Re: bgsmooth (was RE: [APML] Cygnus)

[Matt BenDaniel <matt@starmatt.com>]

At 02:18 PM 9/22/2001 -0500, Howard Edin wrote:
>Would there not be a correlation between average film grain size and the
>radius value used in  bgsmooth (i.e. width of the gaussian filter). 

Howard,

Yes. For the same exposure, I would increase the bgsmooth radius by 50% at 3000 DPI over what I would at 2000 DPI.

With films like LE400, PJ400, PPF400, and Supra400, it seems that 2000 DPI gets all the information. Scanning at higher rez only shows the grain better.

Then there are finer grain films like E200, RG200, and HQ100, that are worth scanning at 3000 DPI.

And tech pan should be scanned at 4000 DPI when possible.

>This would vary with film type and perhaps scanner resolution?
>
>Would there also be a  relationship between the (bgsmooth) cutoff value and
>the sky fog / limiting magnitude recorded?

I don't know, but that's not how I figure out what cutoff to use. In general it is good to keep the cutoff lower, because that loses the least fine detail. But bgsmooth can yield dark pixel artifacts when the cutoff is too small. When I see those, I re-smooth using a higher cutoff. 
--
Matt BenDaniel
http://starmatt.com


--  APML Archives at <http://astro.umsystem.edu/apml/>  ---
             Unsubscribe at <majordomo@seds.org>