First page as image

[Ghitulescu]

I tried to convert a few text PDFs to epub, to spare me the COPY/PASTE effort.
It always converted the first (text) page into an image, which I assume calibre thinks it's the cover, even if I told it that I don't need any cover.

BTW, this explains why many books on archive.org are made this way.

Maybe a hidden setting, somewhere?

[BetterRed]

@Ghitulescu - putting an image of the first page at the start of a PDF is a widely used practice for PDFs that don't have a cover. That said, PDF's do not have a cover as such, if the first thing in a PDF is an image then that is the implied cover.

If you don't want the 'first page image' on the EPUB then replace it, perhaps with a generated cover, before you create the EPUB. If you don't want the EPUB to have a cover use the Editor or the Modify PI to remove it.

IIRC if you do a Save to Disk or Send to Device calibre will create an EPUB with the cover.jpg thats in the library. I don't use those features for that reason (and others). I like to have a cover in the library (most are created via Generate Cover PI), but I don't like/need them on my devices.

BR

[eschwartz]

BR,

I believe the save/send process, which uses the same logic as Embed Metadata, only updates covers. If the EPUB does not declare a cover, none of these processes will add one. Traditionally you needed Modify EPUB or Polish book to force insert a cover.

Also conversion will add one. It will not use the contents of the input file, it uses the cover metadata in calibre.
When using Add books on PDF, calibre will IIRC use an image of the first page to add as cover metadata for the book record. This will be duplicate content.

[BetterRed]

@e-s - That's almost the opposite what I recall of my experience when I first used calibre.

If I had a cover in the book folder (most often created by me using GenCover PI) then when I created the EPUB via conversion it would inject the cover.jpg into the EPUB - thats the way conversion works I was told. However, even if I used Modify, the old Tweak, or Sigil to remove the cover image and then did a save to disk or send to device the cover would be back in the output EPUB

The answer I got from various 'experts' was - if you don't want covers on your device you'll have to use your file mangler instead of save to disk or send to device, but you wont be able to use plugboards etc. Which was fine by me because that's exactly what I do with all other file types - xls, flac, avi, jpeg, cr2 etc.

I may have been getting confused/sidetracked/bamboozled by the ORIGINAL_XXXX files, which at the time was another thing I couldn't get my head around. I eventually found the Tweak to prevent their creation, but I think that was after the "I don't want covers on my books when they 'leave the library'" to and fro.

Traditionally you needed Modify EPUB or Polish book to force insert a cover.

Aaah, but I travel the other road - being an iconoclast I use Modify to rip the covers off and clear all the non dc metadata

I've since come to realise that many people are as OCD about covers as I am about dangling participles :lol:

BR

[Ghitulescu]

Thank you for answers.

The point is that the first page of the PDF becomes the "cover" and the text disappears, which makes me either type it back or COPY/PASTE it from the PDF, then redo the formating (which is usually different for a first page, because of title, author, chapter etc elements.).

[cybmole]

the text should not disappear.
I have many PDFs in my library ( via a simple add books) where calibre has made a "cover" from a 1st page image, but the 1st page text is always still intact.

maybe you can post an example ?

PS I usually then use "generate cover" & calibre makes one, with the title, author, series etc info on it. it takes that data from its' own metadata, whcih you can modify beforehand.

[BetterRed]

Originally Posted by cybmole

the text should not disappear.
I have many PDFs in my library ( via a simple add books) where calibre has made a "cover" from a 1st page image, but the 1st page text is always still intact.

maybe you can post an example ?

PS I usually then use "generate cover" & calibre makes one, with the title, author, series etc info on it. it takes that data from its' own metadata, whcih you can modify beforehand.

cybmole - I convert at least 40 PDFs a week, some already have an image of the first page as the cover, others don't.

I may be wrong, but I suspect slapping an image of the first page at the front of PDFs is not a calibre convention. A well-known global financial institution does it with some of their routine PD reports, and they've been doing so for 20 odd years.

BR