Noticeable latency

Danoc wrote:Triplet roll wouldn't cause latency, they themselves would be on time when done. I do it all the time.

He needs to adjust the pad sensitivity to 8 and fix his samples.

It might. If you have note repeat enabled, and you come down early on the pad, you’re going to get that delay, though I can’t imagine not realizing that it’s enabled.

I doubt its that

[youtube][/youtube]

Cockdiesel wrote:With the updates some you have to shift timing to -1.

Edit: i don’t think it’s shift timing that needs to be set to -1, but something else particular to the samples. I remember reading about it on gs when this topic was brought up. Hope that helps.

I'm going to make a video showing latency timing examples With a keyboard machine vs MPC 4000/2500 very soon




Video is for people wondering how I'm using the MPC X and what's causing the latency.

Caustic Yoda wrote:I have the live and 2500, im a percussionist or urgh "finger drummer", latency on them is both pretty unnocticable, are you running heaps of fx or maxxing out your ram, might make it a bit laggy but best pads and drumming surface so far on the live

I'm not running any fx and my ram isn't maxed out. I just try to use different sounds from ableton via midi and it seems like theres a lot time between when i press the pad and the sample plays. I've tried changing my buffer size to 64/32 samples. I can get the in/output latency down pretty good but it still feels like its off. im probably better off recording those notes while playing them seperately? idk i ended up just saying **** it. any suggestions? what audio interface do you use?

well if the latency is over midi control of abelton thats different to mpc having some sort of internal latency. i dont use midi on my computer, my interface is motu828mk3 fw latency round trip is 5.3ms at 44100. maybe your computers not up to it, have you tried a different midi controller than the x?

Caustic Yoda wrote:well if the latency is over midi control of abelton thats different to mpc having some sort of internal latency. i dont use midi on my computer, my interface is motu828mk3 fw latency round trip is 5.3ms at 44100. maybe your computers not up to it, have you tried a different midi controller than the x?

i have a 2500 and my imac is like from 2011 with 4gb ram so maybe that’s the issue.

I got a used mpc live about 5 months ago. Worked great but one day i was having huge latency problems on the pads. I rebooted and it worked fine until last night. Now every project has this issue. I have rebooted many times. Installed the newest 2.5 firmware update nothing has changed. Shift timing is set to 0. Shifting it to -1 seems to make no difference. Pad sensitivity and all of that are at 8. This is really frustrating because i used to be able to finger drum and overdub loops so perfectly and now its creates huge trainwrecks. My projects use less than 75 % of the Ram.

Mr Mizu wrote:I got a used mpc live about 5 months ago. Worked great but one day i was having huge latency problems on the pads. I rebooted and it worked fine until last night. Now every project has this issue. I have rebooted many times. Installed the newest 2.5 firmware update nothing has changed. Shift timing is set to 0. Shifting it to -1 seems to make no difference. Pad sensitivity and all of that are at 8. This is really frustrating because i used to be able to finger drum and overdub loops so perfectly and now its creates huge trainwrecks. My projects use less than 75 % of the Ram.

Do a overall hardware reset plus firmware hard key combo update. there is a hardware (Easter egg) update function on the units. I'm not sure how to do it. ( I forgot the the button combo) But call Akai. Ask how to do it on the live. It will fix the lag.

I was having some issues with this latency and clearing out the hard drive seemed to help. I had to clean out the project data files which were a little tricky to find in the browser. I had to use the mpc toggle button

I also definitely feel some latency when playing the pads on my MPC Live. As you’d expect, it’s particularly obvious when playing fast fills etc. It’s not horrible, but it’s enough to make it hard to “groove” with the thing.

Making sure the sample start sits right on the start of the transient helps to minimise the problem - if there’s latency you def don’t want to add any extra.

Im a drummer, and back in the day I was doing a lot with pads and midi triggering. I know from tests that when latencies get beyond 4 or 5 ms it gets hard to play. Some time soon I’ll measure what it is on the MPC Live, but I’m betting it’s over 5ms. I’ll let you know.

Of course, tapping pads on the MPC is mostly just a method of inputting data, you’re not so much recording a performance because it gets quantised on playback. So it’s not a show-stopper for me.

By comparison, the SP12 was killer for playability and dynamics. It felt like tapping a groove out on a tabletop. That machine is the gold standard for me in that respect.

FWIW, the MPC Live lags quite a bit when synced to midi clock too, certainly more than other drum machines. I’m guessing this is a related issue, but it’s fixable in most DAWs (tho not Cubase/Nuendo)

I feel some latency too. Comparing feelings from MPC1000 and Live - is like a hardware synth vs VST plugin. There is not so big amount of latency, but it is annoying and spoil the groove(
I feel the latency when hit the pad, or hit "play start" or release momentary knob etc. Even when send midi from 1K to the Live and both plays simultaneously same samples (drum track) - there is a noticable latency in Live's sound.
Unfortunately, with its plastic, glitches, latency and touchscreen glitches for me Live now feels more like a toy then an instrument(
Hope they will fix it with the coming update..

And the results are in. The average latency between tapping a pad on an MPC Live (running software v2.5) and sound coming out the machine is...

wait for it (no pun intended)...

17.3 ms

That's not good at all.

What's more, the latency isn't consistent. Over the taps I measured, the fastest response was 16.1 ms and the slowest was 19.5 ms.

This test was done with the unit idle, just tapping one pad which was playing back one snare sound (and yes, the sample was edited so that it started well into the initial transient).

Maybe I'll do some more tests with the machine actually doing some work, playing a sequence etc, or perhaps triggering it via MIDI. But unless anyone knows any magic solutions I may have missed, I don't anticipate results getting any better than 17 ms.

Make of this what you will.

popelife wrote:And the results are in. The average latency between tapping a pad on an MPC Live (running software v2.5) and sound coming out the machine is...

wait for it (no pun intended)...

17.3 ms

That's not good at all.

What's more, the latency isn't consistent. Over the taps I measured, the fastest response was 16.1 ms and the slowest was 19.5 ms.

This test was done with the unit idle, just tapping one pad which was playing back one snare sound (and yes, the sample was edited so that it started well into the initial transient).

Maybe I'll do some more tests with the machine actually doing some work, playing a sequence etc, or perhaps triggering it via MIDI. But unless anyone knows any magic solutions I may have missed, I don't anticipate results getting any better than 17 ms.

Make of this what you will.

Wow thats crazy. We have plenty of Mpc users that can't even detect or tell their units have latency.

How do you do your test?

Pad latency should be easily detectable by ear

With guitar and keys into mic preamp and midi input on my computer interface using ASIO drivers the latency is about 3ms for monitoring the sound and undetectable latency playing.

In contrast to that, the first USB audio interface I bought in 2000 was the Tascam US428 and the monitoring latency was about 18ms and made it near impossible to play and monitor the sound simultaneously. Really threw me off playing and hearing a delay.

The M-Audio interface after the tascam was better giving me about 6ms latency at 128 samples in ASIO settings.

The next interface and my last was the Focusrite Scarlett 2i2 and gave me 3ms at 64 samples but was capable with last drivers I used it on with win 7 and Ableton Live 8 in 2014 of 32 samples and 1.2ms latency but would crackle on audio so backing off to 3ms was perfectly acceptable and no distinguishable monitoring latency playing my guitar into Guitar Rig or playing my keys controlling a vst.

I also used a Korg padKontrol with Geist back then and no noticeable latency hitting pads and monitoring audio output.

In contrast to that, playing the pads on my mv8800 and sampling my guitar and hearing the monitored sound I can detect no latency at all. Everything snappy and playing feels no different than plugging into a head connected to a cabinet for guitar or playing the direct keys on a hardware synthesizer. My master keyboard controls 10 hardware synths through the mv8800 and focus is changed by selecting the track on the mv8800. Zero noticeable latency playing any of the 10 synths from the master keyboard routed through the mv8800 by midi (no midi latency) and sampling a module through the audio inputs directly has no noticeable latency either.

The same for Korg padKontrol with BeatMaker 2/3 on iPad through the focusrite Scarlett 2i2

The same for Synthstrom Audible Deluge with midi controller keyboard or playing pads on the 128 pad grid and monitoring audio through mixer and PA

The same for Zoom Sampletrak ST224 or Gotharman LD2 sampler

Hardware samplers and synths should have zero latency noticeable. Particularly if just triggering samples from pads and just directly monitoring audio output. If there was support for third party plugins on the standalone and many fx being used, you might experience audio latency which would show as delay from hitting a pad to hearing its sound.

I would check what fx are on and strip the sample or program of global fx and individual fx and see if there is still audio latency.

Simple. To be clear, the idea is to measure the time between tapping a pad and the MPC playing the sample. So I did it the most obvious way. I close-mic’d a pad, and recorded the sound of me hitting the pad, and the MPC output simultaneously on a Sound Devices recorder (could have been anything of course). Played the pad with a drumstick to get nice clean hits. Then worked out the delay between the two transients by viewing the waveforms in a DAW.