Subject: Re: [nmr_sparky] Re: hello....slow spectra refreshing
From: Thomas Goddard
Date: Oct 20, 2006

Previous: 245 Next: 249


Hi PP,

I forgot to discuss the most important aspect of slow spectrum display
in my previous posting.

The test case I described, a 2-D noesy spectrum of size 2048 by 4196
has about 1000 peaks. Typical Sparky use does not involve displaying
the full spectrum. It is not often useful to see 1000 miniscule peak
markers and contours. Typical usage involves scrolling around, or
flipping to cross diagonal positions (xx), or centering on a peak
selected from a list (vc), or double clicking a peak in listed in some
dialog, or using named regions (rt)..., while viewing a region with
a modest number of peaks (10-100). Perhaps you are accustomed to
viewing the whole spectrum as a way to navigate to particular regions.
Unfortunately Sparky may be too slow for that method of working.

If your spectrum is much smaller than my example in terms of grid
points then there really is some performance problem specific to your
computer. If your spectrum is large in number of grid points but
covers only a small PPM region with a reasonable number of peaks
to look at all at once, then you may have an unreasonably large amount
of zero filling. I would like to understand your case. It is a rare
complaint that Sparky is too slow. It certainly could be made much
faster but there has been little interest expressed by others.

Tom