Coolscan 9000 ED Diffuser Tests

So, gang. In the interests of science, I have been researching whether or not I am wasting my time trying to improve the apparent grain coming from my 9000. I was frustrated by the fact that I had scanned some Fuji Press 800 film containing people shots and was impressed by the improvement obtained by taping some AN glass to the top of my 35mm carrier (AN side up) acting as a diffuser. The frustration was that the overall picture looked better, more contrasty and cleaner in the diffuser scan but the 100% crops looked worse. Since attaching the glass was a pain I didn't do any more 1 to 1 comparisons until I finished scanning that series.

I have returned to my standard test image. It is a night shot on (I believe) Konica Minolta Centuria 100S negative film. Though it is ISO 100 the sky often looks grainier than I would like because there is such a contrast range in the image. I scanned it using no diffuser, the AN glass as well as two different textured white plastics that I cut from various folders and filers I found in the stationary shop.

I present the results below. Executive summary: You need a microscope to tell the difference and when you get out the microscope the plain scan looks best.

Plain:
http://www.flickr.com/photos/samagnew/5226798140/

AN Glass Diffuser:
http://www.flickr.com/photos/samagnew/5226793220/

Thick Plastic Diffuser:
http://www.flickr.com/photos/samagnew/5226788130/

Thin Plastic Diffuser:
http://www.flickr.com/photos/samagnew/5226189183/

I was going to add wet scanning to this test but before I could get the bubbles out I had already scratched my test negative pretty badly (I only wet the area of the one frame I was interested in. It bonded the film to the glass strongly and in rotating it I ended up scratching the dry area of the film even though the glass appeared perfectly clean). I was fed up and aborted that test.

Sam

My understanding is that the 9000 has a more diffuse light source compared to, say the 5000, or a 5400 without the grain diffuser. So adding extra diffusion might not contribute much. If anyone cared to make a similar comparison on a 5000, I suspect you would find a bigger difference. And I think the impact of the GD on a 5400 is well-documented.

Thanks for sacrificing a negative in the interest of science...

--Greg

I've read the same thing but my personal experience is that I see more grain in scans than, say, in projection. I still feel that there should be a way to improve the situation. My Konica Minolta Scan Dual IV seemed to manage a trick of looking sharper and seeming less grainy. However, that scanner had many worse problems -- mainly a really crummy sensor.