Scanner speed - what is reasonable?

[jhitchcock]

Not feeling my creative juices flowing as freely as it seems everyone else here, I shamelessly built Daniel's new standard scanner. When I ordered my cameras I had to send one back for being the wrong model and unsupported by SDM. As it stands, I've made 2 complete book scans using the single camera and paging through each book twice, once for each side. Now that I found one on ebay for $50, that should cut my scanning time in half. Seeing many of the different strategy builds around here has gotten me thinking, I don't want to spend my life turning pages. I'd love to get a wide mix of answers and "scanner speed profiles" here.

1. Tell the basic design of your scanner
2. If possible, please give an averaged pages/minute scanning speed
3. If possible, please tell your processing speed as well. How long does it take to go from scans to completed product with an average sized paperback?
4. What makes your scanning slow? (or fast!) Ideas for speeding it up a notch without having to spend more time processing b/c you were careless...

[jhitchcock]

After making this post, I realized there are already some answers to this question in the software section. Why don't we try to keep this hardware focused?

[StevePoling]

a while back I suggested we hold "scanner races" where multiple teams would scan the same book to learn which of competing design paradigms performed better than others. This would require each team to document its workflows and equipment so that subsequent rounds could incorporate lessons learned by winning teams. Maybe the notion of racing is a bit more competitive than some like. Instead, it may suffice to document our respective workflows, instrument them, and publish them so that we can establish reasonable expectations for each of the performance parameters we deem significant.

[you1]

Pages: 1200
Setup: 2-5 min
Phsical Scan: 45 min (going as fast as possible; I was tired towards the end)
Software activity: 15 min
Software post processing: over night
Scanner: viewtopic.php?f=1&t=74&p=724&hilit=1200#p724

Edvin

[spamsickle]

I just did 1200 pages -- 3 books -- in 72 minutes, or 1000 pages per hour. It was an easy set, not tiring, and I wasn't rushing. This is with my version of Rob's new platen design (http://diybookscanner.org/forum/viewtopic.php?f=1&t=376&start=10#p4213), which I really like. I've never used a "rails" design, but my "floating box" design has a tendency to bind occasionally, which I understand is a problem people sometimes have with rails as well. This "leveraged" design doesn't have a binding problem at all, and the counterweights make it almost operate itself. Raise it a little, and the weights take over, raising it without muscle power. Most of the energy I expend is in lowering the platen.

[jck57]

Most of the energy I expend is in lowering the platen.

How about a damper that would gently lower the platen after you take your foot off the treadle? It could be an air dashpot or something hydraulic. For a previous project I made an air dashpot out of an air bed hand pump. pushing the piston in is easy because air is exhausted through a check valve but pulling the piston handle out is slow because metered air is sucked in as the piston is withdrawn. By adjusting the air inlet ball valve, you can set at what rate the platen will descend to the book.

[spamsickle]

I'm not using a treadle, and there's probably not a comfortable way to add one to the system I'm using (all three of mine have been on the floor, with me sitting on a small stool to operate them), but that might be a good idea for someone else. I like lowering the platen by hand, because sometimes I'll want to kind of wriggle it down into the pages to make sure the inner margins are visible. This is especially true when I use the 1/4" plexiglass platen rather than the 3/16 glass one. Plus, I like the control I get by lowering it myself. I'd hate to be slowed down waiting for something automatic to do it slowly and gently when it's really very little effort to do it manually at a rate which gives good throughput and doesn't damage the book.

[jck57]

I'm not using a treadle, and there's probably not a comfortable way to add one to the system I'm using (all three of mine have been on the floor, with me sitting on a small stool to operate them), but that might be a good idea for someone else. I like lowering the platen by hand, because sometimes I'll want to kind of wriggle it down into the pages to make sure the inner margins are visible. This is especially true when I use the 1/4" plexiglass platen rather than the 3/16 glass one. Plus, I like the control I get by lowering it myself. I'd hate to be slowed down waiting for something automatic to do it slowly and gently when it's really very little effort to do it manually at a rate which gives good throughput and doesn't damage the book.

OK. I understand that you prefer to lower the platen manually. But as far as "slowly and gently" goes, with a dashpot you could adjust the lowering speed to whatever you want, matching it to the speed you prefer when lowering by hand.

I also wanted to comment that your idea for pivoting your platen arms with S hooks is IMO much better than trying to use ball bearings or improvised bearings. Anybody else thinking about building a pivoting arm platen should take note.

[Gaaren-gaargle]

I currently use OpticBook 3600. Scanning speed on that scanner depends on the size of the book and weight. I recently scanned a book that measured 11 x 9 inches and weighted 8 pounds, it was a pain to scan and the speed was about 3,5 pages per minute. I had to scan some before and after work to get through the beast, I got done this night. Now it's time to sleep