a few questions about ebook fonts

[tatagi]

a newbie myself, I am glad there's forum solely for ebook related topics.

I am trying to make collection of ebooks, turning my pc into privary library.
But as collection grows bigger, there comes storage limitation issue.

most of my ebooks are epub format, and each file is sized ranging from 10 to 40MBs, all with searchable text. (I don't like scanned version anyway, the quality will be compromised after compressing them to reduce sizes)

I heard that by removing embedded fonts you can expected to dramatically shrink the ebook size, regardless of page counts.

now, the question part.

1. how much can you expect the file size to be reduced? especially epub and pdf formats

2. will this affect original document in a bad way? like dislocated contents inside the charts or overall layouts, or some characters breakage (like empty square replacing the original character)

3. does Calibre support removing fonts? I have little to no knowledge of html or css editing. wish it's not too difficult and offer step by step solution.

4. any tips and advices while doing this.


Thank you so much in advance!

[retiredbiker]

If you have epub files with sizes of 10 to 40 Mb, they must be chock full of fonts and, more probably, images. Images usually take up more space than fonts in huge epub books. Pdfs are different, see below.

For a baseline, a simple epub novel without illustrations other than the cover, but with lot of fonts, might be, say 5 or 6 Mb, extreme case. Removing the fonts and compressing the cover might bring it down to 0.2 to1.0 Mb depending on the text length -- short novel at one end to War and Peace at the other.

All epubs are searchable, as the text in them is in html format...in other words, coded plain text. A "searchable pdf" on the other hand, is a pdf with text in it that is searchable...this is an option for a pdf.

To get started with the editor, Kovid has an excellent video tutorial: https://calibre-ebook.com/demo#tutorials. Ebooks are complex critters, so there is a learning curve, but once you see the basics it is not too difficult.

So if you have a large epub, you can open it in the editor. There is a file browser that will let you see what is in it in the way of fonts and images, and how big each file is. You can remove fonts or subset them to reduce their size. You can compress images, or remove them. Explore the "Tools" section of the menu to find what you need.

Yes, you can really and truly mess up a book doing this. If you have school or scientific textbooks, you are best to just leave them alone. Work on a copy of your book and see what happens.

Calibre is NOT a pdf editor. Taking fonts or images out of a pdf will have to be done with some other tool, and good luck. Pdfs are not meant to be edited.

[Quoth]

if the fonts are subsetted they have little impact on file size on any epub otherwise more than 1M byte.
Some fonts are not so large even without subsetting (like they only have latin letters) then it makes no difference.
If you files are 10M and larger it's usually not fonts but images.

[JSWolf]

The thing is, you should not just rip out the embedded fonts without checking if they should be there or not. I've seen some eBooks where ripping out all of the fonts is not a good idea. Also, you may want to keep some fonts and rip out others in the same eBook. So any eBook with fonts should be taken on a case-by-case basis.

Most large eBooks are due to images. One thing you can do to reduce the eBook size is to Polish book and select Losslessly compress images. That removes the junk in the images without changing the compression. I cannot say how much space this will save, but it will save some.

[tatagi]

thank you all for the reponses..

helped me a lot to get an idea how things are done.

probably I should've mentioned this to avoid the misconception.

most my ebooks are written in korean, and it has 11,172 syllable combinations as separate code in the font file along with thousands of basic chinese characters, unlike latin alphabet with dozens of combination.

I only intend to get rid of fonts in non-graphical docs like novels or essays to mitigate the issue of damaging the layouts.

any recommended software(free the better) for editing ebooks?

It seems like Calibre is too heavy, taking up a lot of rams and storages.

my main viewer is sumatra pdf, now it's time to find lightweight, fast editors.

[JSWolf]

Originally Posted by tatagi

thank you all for the reponses..

helped me a lot to get an idea how things are done.

probably I should've mentioned this to avoid the misconception.

most my ebooks are written in korean, and it has 11,172 syllable combinations as separate code in the font file along with thousands of basic chinese characters, unlike latin alphabet with dozens of combination.

I only intend to get rid of fonts in non-graphical docs like novels or essays to mitigate the issue of damaging the layouts.

any recommended software(free the better) for editing ebooks?

It seems like Calibre is too heavy, taking up a lot of rams and storages.

my main viewer is sumatra pdf, now it's time to find lightweight, fast editors.

To edit ePub and KF8, Calibre is recommended. If you want to edit just ePub, then either Calibre or Sigil. The thing is, Calibre does more then editing eBooks. It allows you to keep track of them (manage them).