Vuescan introduces wild cards (*)

[mendel_leisk]

Part of the "what's new" description of the latest Vuescan:

 

----

 

What's new in version 8.3.09

 

Added support for wild card names in "Input|Files"

 

Added support for wild card output names when scanning from files

 

----

 

If this means what I think it means, it's a fantastic little

addition: Just put all your raws in one directory. Same story with

output. As an example, the following files could be sourced and

output in one operation:

 

By specifying source for scan-from-disk:

 

roll_*_raw_01

 

and output:

 

roll_*_01=

 

Doing scan-from-disk of a single directory containing:

 

roll_01_raw_01

 

roll_01_raw_02

 

roll_01_raw_03

 

roll_02_raw_01

 

roll_02_raw_02

 

roll_02_raw_03

 

could produce the following output files in a single directory:

 

roll_01_01

 

roll_01_02

 

roll_01_03

 

roll_02_01

 

roll_02_02

 

roll_02_03

 

Haven't tried yet, but that _has_ to be the intent.

[jtk]

Mendel, how does that significantly differ from the earlier release?

 

It's been standard to add + after the file name to get the automatic sequential numbering: eg Mendel+ would be Mendel+1, Mendel+2, Mendel+3 etc automatically. Are you describing something else?

 

How could RAW be relevant to scanning workflow, Vs TIFF? If it was really RAW (which I question) how is it better than TIFF?

 

What does "scanning from files" mean? ... Where do you find that option? I only find "enable raw from disc" in Preferences, and have no idea what that means. I wonder if it actually is properly RAW...

 

I think/thought the scanning equivalent of RAW is film, and that Vuescan's notion of RAW was technically improper?

[mendel_leisk]

John, haven't had time to check it out myself. For application of (*) see here:

 

http://www.hamrick.com/vuescan/html/vuesc22.htm#topic21

 

And for raw file concept (I thought we'd gone over this before, maybe I was bending someone else's ear?), see:

 

http://www.hamrick.com/vuescan/html/vuesc4.htm#topic3

 

I'm curious how introducing wild cards (*) effects Input|Frame number and Input|Batch list, will check out later (at work now).

[jtk]

Thanks for the links. I'll review later today..also theoretically at work :-) Maybe I'm having a senior moment about past RAW discussions.

[mendel_leisk]

Further John:

 

Previous releases allowed the following suffix 'switches':

 

(+) - does a straight increment (you mentioned)

 

(=) - matches the number at end of Vuescan raw file, in scan-from-disk

 

This release adds:

 

(*)

 

Cut and pasted from the Output Tab Section of the current users guide, this is the additional option:

 

----

 

Using same file name as raw files with *

 

Use an asterix (*) for the output file name to cause the raw file name to be used. For instance, when scanning raw files with different names that aren't in numerical order, click the @ button to the right of the "Input|Files" option, choose multiple file names, and when you batch scan these, the output files will have the same file names as the input files, albeit in different folders and with different file extensions.

 

----

 

Leads me to think it's a little more simplistic than I first thought, but still some increase in input and output options. As I said, haven't had a chance yet to check it out.

[mendel_leisk]

John, at a spry 54, I have a few of those moments myself ;o

 

Yeah, if you're using Vuescan and have the pro license, but 'Vuescan Raw File' and 'scan-from-disk' is still beyond your horizon, definitely look into it, the sooner the better, you _are_ missing out on something.

 

I create my 'Vuescan Raw File' in a hybrid fashion: through Minolta Scan Utility, by outputting 16 bit linear tiff with auto exposure off, and all exposure sliders (typically) zeroed. All in order to circumvent Vuescan's cleaning, and utilize ICE instead. As I outlined in my thread asking if it was possible to output 'raw' through NikonScan.

 

Did the bell just ring? ;)

[jtk]

Mendel, did you conclude that Ice was better in some way than Vuescan's infared?

 

Seems very unlikely...I use the lowest of Vuescan's three infared settings for all appropriate films, have seen no utility for the higher levels.

 

Also, I've compared those three relatively carefully to see if they damaged scans of B&W silver film, knowing they wouldn't clean it...there was no discernable bad effect with slight or medium infared and if there's one with heavy infared, I may be imagining it.

 

Is your interest in RAW the pursuit of a more discrete set of workflow steps? I don't understand how you'd use Ice AFTER scanning with Vuescan.

 

Oddly, I did just reinstall Vuescan, got the "Pro" confirmation, but don't have the RAW option you mention. Perhaps this is another N Vs KM difference? There MAY be a significant difference in lowest level Ice between current and older Nikons, even sharing the same Nikonscan and Vuescan software.

 

I've sent a note to Hamrick hoping he'll give me some direction if I've somehow not gotten Pro (again) despite the system's confirmation.

[mendel_leisk]

Hi John,

 

Regarding "Mendel, did you conclude that Ice was better in some way than Vuescan's infared?"

 

Yes, absolutely. Much more complete, seemless, and little or no softening, with my 5400. Still not perfect, some half-cleaned or missed artifacts remain, but one h**l of a lot better than Vuescan.

 

Regarding: "I don't understand how you'd use Ice AFTER scanning with Vuescan."

 

I do no 'actual' scanning with Vuescan, just 'virtual' scan-from-disk, sourcing from Minolta Scan Utility 16 bit linear output. In other words, all my actual scans are thru MSU.

 

Regarding: "but don't have the RAW option you mention. Perhaps this is another N Vs KM difference?"

 

Definitely not. It should be available. Look in the output tab. There should be an option to output raw file. Let me know if you don't see it in there.

 

 

Apart from the cleaning issue, which I've resolved to my satisfaction through the hybrid workflow, I'm quite satisfied doing the rest of the steps through Vuescan.

 

For my slide scanning project, I profiled the scanner through Vuescan, using an Ektachrome IT8 target slide and data file supplied through Wolf Faust. After a few 'forays' into alternate profiling software, and profile assigning experiments in PS, I've come back to using the Vuescan generated scanner profile, with scan-from-disk. I'm totally satisfied with the color balance. Spot-on!

 

I thought I'd get clever, scan-from-disk the target (just a 16 bit linear MSU output, like all my slide scans), and do the Vuescan right-click on neutral grey to set Manual color balance. Did this, say a dozen times, fed all the results (red/green/blue) into a spreadsheet, averaged them, and fed _those_ values back into the Vuescan manual color balance fields.

 

Output a few rolls that way, and found something was 'lacking', and the images had a 'cold' feel. Tried Vuescan's white balance instead, and got what I want. The fact that white balance was very close to either manual or neutral speaks volumes to me: saying that I've got it dialed in quite well, with the profile. And white balance just 'took it home', gave skin tones a more natural warmth, cleaned up the highlights. Because of scan-from-disk, I always have the option open, to redo, with some other color tab setting if I want.

 

Also, with the Wess mounts I keep babbling about, the ones that tension the film, I'm finally dialling in my focus, with a very consistant hit ratio. I've come to the conclusion 50% of the problem was film flexing during the scan. This does not happen with the Wess mounts. And the tensioning of these mounts also pulls the center dome down into a dimple, more-or-less level with the edges, leaving only little local pocket of doming in the corners, but with much less severity, and _largely_ within my scanner's depth of focus.

 

Nirvana!

[mendel_leisk]

fed _those_ values

 

should read:

 

fed _that_ value (the average value)

[jtk]

Mendel, I found the RAW. Whew! D'oh. Now I have to admit it to Hamrick.

 

How many Vuescan infared settings do you find with your KM (I)? I've discovered in another thread that Nikon 4000 seems only to offer two, whereas I have three for Nikon V. There may be a hardware issue with some scanners...

 

Using Vuescan's lightest setting, I literally never have infared issues with film suffering light airborn dust. And studying the effect of infared fairly closely, I don't believe the low setting has ANY softening effect...actually, I also doubt it with the mid setting.

 

I'm not surprised about your flex-during-scan hypothesis...but I think a scan-only from your Minolta may be very fast...

 

Can you confirm the following, emailed to me :

 

"I have the original Min 5400... it is 8 to 12 min. with the Minolta and ICE on (less than 1 min. with ICE off). I time mine with a wristwatch also which is good enough for government work."

 

I'd reported to him that Nikon V takes 2.5 minutes total with infared (same with Nikonscan 4 and Ice, same with 5400II and Ice)...1.75 Nikon mins without infared. Note that he says 5400 (I) takes LESS than a minute without Ice. Correct?

 

Nikon-Vuescan-infared/Nikonscan-Ice and 5400II-Ice are FAR faster than original Minolta's Ice if this guy's 8-12 min total is valid.

 

If he's right, and if Ice partially meddles DURING the scan (?) I can appreciate why you'd want to separate that looong Ice further from that scanning process...and why Ice might soften images more in Minolta (I) than I experience in Nikon.

[mendel_leisk]

Hi John. I thought it simplest to just cut and paste your thread, and add comments:

 

john kelly , nov 16, 2005; 04:19 p.m.

 

Mendel, I found the RAW. Whew! D'oh. Now I have to admit it to Hamrick.

 

****

 

Many Vuescan features are "understated". It's there, though. One caveat: whatever you experiment with, back up beforehand, and review the results very carefully. If you can do the exact same thing in PS, that might be better. Case in point is custom downsample, in the input tab. It's convenient, but inferior to PS bicubic.

 

Also, good idea to ensure 'warn on overwrite' is ticked in the prefs tab. I think it is by default.

 

I will never use Vuescan again involving any operation that involves modifying and saving any of original files, my MSU 16 bit linears. Too risky, to many variables. I only use it for outputting finished files.

 

****

 

How many Vuescan infared settings do you find with your KM (I)? I've discovered in another thread that Nikon 4000 seems only to offer two, whereas I have three for Nikon V. There may be a hardware issue with some scanners...

 

****

 

I have (the usual?) 3 settings. Don't use any of them ;o

 

****

 

Using Vuescan's lightest setting, I literally never have infared issues with film suffering light airborn dust. And studying the effect of infared fairly closely, I don't believe the low setting has ANY softening effect...actually, I also doubt it with the mid setting.

 

****

 

Could be hardware related. Ed Hamrick 'likes' Nikons, for various reasons he's cited on the Usenet scanner forum, a while back. Perhaps he's able to implement Vuescan's cleaning better with the Nikons, as well.

 

****

 

I'm not surprised about your flex-during-scan hypothesis...but I think a scan-only from your Minolta may be very fast...

 

Can you confirm the following, emailed to me :

 

"I have the original Min 5400... it is 8 to 12 min. with the Minolta and ICE on (less than 1 min. with ICE off). I time mine with a wristwatch also which is good enough for government work."

 

****

 

That wasn't me was it? Sounds like one of my expressions (gov. work). My time for scans varies for slides and color negs.

 

I have a very esoteric workflow for color negs, and my scan time comes in around 13~14 per. That something out-of-the-ordinary, due to manual exposure coupled with large adjustments to the color channel exposures. I'm probably going to avoid color neg film in future. Just too hard to scan. Well, I'll likely avoid ALL film from now on, haven't shot any since getting a 20D.

 

Now, with slides, my scan time is 5~6 minutes. That's telling msu the media is color slides, with auto exposure off, and all exposure sliders at zero (the middle setting). Scan time is directly related to exposure. My turn-around time is more like 8 minutes, in other words, the time the scan is output, from one to the next. This is largely due to my careful manual focus at each frame, and msu's insistance on doing a fresh pre-scan after the focus.

 

All of the above times are with ICE and Grain Dissolver on. GD is not optional with ICE: turn on ICE, MSU automatically turns on GD. With neither, the times around 1.5~2 minutes, guestimate. With GD alone maybe 3.5~4 minutes, another guestimate.

 

****

 

I'd reported to him that Nikon V takes 2.5 minutes total with infared (same with Nikonscan 4 and Ice, same with 5400II and Ice)...1.75 Nikon mins without infared. Note that he says 5400 (I) takes LESS than a minute without Ice. Correct?

 

****

 

Under 1 minute for a 'bare' scan with the 5400. I dunno. With my PC, the scanner 'pauses', maybe 1/2 dozen times during the course of the scan. I suspect it's waiting for the data transfer to catch up. But I have the next to best Intel chip, a gig of ram, and dedicated drives for receipt of the scans. So, dunno about that, improbable?

 

In any event, the slide scan time with ICE and GD (5~6 min) works for me. I can just keep up, with a little slack, dismount/remounting slides as I scan.

 

****

 

Nikon-Vuescan-infared/Nikonscan-Ice and 5400II-Ice are FAR faster than original Minolta's Ice if this guy's 8-12 min total is valid.

 

If he's right, and if Ice partially meddles DURING the scan (?) I can appreciate why you'd want to separate that looong Ice further from that scanning process...and why Ice might soften images more in Minolta (I) than I experience in Nikon.

 

****

 

You lost me:

 

1) In my experience, Vuescan cleaning is softening more than ICE. I'm not finding any appreciable softening with MSU's implementation of ICE. Well, let me temper that: MSU's ICE operates only with Grain Dissolver on. Grain Dissolver sort of 'mutes' the grain. It also 'takes the edge off' damage. Also, it more-or-less eliminates pepper grain, which you see in films such as Fuji Provia. Per my examples, in this post of yours:

 

http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=006Uxf

 

2) ICE can only be done at time of scan.

 

3) Vuescan cleaning can be done post-scan using 64 bit Vuescan Raw File (the extra 16 bits are the infrared channel info). The Minolta Scan Utility 16 bit linear tiff does not have this capacity. But, I don't miss it. It's more a complication than an aid.

[mendel_leisk]

Getting back on-topic: I could not get the new (*) naming feature working. Tried everything. FYI, when I went to download the version was up to 8.3.10.

 

In Input|Files, you just select multiple files. Vuescan stacks them all in that field, deimintated by semi-colons. That bit seems to work ok. You can scan them all, though the frame order is a little juggled. Not a problem IF you intend to scan all. But then it goes downhill:

 

In Output|Jpeg file name, I tried following the help file guidance(which seems to be disabled in this release, a bug rears it's head). Windows was not happy nameing a file "*". Hmmm, something not right. Tried naming it blank: got the frame number as a name, not too bright. Tried toggling auto file name on/off. Still stymied.

 

Anyone else have any luck? If no, I'll probably just watch the new features list for a few more releases. I'm sure the bugs will get sorted.

[mendel_leisk]

Update: after some some emails with Ed Hamrick, and his supplying me with a beta version, the wild card feature is now working, *almost* 100% as advertised, for me, on Windows XP. I'll post another update, with a walk-through on using this new feature, when a patched release comes out.

[mendel_leisk]

Ok, version 8.3.11 is released, and as reported, wild card capability is fixed.

 

You can enter wild card (*) in file name in either (or both) "Input|Files" (when scanning from disk) or "Output|Jpeg(Tiff) file name".

 

The name can be stand-alone (relying on "Output|Default folder"), or have a path prefix.

 

If for example, you have a raw file directory containing:

 

Directory of E:\os-raws\test:

 

osraw001-0001.tif

 

osraw001-0002.tif

 

osraw002-0001.tif

 

osraw002-0002.tif

 

 

And:

 

For Output|Default folder you specify: E:\os-raws\test

 

For Input|Files you specify: *.tif

 

For Output|Jpeg file name you specify: c:\my documents\*.jpg (for example)

 

You will get jpeg output of all the raws in the default directory, placed in the "my documents" directory.

 

If you revise "Input|Files" to: osraw001-*.tif

 

You will get jpeg output of only the first "roll".

 

The Output|Default folder can be set so as to avoid having to put path statement in either of the file description fields, or for a third directory (if you put path statement in both the file fields).

 

Very nice new convenience. You could literally crank out 1000's of scan-from-disk images, in an overnight operation, with names specifying different rolls.