Apache-Talk @lexa.ru 

Inet-Admins @info.east.ru 

Filmscanners @halftone.co.uk 

Security-alerts @yandex-team.ru 

nginx-ru @sysoev.ru 




      :: Filmscanners
Filmscanners mailing list archive (filmscanners@halftone.co.uk)

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: filmscanners: Link to Nikon 8000 banding example...

On Fri, 20 Jul 2001, Austin Franklin wrote:

> Can anyone give a reasonable explanation of how resonance can manifest it
> self in the actual data from the scanner being incorrect?  Resonance
> certainly could cause micro distortion, but that is not what I believe we're
> seeing.  I'm not convinced it's resonance, but certainly can't rule it out.

I can see how resonance could cause positional 
errors, but not (now that you mention it) 
variations in CCD output.  However...

Electret (condenser) microphones create 
alternating current by allowing sound waves 
to impinge on a charged capacitor. 
Q=CV, and when C changes, current flows.
(aka "microphonics") This might come from 
vibration of the CCD itself, or of any 
wiring between the CCD and analog input 
circuitry.  I've designed picoammeter 
circuits where this effect was a major 
source of noise.

> I believe this is an analog manifestation.  My guess is it is either in the
> analog gain of the CCD, power supply/grounding of the CCD or A/D (causing
> conversion offset problems) or the illumination.
> I believe these are negative scans that people have used for banding
> examples.  Can someone scan a positive and show the banding to see if it
> gets lighter, instead of darker?

I saw some very subtle banding in the deep-
blue regions of some slides scanned last 
night. As I recall, the effect was 
"reversed" from negative scans.  In 
any case, on-screen, you really had to 
squint to see the banding.

> For negatives, clear on the film represents black, and will give higher
> numbers, and black on the film is white or will give low numbers.  When it's
> inverted, white is high numbers and black is low numbers.  The banding in
> the image appears to be pixel values are darker.  This would mean a higher
> value out of the A/D in the scanner.
> People have said that it appears that the problem isn't consistent.  I would
> tend to suspect a grounding problem then.  Does there seem to be a
> temperature component to this, ie, if it's cool, does the problem manifest
> it self, and warm, not?  Material will, of course, expand when warmed, and
> contract when cooled.

I have suspected a thermal component 
for a while.  The first time I saw 
banding was when it was around 85 
degrees F outdoors, and about that 
same temperature in my study, where 
the scanner sits.  It's been reasonably 
cool since then, which might explain 
the relative absence of banding.

> It could also be a ground loop...  As a thought, try floating the scanner
> (using a two prong A/C adapter), and see if that affects anything.  Also,
> the SCSI cable, if not made correctly, can cause ground loops...

Except it's a Firewire connection, not SCSI.

Given that this problem has been fairly 
repeatable among several 8000s, I'm guessing 
Nikon will figure out the problem after a 
few of these have been sent in and serviced.
I'm just a bit wary of having my 8000 be 
the guinea pig.

rafe b.


Copyright © Lexa Software, 1996-2009.