Vuescan Q for the Coolscan 9000 users out there

[erl]

Advertisement (gone after registration)

Pete, that's an interesting POV. I would need more info to be convinced of it though. In my simplistic POV, the final scan does the work of collecting the data I finally use. If it differs from the pre-scan info, I would expect the pre-scan to be ignored. If I change anything in the pre. eg. density, I would expect tyhat to be followed through despite the original pre-scan.

 

However, I am no guru and may easily be wrong.

 

basically, I expect the scanner to do just that, with some minor adjustments by me maybe. Essentially, 'heavy lifters' like PS should be used to do more accurate adjustments, IMO. I also acknowledge that individual techniques can be quite valid if they work.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[Stealth3kpl]

For the idiot of the group, can you please explain why is helps to set the preview to high resolution?

 

Here's where I saw it:

http://www.l-camera-forum.com/leica-forum/film-forum/228394-plustek-scanner.html#post2010551

 

Pete

[Stealth3kpl]

In my simplistic POV, the final scan does the work of collecting the data I finally use. If it differs from the pre-scan info, I would expect the pre-scan to be ignored. If I change anything in the pre. eg. density, I would expect tyhat to be followed through despite the original pre-scan.

.

 

This seems intuitive, it was certainly my thinking, but Steve has experienced a difference so I defer to him.

Pete

[erl]

This seems intuitive, it was certainly my thinking, but Steve has experienced a difference so I defer to him.

Pete

 

I too respects Steve,'s opinions and his observations.

 

Curiously, my default method is to prescan at 2000 and scan at 4000. That complies with Steve's theory. I will now experiment with lower res for Prescan to see what I can notice.

[plasticman]

I might be wrong, but are we in danger of over-complicating things?

 

For a long time now I've been sticking to a very automated approach to my scanning, which relies on presets which I've configured for each type of media to first capture 12 (or 3 for 120 film) thumbnails with Vuescan's interpretation of the film used, which I save to an indexed folder for each roll, and then load other presets which are configured purely on the basis of image-spacing but which otherwise capture a full-scale (4000dpi) linear negative with no color settings or any sort of preprocessing in Vuescan.

 

I find that a roll of film often includes images that are really different in exposure values, subject, lighting and so on. I can't imagine what this would involve for tweaking in Vuescan if I were to adjust on an image-by-image basis - it would surely take half a day to get one scan right?

 

I have to qualify this by saying that none of my images are critical these days - I've been on paternity leave for eight or nine months now, so apart from deadpan landscapes that I occasionally get the chance to photograph, everything else I do is purely family snaps.

 

I'm also convinced of another critical fact: in my view, if an image looks right to you, then it is right. My final images often have a color-cast (most often a warm cast). I don't care about this: if I wanted a scientifically accurate image I wouldn't be using film. ymmv.

 

The main key factor in my workflow is undoubtedly ColorPerfect - opening the negative scan in CP usually gives me a pretty good starting-point, and I'll usually just pull down the highlights and pull up the shadows a little (much better with the latest version), and if necessary I'll color balance with a WB adjustment. Each image will rarely take more than 5 minutes of my time*.

 

Together with the automated scanning process (which I just leave to tick along while I'm doing other stuff), I get images that make ME happy without dedicating my life to sitting in front of the computer screen (which already takes too large a chunk of my waking hours).

 

My workflow is NOT a perfectionist's model, though - but it works for my purposes.

 

* to be fair, sometimes sitting in front of the glowing screen after the kids have been put to bed and tweaking images in Ps for hours on end is really fun.

[plasticman]

That post was v long in retrospect, so here's the TLTR version:

 

Vuescan / Coolscan 9000 workflow:

 

- I've saved a set of vuescan.ini files with presets (see image)

- I load each preset as required - low-res presets save 150dpi thumbnails to specified folders, hi-res presets save 4000dpi linear negatives to other folders

- I scan 12 (or 3) thumbnails with Vuescan's presets for the film used, then review the images to see which images I want to scan hi-res without any film settings.

- Post-process in ColorPerfect with (usually) minimal manipulation

 

PS: apart from the grayscale/color settings, the only differences in the hi-res presets are based on negative size and positioning (for example, negatives are differently spaced for the Mamiya 6 and the Hasselblad).

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[plasticman]

Advertisement (gone after registration)

Just one more point to add (while doing the laundry) I always do the large scan at the scanner's highest resolution. Storage is cheap these days, and for me the most time-consuming part of scanning is actually getting the negs in and out of sleeves and into the film holders, and then the previews etc. Scanning at 4000dpi may take slightly longer, but I know that I've captured the very best base file I can, once and for all.

 

Combined with this, I never overwrite the original negative scans - I treat these like digital raws, and use 'save-as' when converting and post-processing. (This probably strikes most people as a lot of trouble to go to, for a few family snaps).

[philipus]

Why, oh, why did I leave Sweden? Incredible benefits. I have just used 1,5 of the 4 paternity weeks I get from my work Still absolutely priceless of course.

 

It is fascinating to compare scanning workflows. I like very much how you do it, plasticman. Mine is very similar. The two exceptions are:

 

1) that I skip the thumbnail step, and

2) scan at "Print" resolution.

 

The reason I don't do thumbnails is that I find Bridge offers good enough overview of my images. If necessary I go back to better images to scan at maximum resolution, but I don't do this as a matter of course.

 

I also have presets (though not as nicely organised ) which save a lot of time. I have two versions of each actually, one for the 9000 and one for the V ED, because Vuescan varies a bit between the scanners.

 

I do agree about colours and find that ColorPerfect really is an easy way to achieve a colour balance that I like. Still, I will occasionally tweak the colour balance in Photoshop, esp. if I didn't manage to set a similar balance in CP across images shot at the same time.

 

I also treat the TIFF files as digital negatives. I usually spot b/w images/develop colour images in CP while scanning and then save them as "originals". Then I use various actions I have set up in Photoshop. For b/w I used the excellent actions Tim Gray posted on his site as a starting point; very helpful, I found. For colour images I have created an action of my own that takes the images into LAB (my fave colours space), makes a few basic adjustments and then saves it as a jpeg. This saves a lot of time and gives me results I am pleased with.

 

That's about it.

[Stealth3kpl]

What sort of changes do you make in LAB?

Pete

[philipus]

The action I have created is very basic.

 

It switches to LAB and then adjusts by equal amounts the a channel's green and magenta points and the b channel's yellow and blue points. I don't remember exactly now the values I have set but it is a few points only. This makes the colours pop a little bit; it is as if a veil is removed from the image (and actually LAB can be used to almost see through mist in images - sounds odd but try it sometime).

 

Then the action adds a very minor S curve to the L channel, applies a minor unsharp mask, and saves the image.

 

I have set the action to display the dialog boxes so that I can tweak each image beyond what the action itself does. The default changes to the a and b channels are not always appropriate for every image.

 

For anyone interested in the LAB colour space, which I personally think is the coolest thing in PS, there's a brilliant book by Dan Margulis called "Photoshop LAB Color: The Canyon Conundrum and Other Adventures in the Most Powerful Colorspace". The basic LAB adjustments my action makes are based on this book. I don't understand even half of what the book covers but it shows some really impressive stuff that can be achieved by going into LAB. Harold Davis has also written a few articles about LAB adjustments over at photo.net, all of which are very good.

[roguewave]

The action I have created is very basic.

 

It switches to LAB and then adjusts by equal amounts the a channel's green and magenta points and the b channel's yellow and blue points. I don't remember exactly now the values I have set but it is a few points only. This makes the colours pop a little bit; it is as if a veil is removed from the image (and actually LAB can be used to almost see through mist in images - sounds odd but try it sometime).

 

Then the action adds a very minor S curve to the L channel, applies a minor unsharp mask, and saves the image.

 

I have set the action to display the dialog boxes so that I can tweak each image beyond what the action itself does. The default changes to the a and b channels are not always appropriate for every image.

 

For anyone interested in the LAB colour space, which I personally think is the coolest thing in PS, there's a brilliant book by Dan Margulis called "Photoshop LAB Color: The Canyon Conundrum and Other Adventures in the Most Powerful Colorspace". The basic LAB adjustments my action makes are based on this book. I don't understand even half of what the book covers but it shows some really impressive stuff that can be achieved by going into LAB. Harold Davis has also written a few articles about LAB adjustments over at photo.net, all of which are very good.

 

 

An excellent reference. Dan is the source of all wisdom in these regards. The most important contribution about processing in the LAB space is that it's the easiest to replicate, as there are numerical values associated with each of the 3 components. It's the closest thing to a scientific approach. Dan's my hero. Don't forget to read up on the best ways to convert the LAB files back into the wide array of colorspaces in current use. Have fun!

[Stealth3kpl]

I'm glad the subject of ColorPerfect came up again. I've been having a play with my new Plustek 8100 and I think I'll settle into a routine with Vuescan and ColorPerfect rather than the bundled Silverfast. I've had some very pleasing results today.

Pete