parallel compression mpc5000

Yes, it has an exact meaning. But you have to stop thinking the term only fits in the two contexts you found articles about on the internet. Sample-accurate fits perfectly in the context where I used it in.. and I'm not the only one using it in that context. Ask Mr. Anderton if you like. He posts regularly at Harmony Central Forums. I think he's also long enough in the biz to remember when audio-midi-sequencers were introduced and had sensationally new sample-accurate internal MIDI.

You know why it's called sample-accurate MIDI? Because IT IS actually sample-accurate within the host.

Ill-Green wrote:No, phasing happens when two signals are of the same frequency therefore canceling out vital sounds that would've otherwise enhanced the sound, has nothing to do with timing.

Two exact same waveforms played at exactly the same time will be louder.
Two exact same waveforms played at exactly the same time with one waveform that has it's phase inverted will be silent.
Two exact same waveforms played at very slightly different times will phase. You will hear the peaks and troughs of each waveform summing (+ve + +ve, +ve + -ve, -ve + -ve) depending on the offset time difference.
Timing IS everything when talking about phasing.

dustymaestro wrote:You got it twisted, two together creates phasing.

No. You have it twisted and very wrong. See above.

Yes I was wrong. Read the later posts. But some folks dont want to admit to their mistakes. Midi accuracy and sample accuracy are different.

Champ, the same article is at harmony central. LOL

jimmie wrote:

dustymaestro wrote:You got it twisted, two together creates phasing.

No. You have it twisted and very wrong. See above.

Actually the part you quoted is true.

dustymaestro wrote:Yes I was wrong. Read the later posts. But some folks dont want to admit to their mistakes. Midi accuracy and sample accuracy are different.

It wasn't a mistake... you're confusing semantics. I'm not talking about MIDI accuracy and compare it to sample accuracy (in the context of digital sync). That's what you do.

I'm talking about the very thight internal MIDI of hosts like Cubase and Pro Tools that are actually sample accurate, i.e. the implementation allows the internal MIDI to be so exact and on point that there is only up to a couple samples (the 1/44100th of a second with a sampling rate of 44100 Hz) of drift when triggering internal synths/samplers depending on tempo and resolution of the sequencer (PPQN). That could be considered a special form of MIDI jitter because of rounding errors. But that jitter is always fix at the same tempo and resolution. And most importantly parallel and overlapping MIDI events are sent parallel and always exactly the same way each pass and not sequentially (or serially) like in the physical MIDI world (like with MPCs) where each pass is a guessing game.
That implementation above is called sample-accurate MIDI. That's not something I've made up.

http://www.gearslutz.com/board/so-much- ... iming.html
http://forum.cockos.com/showthread.php?t=23481
http://www.native-instruments.com/forum ... hp?t=85469
http://forum.cubase.net/phpbb2/viewtopi ... 4fa733ba19

In those threads above they're talking about sample-accurate MIDI, how it compares to the physical world and rounding errors.
The important factor of sample-accurate MIDI in the context we are talking about was that events are sent parallel, not serial (as compared to physical MIDI). That's the only reason why I mentioned how it is in DAWs, what it is called and how it compares to the way MPCs handle MIDI.

I know what I'm talking about contrary to you. It seems other users have noticed it too.

Yes Fung, jitter is when the two sources clocks don't line up correctly so you're not getting a 1-1 transfer of samples.

And thanks again to Otobot for helping clear up the misinformation.

Anything can be 'sample accurate' so as long as you describe it in terms of samples from some source....usually an audio stream.

For the record, 'samples' is not a distinctly audio term.

nogginj wrote:Yes Fung, jitter is when the two sources clocks don't line up correctly so you're not getting a 1-1 transfer of samples.

Jitter can happen with only one clock.

nogginj wrote:Yes Fung, jitter is when the two sources clocks don't line up correctly so you're not getting a 1-1 transfer of samples.

Not exactly. What you describe is latency.

With respect to syncing two digital audio systems, jitter refers to the variable delay between the transmission of a sample from the sending system, and the expected receipt of the sample by the receiving system. Some samples would be lost because the sending system did not send when the receiving system was expecting to receive. In other words, the clocks would line up correctly (and thus correctly transmit some samples) sometimes, and incorrectly (and thus not be able to transmit some samples) at other times.

Note: in this example, 'sample' refers to a single snapshot of the sound at a given instant (I.E. one of the 44100 'samples' per second), NOT a complete digital audio wavefile (I.E. a snare sample).

Jitter is merely the difference in time between when something is supposed to happen and when it actually does happen. It goes hand in hand with latency. Latency is a steady timing difference, jitter is a variable timing difference.

Fung45 wrote:Phase is for some reason one of the hardest things for people to wrap their head around when you first learn it. I still don't quite understand jitter fully. Otobot does jitter refer to the 1/44100th of a second samples can be off when not using a sync like wordclock?

Wordclock and jitter would be another case where digital synchronization is concerned. When a clock (quarz that swings or another pulse source) to which audio is sync'ed isn't exact enough to keep for example 44100 Hz (if that's the set sample rate) and gets 'out of sync' every now and then which causes (sometimes unnoticable) dropouts.

Physical MIDI has jitter too. I think that jitter is the cause for the existence of this thread. It's exactly why the phasing or comb filtering we notice when layering in the MPC keeps comming and going, alternating, being unpredictable.

The jitter I meant regarding the "sample-accurate" MIDI of DAWs isn't really jitter but a fixed rounding error caused by the fact that the sequencer has a fixed resolution of for example 960 PPQN (pulse per quarter note) which is not as fine as the sample rate.
One quarter note at 90 BPM = 60s/90BPM = 2/3 seconds long.
2/3 s / 960 PPQN = 0.69444... ms -> 1 pulse (or tick) is 0.69444... milliseconds long.
1 Sample @ 44100 Hz = 0.022676 ms ( <- that's a finer resolution)
-> 1 PPQN @ 90BPM @ 44100 Hz sampling rate = 30.625 samples long.
-> The finest resolution of the internal MIDI sequencer is 30.625 times bigger than the finest resolution of the sample rate
or in other words 1 tick in the sequencer is 30.625 samples long @ tempo 90 BPM.

Since there are only whole samples the 2nd tick would be at sample 31 not 30.625. The next 16th note however would be at sample 7350 (30.625 * 960 / 4) which is a whole figure... so the 16th note would come out exactly at the right time on a whole sample pass... I hope I could explain that well.

But this is really not a problem for the sequencer since it is still "sample accurate" each pass because there is no real jitter that causes the note to come out on a sample later or earlier the next pass since it is rounded correctly each time.

nogginj wrote:Yes Fung, jitter is when the two sources clocks don't line up correctly so you're not getting a 1-1 transfer of samples.

And thanks again to Otobot for helping clear up the misinformation.

Anything can be 'sample accurate' so as long as you describe it in terms of samples from some source....usually an audio stream.

For the record, 'samples' is not a distinctly audio term.

No problem
You're right about the term sample. I think the confusion started because we talked about sample accuracy in regards to sample rate in the context of hardware samplers that deal with "samples" or as we should call them "sound snippets" or something along those lines

Fung45 wrote:Otobot does jitter refer to the 1/44100th of a second samples can be off when not using a sync like wordclock?

One thing I might add is, Wordclock is a professional way to deliver a really tight clock signal to several digital audio devices (like audio interfaces, standalone AD or DA convertors) when syncing them, so they all keep their sample rate in sync to this one master clock.
It's not essential when you have only one audio interface or one interface connected to an ADAT DA or AD convertor. Then it's enough to tell both sides which one of them is the master.

dustymaestro wrote:

jimmie wrote:

dustymaestro wrote:You got it twisted, two together creates phasing.

No. You have it twisted and very wrong. See above.

Actually the part you quoted is true.

lol. I don't know what you mean. Two exact same waveforms played together at exactly the same time do not phase they are louder in volume.
If you want them to phase you have to shift the time that one of them plays, thereby shifting one waveform's phase in relation to the other.

The mpc2500 that I own either plays two sample at exactly the same time OR plays them offset by a constant amount of time. The mpc does this randomly.