Sort question - even odd problem

[kicker74]

Hi there,

first of all: Scan Tailor is a really nice project: I implemented something similar in Java, but not so professional like Scan Tailor. Really nice tool.

Anyway: can I ask the crowd one question about importing and the sort:

I scan every page unique on my Pulstek Opticbook. After everything is scanned I rename all pages with its page numbers (make it much quicker to check whether a page is missing)
At the end I have even and odd pages. Normally all odd pages are flipped 180 degrees.

So when I import ALL pages (otherwise I have a different size for every second picture) I have to turn every second picture. No problem, but it takes a while for a book with about 1000 pages to mark every
second picture and telling scan tailor to flip "All selected pictures". So it would be better to rename it so all even pictures are packed, I click the first even picture and the last picture.

But I am not sure how the pictures are sorted in the thumb view: it is not the name (I already tested it to rename to EVEN_1000, EVEN_1002, ... and ODD_1001, ODD_1003) - also not the last modified date in an ascending order. Is there a possibility to write something into the picture so all pictures are sorted e.g. by name, creation date, modified date.

Thanks for any advice,


Holm

[Tulon]

First of all, the "Fix Orientation" stage allows you to apply the rotation to every other page. Look for this option in the Apply dialog.

As for sorting, it's actually filename based. It's a smart sorting though, so for example "page3.tif" will go before "page22.tif". If it doesn't work that way, then it's a bug.

[Moonboy242]

Kicker, there may be a second option to 'batch processing". Have you tried processing the odd numbered pages first and then the even numbered pages second? Once processed in Scan Tailor you can then combine the odd and even files into one folder for PDF processing. You do have to keep track of the margin information used in the first set so you can use them in the second and get output files as closely similar as possible (a flatbed scanner shouldn't have too large a variation for each page provided you use the same settings for all pages).

The benefit is that the processing time in Scan Tailor is reduced and you generally won't worry about rotating pages in the proper direction.

[kicker74]

Hi there,

yesterday I had a bit more time so I was able to test it. At the end I follow the idea with the renaming.

Meaning all even pages should not been flipped, rest yes. As example: 02.jpg, 04.jpg, 06.jpg Odd-pages: 01.jpg, 03.jpg,05.jpg
So I rename these pictures into 102.jpg, 104.jpg, 106.jpg. 201.jpg, 203.jpg,205.jpg

So I can mark 201.jpg, 203.jpg,205.jpg (they are now grouped together) and choose "Correct Orientation for all selected pages"

NameWizard helps me doing it ...

Thanks for all tricks,

Holm

[Tulon]

Why not just apply orientation to every other page? That option was created specifically for what you are trying to do.

[spamsickle]

Seriously. I think kicker74 must have just missed this suggestion the first time you made it, so let me emphasize it: There is no reason to be playing these games with renaming to batch the even pages, or attempting to select every other picture. Scan Tailor already gives you the option of rotating every other page, and that's the best solution to the problem you're describing.

Select the first page you want to rotate, then click on whether the rotation will be clockwise or counterclockwise. The first page will rotate. Right below the arrow you just clicked, there is an "Apply to" button. Click on that, and a pop-up dialogue will allow you to specify "every other page". When you click "OK", the dialogue will close, and all of your even (or odd) pages will rotate 90 degrees in the direction you've chosen. If you need to rotate 180 degrees, just do it twice.

[Moonboy242]

Using the Scan Tailor function is the fastest way to rotate pages. No argument there. However, and this is a matter of preference, my processing workflow includes batch renames of the odd and even pages prior to Scan Tailor processing. Why? I do it as a kind of check to ensure that I've actually imaged each page in the book and to also ensure that the quality of each page is adequate for processing (the prime example would be those occasional pages where there is less than a paragraph of text and the image turns up out of focus). By working with my images prior to Scan Tailor I am able to weed out any operator errors that would be magnified by Scan Tailor, such as rotating every other page only to find out that I somehow left out one page and have now rotated half the book in the wrong direction.

Yeah, I am sorta straying away from the original post topic. I think developing an efficient workflow prior to Scan Tailor processing helps ensure the highest quality product output.

[TJ-Shredder]

I would love to see the "apply to every other page" also in other stages. Most important obviously would be the Margins stage, but the Split Pages would make sense as well, as the automatic detection often fails...
I would not mind if the "Apply to..." would be the same for all. (simply reuse of the same code...)

Stefan