[LAD] Tests directly routing pc's midi-in to midi-out

Ralf Mardorf ralf.mardorf at alice-dsl.net
Thu Jul 15 08:40:53 UTC 2010


On Thu, 2010-07-15 at 08:26 +0200, Clemens Ladisch wrote:
> Ralf Mardorf wrote:
> > On Wed, 2010-07-14 at 19:56 +0200, Arnout Engelen wrote:
> > > On Wed, Jul 14, 2010 at 03:23:03PM +0200, Ralf Mardorf wrote:
> > > > Yamaha DX7 --> Alesis D4 results in a 100% musical groove.
> > > > Yamaha DX7 --> PC --> Alesis D4 results in extreme latency
> > > 
> > > So here you're directly routing the MIDI IN to the MIDI OUT, and experiencing
> > > latency. Are you using JACK here, or directly ALSA? In other words, are you 
> > > connecting 'in' to 'out' in the qjackctl 'MIDI' tab or in the 'ALSA' tab?
> > 
> > I'm connecting MIDI in the Qtractor (quasi QjackCtl) ALSA MIDI tab.
> 
> Please make a test without any program using JACK, just connect the
> DX7 port to the D4 port with aconnect(gui), and try that.
> 
> 
> Regards,
> Clemens

I'll test this :).

On Thu, 2010-07-15 at 09:57 +0200, fons at kokkinizita.net wrote: 
> I'm sure they would be sensitive to bad timing.  But that's not
> the question. Would they be able to identify the recordings listed
> above ? Until you try it you won't know, and your claim that 2 ms
> of jitter 'destroys the groove' is pure conjecture.

Who knows? Perhaps 2ms shown by Audacity are 20ms, because Audacity has
a bug? Perhaps 2ms shown by Audacity are 2ms, but when playing JACK, the
sound card or what ever will add additional jitter?

I can't test the list, because on my machine there is something audible.

On Thu, 2010-07-15 at 09:56 +0200, Arnold Krille wrote: 
> On Thursday 15 July 2010 01:14:45 Ralf Mardorf wrote:
> > On Thu, 2010-07-15 at 00:46 +0200, fons at kokkinizita.net wrote:
> > > Apart from that, it remains to be seen if *real* timing errors of
> > > +/- 2 ms do 'destroy the groove'. To test this, make the same
> > > recording
> > > 
> > > - without jitter,
> > > - with 1 ms jitter,
> > > - with 2 ms jitter,
> > > - with 3 ms jitter.
> > > 
> > > and check if listeners are able to identify which is which,
> > > or at least to put them into order.
> > I know very gifted musicians who do like me and they always 'preach'
> > that I should stop using modern computers and I don't know much averaged
> > people. So the listeners in my flat for sure would be able to hear even
> > failure that I'm unable to hear.
> 
> You really should do that test first before speculating about the outcome and 
> your audience.
> 
> You would expect Audiophiles to spot the "super sounding" denon cables by 
> listening, right? Yet a blind test showed the opposite. The test was to 
> identify which audio take was played with denon-cables, el-cheapo cables from 
> walmart and a bended cloth-hanger. If they where as good as they claimed, the 
> denon-cable should get hits with probability significantly better then 1/3, 
> otherwise its just luck.
> Guess what the outcome was: There was a significant hit: But they spotted the 
> cloth-hanger as the denon-cable. Thats what real experts do...
> 
> Do the listening test with as many people as possible and then show the 
> results. And only afterwards start the speculations what the reason and the 
> effects might be. (Thats called science btw.)
> 
> Have fun,
> 
> Arnold

Perhaps it's not that 2ms, but here are audible issues. As I mentioned before. Audacity shows 2ms, but JACK, the driver the hardware might add jitter.

FWIW blind tests aren't scientific,just double-blind studies are meaningful. And if you wish to test cables you need to test the quality after one year, after two years etc.. Anyway, a bad cable might cause a bad sound quality, but not bad timing.
Timing is the meat and potatoes to music.

Regarding to my Linux computer such studies aren't needed. A bad timing is a bad timing.
At least for the USB MIDI that I'm not using anymore, I made tests with a Windows install (I don't have this install on my machine anymore, so I can't test the PCI card with Windows).
The USB MIDI was much better on Windows, even better than the PCI cards at the moment are on Linux. So I guess, yes I don't know, that the hardware is ok.

- Ralf




More information about the Linux-audio-dev mailing list