playback hangs when EQ enabled - iPod mini

[erem]

Hi,

I'm new to the Rockbox software so I hope this issue is commonly reconized and can be solved. So help pls.
I have iPod Mini 2gen 4GB for 4 months now and I was very excited when I found Rockbox on internet - a software which addresses two vital options iPod didn't have - equalizer and crossfeed.

I have installed Rockbox succesfully and enjoyed it very much until I started to tweek with equalizer.

Issue description:
1. If eualizer is enabled and when WPS is viewed Rockboxed iPod plays for 5 to 30 seconds and then it stops for less than second to activate disk (virtual diode), and then palying resumes for next interval (the same length), and so it goes - playback-gap-playback-gap and so on.
2. If I leave WPS and enter any menu - the effect of interruptible play dissapear, when I'm back on WPS - it's back again.
3. If I disable equalizer - everything works fine.

I tried to tweek with some disk buffer sensitive settings (Disk Cache, Anti-Shake Buffer, Directory Buffers) but with no success but some minor changes in the length of the interval betweed succeeding playback gaps.

I should add thad iPod is new and legal, no tweeks applied, device not serviced yet. Rockbox installation done w/o problems.

Is this issue known? Can someone help me with it?

Regards,
Michal Rogowski

[bluebrother]

This is a performance issue and covered in the FAQ. Turn off the EQ and use a theme without peak meters.

[erem]

OK, do you say that it's no use in enabling EQ?   Oooops, I installed Rockbox just for that purpse.

Is there any way to fix this problem? Is this possible at all with iPod mini? Is this because the processor power's not sufficient?
I noticed that with EQ enabled but all gains zeroed it works fine.


BTW no mention about this issue  in faq or I'm blind or dull (or both  )

Regards, Michal

[bascule]

From the iPod FAQ

Playback is glitchy / skips!
Again, it's in early stages. The simplest solution is to choose a WPS without peakmeters, for example iCatcher. Higher bitrate MP3s especially have these problems, as well as higher bitrate AAC files. This is beacause there is still quite a bit of optimization to be done on the iPods. During decoding some files the one core used is at nearly full capacity, which is why you may notice slowness in the menus or adjusting volume. If you have replaygain or equalizer settings enabled, turing them off may improve your playback quality. As well, using a simpler while playing screen, preferably one without images, may also help considerably

[erem]

bascule, thanks. It seems that I'm dull,  anyway  

It doesn't answer my questions, however. In the meantime I tweeked a little, wrote my own simple WPS w/o peakhold meters, turned off anything could load processor, but enabled EQ. At first glance I thought that it's OK, but some more changes in EQ settings (Q and f) led me back to an old issue. Now gaps occour seldom, but still they are. I wonder if someone will cope with this issue in the nearest feature...

[dan_a]

There is work going on to make use of the second core on the iPods (and other machines based on the same chip.)
This should mean that there is enough processor power available to do decoding and EQ at the same time without skips.  Unfortunately we can't say when it will work.

[preglow]

It doesn't answer my questions, however. In the meantime I tweeked a little, wrote my own simple WPS w/o peakhold meters, turned off anything could load processor, but enabled EQ. At first glance I thought that it's OK, but some more changes in EQ settings (Q and f) led me back to an old issue. Now gaps occour seldom, but still they are. I wonder if someone will cope with this issue in the nearest feature...

If you absolutely need to use the EQ, I guess it's worthwhile knowing that each band you enable will add to the CPU. Thus it pays off to make do with as few bands as possible and set the ones you don't use to 0.0 dB. Apart from this, Q/freq setting has absolutely no impact on CPU usage except for the moment you adjust the setting itself. As for fixing, this isn't trivial. The EQ is already coded as tightly as I can make it, which means we'll have to optimise codecs, which isn't easy either. Your best hope of this getting fixed soon lies in the effort to use the second CPU core.

[erem]

Thanks for info, it really helped. I have already noticed that fixing Q/f momentarily influences playback but has no further impact. I have also found that .acc files are much more processor consuming that .mp3. Now with all other options but EQ turned off I can play .mp3 files with no gaps (or so I say after few hours of listening). If I change to .aac - same old story again. So I simply don't use .acc. No big deal to my ears, anyway.

[Davide-NYC]

Erem,

If you want to contribute some Codec Performance numbers to the following wiki page it would be of great help. So far I'm the only one that has bothered with this.

http://www.rockbox.org/twiki/bin/view/Main/CodecPerformanceComparison

Eventually we will have a better test involving a decoder plugin which will spit out a detailed logfile, but until then this is the only information we have.

Everytime I see one of these "playback stops/interrupts/ships.hangs" threads I just want to point people to the wiki page. Maybe I should start doing just that.

[erem]

Erem,

If you want to contribute some Codec Performance numbers to the following wiki page it would be of great help. So far I'm the only one that has bothered with this.

http://www.rockbox.org/twiki/bin/view/Main/CodecPerformanceComparison

Eventually we will have a better test involving a decoder plugin which will spit out a detailed logfile, but until then this is the only information we have.

Everytime I see one of these "playback stops/interrupts/ships.hangs" threads I just want to point people to the wiki page. Maybe I should start doing just that.

Interesting test. I would be helpful and hope I'll be one day. But first I have to gather more information to be a partner for you codec gurus