[Ardour-Users] Ardour's DSP load rising and falling whilst idle [WAS: Denormals]

[Q]

On 01/01/12 16:41, Robin Gareus wrote:
> [Wow, quite an impressive accurate bug summary. -- disclaimer: I'm still
> still hung-over :)]

Thanks. I thought it best to summarise things as there's been a bit of 
back-and-forth and sometimes details can get overlooked. Plus, the more 
info there is, the easier (and quicker) it might be for others to see 
what the cause might be. I wasn't hungover (went to bed early and read, 
what a rock n roll life I lead!) so I managed to be vaguely coherent in 
writing this morning :-)

> How did you 'de-activate' the plugins?
>
> Did you remove them or only bypass them?

Niether bypassed nor removed: secondary click, then selecting 
Deactivate/Deactivate all. I didn't fancy removing things and undoing 
hours of mixing.

> I'm not certain, but..
> Isn't deactivating a plugin/route in ardour2 equivalent to bypassing it?
> Isn't the plugin's processing function executed anyway, also for
> bypassed plugins?
>
> A quick test here (A3) seems to indicate otherwise, but there's quite a
> few options hidden somewhere and I recall from some previous email-list
> communication that plugins are can be run continuously.

I've no idea, but my guess -- going by the different effects on DSP load 
-- would be that deactivating isn't equivalent to bypassing.

With plugins running continuously, are you thinking of the option to 
"Stop plugins with transport"? Carl Hetherington said he believed that 
was broken in Ardour 2 and doesn't do anything, which fits with the high 
load I see when Ardour is idling.

> All in all, it smells like a Denormal issue in some plugin(s).
>
> Denormals usually show up after some period of silence:
> Once there was some input,.. now the input is silent and the
> [plugin] algorithm does lots of near-zero calculations. ->  Denormals.
>
> Depending on the plugin, it can take quite some time until they show up
> (esp. on delay-lines it may be 90+ seconds). Also depending on the
> plugin (internal feedback) they may or may not settle down after a while.

Interesting, thanks for that. I hadn't realised it could take so long 
for denormals to build up, although thinking about it now it does make 
sense.

> There's a small app called 'bitscope' or 'bitmeter' that can display
> denormals on JACK ports: http://devel.tlrmx.org/audio/bitmeter/
>
> But I'm not sure if it is helpful to debug your problem because, IFF the
> FPU supports it, Ardour will flush the denormals to zero before pushing
> them on the jack-port.

That's the thing, I have no processor handling of denormals in Ardour 2 
as they are all greyed out, only DC bias. Yet A3 indicates that the 
processor will handle denormals (and overall the load was lower, 
although the problem still seemed to be there).

I'm still intrigued about this.

>
> Could you post the ardour-session (minus the audio-files) somewhere?
> Since you say the problem persists even when Ardour is idling this
> should be reproducible on other machines.
>

30 MB archive, session folder with all files except audio ones:

www.quirq.ukfsn.org/ArdourSession.tar.bz2

> Digging in the dark:
>
> http://subversion.ardour.org/svn/ardour2/branches/2.0-ongoing/libs/ardour/route.cc
>
> adds 1.0e-27f; to each sample. Could that be too small a bias?
>
> http://lists.linuxaudio.org/pipermail/linux-audio-dev/2009-August/024822.html
> :
> "I [Fons] use something like 1e-20f, which is 400dB below normal levels"
>
> -=-

Interesting, but not something I could ever shed any light on :-)

> Are you chaining Plugins?
>
> AFAICT the DC bias is only added before the first plugin. A 2nd plugin
> in the chain may actually be prone to Denormals if the first plugin
> removes the DC offset.
>

Again, very interesting. Yes, I do have many tracks where there is a 
chain of plugins.

> Are you using phase-inverts in your ardour session? Ardour3's route.cc
> subtracts the bias if the phase of the route/channel is inverted, A2
> does not. It may be that the bias is effectively canceled out in the
> plugin..
>

No, I've not inverted the phase on any tracks or busses. In some cases I 
have multiple things running into one buss, but where there are two 
tracks of the same instrument, those tracks have been aligned by hand 
(i.e. I haven't got anything 180 degrees out and inverted to prevent 
them from cancelling).

>
> I hope I'm not just adding to the confusion..
> robin
>

I think it has helped clarify things a little towards it being a 
denormal issue. And even if it hasn't, I've still learnt a thing or two, 
so thanks!

Q