Microtek ScanMaker 3600 driver / News: Recent posts

new release with respect to endinaness available (as in SANE 1.0.13)

Since I never found Glenn's hotplugging info again, I just wrote it for sm3600.sf.net.

There are some instructions about how to use the scanner with the linux hotplug system, these included with the release notes of the SANE 1.04 patch available from the files area. You can download the release notes without downloading the whole patch. BTW the 1.04 patch is obselete since the driver is now part of the main SANE distribution.

The sm3600 driver has been included in the SANE distribution since release 1.05 of SANE. You no longer need to download any of the files from this site.

An updated sane backend with brightness compensation is now available. The results are much better than the previous sane-backend-104 release.

Now the background has a constant brightness.
Try the new version of the backend!

We have made a release of a sane backend. It can be downloaded as a patch against sane-backends-1.0.4. Download it from the files area or the summary page.

The backend now supports Gamma LUT processing inside the scanner for lossless exposure control.

And the slider return is far more robust now.

The port from a standalone application to a SANE backend is in progress. This includes *some* architectural changes.

See the mailinglist archive and STAY TUNED!

Hey, we now can correct brightness and contrast.
it is great especially for copy-halftone-scanning.

Have fun with it!

Now You can scan with grayscale and black-white modes.

B/W offers line art (2 color) and error diffusion, what makes it ideal for copy scanning.

The color sensor lines are skeyed by approx. 2/300 inch. The ouput is now passed through a delay pipe.

Thus, the scans now are clean, sharp and still fast.

After Glenn pointed out, that there is a "speed" option with SW5, i optimized a bit on the scan and movement.

Now we can jog the slider at full speed.

color scans can be done with 75-600 dpi with arbitrary scan windows.

We have still problems with the slider end-of-path detection.

Some scans give RGB byte ordering, some give BRG. Relation is unclear. The scans are fully windowed.
The 1200DPI scan is performed at real 600DPI. RGB lines are parallaxed by 1/200 DPI or more and need correction.
Package is modularized now.

I figured out, how the sm3600 makes real scans. It uses an asychronous model. Now we can do the "overview" scan.