Re: To David Soderman: A Little Test

[Kennedy McEwen <rkm@xxxxxxxxxxxxxxxxxxxx>]

In article <3CB4C763.6087.B32E9B@localhost>, rafe.bustin@verizon.net 
writes
>
>I must be getting dense in my old age, as I still don't get it.
>
>I trust you have seen the sample JPG on the Nikon site
>(URL in yesterday's post)

Yes

> or maybe Lawrence Smith's
>sample, posted on his website.

No

>  In either case, you'll see
>that the bands are widely spaced and quite periodic.

That's consistent with what I would expect.

>  I don't
>have a banded image to refer to at the moment, but as
>I recall, the bands are much, much wider than the line
>gap of any CCD array that I know of (say, 8 or 12
>pixels.)  From vague memory, the bands were, say, 1
>mm apart, which is 160 pixels at 4000 dpi.
>
160 pixels on a 4000dpi sensor is not that unusual, less than 1mm on the 
actual chip.  In fact the larger the gap the more likely to be errors in 
the calibration in the first place, so your description of the effect is 
quite consistent with that explanation.

However I am not suggesting that this is the cause, (I don't have an 
8000 to experiment with, so I can't be sure) just explaining that there 
are well known miscalibration mechanisms which can give rise to the 
symptoms you describe and that there are equally tried and tested 
techniques which overcome them.

>The TDI scheme you describe sounds quite a bit like the
>shingling that is used on inkjet printers to smooth the
>output (via integration of errors, more or less.)

No - in the inkjet their is clearly no integration occurring, just a 
redistribution of the fixed errors into a higher spatial frequency band, 
similar to the process of interleaving the lines, but interleaving 
pixels as well.

TDI eliminates the errors by ensuring that each pixel is a derivative of 
all n lines in the CCD.

>Here's my other problem with Ed's theory:  If it were so,
>Nikon should be able to offer a firmware upgrade to fix this
>problem -- but they have not.
>
But, in their opinion, they have - they might even claim to have 
anticipated the problem and the firmware was already available as an 
option, called superfine mode.
-- 
Kennedy
Yes, Socrates himself is particularly missed;
A lovely little thinker, but a bugger when he's pissed.
Python Philosophers
-
Turn off HTML mail features. Keep quoted material short. Use accurate
subject lines. http://www.leben.com/lists for list instructions.