- This topic has 19 replies, 4 voices, and was last updated 14 years, 6 months ago by
.
-
Posts
-
Can you please include the jitter measurement indicator into the test function on BAE?? It looks like it’s in Live and BBP, but it’s missing in Enhancer.
This would be very useful in some very fine tuning of this outstanding tool.
Thank you!
Good idea! Sure.
Until I do, you can use the free AudioResampler to see the jitter values and find the optimum buffer size, as it shares the audio i/o code with Breakaway.
///Leif
I’ve added the feature:
I don’t know when the next release is, but when it is, the feature will be there 🙂.
///Leif
Thanks, Leaf! In repeater, would I just mimic the input (breakaway) and output (audigy)? Make sure theres no audio playing..
[quote author=”timmywa”]Thanks, Leaf! In repeater, would I just mimic the input (breakaway) and output (audigy)? Make sure theres no audio playing..[/quote]
Hey, I don’t grow on trees, buddy 🙂 Actually my name is pronounced leyf (same vowel as in hey ). I would run audio repeater with as close settings to bae as possible, including audio playing.//Leif
U: buffer Underrun
O: buffer Overrun
Rst: buffer ResetLeif or others can probably describe each one more concisely than I.
Timmy is absolutely correct. Overrun or underrun happens when jitter is so bad that the buffers run out, and causes an audible glitch.
Reset happens when the sample rate suddenly changes, either due to extreme system lag, extreme jitter or a change in clock rate (possible for example when using a sound card with a digital input, using the input as the clock source).
Ideally, all three of these should read 0 forever, but it’s okay to have an under/overrun or three right in the beginning, as long as it settles down.
///Leif
[quote author=”timmywa”]I do know how to spell your name, my phones spell checker does not. Sorry.[/quote]
LOL, no worries. 🙂
///Leif
Thanks guys.
Any thoughts on my other post about the jitter being high but the audio was stable, and when I lowered the jitter (by increasing the buffers), the audio popped now and then?Leif,
One further question… On Audio Repeater, the goal is to get below 10% jitter at 2 buffers. I think this goal is different in BAE… 2 buffers may be a bit low for solid audio… My question is if I get the lowest jitter through repeater at 2 buffers, 1024 size, do I actually use 2 buffers at 1024 inside BAE? or should I use 4 buffers at 512, or some other combination?
Or is my logic all messed up and the number of buffers should be the same inside BAE as it is in repeater?
Side question, when I tuned it with repeater, the input on pipeline was pretty high, maybe 1024, it seemed odd that a virtual device would have to have such a large buffer size. I could see an old audio card needing it as the hardware is crap, but this isn’t hardware at all.
I hope all this makes sense. Thanks!
[quote author=”Milky”]Thanks guys.
Any thoughts on my other post about the jitter being high but the audio was stable, and when I lowered the jitter (by increasing the buffers), the audio popped now and then?[/quote]That doesn’t make sense at all 🙂. How high was it when it was high?
///Leif
[quote author=”timmywa”]Leif,
One further question… On Audio Repeater, the goal is to get below 10% jitter at 2 buffers. I think this goal is different in BAE… 2 buffers may be a bit low for solid audio… My question is if I get the lowest jitter through repeater at 2 buffers, 1024 size, do I actually use 2 buffers at 1024 inside BAE? or should I use 4 buffers at 512, or some other combination?
Or is my logic all messed up and the number of buffers should be the same inside BAE as it is in repeater?
Side question, when I tuned it with repeater, the input on pipeline was pretty high, maybe 1024, it seemed odd that a virtual device would have to have such a large buffer size. I could see an old audio card needing it as the hardware is crap, but this isn’t hardware at all.
I hope all this makes sense. Thanks![/quote]
Actually, I would say anything under 50% jitter is potentially acceptable, as long as it runs stably.
The number of buffers should be the same inside BAE as the repeater — it’s the same i/o code, BAE just does more work.
With the East Asian language bugs that were just reported here in the forum, it may not be too long before next release, because that’s something that does warrant going through the process of releasing again.
///Leif
- You must be logged in to reply to this topic.