Daniel Reetz, the founder of the DIY Book Scanner community, has recently started making videos of prototyping and shop tips. If you are tinkering with a book scanner (or any other project) in your home shop, these tips will come in handy. https://www.youtube.com/channel/UCn0gq8 ... g_8K1nfInQ

A simpleScanner setup, could use improvement

Built a scanner? Started to build a scanner? Record your progress here. Doesn't need to be a whole scanner - triggers and other parts are fine. Commercial scanners are fine too.
Post Reply
rogermaris
Posts: 10
Joined: 04 Mar 2014, 00:52

A simpleScanner setup, could use improvement

Post by rogermaris » 27 Feb 2010, 21:17

In comparison, below amounts to not even half of the quality of other designs I've seen posted on this forum. Nevertheless, here's a fully functional setup in its improvement stages
The first two pictures were plans that I forfeited, and the last picture, just hammered together, works. I tested it.
11302009030 - Copy-1.jpg
11302009030 - Copy-1.jpg (140.39 KiB) Viewed 6903 times
My original plan was to have it this way right above. What I wanted was to make use of the scanning device when I am not scanning – that is, to utilize it as book holder with particular intent to hold heavy text books
My thoughts as to how make it work were to make it similar to ordinary drawers (pull out a drawer, and push it back in); and the angle whole device was such that the platen gravitated towards the book holder. So, there are two drawer slides installed--below shows that in more detail
Instructions.jpg
Instructions.jpg (267.6 KiB) Viewed 6903 times
So once again, the picture immediately above shows how it works, (or would work), by loading the book, and centering any book's spline with relative ease.
disadvantage of this design is that it clearly requires, (as it is not installed in the pictures), vertical support to hold the book and that support would have to be adjusted for each particular book's dimensions in order to accommodate a broad range of different page sizes/margins. This was an additional step of which discouraged me from the original plan. Not to mention, given the way I made the platen my design is not really suited for this setup, because ideally it would be easier if the platen was wholly transparent.
What held me back from completing this were mostly noise issues with my neighbors. As a matter of fact, I had all that wood cut at the hardware store, because if i bought a saw it'd be too loud, seeing as even my newly bought drill was annoying my neighbors. Below is what I resorted to, which was basically a big shortcut -- reducing various complexities and keeping it simple.
Solution:
shortcut.jpg
(431.98 KiB) Downloaded 910 times
This is currently my setup, which is geared more towards the design in the diy instructions. It has some defects yes, but nevertheless this scanner is fully functional – tested after scanning 300-page book in roughly 50-60minutes.
Aside from the obvious issues that you can probably observe, I'll mention what i think are the most notable issues in the following reply comment
Last edited by Anonymous on 11 Mar 2010, 14:02, edited 3 times in total.

rogermaris
Posts: 10
Joined: 04 Mar 2014, 00:52

Re: A simpleScanner setup, could use improvement

Post by rogermaris » 27 Feb 2010, 21:25

Problems: You probably can't tell that those are two different cameras. I got one A590is and one cannon g7. This seems to have been a mistake, because the picture quality from each camera is considerably different. My question is – should they be? Perhaps all I need to do is tweak the settings a bit so that both cameras output the same resolution, and overall same format leaving no difference between the two. But that would have to be, other than, the settings presented in instructions posted on Dan's DIY, Such as superfine, size L, tugenstein, were the same for both cameras, but still outputting differently. So, it's possible I need only to make some technical adjustments. Maybe the differences are the result from different firmware from the A590is and the G7, (both support stereodatamaker). I should probably do some more research on that
First, If I can read my scans just fine, why do I bother? The reason this matters is mostly due to OCR issues—I read the book I scanned fully, and every page was perfectly legible—and that's reason enough to ignore this issue. However, my goal now is to have professional quality pdfs—or at least reach a point where the differences are not noticeable.
What are the differences? While reading through the whole book, what stood out the most, for example, was that anyone could tell that every other page was from a camera A , say every odd page, and likewise every even page was from camera B; the differences are discernible not because it follows that numerical pattern, but the visual peculiarity of each camera is noticeable. It is like reading a white page for every even page, and light brown page for every odd page (which may present itself as a nuisance to the reader). The OCR-just exaggerates the differences—question is then, are the differences an issue of software (OCR), or hardware (having two different cameras).
Of course, I could Photoshop, or make each page exactly the same, but what I am thinking is – this is mostly due to shooting with two different cameras.
Main Question: Are there any known issues from having two cameras not of the same kind? Should I have gotten two A590is instead?
Last edited by Anonymous on 07 Mar 2010, 17:52, edited 1 time in total.

spamsickle
Posts: 596
Joined: 06 Jun 2009, 23:57

Re: A simpleScanner setup, could use improvement

Post by spamsickle » 28 Feb 2010, 19:46

If you could post pictures it would be easier to diagnose the difference, but I doubt that it's due to having two different cameras per se. I was originally shooting with two cameras, one DSLR and one point-and-shoot pocket camera, and the only way to tell which camera took which picture was to look at the EXIF (or the default file name). I suspect you may have a white balance or exposure difference, but nothing which couldn't be resolved by adjusting one or both of the cameras.

User avatar
daniel_reetz
Posts: 2786
Joined: 03 Jun 2009, 13:56
E-book readers owned: Used to have a PRS-500
Number of books owned: 600
Country: United States
Contact:

Re: A simpleScanner setup, could use improvement

Post by daniel_reetz » 28 Feb 2010, 20:57

Love the out-of-the-box thinking you have going on here. Please post pics from each cam (ideally from a book with no copyright problems, plz) and we'll see about getting your settings to match up. I agree with Spamsickle (king of non-homogeneous camera setups, BTW) you should be able to get them very, very close.

rogermaris
Posts: 10
Joined: 04 Mar 2014, 00:52

Re: A simpleScanner setup, could use improvement

Post by rogermaris » 02 Mar 2010, 00:22

Thanks for your replies.
My question has been for the most part answered, i.e. two different cams isn't, or shouldn't be, the source of my problem.
About the request: I just took 2 pix and then I thought - copyright - Infinite editions -- but whatever. To adhere to the ideal, I believe that would be a public domain text--before 1923, right /in which case I'll just swing by the library after class sometime later on
In the meantime, I'm going to double check everything later in order to rule out possible mistakes I made.
For now, could it be my lights? - Attached you'll see a pick of the halogen bulbs I'm using.
IMG_5734.JPG
IMG_5734.JPG (120.47 KiB) Viewed 6817 times
The bulb on the left is 50W/120V Flood. The bulb on the right is a outdoor 90W/120V. -- So two different bulbs could be a mistake?
Last edited by Anonymous on 04 Mar 2010, 18:00, edited 1 time in total.

User avatar
daniel_reetz
Posts: 2786
Joined: 03 Jun 2009, 13:56
E-book readers owned: Used to have a PRS-500
Number of books owned: 600
Country: United States
Contact:

Re: A simpleScanner setup, could use improvement

Post by daniel_reetz » 02 Mar 2010, 03:20

Less likely than different camera settings. I use the kind of bulb you have on the right, and they work really well and provide very nice light.
public domain text--before 1923, right
That's the most-safe option, but don't stress it too hard.

rogermaris
Posts: 10
Joined: 04 Mar 2014, 00:52

Re: A simpleScanner setup, could use improvement

Post by rogermaris » 03 Mar 2010, 23:40

I changed the bulb, which now produces better results. :)
Since my last reply, I tampered with my setup a bit and I think I narrowed down some differences between the non-homogeneous cam setup simply by making more space and repositioning the tripods, etc. Nonetheless, I suspect the differences, although subtle, could still be existent and would like a second opinion. Even just a general opinion about the output quality would be appreciated.
If you could post pictures it would be easier to diagnose the difference
The following pictures are from a book whose copyright is 1912, public-domain-worthy. and this was easy to find at the library, and fun to get a hold of a near one hundred year old book.

It might be hard to diagnosis just by looking at these pictures, So i thought i would stress taking a look at the 'photo info.,' contained in each picture file because perhaps this will lead to an explanation of differences--however subtle that might be. either way, im sure the data contained will be better understood by people who know more about these camera settings than I do.

The file contains details about the exact settings I'm using for each camera.
For example

Code: Select all

Filename: 590is (left).JPG
Camera: Canon
Model: Canon PowerShot A590 IS
ISO: 80
Exposure: 1/125 sec
Aperture: 8.0
Focal Length: 19.6mm
Flash Used: No
Latitude: n/a
....
Filename: G7 (Right).JPG
Camera: Canon
Model: Canon PowerShot G7
ISO: n/a
Exposure: 1/125 sec
Aperture: 8.0
Focal Length: 25mm
Flash Used: No
A more thorough photo information should be accessible through windows detailed folder view with more options (you might have to manually selected them, but they should be auto for window7 users)
Anyway, if you're able to access this, then you'll be able to compare/contrast each camera through the pictures you see right below; and that is to say, the details contained in the files could be more revealing than just looking at the pictures themselves.
Left from A590is:
590is (left).JPG
590is (left).JPG (187.7 KiB) Viewed 6763 times
Right from G7:[img]
G7 (Right).JPG
G7 (Right).JPG (175.55 KiB) Viewed 6763 times
After comparing the photo details between the two myself, I noticed that the A590IS says that the ISO speed = ISO-80. Whereas, the G7 doesn't say anything about that (no ISO speed?) – Problems with this?
Any other noticeable things about settings, or the pix themselves that you recommend i change?

Let me know if more/ different pix are needed.
Thanks

User avatar
Misty
Posts: 481
Joined: 06 Nov 2009, 12:20
Number of books owned: 0
Location: Frozen Wasteland

Re: A simpleScanner setup, could use improvement

Post by Misty » 04 Mar 2010, 10:26

I'm surprised the ISO isn't in the EXIF for the G7 - that's a bit weird. It supports 80 - 1600, and if it's anything like my G10 it defaults to an auto mode that's probably giving you ISO ~200. The G7 has a physical ISO dial; make sure that's set to 80. That said, the exposure is very similar between both cameras already.

I did notice that both cameras' shots are underexposed, which is why they're appearing so dark. I think that may pose some trouble for processing the photos later, though if you're going to be using Scan Tailor you might be okay just brightening them in software. Like I explained to Ann in her thread, take a look at this histogram for your second image:
rogermaris-histo.png
rogermaris-histo.png (8.5 KiB) Viewed 6751 times
It's only taking up about half the space, with empty space to the right. The problem is that your shutter speed is too fast for the brightness you have there, so it's not letting enough light in. If you can, increasing the amount of light you're using would be the best solution. Otherwise, you can increase the exposure you're using. If you're using manual settings, make the shutter speed slower to let in more light; if you're using P mode, set the exposure compensation setting to about 1 2/3. You'll have to make sure your tripod/support can keep the cameras very stable, though, to prevent any jittering that would blur the image.

It looks like you're getting very similar images between the two of them! Looks like using two different cameras shouldn't give you any trouble. The only other thing I'd suggest right now is making sure the brightness is even on both pages; it looks like the second page is darker on the upper half, while the first image you posted is more even. Scan Tailor might be able to deal with that as is, though.
The opinions expressed in this post are my own and do not necessarily represent those of the Canadian Museum for Human Rights.

rogermaris
Posts: 10
Joined: 04 Mar 2014, 00:52

Re: A simpleScanner setup, could use improvement

Post by rogermaris » 07 Mar 2010, 22:48

The G7 has a physical ISO dial; make sure that's set to 80
Unfortunately, without changing anything, the dial is already set at iso-80, which is in fact strange. Nonetheless, I'm assuming that doesn't have much influence over the histogram results displayed in the pic you posted. Either way, I will happily direct my attention to other issues clearer than this.

The histogram picture you posted was particularly helpful.
It's only taking up about half the space...If you're using manual settings, make the shutter speed slower to let in more light
That gives me a grasp of the relationship between the two settings. Also, I should say I'm using manual settings.

To correct the problem with the G7--in other words,
The ultimate aim for correcting the G7 will be reflected on a histogram, ensuring that the G7 outputs pictures that occupy the full 'space' on the scale, unlike only half of the space as is displayed in the histogram pic shown in misty's reply above. So if that is the case, then I'll know what to shoot for. thanks misty.

Now the rest of my reply stated below here is concerned with trying to adhere to,
increasing the amount of light you're using would be the best solution
So, this is my current goal, and I just wanted to mention one last thing before I move everything around again. Forgive my lack of photography knowledge, but could increasing light be done somehow without buying more bulbs? What I have in mind is simply moving the G7 closer to the platen--as opposed to leaving it in the same position, which you can see by looking at the picture. [Oh yeah, please notice that I've taken liberty to swap an old photo of my scanner setup with a new and you can find this on the third picture of my first post now edited on this thread.] I'm saying perhaps I left out a valuable piece of information that I failed to include before, which is to say, if you look at the new picture now, I think it may help give a better idea about what's going on because it shows the actual position of the tripods when I took pictures and uploaded the pages as you see on my previous reply. In other words, perhaps my setup displayed in new picture could now explain the results on the histogram. So, I'm hoping to have provided my G7 lighting issue a clearer view now having uploaded a more accurate picture of my setup. by the way, if the picture doesn't provide anything, then at least it looks better /uptodate for the interests of this thread in general.

And yes, it probably would have been a quicker for me just to move the G7 closer and produce another series of tests myself without posting another reply on this thread. It's also the case that the pictures are adequate already. Nevertheless, I find this reply worthwhile (sry it is rather long) because I wanted to understand the suggestions better—namely, it is possible you meant that while adhering to the best solution, the only way I can make G7 better is if I get more bulbs, and if this were the implied suggestion then maybe the pic uploaded may cause alternate suggestions—which if it doesn't I'll have to decide whether buying another light fixture/halogen bulb is worth it. (Prob not).

I think this will be a last attempt to mess with any hardware or camera related factors in my setup. If there are remaining issues, but software-related, I'll keep them independent of this thread.

In summary, bearing in mind, any change that I plan to make, stated in this reply, is aimed at getting better results for specifically the G7; and those results will be reflected via histogram (full, not half). To increase light exposure, the possible changes, seem to be—bulb or no bulb? Or the pic I just uploaded causes alternative suggestions – simple maybe e.g. move the damn g7 tripod closer! Or maybe others. Suggestions welcomed. Thanks again.

User avatar
Misty
Posts: 481
Joined: 06 Nov 2009, 12:20
Number of books owned: 0
Location: Frozen Wasteland

Re: A simpleScanner setup, could use improvement

Post by Misty » 09 Mar 2010, 15:18

I'm not sure that moving the G7 closer would help enough to make up for this.

Using more lights or brighter lights is one option, definitely. Check the lumen rating of your lights, if it has one. If you replace them with lights with a higher lumen rating, you can get more light using the same amount of bulbs. You have another few options too.

- Use slower shutter speeds. The EXIF shows that both cameras fired at 1/125s. If you slowed the shutter speed to something like 0.3s - 0.5s, you can let the sensor gather more light. However, the disadvantage is that the images become much more susceptible to slight jittering. If your cameras aren't perfectly stable and move slightly during the shot, or something else moves just slightly, you will have blurred photos. With stable camera mounts, this can work fine. Since I currently have no control over the lighting where I work, I usually use a shutter speed of 0.4s and get perfectly clear images.

- Use a cloth shroud around your scanner, like Ann did in her thread. Aside from blocking unwanted outside light, it's possible that this could help by keeping the light from your book scanner inside the cloth. However, I don't know if this would help enough; I'm not sure how much light you're losing to it dispersing.
The opinions expressed in this post are my own and do not necessarily represent those of the Canadian Museum for Human Rights.

Post Reply