Re: Document Feeders (was Re: Anyone with a Umax Document Feeder?)

Oliver Rauch (oliver.rauch@Wolfsburg.DE)
Tue, 27 Jul 1999 20:24:10 +0200

Nick Lamb wrote:

> On Tue, 27 Jul 1999, Oliver Rauch wrote:
>
> > Where is the problem when the backend calls "change page"
> > before starting the scan when scan source is set to "ADF" ?
> > This does work for all scanners with ADF!
> > I see no problem with this, because the idea of the ADF is to put in
> > a lot of papers that shall be copied, faxed or archieved, so the user
> > selects starts and comes back half an hour later and everything is
> > finished.
>
> If this works already, I'm not sure what you want changed in SANE ?

Nothing for the backends that do work this way.

The problem is that it is not defined as a standard until now.
So we have to add the source="Automatic Document Feeder"
to the sane api to give the frontend the chance to detect
the ADF mode!

> > If you want a user interactive mode, the frontend does not need to
> > know anything about that because the user does press the buttons.
> > So if the user shall be able to say "next paper" that is not the idea
> > of ADF. I suggest to add an other scan source "Manual Document Feeder"
> > or something like that. But we do not need any well known options for
> > this because the frontend does not do anything automatically!
>
> Mmm. Well, the nice thing about the HP design is that I can write something
> to do this, totally automatically, and without behaviour I don't know
> about behind the scenes. I'd like this preserved, but it's fine if this
> works just when source == MANUAL or whatever.
>
> while (we've got some sheets left) {
> load another sheet
> quick-preview-scan
> auto-select scan area (just call lists) and brightness/ contrast
> high-resolution-scan > file
> }

This does only work in the source=MANUAL mode.
A preview can not be done in ADF mode!

> > > Also, Oliver: Does the Umax way give you separate scan data for each
> > > sheet, or does the user/app have to cut it up post-scan?
> >
> > You get one image per page!
> > You call SANE_START, the scanner loads the page and scans the selected area.
> > Second SANE_START produces next image.
>
> Does this mean that the UMAX can really load and unload pages, but you
> choose not to export this behaviour from the backend?

No. The paper can not be only loaded.
Loading and scanning happen at the same time:
The backend tells the scanner to use the ADF, the
scanner does not move the scanhead, instead it moves the
paper over the fixed scanhead while the image is scanned.

>
> Can I do this (below) with the design you have in mind?
>
> 1. Put phone bill (35 sheets) into hopper, run Xsane
> 2. Scan in first sheet, hand-tune brightness and scan area in Preview
> 3. Do a final scan on that first sheet, check it for problems
> 4. Tell Xsane "Scan all the remaining sheets with those settings"
> 5. Go and have coffee
>

No, you only can scan a page only once. So there is no way
to do two scans (preview and scan) in ADF mode!

But you can do it this way:
Put the first page to the flatbed surface, the rest into the ADF.
Do a flatbed preview scan, select area and enhancement,
do a flatbed scan for the first page. The rest is scanned in
ADF mode... go and have a coffee.

Normally I would do it a bit different.
You will have different kinds of ADF scans.
For each kind you define the settings once
fax: paper size A4, scan mode lineart, threshold,
and save them.
Then you simply press start and go and have a tea.

For your backend I would suggest to check the preview bit
and if it was set while the last scan the page is not loaded for
the next scan, then you will be able to do a preview
and a scan on the same page.

>
> ALSO: Can I do this *without* XSane, in a shell script or OCR program?
>

I think you will not select a preview area or the enhancement values
the way you suggested above. Instead you will say:
scan all pages in the ADF with scansize DIN A5.
No problem for a script.
An OCR program should do it the way like xsane does it.
That is a reason why it has to be specified in the sane standard!

Bye
Oliver

--
EMAIL: Oliver.Rauch@Wolfsburg.DE
WWW: http://www.wolfsburg.de/~rauch

--
Source code, list archive, and docs: http://www.mostang.com/sane/
To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com