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

ST suggestions and feature requests go here.

Don't know where to start, or stuck on a certain problem? Drop by and tell us about it. Feel like helping others? Start here.
Post Reply
User avatar
daniel_reetz
Posts: 2780
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:

ST suggestions and feature requests go here.

Post by daniel_reetz » 12 Dec 2010, 15:18

Hi all,
As the author of an excellent Free Software package, Tulon is inundated with support requests and feature requests. Tulon has repeatedly asked that people stop asking for new features. He has also said that he has officially retired from Scan Tailor development. These two things are not unrelated. Tulon has also said that he is only interested in working for himself, and only interested in working on dewarping in the future (if even that). So these kinds of requests should no longer be directed at Tulon personally for any reason.

From now on, please post your feature requests and new ideas here in this thread. Do not expect a reply from Tulon. Even though you should not expect a reply from Tulon, future Scan Tailor (or whatever program we end up developing) users and developers can look here to see your ideas and possibly implement them. We all want something a little different, and as we pick up developer talent and move forward, there will be plenty of opportunity for new features. That means that you shouldn't worry if you can't have your most-wanted feature right now. In time, it may come.

Thanks for your understanding, and for posting your ideas here. We really do value your ideas, they help us all get to a better place for book scanning.

kasslloyd
Posts: 41
Joined: 19 Dec 2010, 21:25

Re: ST suggestions and feature requests go here.

Post by kasslloyd » 07 Jan 2011, 02:28

From my experience I think the mask editing and content selection for Mixed mode needs some much needed improvement. Maybe a slider bar that increases sensitivity of what is masked for quick adjustments, or click an object and it intelligently selects it, more tools than just a polygon tool... Fixing masks BY FAR takes the longest time.

Biggest is, you should be able to preview the mask BEFORE generating the final output, churning out a 600 dpi output takes WAY to long... anything to speed up the checking the mask to see if it needs edited once it was auto-generated will GREATLY improve postprocessing time of my scans.

My two cents. ;-)

Another thing that would be nice is for the Mixed mode that it automatically detect colored text and color it with the average color of what the text is. So instead of just getting all black text when your doing Mixed mode output you preserve the color text...

User avatar
jimboh
Posts: 82
Joined: 06 Jan 2011, 00:55

Re: ST suggestions and feature requests go here.

Post by jimboh » 07 Jan 2011, 04:16

Are Tulon and jart different people? I ask because jart appears to be actively developing the program over on sourceforge.net I'm not entirely clear on the relationship between this and that site.

In any event, I guess I should have posted http://www.diybookscanner.org/forum/vie ... =793#p7705 here, rather than creating a new thread. Sorry; newbie error.

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

Re: ST suggestions and feature requests go here.

Post by spamsickle » 07 Jan 2011, 07:16

Tulon and jart are the same person. From what he's said here, the only active development he's doing is attempting to perfect the dewarping feature. Once that's done he intends to retire from further development, though he will still be available to answer questions from those who want to continue to enhance and modify the application. He may also fix bugs discovered in existing function.

Tulon's impending retirement shouldn't discourage anyone from posting suggestions and feature requests here, and I hope this will become a place to discuss potential enhancements as well. Whether such enhancements make it into a future branch of Scan Tailor or merely inform the development of a other post-processing applications, honest and respectful feedback from the people who are using the tools can be very helpful to the people creating them.

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

Re: ST suggestions and feature requests go here.

Post by spamsickle » 07 Jan 2011, 07:37

kasslloyd wrote:Another thing that would be nice is for the Mixed mode that it automatically detect colored text and color it with the average color of what the text is. So instead of just getting all black text when your doing Mixed mode output you preserve the color text...
Text is currently all black on an all-white background. This was done deliberately to make it possible to separate easily a text layer from an image layer in post-Scan-Tailor processing. That capability is useful in creating ebook formats like DJVU.

If that were to be expanded to handle color text, rather than replacing text with the average color, it would probably be a good idea to maintain the ability to separate text from image. I'd suggest limiting the number of text colors to the six that can be created with "pure" values. To black (pure 0-0-0, in RGB), one could add red (255-0-0), green (0-255-0), blue (0-0-255), yellow (255-255-0), purple (255-0-255) and blue-green (0-255-255). This should not seriously impact the palette of colors available to create images, and would expand the range of distinguishable text enough to handle anything that wasn't pathological. For the obvious most-common case (red and black text), this could still be done automatically. I can't recall running across "tricky" colored text which would require additional controls in the user interface, though it may exist.

Writing this does bring to mind something I have encountered commonly -- text which appears on a colored background, to distinguish it from the main narrative. This is common in text books; whole paragraphs and even pages may be distinguished this way. While I like the binary text I've been replacing this with, it would be nice to retain the special character of such text, perhaps by framing it with an outlined border.

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

Re: ST suggestions and feature requests go here.

Post by spamsickle » 07 Jan 2011, 16:30

jimboh wrote: It would be a nice enhancement if, in addition to custom shape drawing, several preconfigured shapes -- e.g. a square, rectangle, triangle, circle, and oval -- were offered on a palette and could be dragged, dropped, and resized to create a picture zone.

Similarly, the ability to copy and paste an existing picture zone (i.e. the dimensions, not the content) would be useful, especially if we could center its placement with a mouse click.

To deal specifically with mouse-related tendinitis, some keyboard shortcuts would be nice. Perhaps, for example, we could draw a square or rectangular picture zone by dropping a red dot with the mouse, then use the arrow keys to draw the lines in four directions. Indeed, the arrow keys might be used to draw rather complex boundaries, like an Etch-A-Sketch.
http://diybookscanner.org/forum/viewtop ... =793#p7705

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

Re: ST suggestions and feature requests go here.

Post by spamsickle » 07 Jan 2011, 16:32

jimboh wrote:If a fixed content box is ever implemented, it would be nice if its size were fixed, and its location/placement could be moved up, down, or sideways with keyboard arrow keys. Maybe Ctrl-arrow combinations could be used to expand (or shrink, with Ctrl-Shiftt-arrow?) the box on the side, and in the direction of, the arrow.

Maybe the latter could be implemented apart from fixed-size boxes. Since the code already exists to move all four boundaries in all four directions with the mouse, would it be a huge problem to access the same code with the keyboard? Not really new features, but rather keyboard shortcuts to access existing features.

I'm thinking a lot about mouse-related tendinitis these days. FWIW.
http://diybookscanner.org/forum/viewtop ... t=20#p7733

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

Re: ST suggestions and feature requests go here.

Post by spamsickle » 07 Jan 2011, 16:45

ibr4him wrote: • In 'Split Pages' section, allow click to move page split guides instead of dragging them with mouse; saves a couple seconds per page.
• In 'Select Content' section, allow 'Snap to' feature. So when 'Snap to' is enabled, the area you select almost seems magnetic; This is enormously useful for accurate selection of page content.
• Undo / Redo last action. Nice to have, but not as important as the other two.

These suggestions are based on my practical experience using ST for atleast 8hrs daily now since last 2 weeks. :)
http://diybookscanner.org/forum/viewtop ... =697#p6744

kasslloyd
Posts: 41
Joined: 19 Dec 2010, 21:25

Re: ST suggestions and feature requests go here.

Post by kasslloyd » 07 Jan 2011, 20:23

spamsickle wrote:
kasslloyd wrote:Another thing that would be nice is for the Mixed mode that it automatically detect colored text and color it with the average color of what the text is. So instead of just getting all black text when your doing Mixed mode output you preserve the color text...
Text is currently all black on an all-white background. This was done deliberately to make it possible to separate easily a text layer from an image layer in post-Scan-Tailor processing. That capability is useful in creating ebook formats like DJVU.

If that were to be expanded to handle color text, rather than replacing text with the average color, it would probably be a good idea to maintain the ability to separate text from image. I'd suggest limiting the number of text colors to the six that can be created with "pure" values. To black (pure 0-0-0, in RGB), one could add red (255-0-0), green (0-255-0), blue (0-0-255), yellow (255-255-0), purple (255-0-255) and blue-green (0-255-255). This should not seriously impact the palette of colors available to create images, and would expand the range of distinguishable text enough to handle anything that wasn't pathological. For the obvious most-common case (red and black text), this could still be done automatically. I can't recall running across "tricky" colored text which would require additional controls in the user interface, though it may exist.

Writing this does bring to mind something I have encountered commonly -- text which appears on a colored background, to distinguish it from the main narrative. This is common in text books; whole paragraphs and even pages may be distinguished this way. While I like the binary text I've been replacing this with, it would be nice to retain the special character of such text, perhaps by framing it with an outlined border.
Of course it should be an option, more options are better than less.

I scan all my text books (easier to carry an ipad than a ton of textbooks at college) and yes this is rather an issue of colored backgrounds. I'm sure there is a way to detect colored text boxes and make it pretty by equalizing the background color and text. Another problem with textbooks is that in Mixed mode it has a serious problem detecting diagrams. Virtually every diagram I have to manually edit the mask to make it appear right. It also has issues with very dark images, doesn't properly detect the picture box, even when it's square.

User avatar
jimboh
Posts: 82
Joined: 06 Jan 2011, 00:55

Re: ST suggestions and feature requests go here.

Post by jimboh » 23 Jan 2011, 15:25

... even when it's square and has a solid black line around the outside.

It is great to see several new people contributing code for ST! Maybe one of them could look at the automatic generation of picture zones in Mixed mode, given how crucial it is and how badly it currently works? I'm not a programmer; otherwise I would help with the code.

Post Reply