How is Vuescan working with Elite 5400?

[mendel_leisk]

The Konica-Minolta Scan Elite 5400 is my first choice for upcoming

scanner purchase for several reasons:

 

1. I am not a professional, scan time is not my first priority. As

long as I can do one slide holder (or strip holder with 4~5 frames)

every half hour, I'm satisfied, though not thrilled. That is assuming

ICE and GD enabled.

 

2. The light source has been noted by several reviewers and users as

producing good results with both Kodachrome and silver based black

and white film.

 

3. The high resolution, coupled with the grain dissolver, seem

capable of extremely clean and detailed scans, with experience and

coupled with the right software.

 

I'm concerned by the problems Ed Hamrick has had assimilating this

scanner with Vuescan. I would much prefer to use the scanner with

Vuescan, for the usual reasons. I would like to hear from Elite 5400

owners who are using (or have used) Vuescan if the Vuescan problems

with the Elite 5400 are ongoing, or resolved. Some things I've read:

 

1. Vuescan was having problems "initialising" the Elite 5400, with

various poor scanning results, I think way-off color casts, banding,

and the like. Recently, Ed Hamrick posted on Usenet, saying to start

Minolta sw first, intialize the scanner, close the Minolta sw, then

open Vuescan and run its initialisation. This he termed a work-

around, and he said he is working on the problem.

 

2. Vuescan was having problems dealing with the infrared channel, mis-

cleaning adjacent to the damaged area, softening large areas,

etcetera. Also, and perhaps associated, the Vuescan cleaning was not

as thorough.

 

Anyone using this scanner with Vuescan, please respond as to how it

is working for you, at present, assuming you're using latest Vuescan

version. Thanks in advance.

[mike_zallini]

I find Silverfast is better; but the answer to your question is 'not great'; especially for color negatives. Slides work much better. Have not tried Kodachrome or B&W with Vuescan.

 

I don't like the levels adjustement controls in Vuescan. They are clunky at best. SF on the other hand has the best I've ever seen; better than Photoshop too.

 

But if you're used to Vuescan you'd probably be able to get better results from it than I can...

 

Mike.

[stephen gallimore]

Mendel,

 

I have been unable to get ViewScan to work acceptably with either transparencies or negatives with the DSE5400. This includes after initializing the scanner using the Minolta software. In fairness I haven't tried very hard, unlike some other people, at this and clearly others have managed to make it work for them. As I have mentioned in a couple of other threads I am just using the Minolta software (in its stand alone version) and am getting excellent results for transparencies with films I have profiled using the Wolf Faust IT8 targets. Just for experimentation I decided to go back and scan a couple of old negatives and some rare (for me) B&W to see what the Minolta software made of them, you can see the results here:

 

http://www.photo.net/photodb/member-photos?user_id=888858&include=all

 

The Minolta software actually did a pretty good job with the colour negative stock and for the small amount I have I see no reason to use anything else. Sorry I don't have any Kodachrome.

 

Personally I am not missing VueScan at all, but I always had an on/off relationship with it on my old Coolscan IV.

 

Regards,

-stephen

 

BTW 30mins for a strip of 6 is probably pushing it with ICE enabled once you add in mucking about with the focusing and exposure. It will definitely take longer if you use 2x or 4x multisampling.

[mendel_leisk]

Of interest: Wilfred van der Vegte offers two Photoshop Actions that work with the Vuescan IR channel in Vuescan raw file. I haven't tried it (I don't have a scanner with infrared), but have read through his helpfile and read step-by-step through the actions themselves, in Photoshop. It seems a logical and direct way to work with the IR channel. Of course, it all "open source", you can see what it's doing, modify it, try variations of settings. It dowmloads as a zip with contained html help.

 

http://www.vandervegte.com/Actions.html

[mendel_leisk]

Mike,

 

Thanks for responding, I gather you prefer Silverfast (or Minolta) regardless of the hardware.

 

Stephen,

 

If the initialising work-around is not working, it makes Vuescan (currently, at least) a lame duck with the 5400.

 

I'm scanner shopping. I have just completed scanning of about 1800 mostly Tri-X images with Scan Dual II, utilizing Vuescan with it's Raw File workflow. It has taken almost 2 years: a month or two to acquire the scans in Vuescan Raw File format, the rest of the time cleaning these, in Photoshop.

 

I now want to address 1500~2500 mounted slides (about 50% Kodachrome), and maybe 1000~2000 color negatives, the latter cut in the usual strips of 4. I would like to do them similarly to above paragraph, albeit with ICE.

 

One (rambling) thought I'm trying to hold onto: choose the hardware that best suits ones needs (not an easy decision, considering light sources, holders, speed, dusty mirrors, dynamic range, fringing, banding, etc.), as long as there is currently SOME way to get decent and useful results from it (hopefully getting at least part of the workflow through Vuescan) and pray that the software will evolve to meet ones needs better in future. Hope and pray...

 

Bart van der Wolf, who posts often on Usenet, advocates outputting "16 bit linear" through the Minolta software, for subsequent use as Vuescan Raw File, for now, if you want to use Vuescan. He too was unhappy with Vuescan initialising the 5400. He has a very interesting Java image comparing the 5400 with ls-4000, here:

 

http://www.xs4all.nl/~bvdwolf/main/foto/scan/se5400/se5400-5.htm

 

With ICE enabled (and GD, no choice), I understand the Minolta "16 bit linear" file will be cleaned. In other words, it cannot be 64 bit, having the infrared channel separate. This is not necessarily bad, just limiting. You cannot then work with the infrared data, either with Vuescan or Photoshop. Might be fine for me. I would be happy to output this file uncropped, open it in Photoshop, carefully crop right out to the usable edges, rotate as necessary to landscape, and then start using it for Vuescan scan-from-disk. FWIW, you can even output a fresh Vuescan Raw File from this file. It will be identical, except better lzw compression is currently possible through Vuescan (actually PS's lzw compression is often inflation!).

 

I found your recent posting regarding your experience with both the Minolta 5400 and Nikon 5000 very interesting. Thanks for your input.

[marcos_rodriguez]

Mendel,

 

I recently purchased a Minolta 5400 and, after several weeks of trying both VueScan and SilverFast, I've finally found a way to achieve good scans (from negatives) using Minolta's software and Photoshop 7 or CS.

 

The key is to scan as positive, and do all the color correction and balancing in Photoshop --which is easier and more straightforward than you may think.

 

I intend to write a tutorial on this, but finding the time to do so will be the real challenge. I'll try to summarize as best as I can the steps involved:

 

* Note: Before you start, you need to locate and install a Minolta color profile that is not installed by default. The file name is "MLTF5400n.icc", and on windows you'll likely find it in:

 

C:\Program Files\DiMAGEScan\DS_Elite5400\Profile

 

Once installed, you can see the profile in Photoshop as "Minolta DiMAGE Scan Elite 5400 (Nega Input)"

 

----

 

SCANNER SETTINGS:

 

. Color Positive, Manual Exposure, 16Bit, ICE on or off as desired.

 

----

 

WORKFLOW:

 

- On DiMAGE Scan:

 

. Do a preview and go to the exposure tab; put the following values for each channel:

Red 0, Green 1.2, Blue 2.0 (this helps eliminate most of the orange mask by balancing the histogram)

 

* Note that these settings will work most of the time. For badly under- or over-exposed frames, however, you may want to fiddle with the Master slide in order to broaden the histogram as much as possible, although it has been my experience that this is seldom needed.

 

 

- On Photoshop (I have version CS, but version 7 will also work. I don't think prior versions will work, though.)

 

. Load file

 

. Invert (from negative to positive)

 

. Assign profile "Minolta DiMAGE Scan Elite 5400 (Nega Input)"

 

. Convert to profile "Adobe RBG (1998)"

 

From here I'll point you to a wonderful Photoshop tutorial by Bruce Fraser (CreativePro), from which you can follow through. The only trick is to first crop any white area that is not part of the original image, this in order to obtain an accurate histogram from which to work:

 

http://www.creativepro.com/story/feature/17164.html

 

 

After you have balanced the histogram and achieved a decent scan with no clipping on either highlights or shadows (which is the main reason not to use Minolta's software), it is only a matter of personal taste to adjust contrast as desired and enjoy a great image without paying $200 dollars for SilverFast (which is quite slow and requires extensive, frame-by-frame tweaking in order to produce acceptable results) or dealing with the bad ICE implementation of VueScan.

 

The main advantage I see with this approach is that it gives the user complete control over how the image is processed, and best of all it provides a very consistent workflow from which little deviation is needed --as opposed to SilverFast, for instance.

 

Anyway, my two cents.

 

Marcos Rodriguez

 

www.aukeramen.com

[mendel_leisk]

Marcos wrote in part:

 

. Load file

 

. Invert (from negative to positive) <comment 1

 

. Assign profile "Minolta DiMAGE Scan Elite 5400 (Nega Input)" <comment 2

 

. Convert to profile "Adobe RBG (1998)"

 

>comment 1: I've done workflow putting this step last

 

>comment 2: I do sim. with Scan Dual II, except using "posi" version

 

Marcos, I've tried this conversion using Vuescan raw files (with above comments), which are very close to Minolta's "16 bit linear" files. Results were "interesting". Most of what I've been scanning is black and white silver based, up to now. What I found with this sort of workflow, compared to my preferred scan-from-disk with Input|Media set to B/W negative and Color|Film profile set to Tmax400 (contrast index D76ci:.55), with Vuescan:

 

1. Contrast was higher, but too much for my taste. Shadow detail was blocking up.

 

2. Brightness of the results had to adjusted, frame by frame, much more so than Vuescan with autoexposure. I'm not happy with the latter, sometimes, and need to adjust, but much less often.

 

3. Highlight tonality was washed out. Not blown, but just very flat.

 

I'm really interested in the interface between Elite 5400 and Vuescan, but thanks anyway for your workflow tips. I have tried similar, and it all helps in understanding the scanning process.

[mendel_leisk]

Marcos,

 

Regarding the link you posted, for PS autocolor tuning:

 

I'll sometimes output a "flat" image in Vuescan, leaving Color|Color balance on "none", and brightness 1.0 (instead of usual 0.7 setting, with b/w). Still scanning as B/W neg with TMax profile, though. Also, I would output 16 bit greyscale, to give more adjustment room.

 

I would then open this in Photoshop and work with autocolor in levels, similiar to the link you posted. Usually, I just rely on Vuescan to balance, but with problem images, say with subjects in deep shade with bright sky over, that sort of thing, the "flat image into PS" approach can help.

[marcos_rodriguez]

Mendel, one thing I would like to make clear is that Posi 16 Linear is not the same as Color Positive, which is what I use for negatives.

 

I've found that using Linear *with negatives* produces scans that are too contrasty and lacking detail, this because this setting is intended for slides.

 

About scans lacking punch (flat images) that's just an indication that no data has been lost while processing the image, which is what is desired, as it gives the user absolute control over how much information should be thrown away.

 

Marcos Rodriguez

 

www.aukeramen.com

[mendel_leisk]

Ok, I just went through the exercise, scanning 16 bit linear with Minolta SW, using a test slide which I've previously used to produce a Vuescan Raw File.

 

I followed through, using this Minolta 16 bit linear file as a Vuescan raw file, and followed the exact same workflow. The two output can be seen here:

 

http://www.photo.net/photodb/folder?folder_id=419706

 

All the histogram statistics are within 1.0~. Download and toggle between them, if you would like to see how close they are.

 

So, I think I've resolved a work-around that will work for me. 16 bit linear output from Minolta can subsitute for Vuescan Raw File, the only downside being the ICE treatment HAS to be incorportated, and Grain Dissolver is on (no choice).

[mendel_leisk]

Marcos, I understand the difference between Minolta's color positive and 16 bit linear. The latter, as far as I can tell, is a dead-ringer for Vuescan Raw File: gamma 1, no modifications etc.

 

The flat highlights I see comparing to my Vuescan output is due to some kind of curve Ed Hamrick is using, with the B/W profiles. I would also get flat highlights when I Vuescan scan-from-disk as image, followed by inversion in PS.

 

My object with this thread is to ensure I can use Vuescan, at least for scan-from-disk. Per my above post, I think I can do that, so I'm happy.

[wilfred_van_der_vegte1]

Mendel, looking at the two images, which seem identical to me, I don't see why you'd want to import 16-bit linear files from the Minolta software into VueScan. If the VueScan results are identical, then what's wrong with them? Personally, VueScan works OK for me with the 5400 - the IR actions I created and which you mention in one of your posts are no longer needed, and the problems with shadow detail can be solved indeed by running the Minolta software before VueScan.

[mendel_leisk]

Wilfred, I neglected to clarify, my experiment was with Scan Dual II. Considering they are 2 separate scans, with slightly different cropping, I too am impressed at how close they are. My concern is that a Vuescan Raw File from the Elite 5400 would be problematic, due to initialization problem, or whatever.

 

From what you say, I gather Vuescan has improved and is working for you with this scanner, albeit with the first-running-Minolta-Software work-around. Thanks for this info.

 

How are your times?

[wilfred_van_der_vegte1]

Times? You mean scan times? I never measured them - the scanner is slow, but the possibility to scan in batches compensates for that. My impression is that VueScan scans faster but it needs a lot of time for post-processing and saving, whereas in the Minolta software the image has been post-processed and saved immediately after scanning. Of course, VueScan makes it possible to use IR cleaning without grain dissolver - this reduces exposure time and scanning time.

[mendel_leisk]

Yes, it is scan times I'm wondering about. There is a lot of reports, already, say: around 6 min with ICE GD 1x on slides. So GD does bog things down, due to light reduction, though I think it is a real plus.

 

I'm thinking to quite waffling, and get one.

 

One simple thing to speed things up a bit, I would order extra holders, for loading while the other's scanning. Or for when one breaks, I hear some closer on the holder is fragile.