CZUR ET18 Pro Book Scanning Vintage Computer Manuals
Вставка
- Опубліковано 8 вер 2024
- Today I'm looking at how I can scan glue bound books using an overhead scanner!
See more about the CZUR Scanner here and use code "CZURTang" for $40 off: bit.ly/3xTEPae
#czur #ET18Pro #bookscanner
Here are some of the scans I got with it:
CBASIC Software Directory: archive.org/de...
HDD Manual And CBASIC Manual: drive.google.c...
(I decided not to put the first two on archive.org yet because I know I can do better now. So these are just here as examples from the video)
Here are some ways to help me make more videos!
Buy some Merch:
Merch: tech-tangents....
Leave a one time tip:
Paypal one-time: paypal.me/AkBKukU
Ko-fi: ko-fi.com/tech...
Amazon Wishlist for Production Gear:
www.amazon.com...
Become a monthly supporter:
Patreon: / akbkuku
UA-cam Membership: / @techtangents
Paypal Monthly: techtangents.n...
Other Links
UA-cam: / akbkuku
Twitch: / techtangents
Github: github.com/AkB...
Thingiverse: www.thingivers...
Discord: / discord
The GPU utilization seen when I was looking at it when doing OCR was because I was using the GPU to do 4k screen recording on the computer I was working with. It is possible it uses the GPU for OCR, but OBS was definitely using the GPU more if so. I suspect it is CPU based still though.
Well, I'm assuming by now you've had a chance to do a similar test without screen recording/OBS running. Is it hitting the GPU?
There are books that are 500 years old..... I do not know of any other storage type so durable as books..
CD's from early 80's already failing.... HDD's even sooner....
"the cloud" is neither eternal or safe... I have downloaded corrupt files from my google drive account...
I have a video upload in youtube that is corrupted... and odiously it was fine when y upload it...
Preserves the paper Shelby.... it is safer...
@@Ramdileo_sys books are not without their own disadvantages. Both water and fire will happily destroy them. Fungi love books, as well as rodents, and insects.
The best form of preservation is to copy it to as many places as possible.
Do you now there using you name in the comments
is that guy fake the one who says text me on telegram ?
As a historian, I truly thank you for all the effort you're taking in your archival work.
It doesn't go unnoticed!
One trick to help with page edge detection is to use a black paper between the previous pages. Adds a bit workload when turning pages, but helps a lot especially when pages want to skew or there is tabs protruding out.
Gosh, this definitely seems to have a fair bit of what I usually call “AI BS”, for both the OCR and the page outline extraction, except here it seems genuinely super useful rather than gimmicky and needing its hands held the whole time. 40 minutes for the entire book on your first time while talking really is amazingly impressive. The software looks really quick to get in a workflow for touching up the occasional egregiously wrong ones, but the rest seem 90-95% as good anyway and thus just as well to leave it to itself most of the time.
Oh, my dad got one of these last year to scan books with (he's an illustrator, blah blah lawyer reasons wanted copies of everything with his art) and after scanning thousands of pages with these I'd say they're pretty good, but a little underwhelming. Straightening doesnt work as well as it should IMO and the quality is a little underwhelming. Also the software was on the buggy side but maybe it's improved in the past year.
Edit: in all I'd say it's a fairly good product. The software + foot pedal did make it pretty convenient, there's just room for improvement.
Chinese products like this have almost caught up in terms of engineering, a traditional library scanner is much more expensive.
An excellent presentation for a first time user. Plus, you have to edit dialog, read the documentation, and , become semi-familiar with the product which I thought was short and to the point. My CZUR ET 24 Pro arrived yesterday. I ordered it as a Black Friday purchase two days before Thanksgiving and it arrived by FedEx and was handed to me in person the day after Thanksgiving. I was pleasantly surprised that it came to me in 3 days. The ET24 has features that the ET18 does not have and if you have an opportunity to work with the newer model your subscribers will be happy that you put in the time. I found your Channel interesting and professionally produced. I thank you for sharing with the rest of us out here as it is most appreciated. 👍👍
I've recently ordered one and I'm excited to try it out. There's no competition on the market for this type of scanner at this price point as CZUR seems to be the only company. Despite that, I'm glad it seems like a competent product.
I had hundreds of pounds of books. Every time I moved I would have a major project to move the hundreds and hundreds of pounds of paper.
So I finally decided to take a utility knife and cut them all up and send them through a document scanner, output as PDF.
Took me 2 years, but I now can fit all of my books in my shirt pocket. 🙂
Biggest issue I had was checking the scans to make sure all the pages got scanned as some of my books were 900 pages.
I recycled all of the paper. :-)
CuriousMarc UA-cam Channel did a few reviews of this brand along with him scanning old HP technical books and schematics. Wanted to mention in case you weren't aware of his channel.
I'm pretty sure Shelby is well aware of Marc, at least he follows him on Twitter.
And this has been recommended in the previous videos. A lot.
I think I remember Shelby even mentioned that when he first talked about archiving these manuals. Yes, he's aware of Marc trying this device, or a previous version thereof.
FYI, for really small runs and even things like snapping a whiteboard or posters off a wall, Microsoft has a free app for your phone called Lens. It does an ok job at framing and capturing in a pinch. It has replaced any need for a flatbed scanner I’ve had for a few years now and has the advantage of being in my pocket.
My uncle used to work at compaq in houston, thanks for the nostalgia and cool hat!
Things I learned from using one of these machines last year or at least some of them. The software has improved a lot I just watched somebody use the exact same machine with different software and finished a book a small book in a few hours. The Software is a bit chugly on older systems. If you're gonna buy one make sure you get one with a laser correction system otherwise it's all optical detection and it's totally worth the time and effort. Oh and some models are sold as having a lamp function for when not in use of their normal function.
I want to see you scan some BASIC listings, copy them to a disc, and run them on a real machine.
I used to rent time on a large format Cruz scanner for capturing pinball playfields and backglasses for repro purposes. Awesome machine.
I like to see these videos where someone has a sponsor for an item that they will actually use and that works well. Thank you for the video, I do not have a need for one of these but I am sure that this will be a helpful video for anyone that does need/want one.
Great video. As someone who has scanned hundreds of classic computer books and documents, I can see that some sort of raised edge in any corner of your desk would really help. Not only for workflow but also so the software doesn't need to do as much correcting. Trying to keep the book in the exact same spot every time you turn a page is a pain and will slow down work flow. If you had some raised surface to quickly line up the book after turning a page, it would make it much easier and quicker to scan. A L-shaped piece of wood clamped to the desk would probably work depending on your set-up.
Scam!
Seems you could do this with an SLR camera, a decent macro lens and a couple of diffused flash units. However, the software is neat!
After experiencing it, the software really is the secret sauce that makes this all work. I was thinking about ways I could do this with a DSLR and imagmagick for distortion correction but seeing this in action kinda hit home how difficult a fully DIY solution would be.
You *could* , but with this device, you *don't have to*
lots of video's on how to do is like that I was watching them year's ago.
Work smarter not harder
All this can be used to make a VR room where you can look at this stuff like you have it in your hand.
@CuriousMarc reviewed the ET18 almost a year ago (and the ET24 two months ago) - which was supposed to be the "updated" model at the time - hopefully I'm in a position to purchase one sometime.
Great work on historical preservation. I have a vision for practical use of the vintage and historical computer field of knowledge and hardware. I want to do systematic studies of historic hardware and software development, and then develop that into practical lessons on how to improve the design and development of hardware and software today.
Who got the bright idea to give their product a name that sounds like "seizure"!? That's almost as bad as the MP3 player I saw named Ainol.
My thoughts exactly!
I read it as "Caesar"
Or the RCA Dimensia :P
@@Fuzy2K Or radio stations owner Entercom changing their name to Audacy, which sounds exactly like Odyssey when said on the air, forcing them to spell out their web site address. I bet they paid some "corporate synergy consultant" to come up with that name!
@@vwestlife I keep misreading that as "Audacity" 😆
Nice demo! Shelby, you might want to take a look at the open source project ScanTailor / ScanTailor Advanced. It might help you a lot with its fully automated workflow for cropping the pages evenly and eliminating remaining shadows. If you are also going to try out an SLR camera setup as an alternative, as is suggested in another comment, BookSorber (albeit not free) is a good companion, delivering very nice postprocessing results with regards to curvature correction, page splitting, cropping and finger removal.
Excellent video. I'm interested in magazine content like photos in b/w and color, typical articles and cartoons, the full page ads, the detail that is preserved. Thanks a lot for the overview.
First!
I'm surprised they are having you review the older model. Weird...
I was told they are out of stock of the 24MP models out the moment because they are fulfilling all the indiegogo orders. The door has been left open to potentially take a look at that one later though.
@@TechTangents Definitely... I can't wait to have my ET24 Pro. I've ordered one 2 months ago after seeing curiousmarc's video about the new model.
8:36: I couldn't quite make it out, but I'm not sure the character you deleted was superfluous. Could it be that that's just a visually rendered soft hyphen? Because it might be smart enough to recognise that.
I looked at the link for the price of this thing, expecting it to be super expensive. But it is actually really inexpensive for such a professional product, I have to say!
Did not know this existed. Cool.
That's a really impressive piece of hardware.
Glad CZUR sponsored you! I saw Dave Jones use one of these too. Seems pretty nifty!
I have been looking at such a solution for a while now. My parents have about 20 annotated photo-albums. Especially the annotations are important as the images are already digitized (scanned negatives). With such a scanner I can photograph all pages, combine them into a PDF and add the (high-quality) images as attachments. This looks way better then my DIY attempt with a lightbox and a Sony photocamera. Would you recommend this tool for this workload?
Incredible software. holy crap.
This scanner is quite impressive.
Damn this looks good. My only complaint is that unlike Zeta’s large format book scanners, this doesn’t have a way to hold the cover of the book at a different level, useful when dealing with really old or fragile material. Seems to scan way faster than the Zeta ones and interfaces with a laptop, rather than a thumb drive so UX is a bit nicer.
Quite impressive.
I'm already looking into getting something like this for exactly that reason: amzn.to/3yb9IXL I need to test that the dewarping can handle angles but I already had one book with a weak spine start to fall apart while scanning. I think it should be possible to use with this though from what I've seen.
I saw a similar camera in the medical industry used to verify proper container sizes when filling prescriptions in the hospital's pharmacy. It's database must've known the pill size or expected bottle to be used for the number of pills issued
Hi. The scanner looks pretty impressive but I have a few concerns. My main concern is damaging the spine. I've scanned a lot of books for preservation and some of them are a bit rare/old. Some of them I've bought myself for hundreds of dollars and some of them were from libraries/private collectors so I'm not super keen on messing them up.
The usual way to prevent spine strain is to rest the book in a cradle so the book isn't laid flat. Then the picture is taken from the sides instead of top down. Most people also use a V-shaped glass plate to keep Tha pages flat.
I managed this at home with some plywood and acrylic glass. To capture the pages I use my DSLR on a stand together with a desk lamp.
I would say that I can scan the book in about the same time as you did here but it would be faster if I could capture two pages at a time.
But some things about your scanner that I think looks really good.
First is the ease of use. For me it's a bit of a hassle to set up even if it's fast to capture the books so I often procrastinate until I have a enough books to make a day out of it. And the fixed lights on the thing seem so much better for getting consistent good lighting conditions.
The foot pedal seems pretty neat also (I use a camera remote).
The thing I'm really impressed by is the software. Both its performance and content. I use Scantailor advanced and JASW to process the pages. They are really good (especially Scantailor) but I find them not that forgiving when it comes to capture quality. Especially white balance. And AI warping correction is still not good enough for me to find it useful.
The czur software seems really intuitive and packs quite a punch. The built in OCR is also something that looks really good. I use Foxit Phantom PDF to compile and use OCR. It's an excellent app but it adds an extra cost and is subscription based. And the conformability of having an all in one software gives the Czur app an advantage.
If Czur could make an adjustable cradle and a glass plate I would seriously consider buying it just for the ease of use and consistent conditions.
I can't help but notice that the scanner looks like it came out from the War of the Worlds :D
Can you use a glassplate with it to scan?
Called it! Really seems like a good device.
OCR seems to be single threaded, if it is running on the CPU that is. Would probably be 4x faster with multi threading
You are doing the... excuse me for this, but the Machine God's work!
Oh shoot, didn't know that it's so amazing!
17:40 You didn't comment on the high GPU usage. Is it because the processing is hardware accelerated? Hence your surprise of how quick it is? Or the high GPU usage is simply because you are screen-capturing at the same time, for the sake of making this video?
If CZUR uses GPU acceleration for image processing, that's pretty awesome!!
The high GPU usage was probably OBS saving the high quality 4K screen capture. I was doing everything on the same system.
Tesseract pegs all CPU cores at 100% instantly, so that's what I was looking for but it didn't do it which is why I was so surprised.
thats a very nice piece of kit .
I'm doing something very similar with my own collections of manuals and such, though they're for a very much different (though no less historically significant) field - patternmaking and fashion design manuals. I'm currently restricted to the flat media and media that I can lay out flat in the A3 scanner that I recently picked up inexpensively. I would love something like this to help with scanning what I can't/don't want to unbind for digitising purposes, but it is very much not in my budget right now.
Thanks, can't wait for mine to arrive now. I thought they would probably be using Tesseract, but maybe not if it's a lot faster. Tesseract is good but has issues, especially with Japanese text.
17:43 CPU usage during OCR, 18%. GPU usage, 75% on a RTX 3070(3060?). GPU accelerated OCR makes sense, but damn!
That was most likely OBS for the screen capture
@@TechTangents Oh, I see. Having GPU accelerated OCR would be a nice feature. There is a Nvidia PDF from 2014 about GPU accelerated OCR that is an interesting read.
I wish there was a version for big format plans and drawings- I always have to use a A3 scanner and stitch everything together manually after. Bigger scanners are just way too expensive to justify using 5 times a year and those things are giant... they also tend to distort the image a lot.
17:50 That's an 8 core 16 thread CPU, 1 16th of 100 is 6.5%, assuming you're computer isn't doing literally nothing else, it wouldn't surprise me if it's maxing out a single thread.
Not a single word about the file sizes this thing generates and how to manage it by adjusting the resolution. That for me is a major problem, with scans generating huge file sizes.
Given the speed and quality, I'd expect that this professional-grade archival scanner is probably using custom ASICs for most o the heavy lifting.
It seems like you would have gotten better results with the last pair of books if you had used B&W (I assume this is the software's 1 bit per pixel setting) instead of grayscale. I suppose you could argue text would be easier to read, but if your DPI is high enough I don't think that would be an issue for an end user unless they were super zoomed in on the document.
Well, the obvious downside is the lack of control of recognized text, page segmentation, etc. It's like the Sprint version that came with some old scanners.
The software works under windows. But does it work with wine under linux or is there a linux version of the software?
You didn't address that.
Oops, I did edit that part out after making some corrections. I did test it on linux but the scanner is a UVC device and there are apparently some issues with using those in wine and I couldn't get the scanner to show up in the software. The software did run though, so if the UVC issue could be solved I bet it would work.
Czur doing you a solid!! CuriousMarc has one and he uses it a lot.
It looks like there's a little screen on top, so I'm guessing it's designed for you to stand over it to do the scanning?
Also, loved the cat cameo (catmeo?) for the hard copy.
You could look at the screen on it, that is probably more useful when you are using it with the mobile app. But I didn't end up using it when I had it connected to a PC since you get live video.
And that was very unexpected! I didn't put him there, he ran up to the printer himself when it started!
Is there any Linux open source solutions for this type of scanner?
I think a little better diffusing would be better for the lighting.
Neat. Not cheap, but neat.
Impressive.
These things make me think I can get by with my phone camera.
So this won't actually do any global type setting after the scan or can it? I'll need to reformat all pages at once, change fonts... and remove double spacing.. I have dozens of typewriter typed journals of my ancestors that I want to reformat and print for home publication. Sorry for the question but it wasn't clear if it just did the ocr and then saved it out for use in another processor or whether it could do the editing as well.
Nice demo! Just wondering... Did you have any issues getting that ET18? Asking because I have a 24 "shipping" to me and they're apparently having a bunch of issues getting the correct shipping information to people (inclusive). Plus, their customer service is being pretty evasive across the board. Looks like a great product line otherwise.
I wonder if it can handle OCR..
welp i answered my own question.
What is your external keyboard model name? Thanks
Can you tell the make name and model of your laptop stand?
How large is that mat, i.e. how big is the maximum scannable surface, and does it handle really thick tomes on top of that?
Nice!
Why czur18pro but czur48pro one?
Does it also work for taking pictures of PCBs?
Czur has windows software only?
$400 seems like a really reasonable price too. People pay that for a phone these days.
For _half_ a phone even… definitely a very decent price for what is quite a niche product. This type of systems have been used for quite some time (if we include silver film captures, more than a century by now) by historians, archivists, museums, etc. It is nice to see a system that is geared towards also hobbyists and smaller institutions that don’t have the funds for a top end system. With that capable hard- and software, they probably have a customer even in the most serious archives and museums for many documents.
@@DrBovdin I love the future.
Holy schnikies! $400, nice but little too rich for my blood. Great video
17:54 looks like it is using 80% of the GPU though. So would need a decent GPU for this to be kinda fast.
I use Linux. Every scanner I have come across only supports windows and in some cases Mac os. Czurs more expensive scanners are not for me. I'm looking for something more reasonable that scans books properly. $400 dollars and below CAN funds. Maybe I'll by a scanner wand that might support Linux.
What do you use to edit your videos?
I am interested in your computer stand, can you tell me the brand?
...and its missing wireless?
What is that laptop stand?
What are the system requirements, i.e. how much computer does this need to operate well?
(I'm looking for minimum and recommended tech specs, preferably no hand-wavy "anything decent" comments.)
the OCR seems to be Cuda accelerated as the GPU sat at ~80% when you looked at task manager during the process
Anything way less expensive
I thought it was pronounced churrrr, hah. I think i prefer that anyway over Caesar. Sounds too fancy.
Have you tried the software on linux?
I did test it on linux but the scanner is a UVC device and there are apparently some issues with using those in wine and I couldn't get the scanner to show up in the software. The software did run though, so if the UVC issue could be solved I bet it would work.
The one thing I hate about all of the semi-professional scanning solutions is the lack of Linux/open source software. I would love a nice batch photo scanner, but all the ones I have tried are hard tied to Windows and very clunky software.
15:30 a bit overreacting , like you can totally notice it’s a sponsored video 😂
I don't see a reason to ever use a flatbed scanner ever again for scanning documents!
Love your videos. If I could suggest one thing to instantly make your videos far better: train yourself to not say “uh” and “um.” I’ve been subscribed for a long time, but it’s getting pretty hard for me to watch. Part of my frustration is not just because I have a dumb pet peeve; it’s because I know you’re a lot smarter than the impression left by this vocal tic. For many, this common habit is a result of a lack of confidence. If that is the case with you, then I can tell you that you’ve got no reason to lack confidence. You have interesting content, your love for what you’re doing is infectious, and you do it increasingly well all the time. Changing vocal habits is not easy, but I think it would be worth the effort.
It's probably not going to happen because like you mentioned things like that are difficult to train yourself to drop. But I did want to mention that confidence isn't the main reason people use "uh"s and "um"s like this. It is subconscious tactic to keep "talking" to maintain attention from those listening and buy yourself time to form your next thought. This video was not scripted and I was coming up with what I was saying "live" which is where you will see this happen most.
So it might help to think about it that way, I have zero problems being on camera, but I try to be deliberate in what I say which will lead to me thinking statements through more.
Guess your not watching curious Marc, that's too bad...
I want one but fuck that shit 400$
I read scanning as scamming.
Looks like your channel has become big enough to attract the scammers replying to everyone that they’ve won something.
they literally could not pick any another name that doesn't sound like "SEIZURE"? lmfao.... i suppose it's better than PRLAPS....
There are books that are 500 years old..... I do not know of any other storage type so durable as books..
CD's from early 80's already failing.... HDD's even sooner....
"the cloud" is neither eternal or safe... I have downloaded corrupt files from my google drive account...
I have a video upload in youtube that is corrupted... and odiously it was fine when y upload it...
Preserves the paper Shelby.... it is safer...
Paper WILL yellow, deteriorate and disintegrate over time, especially the acid-based paper commonly in use.
It's really useless
Too bad it's not good enough for picture books. The color sucks.
ARGH. Another youtuber that doesn't use high pass filter on his mic audio!!! Every time you bump/move the table/things on the table, my subwoofers goes completely bananas. (Do you have a mic arm mounted to your table or something?) LEARN HOW TO RECORD/POST PROCESS AUDIO PROPERLY PLEASE.
Not everyone can be an expert at everything. No need to be so harsh when simple actually constructive feedback is more useful. Consider that your subs may also be high, and you can turn them down. I have a nice system myself, and I'm a professional live audio tech, and while it was noticeable, I didn't think it was that bad. Chill out.
Doors over to the left there bub