filmscanners: Scan pass alignment, how does it smoke? (important to scanwit users) RE: filmscanners: Canoscan 4000

["Oostrom, Jerry" <Jerry.Oostrom@Alcatel.nl>]

I just wonder how this scan alignment works:

does the software give a command: "advance to frame X" or
"advance to position "offset + (distance between frames) * X",

because if it is the latter, it would perhaps be possible to test the
alignment of frames by looking at the boundary between frame hole and the
surrounding film / slide holder, there should be a transition between black
and not so black. Problems would perhaps be with very dark slides, but with
infrared those slides may not be that black.

And if that transition spot is not so clear, then one could just test on the
first clear transition of the film frame border or filmholder frame border
to the film frame. 

Of course a drawback is that the 'raw' scan would be done on more than the
complete frame and the cropping is done only at the end.

Since I sometimes do batchscanning of raw scans, such a test would be handy,
even if it would just give an indication to the raw scan (e.g. in the name)
that it may be misaligned. If it could be used to correct the alignment:
even better. 

Ah well, if you think about at without knowing the ins and outs it all seems
simple.

I hope I got the attention of Mr. Lo and Mr. Hamrick. Nice to have them
here.

Jerry.

> -----Original Message-----
> From: EdHamrick@aol.com [SMTP:EdHamrick@aol.com]
> Sent: Monday, April 23, 2001 7:49 PM
> To:   filmscanners@halftone.co.uk
> Subject:      Re: filmscanners: Canoscan 4000
> 
> 
> The one thing to be cautious of is that it looks like it makes
> two passes over the film (like the ScanWit 2740S), once for
> scanning RGB and once for scanning infrared (i'm guessing
> this from reading the specs - I've never seen one).  It can
> take almost twice as long to do two scan passes and there are
> potential problems with the two passes being perfectly
> aligned (especially at 4000 dpi).
> 
>