Remove all images during individual/mass conversion

[eduard93]

Device: Kindle Paperwhite
Use case: When I convert books to mobi I'd like to be able to cut all images from the result, because it saves space and it's not like Kindle can display images anyway.

Is it possible to automate this with Calibre?

[theducks]

My K4 (e-ink) and my Kobo Aura2 both display images.

True, they are grey scale, but they are displayed. Maps are visible

SAVE SPACE (And you don't claim to read Manga) ???? most (~65%) of my library is made up of novels under 400K.
Very few go past 1M and those tend to have embedded fonts and additional images (Image only Chapter heads/numbers. your plan would destroy the book).
With 4-8G of device storage, I have almost 100 books (almost a half-years reading) and not even using 1/2 the storage.

OTOH I do reduce HiRes covers to 90% of my screen size (600x800). That saves PROCESSING (scaling to device screen) using the resize cover plugin (this changes the Library copy, which will be also become embedded on demand)

Please reconsider your options or you may end up with a lot of damaged content when the cover is not the only image
This is not to say that the use of the image tools in the Calibre editor should not be SELECTIVELY applied Nor the removal of excessive advertising (Authors: do we really need HiRes covers from all your published works included? Forthcoming )

[retiredbiker]

Kindles will display the images, you probably mean they just look terrible. Mostly true, in my experience. But it is usually the fault of the image itself, or how it is coded, GIGO. How can publishers be so bad at this??? Often the images in the text are awful, but the ads are great, how about that! Maybe that's author versus publisher.

I keep everything in Calibre as epub only...partly for ease of editing, also because I have multiple devices, and yes, to save space.

If I get a new book, say a novel from Amazon, that is several megabytes large, it's usually due to many and/or huge images, and sometimes fonts. So I automatically edit these books. I take out unwanted images, especially ads, and also any font files (your choice, I'm happy with my reader's fonts). Keep important images like the cover, maps, stuff that is story-helpful or just pleasant.

Then I use the compression tool (Tools->Compress images losslessly), but I tick the box that says 'Enable lossy compression' and set it to 50. That results in a huge compression for jpg files, and I absolutely can't see any difference in viewing on a Kindle or a Kobo. A bad image stays bad, but a good image still looks good. Yes, I've degraded the images, but if I can't tell on my reader...so what?

I usually do a lot more, but if you just want to cut the size, it takes about one minute to open the editor, use the font manager to remove fonts, use the compression tool, save and exit. Typical novel that starts at 4.5mb usually drops to between .4 and .6 -- unless it is just a huge book in text, like over 1,000 pages.

[JSWolf]

Sometimes all the wasted space is due to cover images to advertise other books by the author. In Penguin eBooks, there have a page that's worthless that they embed fonts that are again worthless. Also, use Calibre to losslessly compress the images. That will save space. Delete every HTML file you don't need like previews, list of books by the author, reviews, etc. And as for images, when I do a lossless compression in Calibre, I also recompress using the default value of 80. This does reduce the size of images.

SO you can reduce the size of the eBooks if you do it by hand without ruining them.

[Brett Merkey]

@eduard93:
I agree with the excellent advice offered above. However, to answer your question, yes, it is possible to remove images on conversion. Look into the feature in Preferences > Conversion > Common Options > Search & Replace.

With a bit of trial and error, you should be able to find a regular expression to find <img> tags and replace them with nothing. In fact, I just tested with expression <img.*?>
Calibre removed the tag and the corresponding image as I converted from ePub to AZW3. Consider instead of replacing with nothing, it might be helpful to replace with a text box "image removed here" to clarify possible gaps in meaning.

[eduard93]

et al: Obviously Kindle can display images, but I mainly read fiction and full-color illustrations on 6'' monochrome display are useless for me.

@Brett would it work from any format? Obviously it works with epub which is zipped html, but what about other non html-based formats? Or does Calibre converts into intermediary html and runs regexp on it?

[JSWolf]

Are you close to running out of space on your Kindle?

[Brett Merkey]

Obviously it works with epub...

If the Calibre feature is so obvious to you, one wonders why you asked the question in the first place.

[theducks]

Originally Posted by eduard93

et al: Obviously Kindle can display images, but I mainly read fiction and full-color illustrations on 6'' monochrome display are useless for me.

@Brett would it work from any format? Obviously it works with epub which is zipped html, but what about other non html-based formats? Or does Calibre converts into intermediary html and runs regexp on it?

I also read fiction on my Kindle. there ar lots more image uses other than full color. (I use my TAB A, when color or panning is needed. I just whit THAT work in bright sun when I am trying to see the details of a color coded wiring diagram )

[hobnail]

Originally Posted by Brett Merkey

With a bit of trial and error, you should be able to find a regular expression to find <img> tags and replace them with nothing. In fact, I just tested with expression <img.*?>

What about instead adding this to the css: img { display: none; }

If that works then better would be to examine the html and see if images are wrapped in a figure tag and put the display none on it (or both figure and img as well as caption?).

[theducks]

Originally Posted by hobnail

What about instead adding this to the css: img { display: none; }

If that works then better would be to examine the html and see if images are wrapped in a figure tag and put the display none on it (or both figure and img as well as caption?).

The OP wanted to save SPACE. The file still would occupy space, just not offend their vision

[JSWolf]

I suggested how to save space. It seems the OP doesn't want to try my suggestions.

[hobnail]

Originally Posted by theducks

The OP wanted to save SPACE. The file still would occupy space, just not offend their vision

Yes, very true. I should have added that after the css hack they could then delete the images in the images folder, after first determining which one is the cover image and not delete it. If you simply delete the images then the result in the e-reader is ugly.

[Bookstooge]

Originally Posted by JSWolf

I suggested how to save space. It seems the OP doesn't want to try my suggestions.

And then you turn around and insulted them in your next post. Is it any wonder they didn't want to try your suggestion? Try not being silly in your own posts...

[eduard93]

Originally Posted by JSWolf

SO you can reduce the size of the eBooks if you do it by hand without ruining them.

I'm searching for automatic solution.

Originally Posted by JSWolf

Are you close to running out of space on your Kindle?

Yes, I have ~2 mb free rn so it's been a struggle for a last few months.

Originally Posted by Brett Merkey

If the Calibre feature is so obvious to you, one wonders why you asked the question in the first place.

epub being zipped html is obvious to me, not how Calibre processes files. Once again, what does Calibre runs search&replace on? Source book (in that case <img.*?> would work for epub, html) or some internal representation (in which case <img.*?> would work for everything)?