OK, here is the set-up that I’m partly involved in putting together:
It’s in a record office, and we want to make it possible for the public to request scans of any document (within reason). We’ve got the hardware awaiting setup (powerful computer, top-notch scanner and printer, Photoshop CS). What would be ideal would be for the procedures to be as easy as possible for the people operating it, also simplifying training. From what I can see, the ‘Actions’ system in CS is vastly improved, with the opportunities for scripting and stuff. I’m hoping that this will make this simplification possible, so much of the procedure is as near as possible to a one-click operation.
The fixed procedure that will happen with every document is:
Scan at one of two set resolutions (one for paper, one for glass negatives).
Auto-levels
Save to designated location on the hard drive as TIF format
Print
Resize to 650 pixels on the shorter dimension
Save as JPEG (fixed compression settings) to hard drive, and also to a network location
How feasible will it be to aim to make this work as a one-click (or few-clicks) process? Or to put it another way, what can’t be done with scripting? eg is it possible for the process to prompt for a filename, which is to be used for both the image formats? And if anybody can point me to any decent resources about scripts in CS, I’d be very appreciative.
Thanks in advance…
i typically do everything manually in Photoshop so I can’t give you any direct advice for the scripting. However, a good place to begin looking is the Adobe Studio.
Given they’d already forked out for Photoshop before I was involved, I guess it’s too late for me to recommend Gimp instead. Which is certainly a shame. In any case, I suspect Photoshop would still win out because of staff familiarity with earlier versions.
I am not an expert on scripting, but everything you want to do is possible. You simply set script to record and then do your steps. You can have script pause for input such as file names and parameter settings.
I can’t figure out your workflow at all. It seems that your records are not computerized, or you would not have to scan them. Why then do you want to save them once a customer requests a copy? Why save them both as .tif and .jpg? If you are printing on a laser at legal or letter size, a high quality .jpg will give you identical resolution to a .tif and save a lot of space on the harddrive.
They’re original documents (wills, parish records, birth certificates, old photographs, all sorts of stuff). There’s various reasons that they’ll be needed in electronic form, but basically it’s because no original documents ever leave the building. Archiving the scanned images means that firstly no document needs to be scanned more than once, and secondly in many cases future consultations of the document will often be done with the electronic version, further helping preserve the originals. The TIFs are being kept for this purpose - space isn’t an issue, they’re to be periodically put onto CDs and stored. With these documents, there’s a lot of detail lost when reduced to 650 pixels - and in any case, it seems sensible to keep a lossless version to futureproof the collection. The JPEG has two functions: firstly it’ll be linked to from the original archive catalogue entry for the document, and secondly many of them will be used online (such as here). And as additional point, the standards for image type/size/resolution have been chosen as a standard between various museums and archives, so we’re not ina position to alter them.
Anyway, hopefully once I get my hands on CS I’ll be able to figure out the scripting.
Haven’t you answered your own question with using the actions box?
Simply do a scan, throw it into CS, hit record on a new action, do what you need to do. The next scan, hit the play button on the actions box and it will repeat all your well, actions. You’ll have to stop it before you save though, to prevent it saving over the previous file - IME it won’t come up with the save as… option.
The problem with actions, as I’m used to in earlier version, is there’s no opportunity for the if-then-else situations needed for dealing with both portrait and landscape images, for example. Because of this, I ended up with so many different actions it’s been no improvment on doing everything manually. I’m hoping that CS will get around that problem.
I haven’t seen Photoshop CS yet; I’m still in the wilderness of 5.5 (at home) and 6 (at work). The automation of 5.5 seems to be very limited. It’s unclear to me whether you couldd put a “front end” on an action to give it a user interface, for instance. 6 seems to have more options, including the ability to make “droplets”, standalone executables that execute actions, sort of like self-extracting zip files.
The Photoshop help file gives some very limited information about external automation:
Just bumping this, for anyone who’s interested, to let you know the outcome:
Yes, Photoshop CS has extensive scripting capabilities. I managed to do everything I needed, surprisingly easily, and only needing to use ‘alert’ and ‘prompt’ dialogs. The documentation installed with the program is top-notch. There’s also a ScriptingListener plugin, which functions very much like a macro recorder, letting you convert actual operations into scripts.
Sounds like an ideal situation for a bespoke application (SAYS Mangetout, momentarily lapsing into developer mode) - You can get activex objects that will launch the scanner’s TWAIN dialog and can handle the incoming image, which you can also then resize and save in code (I’d make it a database application too, so that next time someone requests the same document, you don’t have to scan it again).
I’m writing a warehouse booking in application at the moment that captures delivery note scans - I’m writing it in Delphi (and version 3! = not my choice), using a component set called ImageEn to do the image capture and handling stuff.