Improved scanner software

On Nov 7, 2006, at 7:20 AM, Paul Atroshenko wrote:

> Is there other software that can bring better results, especially
> in terms
> of sharper images? Or does this basically depend upon the sharp
> qualities of
> the original slides?

Well, of course, the better the original, the better the potential
output. I use two 3rd party scan utilities for my work: Silverfast Ai
Studio from www.silverfast.com, and Vuescan from www.hamrick.com. For
batch processing of slides/transparencies I generally wind up
preferring Vuescan on my particular setup, which isn't the same as
yours. For batch processing of prints, particularly damaged, older
photos, I wind up using Silverfast Ai Studio most of the time. There
are cheaper versions of the Silverfast product that compete price-
wise with Vuescan. But in the final analysis, it will depend upon
which product works best with your scanner and which product fits
best into your work flow methods. Silverfast has some outstanding
tools for the digital photographer and a new print management
software that's getting very good reviews at this time. Since I use
more than one scanner, and more than one printer, and more than one
piece of software, I really needed a 3rd party solutions over the
tools generally provided by the scanner manufacturers, and
particularly since I don't use the same manufacturer of printers or
scanners.

I would also note that there are various websites with helps and
tutorials for managing scan projects using either of these tools, and
they both enjoy a large customer base. Everything from tweaking your
color matching to print output can be found on various websites and
forums. The learning curve for either can be rather long, but I don't
consider them very steep, really. Both work well as a front end to
Photoshop, particularly if you are going to scan them as 48bit raw
images.

Jeffrey

On Nov 7, 2006, at 7:20 AM, Paul Atroshenko wrote:

> Is there other software that can bring better results, especially
> in terms
> of sharper images? Or does this basically depend upon the sharp
> qualities of
> the original slides?

Well, of course, the better the original, the better the potential
output. I use two 3rd party scan utilities for my work: Silverfast Ai
Studio from www.silverfast.com, and Vuescan from www.hamrick.com. For
batch processing of slides/transparencies I generally wind up
preferring Vuescan on my particular setup, which isn't the same as
yours. For batch processing of prints, particularly damaged, older
photos, I wind up using Silverfast Ai Studio most of the time. There
are cheaper versions of the Silverfast product that compete price-
wise with Vuescan. But in the final analysis, it will depend upon
which product works best with your scanner and which product fits
best into your work flow methods. Silverfast has some outstanding
tools for the digital photographer and a new print management
software that's getting very good reviews at this time. Since I use
more than one scanner, and more than one printer, and more than one
piece of software, I really needed a 3rd party solutions over the
tools generally provided by the scanner manufacturers, and
particularly since I don't use the same manufacturer of printers or
scanners.

I would also note that there are various websites with helps and
tutorials for managing scan projects using either of these tools, and
they both enjoy a large customer base. Everything from tweaking your
color matching to print output can be found on various websites and
forums. The learning curve for either can be rather long, but I don't
consider them very steep, really. Both work well as a front end to
Photoshop, particularly if you are going to scan them as 48bit raw
images.

Jeffrey

On 11/7/06 7:20 AM, "Paul Atroshenko" <patroshhotmail.com> wrote:

> Hi one and all,
>
> I have recently bought a HP Scanjet G3010 and I am very pleased with its
> scans of colour slides (main reason for purchasing this device and
> sidelining my old Canon scanner... which doesn't do that).
>
> I would really appreciate advice on getting additional sortware to help me
> scan slides with the Scanjet. At the moment I use the given software from HP
> and work mainly through Photoshop.
>
> Is there other software that can bring better results, especially in terms
> of sharper images? Or does this basically depend upon the sharp qualities of
> the original slides?

I really don't like HPs software, not so much in the scanner driving
department, but in its overhead (4-5% CPU on my iMac G5).

Plus its tendency to nag about updates that don't actually exist.

That said, it's hard to make gold from dross--blurry photos are blurry.
Check the scanning settings such as the Sharpen setting (I'm dubious...) and
the Dust and Scratch removal (causes blurriness, inherently).

Good luck!

Paul Atroshenko wrote:
> I would really appreciate advice on getting additional sortware to help me
> scan slides with the Scanjet. At the moment I use the given software
> from HP
> and work mainly through Photoshop.
>
> Is there other software that can bring better results, especially in terms
> of sharper images? Or does this basically depend upon the sharp
> qualities of
> the original slides?
>
You should check out VueScan. It works fine in demo mode. Just adds a
watermark until you buy a license. See if it helps.

> You should check out VueScan. It works fine in demo mode. Just adds a
> watermark until you buy a license. See if it helps.

I couldn't agree more! I love VueScan. I opted for the Pro version.
It's a great product!

Rick

Is there other software that can bring better results, especially in >terms of sharper images

If your slide or picture is sharp, scan at a high resolution, say, 600 ppi at the area size (eg 5"x7") that you want. Then, when you open the image in a image-editting program like Adobe Photoshop, you can resize or strip pixels to bring your scan into an appropriate size (for inkjets, 150pp at 100% size).

So your scan is sharper because it was sampled (scanned) at a higher pixel density.

I looked at your web site. It's stunning. You look like you are a lot pickier in nature than I am (and I'm not going to let you see my crappy sites). Because I saw your site, your question did not make much sense. So I'm giving a long answer. The conclusion is the last sentence, all of the intermediate stuff is to give perspective. As Lincoln once said, I'm sorry this was so long but I'm in a hurry.

You say that you are pleased, but you want the pictures to be more sharp. It's not clear what you mean by sharp: contrast, resolution or that thing that the unsharp mask does?

But here is the problem. It is almost impossible to figure out how to get a better scan without really serious research and experimentation. I have used an HP scanner with a negative/slide thingy, a Nikon coolscan 2000, and I am finally getting around to playing with a Canon 9950 I purchased last spring for doing family stuff in the old family mansion, including large glass plates--as well as scanners at local establishments and scanners owned by other people and samples I have seen on the net. Here is something I have learned: Don't pay any attention to the specs for the scanner--they are all lying. (Popular Photography says so--but in a different way. You might want to read some of their articles.)

The Nikon Coolscans are considered to be really good scanners, but get this: When I (and many people) do a scan on the Coolscan they use 4 passes to get a clear scan (eliminating noise). The coolscan allows 16 passes and I swear that you can see a difference between 4 and 16 passes when you are looking to get detail after you really blow up the slide (like a soccer photo where you really want to zoom in and get some definition on a face in the middle of a scrum heading a corner kick). (by the way--4 passes on the 2000 takes 5 minutes, 16 takes 20).

I did exhaustive research before I bought the Canon 9950. I don't know what the state of the art is this month, but last spring, there was only one 4800ppi flatbed scanner under $400 street that produced a better image at 4800ppi than 2400ppi and I think that's the one I bought. An enormous number of scanners out there produce garbage and almost none of them produce a scan that meets their published specs. Look at the comparative reviews that show comparisons of actual scans and you will see what I mean.

None of the flatbeds can measure up to a coolscan--even though it is working at a lower resolution. The reason for this is that their sensor rating is based upon the step-size of the motor, rather than the actual sample size of the sampling sensor. (that's also where they get that silly 9600ppi rating for one of the resolutions--are they saying their pixels are rectangles?) Also, they don't have any ability to focus on the film (which is probably not perfectly flat). Also, sampling noise is overwhelming on many of these scanners.

Since your scanner is a brand new model, I have seen no actual comparison reviews of the scanner so it is hard to compare. My really old Nikon 2000 gives a much better scan than my newer HP at a similar dpi--a whole lot better in many ways. Some people say doesn't work so well with slides because its cool lights don't like chrome, but I have found, that although my slides come out dark, I can pull out good images with photoshop--and I have seen accounts of professionals who have found settings that work on Kodachrome.

Just this week I started working with the 9950. It does produce a better image at 4800 than at 2400, but guess what? Although it says that it gives you 16 bits, I can't get it to do it--Photoshop opens these at 8 bits. And at this high setting it takes about 8 minutes a negative. (like the 3010 the 9950 has a full platen light that allows 30 negatives to be put in a bracket at once and the software automatically pulls out the individual negatives on preview and you can select those to scan). So I can load the thing up to work for a couple of hours overnight--but then I have to save each one individually in the morning. Also it cannot do Digital Ice like a true negative or slide scanner--Ice works by using infrared light to detect where the dust specks and scratches are--and while even this hardware based technique fuzzes the picture a little, other noise reduction techniques just use fancy blurring. Even something like GEM, noise ninja or noiseware is not as good as hardware-based solutions.

The 9950 gives better resolution at 4800 than 2400, but so far, I have had to use auto color, because, for some reason when I try to set a mapped color space it really screws up the midpoint of the gamut-and will not let me alter it with that setting. (I'm going to try Vuescan)

Anyway, what I have learned over the last 5 years is this: Your slide has a greater dynamic range than your scanner so make sure you are using as much of the important dynamic range as you can and make sure your HP software isn't cutting it off (like my HP does). (Try Vue Scan---read their scanning articles). NEVER use auto-anything in the scanner--you have photoshop and that is better software than your scanner software. My HP constantly tries to do an autolevels at the scan--even when I turn it off, so i have to reset the levels manually for each scan. Also, try scan in different resolutions-- a lower resolution may work better. Try two scans at different resolutions or with a different midpoint and then layer them in Photoshop (OK now I'm going overboard).

Anyway, I find that, when scanning, the color range within the gamut is more important than the dpi. (and don't confuse bits per pixel with dynamic range--bits are only how many divisions you get within the range). Photoshop can cure some things but it works better by having a good color relationship at the pixel level (I don't know what I mean by that, really, but I have used Photoshop on several thousand pictures over the last 5 years and there is no substitute for a better scan). Apparently, the Nikon gives a better color rendition by using three colors of lights--makes sense. Before you start, be aware that, if your slides are taken with adequate light, they will have much finer grain and greater dynamic range than a typical scanner can capture. So don't get too frustrated by your scanner.

Also, if you have a lot of slides you may not want to kill yourself for hours a day for months, for a bunch of inferior scans. I did a lot of family slides because I got a slide feeder cheap, and I would scan 50 a night into an old dedicated G3 from my old Nikon. Again, the Nikon doesn't like chrome, but it did OK, and it simply would have been stupid to do it without a feeder to do this overnight (Also Nikon software stores the files on disk as cmp.TIFF files with a naming scheme so this was all automated. Don't pay attention to anyone who tells you to save these as jpegs--do that after photoshop.)

Also, although the Coolscan is the leader as the best film scanner, it's not even in the same league as a drum scanner.

The thing is this: You would think that there would be lots of solutions out there to solve this slide problem that people have--but it turns out that while there are lots of people claiming to have solutions, most of them are lying. (I'm shocked, SHOCKED that a computer salesman would lie).

Software can't solve hardware problems--and you already have the best software, which is Photoshop. You should not try to use your scanning software to do a bad job of what Photoshop can do better. But I would echo the above and suggest you check out Vuescan to see if it can squeeze something out of your scanner that your HP software is likely preventing you from getting to. Vuescan supports your scanner. (Also check out Silverfast).

Vuescan can probably get better color out of your scanner, but at the same time, it probably can't improve the resolution.

[Great post, but let's not diverge further from the thread... -Adam]


On 20-Nov-2006, at 12:13, Clyde Kahrl wrote:
> As Lincoln once said, I'm sorry this was so long but I'm in a hurry.

Pretty sure that was Blaise Pascal. Something like "I made this
letter long, but I didn't have time to write a shorter one."

<http://en.wikiquote.org/wiki/Blaise_Pascal> has it as:
> Je n'ai fait celle-ci plus longue que parceque je n'ai pas eu le
> loisir de la faire plus courte.
>
> * Literally: I made this [letter] very long, because I did not
> have the leisure to make it shorter.
> * Translation: I would have written a shorter letter, but I did
> not have the time.

However, there is a quote from Augustine that is similar:

"In regard to the questions which you have asked me, I would like to
have known what your own answers would have been; for thus I might
have made my reply in fewer words, and might most easily confirm or
correct your opinions, by approving or amending the answers which you
had given. This I would have greatly preferred. But desiring to
answer you at once, I think it better to write a long letter than
incur loss of time....."

So, the idea is old, but the actual quote that we use originates from
Pascal.

I like the Augustine quote only because it is so wordy and long that
it is very much a self-definition sort of quote.

What scanner driver for CanonScan 5000F can I install that will be recognized by Photoshop CS2 in my iMac?

On Nov 30, 2006, at 3:14 AM, <tidbits-talktidbits.com> wrote:
> What scanner driver for CanonScan 5000F can I install that will be
> recognized by Photoshop CS2 in my iMac?

VueScan. It works with loads of scanners, Photoshop,
GraphicConverter, etc. Much better than most bundled scanning software.

Steven Lunetta