Latency Setting Issues Please help

[sheep]

Hello Every one!


I just have a quick question that may need a long answer I just need a understanding of what the lattency settings effect?


I understand that if you have the wrong settings you get clips of sound drop outs and just bad sound quality but I need an explination what is the over all goal for the best audio sound and quality

Can some one please explain what the buffer size effects along with the driver error compensation?

I dont know what they really mean and what type of settings I should have for them so can some one please explain?

Thanks in adavance

[tlennon]

The CPU Preferences is a Plug-In Buffer Size setting for balancing plug-in latency and performance. You can set the number of samples processed at any one time by the plug-in. Higher settings may result in a noticeable performance increase but will also result in higher latencies. You would set this high after you have completed tracking and have inserted effects when you notice the performance is suffering.

Live automatically compensates for delays caused by Live and plug-in instruments and effects, including those on the return tracks. These delays can arise from the time taken by devices to process an input signal and output a result. The compensation algorithm keeps all of Live’s tracks in sync, regardless of what their devices are doing, while minimizing delay between the player’s actions and the audible result. Device delay compensation is on by default and does not normally have to be adjusted in any way. However, Live Sets that were created with Live 4 or earlier will open without device delay compensation. To manually turn latency compensation on (or off), use the Delay Compensation option in the Options menu. Unusually high individual track delays or reported latencies from plug-ins may cause noticeable sluggishness in the software. If you are having latency-related difficulties while recording and playing back instruments, you may want to try turning off device delay compensation, however this is not normally recommended. You may also find that adjusting the individual track delays is useful in these cases. Note that device delay compensation can, depending on the number of tracks and devices in use, increase the CPU load.

[Dexes]

^^talk about unnecessarily complicated explanations.

To keep it simple:

If your soundcard is playing the sound your computer generates immediatly and your computer is too slow the sound will stop playing untill the computer catches up again.
The buffer prevents this from happening by "thinking ahead"
A lange buffer means the computer has time to calculate everything in peace and so reduces the load on the CPU and the possibility of audio drop outs. The down side is, if you want to record audio or midi into live it is delayed by the buffer (this is where the driver error compensation comes in - there's a tutorial in live that shows you how to set it up)

[Jajah]

Small buffer => high CPU load => dropouts / crackling.
Big buffer => high latency => what you play or tweak is heard a bit later.
Adjust buffer size to your system so :
- CPU load stays just above 50% in a live situation
- No more than 80% during a whole song while recording/writing/mixing.

As said above, latency is only a real issue for live playing / tweaking / monitoring, otherwise the software will try to adjust the playhead/recordhead to what you hear, that's latency compensation.

How does it work?
The soundcard hands out an empty buffer (think of a basket) to the software at a regular rate.
The soft must fill it up with a bit of sound that it has generated. That is before the next buffer arrives, or there will be a dropout. Dropout means the soundcard did not get the buffer back from the soft soon enough to produce a continuous wave of sound.
The smaller the buffer, the faster the rate.
Faster rate means the software must work faster/harder to compute and deliver the sound buffer in time.
Your direct action to the sound (playing, tweaking) will be heard when the buffer has been computed by the software and handled back to the sound card.
So a slower rate means your (human) action will be heard later, once the previous (bigger) buffer has finished playing.

Hope this helps.
--
J.

[sheep]

OK well I am a live "Live" performer.....I have in live set maybe like 100 Full length songs about 6 mins each....Then I have about 50 sound samples bells horns lyric loops etc. So My sets are pretty massive and i put on a pretty massive performance I also have a Korg zero 4 in my firewire A Edirol PCR 300 In one USB and Maschine in the other... Letterally every thing is mapped to everything

But I never messed with these setting.... so the question remains is this (and sorry im a dumbass so like make it really like 1+1=2)

If i understand correctly

I want to have the Lowest # on buffer size and lowest # of Driver Error that will allow me to do the most complex thing i need to do while performing with out Crack/dropping out Is that like the Goal?

And how does this affect the overall latency and out/in put latency and all that and weres that video I cant find it.

Thanks and sorry I ride the short buss to school....Like seriously