[daip] Image size (fwd)

[Eric Greisen]

Rob Ivison wrote:
> hi Eric et al.
> 
> is the image size limit an insurmountable problem in RMSD, or could 
> the code be altered to take us marginally above 10kx10k? (see message 
> from my student, below).
> 
> cheers,
> 
> 
> Rob
> 
> ---------- Forwarded message ----------
> Date: Thu, 29 Oct 2009 22:32:46 +0000 (GMT)
> From: Vinodiran Arumugam <va at roe.ac.uk>
> To: Rob Ivison <rji at roe.ac.uk>
> Subject: Image size
> 
> Hi Rob,
> 
> It was indeed the image size that was affecting the way in which RMSD produced 
> an output. Don't seem to be able to go above 9500 (9600 onwards) as this seems 
> to be the maximum that rmsd can handle without producing an incomplete noise 
> map. I'm using the 9500 image to make the choice between pbparm=0.3, 0.1 and 
> 0.5 images. If we can use the bigger images later, this choioce should still be 
> valid right (despite using a smaller image)?
>
I have looked at the code and it certainly attempts to work on an image 
32768 in the X direction with even less limitation in the Y direction. 
The size of the area over which the central rms is computed is limited 
to 1024 times 1024 pixels.  Since AIPS cannot handle images > 32K 
anywhere and the 1024^2 box would be very expensive to run, these limits 
seem reasonable.  I will attempt to test this to see if there is 
something ion the code stoping it from doing what its outer intenstions 
indicate.  Is the student attempting some very large box?

Eric Greisen