Image Analysis - causes of unscannable result?

[exMSW4319]

I've noticed no particular pattern to the images that IA rates "unscannable" - it can anything from a simple JPEG in a sig up to monstrous embeddings in multiple attachments, the appliance can be quiet or relatively busy, and the problem images themselves can be very simple. The only common element that I can think of is that none of the samples I've actually looked at could be deemed "inappropriate" in any way.

The Advanced Guide really only has the basics on IA - is there another source of information I should be looking at before hassling support?

Alternatively, are unscannables something that I should just expect and deal with when using IA? If so, does the forum tend to drop, notify or simply wait for end-user bleatings?

Our Asyncos version is 6.40-273 on a C150, if that's relevant.

[exMSW4319]

To answer my own question, the only thing that's come out of a support ticket with our vendor is that any monochrome image will give the Unscannable result. This of course includes many PDF and TIF files churned out by modern commercial workflow systems and the output of any mono photocopier with a scan-to-file or scan-to-mail facility.

Unless the image analysis engine works solely by chromatic evaluation, I can't see why it can't have logic that allows it to recognise that an image is mostly text and returns a more appropriate result. Failing that, would a simple fourth IA verdict "Mono" be a step forward?

We are now on Asyncos 6.5.3-007, which may also have helped though I'm not keeping any exact stats at this time.