Re: RECOMMENDATIONS for EFFECTS PROCESSING on upcoming tour

[andy butler <akbutler@tiscali.co.uk>]

Per Boysen wrote:
> On Wed, Feb 25, 2009 at 10:20 AM, andy butler <akbutler@tiscali.co.uk> 
>wrote:
>> Latency compensation isn't a cure-all.
>>
>>
>> Say you're re-triggering a loop with a note right at the beginning of 
>it,
>> ...no way to compensate.
> 
> So, your are talking about MIDI latency here? That would make sense.
> I've never had a problem with that though.

here's the sequence of events

hit midi controller
midi latency
latency for pc to understand the midi
Software starts playback
Audio interface output latency.
Sound comes out of speaker.


I'm kind of guessing you miss the point about latency comp only
being a partial solution.

If you have latency compensation, it only changes the playback
position from the audio buffer.
Response time isn't improved at all.

So, if you're using latency comp, and you re-trigger a loop then
playback won't start any earlier. All that happens is that playback
starts at the "non-compensated" time, but it starts from a point that
keeps the audio in sync....so you lose the first note to some extent.



> 
>> What happens is that playback can't start until after the latency 
>period,
> 
> Are talking about a third sort of latency, the time it takes for the
> looper to implement the command?
> 
> The worst latency in this example must be the DA conversion (making
> the sound analog so it can be sent to loudspeakers). But I'm sure Rick
> "plays" that latency. Every instrument answers back differently. A
> temple block has many times less "latency" than a big taiko drum.
> 
> Per

yep, and as brain latency is in 100s of mS it's possible to compensate.
Not only that, but foot latency is longer than hand latency :-)


Best scenario for live pc looping is to have an instrument that is
acoustically silent, so that you monitor the sound with the full
audio interface latency. Then don't use any latency compensation,
because all latencies are equal.

As Rick's using acoustic instruments, he's  going to have
deal with varying latencies. Well, unless he remembers his
in ear monitors ;-)
Compensating (with the brain) for one fixed latency on everything is 
automatic,
as long as you can focus on the sound and not the mechanics of producing 
it.
If latency is different for live playing and loop triggering then it
gets much more difficult. 
However, with the lowest possible latency, I reckon it's liveable.



I really value all the work Jeff put into latency compensation,
and in some circumstances it really works.
However, the idea that latency is done away with by that method,
however popular, is just totally misleading (as Jeff would tell you,
he's the absolute expert).
Essentially, the response time is still the same.

andy butler