrating our capture files from third party software into PixInsight.  I have had this issue with MaximDL, but I think others have had this issue as well with other software.  As you may have seen in some of my earlier posts, there are solutions that fix the problem but someone suggested an alternate fix.  He suggested it may make this whole issue go away by capturing images using the unsigned TIFFs format instead of FITs.  Since images are only 16 bits when captured, the limits on TIFF files wouldn't matter and they would naturally be converted into a PI compliant format (either FITs or .xisf), at the first processing stage in PI, normally in ImageCalibration.

Question to the group is, are there any downsides to using TIFFs for image capture that I am missing?

Thanks,

Jim 

[topboxman]

The downside I see when capturing in TIFF format is it will not save important header information like like temperature of CCD, exposure times, name of DSOs, etc. FIT format saves these information. Several processing software rely on certain FIT headers. You probably will have a hard time running BPP with TIFF formats since BPP may rely on certain FIT header information.

Peter

[jkmorse]

Peter,

Great point, though I have never used BPP (an old hands on manual guy).  As a follow-up to Peter's point, will the lack of FITs header info be an issue if I manually process my images?

Thanks,

Jim

[topboxman]

I don't think you will have any issues manually processing TIF images without FIT header. It's that many times I go back to my old captured images to look up FIT header information and I find that a very valuable feature. I am surprised you are having issues between Maxim DL and other processing software for 16 bit signed or unsigned integer. I would make sure to capture only 16 bit unsigned integer. I have had issues with 16 bit signed integers.

Peter

[jkmorse]

Peter,

Its a long sad (and boring) story, but others have suffered it as well so I know I am not crazy.  Other thing that gets weird between MaximDL and PI is even if I fix the unsigned issue, often I have to convert the captures to 32 bit floating point or the centers of stars are black and read zero counts.  Just helped a guy the other day here who had the same thing show up importing MaximDL captures.  Something wonky for sure.  I have work arounds for it all now, but hoping for a magic bullet.  One thought is to capture in FITs, thereby saving the information, then do a batch convert to TIFF to bring them into PI.  No worse than having to convert to 32 bit floating point and avoids the other incompatibilities.

Thanks,

Jim