Camera shake undo?

[Gene Heskett]

Gene Heskett wrote:

Yeah, I talk to myself from time to time, old farts have that tendency 
you know. :-)

> Bruno Postle wrote:
>> Gene Heskett wrote:
>>>
>>> I just now downloaded this refocus-it rpm and installed it.  While 
>>> the procedure and plugins browsers can find it, there appears to be 
>>> no way to actually execute it.  It doesn't show up in the tools or 
>>> transforms menu's nor in the plugin-fu  or script-fu pulldowns.
>>
>> The plug-in has a different name in the menu, but I can't remember 
>> exactly or check from here.
>>
>> If you query the RPM it will tell you:
>>
>>   rpm -qi gimp-refocus-it
>>
> which returns:
> Filters/Enhance/Iterative refocus...
> 
> And by golly there it is.  I'd swear it wasn't there last night.  Its 
> also a wee bit slow, taking what appears to be about 20 minutes to 
> process a 24 megabyte image I preloaded.  The image, which I can't 
> duplicate again without several days reassembly work, had my camera 
> focused about 9 feet away on the grass instead of the object about a 
> foot away I wanted a record pix of.
> 
> The first pass seemed to indicate I needed to expand the top variable, 
> so I ran that up from 6 to 30, and brought the interations down to 4, 
> and I can't tell if its working or not, no progress bar because the 
> screenblanker kicks in before the refresh.  2.5 hours so far.  But, if 
> it will do the job, its worth it.
> 
Ok, that adjustment, at 90% done & 4+ hours into the adjustment, would 
seem to indicate the adjustments I made were apparently in reverse of 
the desired effect, and for some reason the contrast seems to have gone 
up quite a bit too.

So my question now is, if I can measure the diameter of a specular 
highlight thats supposed to be a near point src, is there a way to 
translate that diameter into a set of desired variables to be applied to 
this refocus routine to achieve near optimum results from a first pass?

Trial and error doesn't seem to be very productive so far.

-- 
Cheers, Gene