How do I find the sweet spot at the intersection of speedy and spiffy?

January 6, 2011 7:25 AM

I would really appreciate any assistance in establishing a workflow for scanning images from the books I have piled on my desk as I prepare a series of "slide lectures" for an art history class I will be teaching from April. I'm sure there are AskMefites out there who are experienced enough to be doing this kind of thing in their sleep; I'd love to hear from you! Any help is appreciated!

What tips or tricks can you give me about getting good quality images that will project clearly and (relatively) true to the "original", along with keeping notes about the image itself somehow "attached" to the digital file?

Is there special software for the note-keeping part of this? Should I just dump all my notes and such into the Spotlight Comments box in the Get Info dialogue (I'm on a Mac)? Or should I use iPhoto and its tagging features? Maybe there is something in Keynote I don't know about (and that is likely what I'll be using to do the actual presentation). Just use a paper notebook?

What settings should I use in the scan software (resolution, etc. – pretend I'm a 6-year old, please)? What's too much and what's too little? Since we're talking hundreds of images here, what corners can I cut, and where am I advised to bide my time? Will it depend on what projector I'm using or the size of the screen (yes, I'm that clueless)? I'd like to get as efficient at this as I can, and I know I will learn a lot just by throwing myself into it, but I also would rather not re-invent the wheel.

What else haven't I thought of, that I'll be kicking myself for 3 months from now?

I've searched online – where I presume there must be some info – and here on AskMe, but I must not be using the right criteria. Please feel free to point me to other questions or web resources.

Thanks kindly, in advance!

posted by segatakai to computers & internet (8 answers total)

I would suggest using a digital camera to capture the images. It is much, much faster than using a flatbed scanner and with a little setup can have equivalent or better results. If you set up a copystand (or a DIY copystand, or a tripod copystand which is really just a tripod) you can digitize thousands of images in a single hour, leaving you lots of time to do the metadata properly. You could do all cropping and color-correction in iPhoto, too.

I run a forum dedicated to DIY digitization. If you asked this question there you would likely get a bunch of informed answers.
posted by fake at 7:49 AM on January 6, 2011

I manage a document imaging department, so hopefully I'm leading you in the right direction :)

We use document-management software, which lets us scan directly into the software and then allows us to "index" the images, adding names and dates and such, so we can quickly retrieve them later. If you can scan right into iPhoto and use is tagging features to collect the information, and keep it searchable, would be my best bet, although I'm not familiar with iPhoto or Keynote.

The main objective of scanning documents and indexing them is for retrieval. You want to be able to search and find what you want quickly. So, you need a "key", which is as unique and specific as you can get for individual retrieval, and then tag-like information for semantic searching (eras, artists, styles, etc.) . You don't want lumps of text, which makes searching harder. Think about how your course is planned; you want to be able to retrieve one day's references with as little effort as possible, so put that effort into it now when you're scanning.

Quality: If you are only going to put them up on a projector from a computer, 100dpi is all you'll need. If you scan at 300dpi, your monitor (maybe at most 1680x1050 resolution) is going to shrink down your image to fit, losing quality anyway.

If you're planning ahead for printing things out later, look at 600dpi, and use any descreening settings your software has -- printed documents in books are made up of tiny dots, and the scanner is trying to convert the image into tiny dots, so the two don't overlap perfectly and it results in wavy moire patterns.

Lastly: Scanning and indexing always takes way more time than anyone thinks it will be. You think: pulling things up on the computer will be so much easier. Yes, but the effort of scanning and saving is going to take as much time as it does to look it up in the first place. Come up with a system, and then you could even have your workstudy students do it for you.

On Preview: Fake's camera idea is a good one -- if you're aiming for screen-resolution quality images, a camera will do a good job for you.
posted by AzraelBrown at 7:58 AM on January 6, 2011

The most relevant issue is that you'll be scanning images that are coming from a printed/screened source, so they're show up with Moire patterns once scanned. The trick to minimize this, is to scan at a larger physical resolution that you'll need, add some Gaussian blurring (between 0.8 and 2 pixels in radius), resize down to target dimensions (your final resolution should be 72 DPI), and apply some Unsharp Masking to get back lost details from blurring.
posted by dbiedny at 10:03 AM on January 6, 2011

nthing fake's suggestion. If you plan accordingly the camera sequence can be used to numerically order your slides for presentation.
posted by effluvia at 10:12 AM on January 6, 2011

Whether photographing or scanning, there is a trick to prevent the words printed in the other side of the page from showing through your image. Take a piece of construction paper, ideally a middle grey color, and put it underneath the page you're photographing/scanning. Voilà! No more text bleeding through from the back.
posted by xo at 10:22 AM on January 6, 2011

Build a BookLiberator. (Disclaimer: I have worked with some of the people involved with this project, although I have never tried it myself).

Ordinary scanners are less nice to the spines of books, or result in distorted images.
posted by novalis_dt at 1:38 PM on January 6, 2011

wow. I'm really surprised that a camera is being recommended. I understand that the capture is instantaneous, but the scanner is already built, and is flat, and has essentially nothing between the "lens" and the object. With a copystand, how do you deal with the curvature of the pages and glare from the lamps (without building a BookLiberator)? Please don't get me wrong; I appreciate the advice, I just never would've thunk it. It may take me a while to get my head around it... my head is swimming with questions, but I thnk I may have missed my window on this question's halflife... I'm going to be looking around at some of the resources suggested, but in the meantime I've got a few immediate questions:

fake, do you use a piece (or pieces) of glass with the copystand setup? If not, how do you deal with page curvature? if so, do you not get glare or reflections?

AzraelBrown: are you referring to "image-indexing" software? Can you give me an example (a name)? Are you talking about something like Lightroom or Aperture? I have something called ViewNX that came with my Nikon camera -- is this the same idea? A quick look seems to indicate that it will do so...

About the whole dpi thing, what do I really need, and why? AzraelBrown, why do you recommend 100dpi as opposed to 72? Is this guy making sense here? If I take a pic with a camera, should I then downsample (?) the image to 72dpi (is the only reason I would do this to save space on my hard drive [which is not very big])?

About descreening, my scan software has three settings: newspaper, magazine and fine art print. I guess a coffee table type book would qualify as the latter??

Thanks again, all.
posted by segatakai at 7:49 PM on January 6, 2011

sorry, i just scrolled further down to see that glass is involved...
posted by segatakai at 7:54 PM on January 6, 2011