Hello all,

I've got an Agfa Duoscan HiD for scanning 4x5 and 6x9 chromes and some negative. The main output is a web portfolio. 1500px on the long side and 90-100 dpi. So the resolution of the scanner is more than enough for my purposes. I can get Imacon or drum scans when needed from my pro lab so I didn't considered the current Epson's. Just needed something cheap to be able to upload low res scans.

Because of this initial requirements, the low price of the Agfa and the good reviews I've read all around I thought I would be happy with the Duoscan. I can live with the limited D-Max. I noticed that in contrasty chromes when highlights are close to the limit of the film latitude I got it burned in the scan, no matter how l adjust Vuescan. Maybe this can be less of a problem if I can adjust 1/3 or 2/3 less exposure on constrasty situations, as I see that I can raise the shadows a bit with reasonably low noise. But I have to test it yet.

But I found that my unit has a banding problem.

With 4x5 chromes I get 4 or 5 equidistant bands all the way (always across the way the unit travels) of the chrome, barely noticeable on most chromes, but specially visible on skys. Most of the times I'm able to fix it in PS, by healthing the lines in the blue channel. Here is an example where I got this fixed: http://payload270.cargocollective.co...002web_o_o.jpg . But you can still see some irregularities
On 6x7 and 6x9 chromes I get cleaner results.

It's worst with negatives of any size, the bands are all over the surface, showing different densities and widths. It doesn't matter if I scan in portrait or landscape position. Lately, after some cleaning a new wider magenta line started to appear, so I thought that dust were moving from one place to another.

I've got it dust it off several times. Dissasembled the whole thing, cleaned the mirrors after guessing what the light path was, and got access to the CCD lens which I blew with compressed air and cleaned carefully.
I found out that two mirrors were very dirty, with a dust haze all over the surface. So I thought that my problem would dissapear after cleaning. CCD lens was not but I cleaned it anyway. Nothing changed with chromes and with negatives I still get a lot of small lines. I cannot tell if they have diminished after the cleaning.

Is there a common diagnose for those banding cases when negatives show more banding than chromes? I think that If it were a CCD failure (dead pixels or whatever) both chromes and negatives would show the same. But lines are completely different with each different media. I've tried to scan at 8 and 16 bits: no difference, with every Vuescan setting possible.

I've learned to let lamps to heat up for a while but I still got to stabilize current/voltage and get my PC tower far from the scanner, as now the cable is short and they are too close. Any advice on what else could I try beside trashing the unit?
Changing software, calibrating with a target?

Do you think I still have to clean further?. I got some knowledge about the unit and I think I can dissasemble the CCD unit to have complete access but this is risky as I have to plug off a lot of things.

Cheers!