Beautified a scan I found from IMSLP - how to contribute?

[vaapo]

Hi!

I am new to IMSLP contributions and I need advice how to do properly contribution on based on scan made by others:
https://imslp.org/wiki/File:PMLP116645- ... 00cacc.pdf

I removed the colored background and fixed the blacks on the above scan, so that it can be better printed in black and white.
I have the PDF file, PNG exports and the gimp xcf-files.


rgrds Paavo

[vaapo]

FYI: upload is now pending there for copyright acceptance and one thought came to my mind.
Has it been practice to split the original scan to smaller pdf's e.g. pdf per song, or is it bad practice?

rgrds Paavo

[coulonnus]

If you've beautifyed scan #56498 of https://imslp.org/wiki/Le_nuove_musiche ... i,_Giulio) read https://imslpforums.org/viewtopic.php?t=8977

Once you have the set of 400 dpi jpg's read https://imslp.org/wiki/Image_Conversion

[coulonnus]

Since your beautifyed scan was grayscale, low resolution I thought I had to replace it with a monochrome one, without changing the resolution of US-Bpr. Compare the sizes.

[vaapo]

Indeed - thank you for your help in rolling this out! The sizes is 1/5 from the one I made and potentially much easier to print.

I need to learn the linux tools for this process and see if there is anything to do with gimp to automate certain 'mappings' towards white background or towards black. From the process I can now say that it is good to have color scan, because then cleaning can be made pretty precisely with picking color by color.

Code: Select all

#So in gimp I exported each page to separate file like follows:
file caccini000.png 
caccini000.png:                             PNG image data, 795 x 1229, 8-bit/color RGBA, non-interlaced
#...etc..
# and then simply with convert:
convert *.png caccini.pdf

So need to find more shell trickery to the process.

rgrds Paavo

[coulonnus]

But you have to specify the density when you create the png's. In this case it was 350. And when you convert from png to pdf you have to specify a threshold to obtain a monochrome pdf. In this case it was 30% for a compromise between removing sepia and letting the printed characters intact. Without this threshold the pdf will be bigger.

[LuciaEmari]

This a great piece of technology, but why, when Im running a very large scan -- say 7500 pages -- does eValid get "stuck" part way through?