VB-Audio's Forums

I have made some round trip latency mesurements with "RTL Utility" on different sound cards, using the device in windows audio and in asio when available, all configured to 44.1KHz.

Asus Xonar DX, Asus Xonar U7, Sound Blaster Recon3di : jack-jack cable from headphone output to line input.
RME Multiface 2 : internal buffer : 32 samples, spdif loopback
VoiceMeeter : VBCable max latency 1536 samples, output A on RME Multiface 2 using ASIO Hammerfall DSP


Here are my best results (the latency is almost 20ms varying with windows audio, I just retained my best mesurement) :

Asus Xonar DX windows audio 91ms
Asus Xonar DX XONAR DX ASIO 32ms
Asus Xonar DX ASIO4ALL v2 buffer 512 52ms

Asus Xonar U7 windows audio 58ms
Asus Xonar U7 ASUS U7 ASIO 47ms

Sound Blaster Recon3di windows audio 43ms
Sound Blaster Recon3di (no asio driver)

RME Multiface 2 windows audio 65ms
RME Multiface 2 ASIO Hammerfall DSP 2ms

VoiceMeeter : internal virtual loop back in VoiceMeeter
VoiceMeeter windows audio 76ms
VoiceMeeter VoiceMeeter virtual ASIO 2ms

VoiceMeeter : spdif loopback on the RME Multiface 2
VoiceMeeter windows audio 102ms
VoiceMeeter VoiceMeeter virtual ASIO 5ms


I am a little confused with those results.

I expected VoiceMeeter virtual loop back to have the lowest latency, followed or equaled by the RME Multiface 2. And it's true in ASIO.
Unfortunately, I need to use the windows mixer (videos and games), and the best windows audio latency is achieved by my cheapest, integrated sound card : Sound Blaster Recon3di.

I have tried to reduce the VBCable max latency below 1536 samples, but then I hear cracks.



I found this article from Vincent Burel :
https://www.linkedin.com/pulse/finally- ... ent-burel/
"We can say that finally WIN10 could work correct for real time audio application maybe if using VirtualLock in application and/or maybe with the next RS4 update"

And this article from microsoft :
https://docs.microsoft.com/en-us/window ... io#samples
"In Windows 10, the latency has been reduced to 1.3ms for all applications"

"By default, all applications in Windows 10 will use 10ms buffers to render and capture audio. If an application needs to use small buffers, then it needs to use the new AudioGraph settings or the WASAPI IAudioClient3 interface, in order to do so. However, if one application in Windows 10 requests the usage of small buffers, then the Audio Engine will start transferring audio using that particular buffer size. In that case, all applications that use the same endpoint and mode will automatically switch to that small buffer size. When the low latency application exits, the Audio Engine will switch to 10ms buffers again."


And finaly my questions :
- Is there anything I could do to get a better latency with voicemeeter ?
- Would it be possible for voicemeeter to force the windows audio engine to enter (and stay) into the "small buffer mode" ?

Using WDM device should be optimal under WIN10...
but only if driver are new WIN10 driver (and not USB).

the 1.3 ms announced by Windows is a bit thoeritical, but < 10ms is trully possible with new Windows 10 driver architecture...

but as i explained in this article:
https://www.facebook.com/notes/vb-audio ... 880540508/
most of audio driver are still using old WIN7 driver model, because Microsoft did not finished the job and nothing is working properly.

So I'm very interested in low audio delay. I mean, it is astounding that in 2019, we have 144hz displays with a latency less than 10ms and for the relatively low-bandwidth audio stream the standard is often 50 or more ms?

I got a dedicated usb sound card (sound blasterx g6) and with its ASIO driver, the main output device latency is 1ms (I think 48 sample buffer). So on this end, I should already be optimal, but now it looks like the main bottleneck is the application interface with Voicemeter?
In the VAIO interface, I see we can select 1024 samples as the lowest sample buffer (but then I get dropouts). 2048 works for my setup, but at 48khz that is still >20ms audio delay - right?

I mean, this is already pretty good and with the current settings I run into problems where audio appears before video (i.e. in Chrome youtube/twitch). Does anybody know if Chrome by default adds video delay to account for bad audio latency and if so, how to disable?