Calibre won't side load a zip containing images of different formats to a Kobo

[ceridwen]

The usual way I side load books to my Kobo Aura H2O is with Calibre.

I have zip files containing images with different formats, for instance JPEG, GIF, or PNG. The exact formats don't seem to matter. When I try to get Calibre to side load one of these zip files to my Kobo, Calibre will not give me an option to send the file to my ereader, it will instead pop up a dialog that offers to convert the book, but if I tell it to attempt the conversion, it will then fail. Note that Calibre will place zip files containing images that are all of the same format on the Kobo without any problems, and that works for any zip file containing images all of the same format as long as that format is one the Kobo will read.

This isn't the first time I've seen wonky behavior from Calibre on zip files: at one point when I attempted to side load a zip containing images and an extra file (thumbs.db or something named like __MacOSX that apparently Macs add to image directories), Calibre would attempt to interpret the file as an epub and refuse to put it on the Kobo. That also caused it to pop up a conversion dialog that would then fail if I attempted the conversion.

I'm trying to figure out whether this is a limitation in Calibre or a limitation of the Kobo ereader. If I could put a zip containing images of mixed formats without using Calibre, would it work? Is there a way around this problem that doesn't involve needing to convert all the images to the same format first?

[stumped]

"I'm trying to figure out whether this is a limitation in Calibre or a limitation of the Kobo ereader. If I could put a zip containing images of mixed formats without using Calibre, would it work?"

well lets see- you could test that yourself very easily- or you could hang about hoping someone here will test it for you ... or perform a quick LMGTFY

[davidfor]

@ceridwen: You need to have the books in calibre in a format the device supports, or in a format that calibre can convert to a supported format. "zip" is not a supported format on most devices and definitely not on a Kobo device.

If you do try to send a zip file to a Kobo device, calibre will try to convert it. That conversion will be to look inside the zip file for a valid book. If there is one, it will be extracted and converted if needed. If there isn't one, you will get an error that calibre cannot convert the zip file. If the zip file is a collection of images, then calibre will not convert it.

But, if the zip file is a collection of images, change the extension to "cbz". Calibre and the Kobo devices will recognise this as a comic book and handle this.

And none of this is "wonky". It is all consistent. Maybe what calibre does when converting a zip file isn't obvious, but it is consistent.

[ceridwen]

Originally Posted by stumped

"I'm trying to figure out whether this is a limitation in Calibre or a limitation of the Kobo ereader. If I could put a zip containing images of mixed formats without using Calibre, would it work?"

well lets see- you could test that yourself very easily- or you could hang about hoping someone here will test it for you ... or perform a quick LMGTFY

I tried that first, but simply copying files to my ereader doesn't cause it to register them as available to read. The only way I've been able to side load anything is Calibre. If there's a way to get it to do that, I'd love to know that, too .

[DNSB]

Originally Posted by ceridwen

I tried that first, but simply copying files to my ereader doesn't cause it to register them as available to read. The only way I've been able to side load anything is Calibre. If there's a way to get it to do that, I'd love to know that, too .

As davidfor mentioned, .zip is not a recognized ebook filetype on a Kobo ereader. CBZ and CBR are so you might want to rename your .zip file to .cbz which calibre and Kobo's ereaders are both happy with. OTOH, you may run into issues with the file display order given the filenames used in a cbz/cbr are fairly rigid.

[davidfor]

Originally Posted by ceridwen

I tried that first, but simply copying files to my ereader doesn't cause it to register them as available to read. The only way I've been able to side load anything is Calibre. If there's a way to get it to do that, I'd love to know that, too .

There is absolutely no need to use calibre to sideload books. If you copy files of the right format to the device, they will be imported when the device is disconnected. What you did was to copy a file to the device that it did not consider was a book. And hence it ignored the book.

Calibre is useful when you have lots of books, or books in multiple formats and need to convert them. But, it can all be done without calibre.

[DNSB]

Just for the heck of it, I stuffed some images renamed to 00001.jpg to 00008.jpg into a zip archive, renamed it to tester.cbz. Calibre was happy with it and calibre's viewer displayed it. A default epub conversion split each image into two chunks so not so great and I had to fiddle with the comic parameters on the conversion. Copying the file to my Aura One displayed the images though switching to landscape was definitely an improvement.

I've attached that sample file to this message. Due to .cbz being an unsupported filetype, I added tester.cbz to a .zip file. You'll need to extract the .cbz file from the .zip file to use it.

[ceridwen]

Originally Posted by davidfor

@ceridwen: You need to have the books in calibre in a format the device supports, or in a format that calibre can convert to a supported format. "zip" is not a supported format on most devices and definitely not on a Kobo device.

If you do try to send a zip file to a Kobo device, calibre will try to convert it. That conversion will be to look inside the zip file for a valid book. If there is one, it will be extracted and converted if needed. If there isn't one, you will get an error that calibre cannot convert the zip file. If the zip file is a collection of images, then calibre will not convert it.

But, if the zip file is a collection of images, change the extension to "cbz". Calibre and the Kobo devices will recognise this as a comic book and handle this.

And none of this is "wonky". It is all consistent. Maybe what calibre does when converting a zip file isn't obvious, but it is consistent.

Changing the extensions of mixed-image zip files does work.

It's definitely not obvious to me, because fundamentally these are all the same type of file. Making the behavior depend on the extension doesn't make sense to me, but whatever. It's not something I would do in my day job, but obviously other people have different standards.

[ceridwen]

Originally Posted by DNSB

OTOH, you may run into issues with the file display order given the filenames used in a cbz/cbr are fairly rigid.

I've seen this issue or something similar before. What are the rules?

[davidfor]

Originally Posted by ceridwen

Changing the extensions of mixed-image zip files does work.

It's definitely not obvious to me, because fundamentally these are all the same type of file. Making the behavior depend on the extension doesn't make sense to me, but whatever. It's not something I would do in my day job, but obviously other people have different standards.

There has to be something about the file that tells you or a program what to do with it. The most common method is the file extension. "zip" means one thing. "cbz" means another. And the Kobo software, just like nearly everything else out there, does it this way.

And if changing the extension didn't work, I'd need to see the file to try and see why. I don't remember if I have tested a cbz with different image formats in it.

[davidfor]

Originally Posted by ceridwen

I've seen this issue or something similar before. What are the rules?

Actually, it is simple: Expand the path of the file inside the CBZ, and then alphabetical.

[rashkae]

Originally Posted by ceridwen

Changing the extensions of mixed-image zip files does work.

It's definitely not obvious to me, because fundamentally these are all the same type of file. Making the behavior depend on the extension doesn't make sense to me, but whatever. It's not something I would do in my day job, but obviously other people have different standards.

Fundamentally, yes, they are the same thing at a very low level. Epub, CBZ and Zipped HTML (which is probably what Calbire assumes when you import a .zip file.) are all ... umm, zip files. The extension is how you associate the correct action to take with each one.

[DNSB]

Originally Posted by ceridwen

Changing the extensions of mixed-image zip files does work.

It's definitely not obvious to me, because fundamentally these are all the same type of file. Making the behavior depend on the extension doesn't make sense to me, but whatever. It's not something I would do in my day job, but obviously other people have different standards.

I'm not sure how to take your comment. Are you trying to say that a .epub file is the same as a .cbz file? Or are you saying that your ereader should be able to open any file(s) stored in a zip container and display the contents regardless of how the contents were created?

In the real world, the extension is used to indicate how to handle the file. While theoretically a better solution, analyzing the file contents to determine how to handle the file is a quick trip to code bloat and programmers beating their heads against the wall. So how would you recommend to tell an epub from a cbz? They both use a zip format container. There are quite a few other filetypes that use a zip container as can be told from the the sheer number of times the first two bytes of a file are PK or 0x50, 0x4B.

So in your day job, if you happen to rename a Microsoft Word document with a .txt extension, this would not cause any problems? Or take a compressed log file saved as a .tgz file and rename it as .doc? Try to smuggle a MP4 video onto the network by renaming it as .asc?

[ceridwen]

Originally Posted by davidfor

Actually, it is simple: Expand the path of the file inside the CBZ, and then alphabetical.

At one point, I tried creating CBZs with chapters in subdirectories like 01/, 02/, ... and I could swear that it seemed like the subdirectory name didn't matter, only the name of the file itself. I switched to sticking an absolute number on the front of the file name (like 0001, 0002, ...) and that forces the correct (lexicographic) order.

[ceridwen]

Originally Posted by DNSB

I'm not sure how to take your comment. Are you trying to say that a .epub file is the same as a .cbz file? Or are you saying that your ereader should be able to open any file(s) stored in a zip container and display the contents regardless of how the contents were created?

In the real world, the extension is used to indicate how to handle the file. While theoretically a better solution, analyzing the file contents to determine how to handle the file is a quick trip to code bloat and programmers beating their heads against the wall. So how would you recommend to tell an epub from a cbz? They both use a zip format container. There are quite a few other filetypes that use a zip container as can be told from the the sheer number of times the first two bytes of a file are PK or 0x50, 0x4B.

So in your day job, if you happen to rename a Microsoft Word document with a .txt extension, this would not cause any problems? Or take a compressed log file saved as a .tgz file and rename it as .doc? Try to smuggle a MP4 video onto the network by renaming it as .asc?

Mostly I use Linux's file or bindings. This is more or less an open-source compilation of magic bytes and and heuristics for determining file types, and it can tell the difference between an EPUB and a non-EPUB zip. I'd have to dig through the code to find out what heuristics it uses for that, but for only two cases I would suspect there couldn't be that much code.