Is Minolta 5400 banding caused by software?

Hello to all. This is my first post.

 

I started reading this forum regularly about a month ago when I began

shopping for a 35mm scanner. I read many posts and thanks to this

forum I was able to learn alot about scanners, color management and

workflow. Thanks to all posters who take the time to share their

knowledge.

 

I ended up purchasing the Minolta 5400, mostly because I need to

print big (36"x24") and its the only scanner whose native resolution

would allow me to do that at 200dpi.

 

That being said, the purpose of this post is to share an experience I

just had, and who knows, it might help somebody out.

 

Out of the box, I scanned about 10 slides at different resolutions,

with/without ice, 16 bit/16 bit linear, etc. and was happy with the

results (using DiMage software). I finally decided 16 bit linear with

ice was the way to go. Then assign pos lin profile and convert to

Adobe RGB (1998) in PS CS, color correction and USM.

 

A couple of days ago I noticed a newer version of the software on the

Minolta site (ver. 1.1.2) and installed it. From that point on, I

started noticing horizontal lines in the dark areas of the images. I

tried many possible solutions (isolating the scanner, with/without

ice, 16x on and off, etc.) to no avail. I had deleted all of my

previous scans (cause they were all tests) and could not check if the

problem was new or not). I was getting pretty pissed off about these

lines (they were visible at 100% image size and they seemed to

increase with the 16x multi-sample). I thought I had a defective

scanner.

 

Then I simply reinstalled the DiMage software that came with the unit

(ver. 1.1.1) and low and behold, the lines went away! So far, I've

made 4 scans at 5400, 16 bit linear, 16x, with ice (over an hour per

scan!) with no lines and I hope never to see them again.

 

I've read a couple of posts about these lines with the 5400, I don't

know if the problem is the software fo others as well, but I invite

anybody experiencing this problem to consider their software as a

possiblity.

 

Cheers,

 

Richard

Hi Richard,

 

I am also using the scanner, no line problems, but I wondered what is the difference with 16 bit linear from 16 bit?

I have alot to learn about scanners and user forums.

 

1. Make many tests before coming to conclusions: after more testing, I realised that I still have the horizontal lines (banding). I just wasn't looking careful enough before. The lines do increase with multisampling.

 

2. Do not announce results to the world (ie, this forum) until you are 100% certain of what you are saying. That way, you won't have to retract anything. So here goes, I retract what I first said about having solved the banding.

 

About 16 bits linear:

From what I understand, 16 bit linear is the raw, unprocessed data from the scanner's ccd's.

 

Here is an example of the banding. It is scanned at 5400, 16 bit linear, 16x, no ice, 100% with no magnification.

 

If anybody has any suggestions, I would love to hear from you.

 

Richard<div></div>

Can you try a scanner reset from Minolta softwware ?

<br>

Push CTRL+SHIFT+I

<p>

I have the same banding problems with an older Minolta Elite (tests here: <a href="http://christian.tsotras.free.fr">http://christian.tsotras.free.fr/</a>).

Yes, I have tried the reset within the Minolta software. Still banding. If the reset reduced the banding (as in your examples, well documented, I might add) I am not sure but they are still prominant.

 

Does anybody scanning with the Minolta 5400 not have this banding problem? I was planning on sending these scans to get printed, will these lines be seen in the prints?

 

If anybody had any other suggestions, please let me know.

 

Richard

I have had similar hoizontal line problems and sent my 5400 scanner back to Minolta for repair. No real improvement after losing the use of my scanner for a few weeks!

 

This problem seems to be very common with the 5400. You should not see any problems with small prints but with enlargements (especially of cropped images) this may ruin the prints.

 

There should be some way to get Minolta's attention to fix this problem. If not, I may have to return my scanner and buy the new Nikon one.

Hello Richard,

 

Making three variations of the same scan

 

1. 16 bit -> assign Adobe RGB (1998)

 

2. 16 bit -> assign Dimage 16 bit ->convert to Adobe RGB (1998)

 

3. 16 bit linear -> assign Dimage 16 bit linear ->convert to Adobe RGB (1998)

 

I have noticed strong increase of the noise in the raw 1<2<3.

The idea is to make scan as in the procedure '1'. Colors are slightly different in all procedures, but so is noise level.

I see a lot of noise in shadows in blue channel (inverted->blue) when scanning negatives in the procedure '2' and '3', but can hardly notice it in '1'. I hope this can affect the lines level as it decreases overall noise, in negatives and in slides as well.

 

Your lines are too strong, and they hopefully can be reduced by not using 16 bit linear procedure and following conversions.

 

Still, it is only my observations...

 

Hope this helps.

 

Ruslan.

Thanks for your answers.

 

Mark and Ruslan: I don't think you should put up with those lines. My understanding (after reading several threads on this forum) is that with respect to the lines and the Minolta 5400, you either got them or you don't. Many people seem to have no problems with their scanners, so why put up with one that does not do what it should?

 

 

Ruslan: I tried the three variations you suggested but found little difference (if at all) between them.

 

Richard

Hmmm.... I would be very surprised to hear that 5400 unit without lines really does exist:) Some people (of course, you know) exchanged their units several times with no effect.

 

The question is how strong the lines are. Everything else in this

scanner is as proposed and amazing.

And lines should be an issue to change your mind before buying 5400, not after...

 

Also, exchange is not easy procedure for me now.

With negatives (the only films I am currently using) there is almost

no lines problem.

 

Regards,

Ruslan.

i have gone out to 16 x 24 prints with no sign of lines on print or scan.

Well, I can't compare to too many other film scanners but I can tell you that the Minolta 5400 has way less banding problems than the Kodak RFS3600.

 

I always found (with both of them) that banding occured most when trying to rescue an underexposed slide. This says to me that it's somehow gain related. Sometimes, with my Kodak, if I'd shut it off and let it cool off I could re-scan the same image without the banding.

No lines here and I've inspected full resolution scans at 400%+. I wouldn't settle for lines like you have and I'd take samples and the scanner to Minolta for repair/replacement.

I had a faint greenish pink line running through scans this appeared after I had the dimage 5400 for 4-5 months, after some discussion the unit was returned to Minolta who decided it was at fault. They replaced this with a scanner that failed to operate. Another machine was supplied and this worked fine for about 3 weeks now I think the banding is starting to show again.

It certainly was not there when I started using either of the two scanners but it seems to come in later, may be worse when they are warm? who knows.

 

Martyn