Saturday, September 25, 2010

Latency of the system

A digital system always has some latency. On a forum, someone questioned the use of digital mixing desks because of the latency that is added by the system. Although my experience tells me that the latency is really small, it is always a good idea to measure it.

For this test, I downloaded the 'test tone generator' software an ran this from the tablet. The headphone-output of the tablet is split to two XLRs: one going in the IDR32, the other going in a two channel USB pre-amp. Then, a cable is connected between output 1 of the IDR32 and the second input of the USB pre-amp. Using Audacity, it is possible to capture the exact delay added by the IDR32.


The latency turns out to be 1.2 ms (or 53 samples at 44100 samples a second). This is equivalent to time it takes sound to travel 41 cm through the air. I think this latency is very acceptable for live sound, but things might be different when using in ear monitoring.


(BTW, the gain of the pre-amp was not matched, this is only a test on timing).
Posted by at
Labels: , ,

3 comments:

  1. UnknownSeptember 30, 2010 at 10:03 AM

    Wow, very nicely done!
    Is the latency dependent on the settings of the IDR32? For example, if you turn on or off the compressor, does it change?

    Another thing that I notice is that the waveforms differ significantly (not only in amplitude, but for example also in DC offset after the tone. Did you have any processing turned on that could have caused this?

    ReplyDelete
  2. Guy HarrisOctober 3, 2010 at 2:24 PM

    Interesting point about IEM: we use Aviom and I've never thought about testing the latency. When we move the on-tour show to A&H next week, one of the reasons is to cut out the whole Aviom loop and use PL10s to control individually-tailored monitor mixes, with the wireless IEM senders attached to the outputs from the IDR, as we have in the club installation. Personally, I've not noticed any obvious latency with our current system, which is Mac samples (Ivory & Logic) going through a MOTU 896 into a Yamaha LS9, and out through Aviom A16s. I think the A&H system will be better though.

    ReplyDelete
  3. frankOctober 4, 2010 at 12:51 AM

    The IEM will only be a problem when direct sound is combined with the monitor sound. This is the case when singing or playing some wind instrument (sound travels trough bone). You will have comb-filtering in this case which you will notice in tone.

    When there is no direct sound, there is only latency which should not be a problem at 1.2ms (I think).

    ReplyDelete

Subscribe to: Post Comments (Atom)