yes, all opcodes work on chunks. Changing this is a major task.
But: running at ksmps=10, for my perception is enough. I must say
I can't distinguish events that are less than 1ms from their correct
start time. Csound seems to run fast enough, too, with that k-rate.

Regards

Victor

----- Original Message ----- 
From: "Jonatan Liljedahl" 
To: "Developer discussions" 
Sent: Monday, October 20, 2008 10:40 PM
Subject: Re: [Cs-dev] Sample accurate k-rate was Re: Questions 
aboutmusmon/insert


I actually thought that events could start within a k-rate cycle,
because I saw an "unsigned long start_kcnt" in the eventnode struct and
assumed this was it's usage. But maybe it's actually start time measured
in k-rate cycles?

 typedef struct eventnode {
    struct eventnode  *nxt;
    unsigned long     start_kcnt;
    EVTBLK            evt;
  } EVTNODE;

jitter-free event scheduling would be very nice. But maybe many opcodes
assume they work on whole k-rate chunks?

Victor Lazzarini wrote:
> But the event start times will always be at k-rate boundaries,
> unless we change the whole system.
>
> At 17:09 20/10/2008, you wrote:
>> I understand that all events are rounded to k-periods (e.g. score and
>> MIDI events), so when you increase the size of kr, you are actually
>> adding jitter to the precision of event timing. It would be nice if
>> events had some sort of time stamp that indicated to which sample
>> within the k period the event belonged to, so that opcodes that want
>> to take advantage of this can do so, while others can simply ignore it
>> and continue with the previous behavior. This would enable sample
>> accurate syncing of events even if the opcode receives k-rate
>> information. The opcode would have to keep track of this internally,
>> naturally.
>> My goal is to have score and MIDI events with no jitter, no matter the
>> size of kr. Maybe there's a better way to do this?
>>
>> Cheers,
>> Andrés
>>
>> On Mon, Oct 20, 2008 at 9:23 AM, Victor Lazzarini
>>  wrote:
>>> Explain a little more , please?
>>>
>>> At 15:09 20/10/2008, you wrote:
>>>> How hard would it be to time tag events within a control period? It
>>>> would be great if opcodes could have sample accurate processing of
>>>> k-rate signals (like MIDI for instance).
>>>>
>>>> Cheers,
>>>> Andrés
>>>>
>>>> On Mon, Oct 20, 2008 at 5:43 AM, jpff  wrote:
>>>>> That is one of my current concerns -- I have not looked at Beatmode.
>>>>> Anyone got a simple example I can try?
>>>>> ==John ffitch
>>>>>
>>>>>
>> -------------------------------------------------------------------------
>>>>> This SF.Net email is sponsored by the Moblin
>>>> Your Move Developer's challenge
>>>>> Build the coolest Linux based applications
>>>> with Moblin SDK & win great prizes
>>>>> Grand prize is a trip for two to an Open
>> Source event anywhere in the world
>>>>> http://moblin-contest.org/redirect.php?banner_id=100&url=/
>>>>> _______________________________________________
>>>>> Csound-devel mailing list
>>>>> Csound-devel@lists.sourceforge.net
>>>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>>>
>>>>
>>>>
>>>> --
>>>>
>>>>
>>>> Andrés
>>>>
>>>> -------------------------------------------------------------------------
>>>> This SF.Net email is sponsored by the Moblin
>> Your Move Developer's challenge
>>>> Build the coolest Linux based applications
>> with Moblin SDK & win great prizes
>>>> Grand prize is a trip for two to an Open Source event anywhere in the 
>>>> world
>>>> http://moblin-contest.org/redirect.php?banner_id=100&url=/
>>>> _______________________________________________
>>>> Csound-devel mailing list
>>>> Csound-devel@lists.sourceforge.net
>>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>> Victor Lazzarini
>>> Music Technology Laboratory
>>> Music Department
>>> National University of Ireland, Maynooth
>>>
>>>
>>> -------------------------------------------------------------------------
>>> This SF.Net email is sponsored by the Moblin
>> Your Move Developer's challenge
>>> Build the coolest Linux based applications
>> with Moblin SDK & win great prizes
>>> Grand prize is a trip for two to an Open Source event anywhere in the 
>>> world
>>> http://moblin-contest.org/redirect.php?banner_id=100&url=/
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>
>>
>>
>> --
>>
>>
>> Andrés
>>
>> -------------------------------------------------------------------------
>> This SF.Net email is sponsored by the Moblin Your Move Developer's 
>> challenge
>> Build the coolest Linux based applications with Moblin SDK & win great 
>> prizes
>> Grand prize is a trip for two to an Open Source event anywhere in the 
>> world
>> http://moblin-contest.org/redirect.php?banner_id=100&url=/
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
> Victor Lazzarini
> Music Technology Laboratory
> Music Department
> National University of Ireland, Maynooth
>
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's 
> challenge
> Build the coolest Linux based applications with Moblin SDK & win great 
> prizes
> Grand prize is a trip for two to an Open Source event anywhere in the 
> world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/csound-devel


-- 
/Jonatan         [ http://kymatica.com ]

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great 
prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/csound-devel 



-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net

Oeyvind Brandtsegg wrote:
> I do agree that for most normal uses, 1 ms resolution would be enough.
> Still, I think it would be valid to research the possibility of
> increased timing resolution. If it will be very expensive CPU-wise it
> should be done as some sort of special case so as not to decrease
> efficiency in (most normal) other cases.
> 
> One case where increased resolution would be needed is if we want to
> do "scored" granular textures. With 1ms resolution, it is possible to
> generate a synchronous granular stream at 1000Hz. But no way of
> generating the same at 999Hz....

Another case might be situations where a combination of sounds make the
jitter audible... I'm thinking about interference patterns, for example
where the phase of an onset note happens at regular intervals against
the phase of an ongoing note. I'm not sure I would hear it, but
theoretically it could be a problem, don't you think? :)

-- 
/Jonatan         [ http://kymatica.com ]

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
Csound-devel mailing list
Csound-devel@lists.sourceforge.net