Sampling rate for live gigs

Hi guys this topic might have been touched on before but i would just like to know opinions on whether to run 48 or 44.1 at live gigs as a keys player (not recording) my setup runs 48 just fine but is it necessary (can you hear the difference). Right now im running 48 with a 128 buffer with no probs. But certain rack spaces do push the cpu but no cracks or pops yet.
Thanks

For more info im using a motu ultralite mk4 for audio interface on custom rack mounted windows pc. Works great

Here’s a nice explanation :wink:
https://gigperformer.com/audio-latency-buffer-size-and-sample-rate-explained/

I would worry more about the quality of my audio cable than about the audio interface frequency. :crazy_face:
No earable difference to me between 44.1kHz and 48kHz, and no win to use ultra short buffer.

1 Like

Lucky i use neutrik connectors and van damme cable then lol

Then you can sleep on both ears :sleeping: :ear: :notes:

I run at 44.1k and 256 buffer size for all live shows.

A resulting latency of about 5 msec is okay for live

Thanks everyone I ended up changing to 44.1 at last nights gig and believe it or not nobody in the audience picked up on it. :sweat_smile:

1 Like

Ah. If you want the audience to notice, then try 22k Hz with a sample buffer size of 1024 :stuck_out_tongue:

2 Likes

:sweat_smile: :joy: :rofl:

Lol i wish it would allow me to do that

44.1K 16bit is CD quality audio. 48K is the standard for film/TV, and it’s not an even upsample from 44.1K anyways. I was actually experiencing some pops/etc. with 44.1/24bit so I went to 44.1/16bit and I’m not experiencing that anymore, and it sounds just fine for live performance.

Couldn’t agree more. There’s a reason 44.1kHz was chosen as CD sample rate. Human hearing is approximately between 20Hz and 20kHz and there is a theorem that says that the sampling frequency must be twice the maximum hearing frequency or, in our case, 40kHz or higher.
Anything beyond that is most likely going to get lost in the live setup.

Recoding audio is a bit different as you will manipulate the recording later and the more information you have the better, but for live setup - nobody can hear the difference between 44.1kHz or 48kHz. When you bounce your recording to an mp3 file or even a wav file you’d most likely use 44.1kHz as well.

My current live setup is running at 48kHz, but the only reason is that the impulse response files I use are sampled at 48kHz and I’m a bit lazy to get the 44.1kHz versions. I should do that though one of these days - no reason to tax the CPU one cycle more than it needs to be.

…but enjoy so much to see the bats banging on the walls when i play my ultra-sonic tunes at night! :sunglasses::innocent::face_with_hand_over_mouth:

2 Likes

I deffinately notice the difference between 256 and 128 when playing and i just cant get use to 256 after using 128 for so long the lag is noticable to me. I use in ears aswell so the sound is instant no floor monitors.

As @djogon said in another topic, it would be interesting to look at the total latency and not just at the latency induced by the buffer size. Maybe adding 128 samples to your buffer is the straw that broke the camel’s back for you. The buffer induced latency, is perhaps only the tip of the iceberg. I never checked this, but I will try to test it using the GP tool for measuring the total latency… I am curious to know…

My setup runs 128 with ease so i will continue to use that setting as its feels alot better to me than 256 buffer

Of course :wink:

Holy crap, what are you running gp on, I got a MS Surface i5 8gig ram that can barely run 2 instances of Dive at 256