Irregular/very slow page turns (kepub)

[Mrs_Often]

What happened to the page turn speed of kepubs? (Solved)

I haven't really been paying much attention to new firmwares in the past few years, I've been very happy with 4.9.11311. Likewise I haven't been paying much attention to this forum either.* Until recently: I gave 4.24.15676 a try because it has some features that I was interested in, like the header with chapter progress and the progress bar at the bottom, among others.

The first book I read with firmware 4.24.15676 drove me crazy. It was a continuous battle between self control and throw-the-device-against-the-wall frustration every time I had to 'turn' the page. Each of the 1432 page turns (or however many there were) was a gamble: some where suddenly as fast as I hoped, so i did have to read to the end of the last line before turning a page, but most were so slow that I had to stop my eye movement and wait a few beats before I could continue reading. Extremely distracting from the book content. And very irritating. I thought it was due to the book, because it had ridiculously long chapters. I tried cutting the chapters into smaller files, but there was no difference. The second book I read (and am still reading) turned out to be just as annoying. So then I thought it might be due to one or several of the patches I had enabled. So I disabled all patches. Still horrible irregular and slow page turn speeds. So now I think it's the firmware.

I just downgraded (downdated?) to 4.9.11311 (and enabled the patches I was using there)... dear me the screen flashes to the next page so fast, I have to do a double-take at the first line to be sure it really is the next page! I'm loving it, and will be used the continuous reading flow again very soon.

What's going on? Why did page turns get so much worse in newer firmware? I understand there's probably more to be calculated if you're showing both book progress and chapter progress and a progress bar on screen all the time, but surely there's enough time to do calculations while the current page is being read, to give 'instant' page turn when requested by the reader...? Is there any hope of the page turns getting back to normal in newer firmware (soon)?

If the page turns were only just slow it would at least be manageable by training my page turning fingers to request a page turn early on, to time it perfectly with when I want to continue reading on the next page. But it's so irregular that perfect timing ins't possible.



*I did do a quick search for complaints about slow page turns before I posted this message, but could only find a few questions asking whether page turn speed had improved yet. I couldn't find the start of the issue or whether it was referring to epub or kepub.

[jackie_w]

Obviously I can only relate my own experiences ... I haven't noticed any problems at all with kepub page turning.

ETA: With the current firmware nearly all of my reading has been done on a Forma.

[NiLuJe]

I don't recall the last time I actually timed this, but I didn't nothing anything out of the ordinary when I did, and it was probably >= 4.2x (as in, epub was slow, but getting better, kepub was decent and also possibly getting better).

Granted, that was on a tame novel with sane chapters and sane-ish code, not a gigantic omnibus .

I do tend to notice and keep track of these things, to make sure KOReader still beats everyone, Kindle included ;p.

I'll run a test the next time I can and get back to you.

On the off-chance it's really book-specific, feel free to PM me a scrambled copy of something where you notice something weird .

[NiLuJe]

One easy thing you can try is to check if you can replicate the issue when the device is plugged in (as it *might* help the CPU governor let the CPU clock up more consistently).

[Mrs_Often]

Originally Posted by jackie_w

Obviously I can only relate my own experiences ... I haven't noticed any problems at all with kepub page turning.

ETA: With the current firmware nearly all of my reading has been done on a Forma.

What!!! How is that possible I'm so jealous
Mine is the Clara HD...

Hmmm... do you read really really really fast? It seems that if I turn several pages consecutively just for the page turning of it without waiting to read a page, there is no lag for the next screen to be rendered... maybe I read so slowly that the Clara gets bored and starts doing other things in the background... My WiFi is on... is yours? That's something I didn't check. I will next time I have some time to go back and forth between firmwares again.

Originally Posted by NiLuJe

I don't recall the last time I actually timed this, but I didn't nothing anything out of the ordinary when I did, and it was probably >= 4.2x (as in, epub was slow, but getting better, kepub was decent and also possibly getting better).

Huh...

Granted, that was on a tame novel with sane chapters and sane-ish code, not a gigantic omnibus .

Haha yes well... the book wasn't actually a gigantic omnibus as far as I know*, it's just the number of page turns that was high... (How the total number of pages in a book is displayed is something that actually annoys me in the 4.24 firmware: the number of ereader screen size 'pages' instead of approximated paper book pages. Now I have no idea how to compare what I'm reading to the length of the actual paper book )

*Just looked the book up on Goodreads.com... apparently the paper book has 536 pages. So not a short novella...

I do tend to notice and keep track of these things, to make sure KOReader still beats everyone, Kindle included ;p.

I'll run a test the next time I can and get back to you.

Thank you!!

On the off-chance it's really book-specific, feel free to PM me a scrambled copy of something where you notice something weird .

Oh that's very kind, I will do**! Would be strange if it's book specific, that would mean that it's something that firmware 4.24 can't handle, but 4.9 could...

**EDIT: Done!

[Mrs_Often]

Originally Posted by NiLuJe

One easy thing you can try is to check if you can replicate the issue when the device is plugged in (as it *might* help the CPU governor let the CPU clock up more consistently).

Oh really! I wonder how that works... I will try it.

[NiLuJe]

Originally Posted by Mrs_Often

It seems that if I turn several pages consecutively just for the page turning of it without waiting to read a page, there is no lag for the next screen to be rendered...

That would jive with my "the cpu governor is being stupid" theory ^^.

[JSWolf]

I suggest you try ePub and see how that goes. But you would have to reinstall the current firmware.

[koboy]

trying the same kepub book (regardless of which i try) on my Kobo Mini and one of my original Kobo Aura's 6" both with firmware 3.16.0 vs another of my original Kobo Aura's 6" with firmware 4.24.15676 the page turns on the latter firmware are as the OP describes=slower.

having (only) used firmware 3.16.0 since it was released up until very recently i noticed the slightly slower page turns straight away on firmware 4.24.15676.

can't comment on any of the other 4 firmware's i didn't use them long enough to find out.

best wishes koboy

[NiLuJe]

I mean, you can certainly try a straight ePub, but, spoiler alert: it's going to be slower. Much, much slower .

[NiLuJe]

Okay, finally had time to ran some minimal tests with the (scrambled, Calibre'd) books @Mrs_Often sent me.

As a reminder, I'm on a Forma, on FW 4.24. It was unplugged, with Wi-Fi enabled and an SSH shell session open, and at the battery sweet spot of ~70%.

Given that the sources were scrambled (straight) ePubs, I tested both those and a kepubified copy.

The "good" news is that they behave roughly like any other book I'm used to.

In terms of purely computational power required by (forward) page-turn on *pure text* pages (via the HW buttons) [i.e., that's plain old (real) CPU time, for you *nix nerds out there]:

* KOReader (CRe, on the ePub, not that it matters much): ~60ms
* ACCESS (KePub): ~200-250ms
* RMSDK (ePub): ~450-600ms (a.k.a., Sleepy ;p)

The "faster" the renderer is, the easier it is to catch outliers in actual perceived latency (outside of flashing pages).
There *may* indeed be instances where the CPU is veeeeeery slightly slower to clock up. Or if not that exactly, where there's a more noticeable bottleneck somewhere fairly low-level.

(Again, text only, that means no image, and keeping to the same document shard, because ACCESS is noticeably hitchy when crossing a file border, as is RMSDK, to a lesser extent (mainly because it's so slow in general that you don't really notice the difference ;p)).

(A Kindle with the KF8 renderer used to sit in the sub-100ms range, last I checked, sometimes going neck to neck with KOReader. Which is exceedingly good, as the Kindle *is* a chunked renderer, unlike CRe, which requires a longer (a couple of seconds here) caching step on *first* open to be able to deliver such runtime performance).

Massive grain of sand: Actual latency is *extremely* hard to measure accurately. I don't have the hardware nor skills to make a definitive assertion, so, this is more of a very vague feeling. I tend to notice it more in ACCESS, but I won't lie by saying I've never felt it in KOReader .

I have no actual clue how well the SoC itself is supposed to clock up and down (in terms of latency), and the CPU governor in use (interactive) is notoriously tricky to tune right, and as theses devices run an old-ish kernel, it makes the whole thing even murkier.

I *do* notice it having clocked up (and to the max speed, which is typical of interactive) in all cases, so, that's at least a good thing .

And, yes, if you're sensitive to latency issues, as far as I'm concerned, a 50ms difference is noticeable, and a 100ms one starts to get annoying.
(Fair warning: I've done a fair share of competitive first person shooter gaming (or, hell, playing a healer in a mythic raider group in WoW, which might actually be worse ;p) in my days, so I may be more sensitive than most).

[Nightflyer]

Wow. I just get a book, read a book. No big problems so far.

Just love my Kobo.

[Mrs_Often]

Originally Posted by koboy

having (only) used firmware 3.16.0 since it was released up until very recently i noticed the slightly slower page turns straight away on firmware 4.24.15676.

Hah, so I'm not crazy! Seems so strange to me to have newer firmware be slower Not good.

Originally Posted by Mrs_Often

Oh really! I wonder how that works... I will try it.

Tried it... and indeed there does seem to be less irregularity, and less of the veeeeery long taking page turns.

Originally Posted by NiLuJe

Okay, finally had time to ran some minimal tests with the (scrambled, Calibre'd) books @Mrs_Often sent me.
[...]
* KOReader (CRe, on the ePub, not that it matters much): ~60ms
* ACCESS (KePub): ~200-250ms
* RMSDK (ePub): ~450-600ms (a.k.a., Sleepy ;p)
[...]

So detailed!! Thank you!!

60ms on KOReader?! Maybe I have to start thinking about using it... but I really do like kepubs and the integration with Kobo (affiliates).

Is 'ACCESS' the thing/software/whatever/I'm-stupid that the Kobo's firmware uses to read kepubs? 200-250ms sounds long. Do you by chance still have details on a test you did from 4.9.11311 or similar firmware version? Just to compare.


I've attached a recording I did in unpatched 4.24.15676. I opened a kepub and started turning pages every 30 seconds. This 5 second clip is from one of the worst (slowest) page turns.

Does this mean that there's nothing I can do about it and I'm either stuck with frustrated reading if I want to use 4.24.15676, passively waiting for improvement from Kobo; or stuck with an earlier firmware versions/KOReader?

[Mrs_Often]

Originally Posted by Nightflyer

Wow. I just get a book, read a book. No big problems so far.

Just love my Kobo.

I'm glad!! Lucky you.

[NiLuJe]

Originally Posted by Mrs_Often

Is 'ACCESS' the thing/software/whatever/I'm-stupid that the Kobo's firmware uses to read kepubs?

Yep, AFAIK, it's what the KePub renderer is based on.

Originally Posted by Mrs_Often

Do you by chance still have details on a test you did from 4.9.11311 or similar firmware version? Just to compare.

Not really, unfortunately, and I'm unwilling to downgrade my main device .
I'd, ideally, need two identical devices in front of me for a real test, anyway.

FWIW, I could have sworn I saw numbers around ~150-200ms at one point in time (as in, earlier FW 4.x FWs), though, but, take that with a very large grain of salt, because I stupidly didn't really jot down the results of these kinds of spur of the moment tests ^^.

Originally Posted by Mrs_Often

I've attached a recording I did in unpatched 4.24.15676. I opened a kepub and started turning pages every 30 seconds. This 5 second clip is from one of the worst (slowest) page turns.

Unfortunately, that doesn't necessarily paint a clear picture of what's happening: capturing an eInk screen on video with any semblance of accuracy is *extremely* tricky .

But, yeah, that one doesn't look good ^^. It roughly feels like crossing a file boundary would, except it's not :s.

Originally Posted by Mrs_Often

Does this mean that there's nothing I can do about it and I'm either stuck with frustrated reading if I want to use 4.24.15676, passively waiting for improvement from Kobo; or stuck with an earlier firmware versions/KOReader?

Yup.