internal latency

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

internal latency

barry threw
A question I haven't thought about in relation to Csound, perhaps
related to using Csound for production purposes.

If I were to do the following:

<CsoundSynthesizer>
<CsInstruments>

instr 1

        asig oscili 10000, 440, 1
        asig2 butlp asig, 1000
        asig = asig + asig2

        out asig

endin
</CsInstruments>
<CsScore>
f1 0 512 10 1

i1 0 8888
</CsScore>
</CsoundSynthesizer>

What is the latency incurred by the filter?  I assume some multiple of
kr.  Does Csound correct this?  Is it different in realtime and file render?

This isn't a very good example, but I am thinking of a situation with a
lot of parts that are supposed to be time synced, which are then put
through different processing, and I want to compensate for "plugin"
(opcode) delay.  Is there are simple way to gain this information?

If I'm going through all of the "trouble" of using Csound in terms of
control, I would like to be able to compensate for delay.

Thanks everyone,

Barry


bthrew.vcf (227 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: internal latency

Istvan Varga
barry threw wrote:

> What is the latency incurred by the filter?  I assume some multiple of
> kr.  Does Csound correct this?  Is it different in realtime and file
> render?

There is no latency, other than the normal delay of the filter response
due to phase shift which depends only on the cutoff frequency, and not
on kr or whether real time or file rendering is being done.
There is ksmps latency only if you implement feedback in the orchestra,
that is why in such cases ksmps=1 is often used.
--
Send bugs reports to this list.
To unsubscribe, send email to [hidden email]
Reply | Threaded
Open this post in threaded view
|

Re: internal latency

Iain Duncan


Istvan Varga wrote:

> barry threw wrote:
>
>> What is the latency incurred by the filter?  I assume some multiple of
>> kr.  Does Csound correct this?  Is it different in realtime and file
>> render?
>
>
> There is no latency, other than the normal delay of the filter response
> due to phase shift which depends only on the cutoff frequency, and not
> on kr or whether real time or file rendering is being done.
> There is ksmps latency only if you implement feedback in the orchestra,
> that is why in such cases ksmps=1 is often used.

So, if we wrap the filter area in a UDO we can now safely use ksmp=1 for
feedback without bogging down the entire orchestra?

Man, csound has come a LONG way in the last two years. Mad props again. ;)

Iain
--
Send bugs reports to this list.
To unsubscribe, send email to [hidden email]