[Sony PRS-T1] seeking advice on library management and collections

[Ozric]

Hi,

I use Calibre to manage my library and load books onto a Sony PRS-T1. I've decided that I end up with too many Collections when I let Calibre create them based on series, so I'm looking into alternatives.

For the most part, my tags represent fairly broad genres like fantasy, science fiction, horror, etc. Some tags actually represent series that either don't follow a strict numerical order or are written by different authors, and I might be interested in separating them out into a custom tag if it would help me get what I'm hoping for.

Is there some way I can manage my library so that when I browse Collections on my T1, I can select, for example, Fantasy and when I sort by Author, books that happen to belong to a particular series will be sorted by series index? I've tried a couple things I found by searching the forum, but haven't had any luck.

And if anyone has a system they use for organizing libraries containing a large number of series, some of which don't actually have a series index for whatever reason, I'm interested in hearing suggestions--particularly if it's done with a manageable number of collections.

[Pinecone]

What I have done is make a custom column that is series like.

Then number the books in that "series" the way you want them to show up.

To do this in bulk, order the books by Author and Series. Then Bulk Edit Metadata to add them to the "series' Fantasy with auto numbering on.

[Ozric]

So it sounds like you create something like a meta-series, and stick multiple series in it, order it how you want, then auto-number them. That's pretty interesting. I might give it a shot.

[Starson17]

Originally Posted by Ozric

libraries containing a large number of series, some of which don't actually have a series index for whatever reason

It's not possible to have an empty series_index. It's filled with 1.0 even if there's no series (although it will be grayed out until a series name is assigned).

[jackie_w]

Originally Posted by Ozric

Is there some way I can manage my library so that when I browse Collections on my T1, I can select, for example, Fantasy and when I sort by Author, books that happen to belong to a particular series will be sorted by series index?

If you would like to add the series info to the Author and/or Title info shown in the booklists you should experiment with Preferences - Metadata plugboard

There is an example of a simple one in this post. That screencap is from my PRS650 but it looks very similar on the PRST1.

If you want to try it out the Metadata plugboard settings are:

Format: device_db
Device: PRST1
Source Template: {author_sort}{series:| - | }{series_index:0>2s}
Destination Field: authors

If you read up on the Calibre template language you can achieve all sorts of exotic name customisations.

[Ozric]

Coincidentally, I had already dug up another post somewhere that demonstrates that particular plugboard and I'm currently using it. It's pretty great

I'm currently trying out a new organizational scheme that involves the plugboard linked above as well as a custom "Genre" field, which is the only thing I create Collections from on my device. I use the "Tags" column only to group together related books that aren't part of a numbered series, like "Star Wars," "Warhammer," "Dragonlance," etc. When I send books to the device, I have Calibre put the series, series index, and tag at the beginning of the file name, so if I sort a Collection by file name, books grouped by series and/or tag end up clustered with each other and easily located. I also use genre names like "Science Fiction.Space Opera" to break up what used to be large genre collections.

So far it seems to work okay for me, although I haven't had a chance to fully test everything out yet. But so far it's better than what I had before, which was causing me to exceed the maximum number of Collections allowed on the T1.

[Ozric]

Originally Posted by jackie_w

If you would like to add the series info to the Author and/or Title info shown in the booklists you should experiment with Preferences - Metadata plugboard

There is an example of a simple one in this post. That screencap is from my PRS650 but it looks very similar on the PRST1.

If you want to try it out the Metadata plugboard settings are:

Format: device_db
Device: PRST1
Source Template: {author_sort}{series:| - | }{series_index:0>2s}
Destination Field: authors

If you read up on the Calibre template language you can achieve all sorts of exotic name customisations.

Hmmm...I could've sworn that I had this working, but I've just discovered that either I've somehow broken it, or I never really had it working in the first place and I was fooled by the fact that I include series, series_indices, and tags as part of the file name. When I sort by author or title, I just see the author's name on the second line, instead of the additional data I'm trying to plug in.

My plugboard looks like this:

Format: device_db
Device: PRST1
Source Template: {authors}{series:| - | }{series_index:0>2s}{tags:| - | }
Destination Field: authors
Source Template: {series}{series_index:0>2s|| - }{title}
Destination Field: title_sort

When sorting a Collection by author, I was hoping for something that looked like this:

Awesome Book
Awesome Author - Awesome Series 01 - Tag (if any)

But I just get:

Awesome Book
Awesome Author

Did I go wrong somewhere?

[jackie_w]

Originally Posted by Ozric

Did I go wrong somewhere?

Not necessarily. The PRST1 does seem to have a problem with sometimes deciding to re-extract all its metadata from the epubs and the plugboard data gets overwritten. Kolenka (the creater of calibre's PRST1 driver) thought it was something to do with timezones, I think. It's a pity as the plugboards are rock-solid on my PRS650 and PRS505.

You can try 2 ways to get them back:

Plan A - Reconnect the T1 to calibre, let it run its routines, then disconnect in the normal way.

If that doesn't work...

Plan B - Do a full T1 device shutdown and restart. Followed by Plan A

Having a good suitable filenaming convention is a good fallback option for when the above happens and it's not convenient to connect to the PC. Choosing the 'Sort by filename' option always works.

Of course, if you change the metadata in calibre the plugboard option automatically updates the Sony next time you connect. The filename won't change unless you resend the epub. Swings and roundabouts.

[chaley]

Originally Posted by Ozric

Did I go wrong somewhere?

(edit again -- my second answer was also totally wrong.)

I just tried it with my PRS300. The plugboard
device_db:PRS505 = ([foobar {authors}] -> authors)
puts the word foobar in front of every author, and that is what I see on the device.

Be sure that metadata management is set to automatic. After sending the books, disconnect the device then reconnect it so that metadata is sent again. Then check. If it is still incorrect, then there is something odd about the T1 that I can't debug, as I don't have one.

[jackie_w]

@chaley FYI,

There is definitely something different about how well (persistently) the plugboard feature works on the PRST1 and previous Sony models. It works perfectly on my PRS505 and PRS650 - rock solid. It also appears to work well on the PRST1 ...for a time... until the point (maybe days later) at which sometimes when you connect to/disconnect from calibre, the plugboard info is correct... then it slowly blinks out and is replaced with the default simple Author (FN LN) and Title - probably as the T1 re-extracts everything from each epub. I assume the problem is at the T1 end not calibre.

Since Kolenka's most recent upgrade I have been able to correct the situation with another connect/disconnect (whether that's coincidental I don't know). Prior to that I always needed to do a full shutdown followed by connect/disconnect.

[chaley]

Originally Posted by jackie_w

the plugboard info is correct... then it slowly blinks out and is replaced with the default simple Author (FN LN) and Title - probably as the T1 re-extracts everything from each epub. I assume the problem is at the T1 end not calibre.

Sounds like the T1 is deciding to rebuild its database for some reason. What joy.

What happens if you add a EPUB/T1 plugboard that alters the metadata in the epub itself? This should ensure that if the T1 decides to re-read the metadata from the book, you get the same data that device_db tries to send.

Of course, you would need to resend the books once that plugboard is added. Auto metadata management does not affect what is in the book files themselves.

[jackie_w]

Originally Posted by chaley

What happens if you add a EPUB/T1 plugboard that alters the metadata in the epub itself? This should ensure that if the T1 decides to re-read the metadata from the book, you get the same data that device_db tries to send.

Of course, you would need to resend the books once that plugboard is added. Auto metadata management does not affect what is in the book files themselves.

2 questions, please...

Do you mean add rather than replace?

If you want to send all 3 T1 formats (epub, pdf, txt) does it need to be 'any format'/PRST1 rather than EPUB/PRST1

[chaley]

Originally Posted by jackie_w

Do you mean add rather than replace?

I did mean "add". More later as to why.

If you want to send all 3 T1 formats (epub, pdf, txt) does it need to be 'any format'/PRST1 rather than EPUB/PRST1

It can be 'any format', but I am not convinced that it needs to be. TXT does not have embedded metadata, so the plugboard would do nothing. PDF has limited metadata support, so that might or might not work.

From your description, it seems that automatic management is in fact working. The problem seems to be that at some point the T1 decides to rebuild the database. I assume that if you wait for this to finish then plug in the T1, calibre will refresh the database, restoring it to the desired values.

However, the data will be incorrect for the period between the T1's rebuild and the connection. This problem can be minimized if the data retrieved by the T1 from the ebook is the same data that you want to appear in the device DB (modulo changes since you sent the book). If you have plugboards that both update the device DB on connect (auto management) and the book (on send), then this goal should be realized. Automatic management will still work at least as well as it does today, and the T1's DB rebuilds won't have such a dramatic effect.

This is all supposition. Not knowing how the T1 works or why the T1 is deciding to rebuild, I can't really know what it will do or where it gets its information.

[jackie_w]

Thank you. I have added an EPUB/PRST1 we'll see how it goes. Everything looks OK now but I won't know the results until the next 'mysterious books.db rebuild event'. Will report back when it happens.

Incidentally, with TXT files the plugboard works great and seems to persist even after an unwelcome rebuild event - presumably because there is no other place to try to re-extract from. Unfortunately PDFs seem to be afflicted the same way as EPUBs.

[Ozric]

Yeah, after I reported my confusion last night, I ended up reconverting all my EPUBs with using a different output profile because I discovered it gave me more satisfactory results with certain books. Then I replaced all the books that were on my device with the new ones, and without touching any of my plugboard stuff at all, the device is now properly displaying the extended data.

So my guess is that there is something glitchy with the T1 that can caused the extended data to break on my device. I do have Calibre set to automatically update the metadata on my device, by the way.