Topic: Latency of 1.5ms with buffer size of 64

I have just found that I can acheive much lower latencies than I had been by switching to asio4all and using my emu Xmidi USB interface. I had been using a standard midi cable to connect my keyboard to my focusrite saffire pro 24 firewire audio interface, but I was never really satisfied with the latencies I was getting. I don't know if it was the cables or the midi interface within the focusrite that was causing this, but, as I say, I'm getting 1.5ms which is heaps better than 4.0! By the way, the sample rate is set at 48000 / 32000 and there are no glitches. I haven't had the chance to really try it out properly, but so far so good. In addition, I must note that I disabled all my USB ports bar the one my emu Xmidi is plugged into.

hope this helps someone

Re: Latency of 1.5ms with buffer size of 64

Good news Christopher.

I'm using the EMU 1820M (five years now), but it's a PCI.

Surprising that you can get latency that low through USB.

Which unit is that?  0404 USB 2.0 or the 0202?

Glenn

__________________________
Procrastination Week has been postponed.  Again.

Re: Latency of 1.5ms with buffer size of 64

Note that the latency reported by Pianoteq, or perhaps your ASIO control panel, is only the latency due to the ASIO buffer itself. There may be extra latency.

In my case, with my Fast Track Ultra in standard mode, there is a 110 sample overhead. I.e, if my buffer size is set to 128, I need to add 110 to that to get my REAL latency, which is 238 samples.

Sequencers seem to be able to report the true latency, which makes me think that the ASIO software interface has the capability to report this to applications.  I have tried two - Ableton Live, and Samplitude - both report 5.4ms for the output latency @128samples & 44.1kHz, whereas  Pianoteq only reports the raw ASIO buffer latency (2.9ms)  (Samplitude reports the latency very precisely: 238 samples - Ableton just gives the time in milliseconds)

Also, I suspect that the actual latency would vary a bit from note to note, depending on the precise time the note was struck, in relation to the timing of the audio streaming process.

The above is ignoring MIDI latency, of course.

Greg.

Last edited by skip (21-01-2010 23:48)

Re: Latency of 1.5ms with buffer size of 64

Glenn NK wrote:

Good news Christopher.

I'm using the EMU 1820M (five years now), but it's a PCI.

Surprising that you can get latency that low through USB.

Which unit is that?  0404 USB 2.0 or the 0202?

Glenn

its a midi to usb cable (e-mu Xmidi)

Chris

Re: Latency of 1.5ms with buffer size of 64

Which one.

http://www.emu.com/products/welcome.asp...y=610&

Glenn

__________________________
Procrastination Week has been postponed.  Again.

Re: Latency of 1.5ms with buffer size of 64

Xmidi 1x1

Re: Latency of 1.5ms with buffer size of 64

sigasa wrote:

Xmidi 1x1

Christopher:

Wow, that didn't break the bank did it?

Wish I'd been more careful with my purchase:

http://ixbtlabs.com/articles2/creative-emu-1820/

It nearly broke the bank, but it is nice.

Glenn

__________________________
Procrastination Week has been postponed.  Again.

Re: Latency of 1.5ms with buffer size of 64

I've now got the buffer latency down to 1.3 using 96000 and 128 buffer. I use ASIO4ALL (latest version).