Is there a maximum number of subs you can integrate in PI?

[Komet]

Reason I ask is that I got to 250 last night (all 30 seconds exposure - no guiding atm      ) and PI just bogged down big time.

Star alignment was fine but Integration, normally fairly rapid, took about 12 minutes.

[harist]

I believe that the 300 limit error has been fixed.

Tasos

[Juan Conejero]

Hi Komet,

The limit for ImageIntegration working with FITS files is 8192 images on FreeBSD, Linux and Mac OS X. On Windows, the limit is 4096 images. For the rest of processes, including StarAlignment, there is no specific limit.

Of course, the actual limit is always imposed by the size of your RAM. ImageIntegration works by integrating strips of pixel rows. The number of rows per strip depends on the Buffer size parameter (the default value is 16 MB). By reducing Buffer size you can integrate more images, at the cost of less efficiency. If you set Buffer size = 0, then ImageIntegration will work by single rows, which is extremely inefficient (basically that prevents parallel processing completely) but allows you to integrate the maximum number of images possible with your available RAM.

[Komet]

Well the sooner I get the guiding back on line the better then.   

Thanks

[oldwexi]

With image integration i do have a similar issue.

I do have 104 fits images (QHY8). After 77 images integration i get out of memory
on my 2GB 32-bit XP-pc.
Running the same on my friends 4GB 64-bit Vista pc i am getting out of memory
after 75 images...?
On both Pc's i had set buffer size to 16MB.
Should i use 8MB buffer size to get all 104 images integrated???
Thanks for your advice.

Gerald

[Nocturnal]

No amount of speculation will replace the simple accuracy of you simply trying.

[georg.viehoever]

Some time ago, I did more than 260 images with 1MB buffer. Worked... 8)

[oldwexi]

This night i "tried" 8MB buffer on my 2GB XP-PC and 104 images from the QHY8 were integrated without any problem!!!
Thanks Juan for the "Buffer MB" hint...
Gerald