www.AstronomySite.com

[month-index] [thread-index] [prev] [next]

When we added 16 bit TIFF format to CCDOPS a year or so ago the only program that would read 16 bit TIFF was PhotoShop 2.5.1, and then only if you wrote it in IBM format (vs MAC) as I recall. It seems like a lot of TIFF readers try to interpret the 16 bit grey scale values as 16 bit color (5 bits red, 5 bits green, 6 bits blue?).

I assume the new version of Photoshop continues to support 16 bit TIFF and they have fixed the minor bug (I told them about it then), but I dont know.

One thing that we discovered with TIFF is that while its a "standard" there still are several ways to write TIFF files that are per spec. For example in the old days they liked you to save the image as multiple strips of 8K or less per strip to save memory in the TIFF reader. But now adays many people write a single strip for the entire image. Unfortunately many TIFF readers only seem to read a single format.

The problem with 8 bit TIFF is you loose the precision of the images in converting to 8 bit. That works fine for diplay purposes since most displays will handle 256 shades or grey max, but if you really want to llok at the images, or apply nonlinear processing functions to bring out the faint detail the 8 bit format doesnt cut it.

So whats a image jockey to do? For our users we recommend our native file format (Compressed or Uncompressed). Since its been around since the ST-6 its become accepted by various 3rd party packages. For users of other cameras or for compatibility with the high-end image processing packages I recommend 16 bit FITS format.

I welcome any suggestions or comments.

Matt Longmire mattlong@rain.org Santa Barbara Instrument Group

View index by [date] [author] [subject]
Previous Message:: [M]: ADMINISTRATIVE: Off topic posts - PLEASE LIMIT!
Next Message:: Re: LX200 (fwd)