| Author
|
just a small tip...
|
orik
IsraTrance Junior Member
Started Topics :
45
Posts :
317
Posted : Oct 4, 2004 21:53
|
|
Colin OOOD
Moderator
Started Topics :
95
Posts :
5380
Posted : Oct 5, 2004 01:23
|
LOL
Ok, here's my reasoning:
Computers can only do one thing at a time. Granted, they can do it fucking fast, so that it looks like they're doing a lot of things simultaneously but the truth of it is that only one instruction can be executed at a time.
When a sequencer plays an arrangement, it scans the tracks from top to bottom every time it moves forward in time one tick. Any MIDI events due to be played back are placed in a buffer in the order in which they are scanned and output simultaneously on the tick. Therefore events on a track at the top of the page get put in the buffer before events on a track at the bottom of a page and will therefore be output more exactly on the tick than other events.
This is the way I was told Cubase worked back in 1994 when I was a member of the Cubase Club (geek!). I think the information was also in Sound On Sound magazine at the time.
If things have changed then I'm prepared to stand corrected! 
Like the man said, if you're using VSTis then the whole issue becomes irrelevant as VSTi playback is sample-accurate.
  Mastering - http://mastering.OOOD.net :: www.is.gd/mastering
OOOD 5th album 'You Think You Are' - www.is.gd/tobuyoood :: www.OOOD.net
www.facebook.com/OOOD.music :: www.soundcloud.com/oood
Contact for bookings/mastering - colin@oood.net |
|
|
|
Dimitri
Inactive User
Started Topics :
4
Posts :
229
Posted : Oct 5, 2004 03:10
|
Quote:
|
On 2004-10-04 21:53, orik wrote:
|
|
Mate i wanted to undestand you , not fight . |
|
|
|
Boobytrip
IsraTrance Junior Member
Started Topics :
39
Posts :
988
Posted : Oct 5, 2004 10:00
|
You could do a check by exporting short loops of your tracks to audiofiles: one for kick, one for hi-hat, one for snare etc. In Cubase you can then load in (in an audio track) and chop up the audiofiles using an appropriate quantize setting -> Use a 'scissors' tool and click on the first 16th or 32th gridpoint of the loop with 'alt' pushed down. Next, you can use the audio editor to find the start of each drum sample (use a magnifying glass)in the audio file. Of course you have to make sure that the samples didn't have a silence before their their startingpoint themselves: it's always a good idea to check the startingpoints of samples in your sampler, because some have a short silence at the start.
If the samples in the chopped up audiofile don't start at the beginning of the 'chopped' parts (slices), there is some latency. You could change the startingpoints of these slices, quantize them and thus make your drumtrack sample-accurate again.
On the other hand, if you couldn't hear the difference in the first place, why bother ? |
|
|
|
orik
IsraTrance Junior Member
Started Topics :
45
Posts :
317
Posted : Oct 5, 2004 14:01
|
Quote:
--------------------------------------------------------------------------------
On 2004-10-04 21:53, dimitri wrote:
--------------------------------------------------------------------------------
Mate i wanted to undestand you , not fight .
--------------------------------------------------------------------------------
its a joke man, we have this kind of black humor in israel...dont take it personaly!
|
|
|
|
Dimitri
Inactive User
Started Topics :
4
Posts :
229
Posted : Oct 5, 2004 14:26
|
|
np mate , actualy me from Israel . Next time say it in hebrew because smily pantomime is not my taste. |
|
|
|
ZilDoggo
Started Topics :
4
Posts :
663
Posted : Oct 5, 2004 15:00
|
nevertheless, i think colin is right,.
there is some truth about oriks story, but it only applies to external midi devices.,
midi does indeed send it's data one after the othre.,
so when, for instance your sequencer plays a chord on your virus external synth then the notes from the chord are sent one after the other.,
so there is a little time difference between the first note of the chord and the last one.,
so a normal midi interface can only transmit one command at a time (commands like 'note on' or 'pitchbend' or 'controll change')
this is why there is that story that colin told
about good ol cubase on atari
you realy could make percussion tighter by having them on the topmost tracks.,
it realy made a difference.,
but with vsti's its a different story.,
these should react instantaneously to midi messages from the sequencer.,
if they dont then it's just a bad design (or a bad crack, mind you).,
in practice midi communication inside a sequencer should be far better than external midi communication.,
for instance you dont realy have midi latency.,
commands can be qued way before they are actually needed..,
also, multiple midi commands can be sent to different vsti's all in the time of just one sample.,
this SHOULD be the case
if it's not then there is either something very wrong (design wise) with how the sequencer handles its midi events or there is some kind of error or bug in the sequencer.,
anyway, my 2 cents,.,
greets.,
aka., |
|
|
|
UnderTow
Started Topics :
9
Posts :
1448
Posted : Oct 5, 2004 17:34
|
Hi all, (special waves to Ori!
Although Colin is right about the way Cubase scans the data, you can't be 100% sure that all sequencers scan the data in the same way. (I think it usualy is but only the developers really know).
Luckily we don't have to worry about such issues these days as sequencers have MIDI playback buffers. At least Sonar has it. This means that it really doesn't matter anymore in which order the data is read as it is buffered anyway and sent out to the MIDI interface driver at the right time.
There is a slight latency issue with the MIDI interfaces themselves. Some are better than others. On average, you can't hear this slight delay.
If you can hear this delay, you can either get a better interface or you can adjust the MIDI timing in your sequencer so that it plays slightly early to compensate.
On a side note, some MIDI sequencers/interfaces have special code so that MIDI timing becomes even less of an issue. Logic has the AMT (Advanced MIDI Timing) system and Cubase has another system (of which I have forgotten the name). I must add that these enhancements to MIDI only work if you use these combinations of sequencers/interfaces. The Steinberg system will only work with a Steinberg interface and the Emagic system will only work with an Emagic interface.
Although, as others have allready mentioned, the MIDI protocol is a serial protocol, this is only relevant for MIDI signals going through slow hardware interfaces and cables. Note that I am not making a distinction between slow and fast hardware. All MIDI hardware is relatively slow. This technology is more than 20 years old.
On the other hand, within the computer, there is no serial limitation. This means that with VST or DX plugins running in a good sequencer with full PDC (plugin Delay Compensation) like Sonar 1 and above, Cubase 2 and above, Logic 6 and above, ProTools 6.4 and above, there is no MIDI timing issue at all.
Any good MIDI sequencer sends the MIDI data to the VST or DX plugin at exactly the time it should arrive. If the plugin is well written, you have sample accurate playback.
Sample accurate just means that the timing is precise to within a 1 sample margin. At 44.1Khz sampling rate, 1 sample is ... yes you guessed it, 1/441000th of a second. We can't possibly hear any timing issues at this level of precision.
So all in all, these issues are not really relevant any more. If you do hear a difference, and Ori seems to hear a difference in Cubase, it is time to upgrade to a supperior sequencer like Sonar. Just teasing! It is probably a plugin that is badly written and/or that doesn't support PDC.
Note that Ableton Live does NOT have PDC. Beware of timing issues ...
Ciao,
UnderTow |
|
|
|
|
