You are here

MIDI Channel 10 is bad for drums - fact or fiction?

For fans of synths, pianos or keyboard instruments of any sort.

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby BillB » Wed Oct 23, 2019 10:16 am

The Elf wrote:
BillB wrote:
The Elf wrote:This all comes from the early days of MIDI. Nowadays I doubt it's an issue at all.
Even in a hardware-based system?
Buyer beware. I doubt that all hardware is made equal.
So... it may be an issue...

I'm not trying to make a big deal about it, just to understand whether it should be a consideration in designing a channel-based template for the MPC1000.
I have everything from 'probably not' to 'yes definitely' at the moment :headbang:
... which is fine - all interesting comments.
BillB
Frequent Poster
Posts: 1057
Joined: Mon Apr 07, 2003 12:00 am
Location: East Yorkshire

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby BillB » Wed Oct 23, 2019 10:18 am

Martin Walker wrote: bets might be off if you overlay loads of controller & pitch bend messages on your drum MIDI note data stream.

Good point, Martin. It would be sensible to keep percussion channels (ports!) clear of controller info.
BillB
Frequent Poster
Posts: 1057
Joined: Mon Apr 07, 2003 12:00 am
Location: East Yorkshire

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby Wonks » Wed Oct 23, 2019 10:21 am

I think you've misread Martin's comment. The data will be on the drum stream as it's affecting the drum sounds.

If it's a hardware unit, just make the drums channel 1 and stop worrying about it.
User avatar
Wonks
Jedi Poster
Posts: 10889
Joined: Thu May 29, 2003 12:00 am
Location: Reading, UK
Correcting mistakes on the internet since 1853

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby The Elf » Wed Oct 23, 2019 10:23 am

BillB wrote:
The Elf wrote:
BillB wrote:
The Elf wrote:This all comes from the early days of MIDI. Nowadays I doubt it's an issue at all.
Even in a hardware-based system?
Buyer beware. I doubt that all hardware is made equal.
So... it may be an issue...
Well the problem is that unless you know how the software within the hardware is functioning it really is impossible to advise with certainty. When we were all running Commodore 64s and Atari STs then I can tell you that there definitely *was* an issue using higher channel numbers. These days, with ultra-fast processing, there's really no reason it should be an issue. But when cost-cutting and human error enter the fray...

I'm with Wonks - if the potential troubles you then use Channel 1 and end the speculation.
User avatar
The Elf
Jedi Poster
Posts: 15050
Joined: Tue Aug 14, 2001 12:00 am
Location: Sheffield, UK
An Eagle for an Emperor, A Kestrel for a Knave.

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby BillB » Wed Oct 23, 2019 10:43 am

The Elf wrote:unless you know how the software within the hardware is functioning it really is impossible to advise with certainty.

Accepted, in the end all I can expect is people's experience and practice...

The Elf wrote:I'm with Wonks - if the potential troubles you then use Channel 1 and end the speculation.

Hah - it troubles me if it is a problem... thus the reason for asking. The disadvantage of setting drums to Ch1 is that you lose out-of-the-box compatibility with the majority of new hardware or software introduced into the studio (like the Tomcat example I gave - HARDWIRED to Ch10!). That's the conundrum. If it was an entirely isolated system and I was never likely to introduce or connect to anything new :lol: then yes, Ch1.
BillB
Frequent Poster
Posts: 1057
Joined: Mon Apr 07, 2003 12:00 am
Location: East Yorkshire

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby Wonks » Wed Oct 23, 2019 11:06 am

Once you use a standard DIN hardware MIDI connection, then you are constrained to the 31250 bits per second of the MIDI standard. So if you have several notes all being sent at the start of a bar on other channels sending messages out on lower channels before channel 10 will delay the sending of the drum note message until those other lower channel note details have been sent.

'Running Status' MIDI messages allow multiple note on/off messages to be sent in sequence without having to send the MIDI channel number at the start to reduce the number of message bytes sent (a MIDI message byte is 10 bits long as it has a start and stop bit as well as 8 bits of data). So using this 'running status' form, a six note chord would take up 13 bytes, instead of 18 bytes if each note was sent as a separate entity. 13 message bytes is 130 bits, so that message would take 130/31250 = 4.1 ms to send. Add in a few more messages like that on other channels below channel 10, and you can see that you can easily get noticeable delays of 20+ms. What's more, with less messages sent on other channels, that delay time from the beginning of the bar (or 1/4 or 1/8 note position etc.) can change each time so it may be 5ms or it may be 30ms. Very noticeable on sharp transient drum sounds, less so on soft attack pads.

But this all depended on the sequencer used. Some used the channels in order from 1 to 16, others prioritised channel 10, then 1-9 and 11-16.

But within software to software MIDI transfers, the software isn't limited to physical serial transmission speed limitations , so it gets transferred within microseconds worst case.
User avatar
Wonks
Jedi Poster
Posts: 10889
Joined: Thu May 29, 2003 12:00 am
Location: Reading, UK
Correcting mistakes on the internet since 1853

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby BillB » Wed Oct 23, 2019 11:33 am

Wonks wrote:But this all depended on the sequencer used. Some used the channels in order from 1 to 16, others prioritised channel 10, then 1-9 and 11-16.

Thanks, Wonks, that is interesting and pretty scientific! Especially that some devices may prioritise channels. I guess an easy test would be to set a note on exactly the first beat on every channel, run it into MIDI-OX and see if they go 1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16 or 10,1,2,3,4,5,6,7,8,9,11,12,13,14,15,16...

It may not matter if all the drum data goes via a dedicated port, but it would be good to know if it makes any difference in a mixed channel data stream...
BillB
Frequent Poster
Posts: 1057
Joined: Mon Apr 07, 2003 12:00 am
Location: East Yorkshire

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby Mixedup » Wed Oct 23, 2019 12:47 pm

Hmmm... interesting. Back in the arse-end of the '80s, when I started making MIDI-based music on an ancient IBM clone, my sequencer's (Prism) manual was very clear that it assigned a higher timing priority to MIDI channel 10 than any other channel, and recommended using it for drums. I can't say I noticed any timing issues working that way, but I was a young whipper-snapper back then and probably knew no different...

(I acquired an electric guitar in 1990 and promptly eschewed all things MIDI for a good decade or so :beamup: )

Did they have it wrong, or did different software treat things differently back then?
User avatar
Mixedup
Jedi Poster
Posts: 4414
Joined: Wed Sep 03, 2003 12:00 am
Location: Cambridgeshire, UK

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby desmond » Wed Oct 23, 2019 12:58 pm

Mixedup wrote:Hmmm... interesting. Back in the arse-end of the '80s, when I started making MIDI-based music on an ancient IBM clone, my sequencer's (Prism) manual was very clear that it assigned a higher timing priority to MIDI channel 10 than any other channel, and recommended using it for drums.

There was a time *before* GM, you know! :tongue:
(Ie, the MIDI sequencing years 1983-1991 - yes, GM standard came about in 1991)

The whole "MIDI Channel 10 for drums" thing largely came about because of the GM standard. After that time, some sequencers may perhaps have given priority to channel 10 events. Before that, however, in general higher tracks had timing priority, as tracks were usually processed from top to bottom.

Basically, there is no "this channel is best" because it's *all* to do with the sequencer's implementation. As long as you have plenty of bandwidth and aren't trying to put too many events on the same rigid clock position, things are generally fine. The moment things get cluttered is where some event jostling takes place, and the exact behaviour will depend on the events, and how the sequencer works.

For example, some sequencers may go in a loop and work from channel 1 up when outputting events. Some sequencers may place recorded events in a list/queue, and output the events in queue order, rather than channel order. Some sequencers may be designed to say "let's handle MIDI channel 10 events first, as we give those timing priority". It all depends on the implementation.

Testing can be fine (if you want to avoid doing anything productive, and yet still *feel* productive :lol: ) but really, make some sensible decisions based on your use cases (eg, lets put critical timing stuff on port 1, and the washy stuff on port 2), or, if you've got a *lot* of timing critical stuff, spread it over two ports - and make music.

If you notice a problem, then investigate *then*, rather than necessarily try to fix a problem you don't yet have.
User avatar
desmond
Jedi Poster
Posts: 10318
Joined: Tue Jan 10, 2006 1:00 am
mu:zines | music magazine archive | difficultAudio

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby Ben Asaro » Wed Oct 23, 2019 1:03 pm

I often have multiple channels going on simultaneously between my TX7, JV-1080, and DR-550 which is on channel 10, all from an MC-500. I’ve not noticed any lag beyond the normal start/stop lag on the MC-500. I mention this because all of this kit is mid-80s or mid-90s vintage, and I’ve not had any call for complaint.
Ben Asaro
Frequent Poster
Posts: 899
Joined: Thu Feb 26, 2009 1:00 am
Location: NYC

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby nathanscribe » Wed Oct 23, 2019 1:06 pm

You could avoid all this MIDI terror completely by programming all drums on your external hardware and syncing over FSK instead. :tongue:
User avatar
nathanscribe
Frequent Poster
Posts: 1207
Joined: Fri Jan 19, 2007 1:00 am
Location: Wakefield, for my sins.
I have no idea what I'm doing.

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby BillB » Wed Oct 23, 2019 2:53 pm

nathanscribe wrote:...MIDI terror...

Pah, MIDI is my friend. I don't want to learn the TR, EMX, and Volca ways of sequencing and saving, I just want to get to know one rhythm programming system to fire off the other devices. That was going to be the Beatstep Pro before I got the MPC, but the MPC makes more sense as the programmer, as it has 16 pads x 4 banks. So any of 64 notes (from 35 to 98) can be assigned to the pads in a sequence.

Perfect as the central MIDI rhythm sequencer, as it can address any hardware drum source I have - excepting that General MIDI has eight drum sounds set below note 35, which I will just have to live without!
[there is even a way around that, but it is too convoluted :headbang: ]
BillB
Frequent Poster
Posts: 1057
Joined: Mon Apr 07, 2003 12:00 am
Location: East Yorkshire

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby BillB » Wed Oct 23, 2019 3:00 pm

desmond wrote:If you notice a problem, then investigate *then*, rather than necessarily try to fix a problem you don't yet have.

You're right Desmond, but as I wanted to put some effort into setting up useable templates to make future efforts more efficient, I thought I would ask the SOS sages :D

Sometimes, just the discussion process points out a way - like assigning rhythm parts to one port only. However, as I will also sending multi-channel info on the other port, it is still good to understand the potential issues or non-issues.

As always, I'm very grateful for everyone's inputs. :thumbup:
BillB
Frequent Poster
Posts: 1057
Joined: Mon Apr 07, 2003 12:00 am
Location: East Yorkshire

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby The Elf » Wed Oct 23, 2019 5:06 pm

If you're using a multi-port interface and assigning your drum module to its own output (I also can't be bothered programming anything pattern-based outside my DAW...) you avoid the MIDI channel priority problem - then you just have to worry about whether lower numbered MIDI ports are serviced first!
User avatar
The Elf
Jedi Poster
Posts: 15050
Joined: Tue Aug 14, 2001 12:00 am
Location: Sheffield, UK
An Eagle for an Emperor, A Kestrel for a Knave.

Re: MIDI Channel 10 is bad for drums - fact or fiction?

Postby James Perrett » Wed Oct 23, 2019 5:07 pm

Folderol wrote:Probably more relevant is don't use MIDI THRU if you can possibly avoid it - especially not on kit that uses active sensing.

Surely this only applies to gear where the Thru isn't a true hardware Thru. The delay on a hardware thru should be tiny.
User avatar
James Perrett
Moderator
Posts: 9807
Joined: Mon Sep 10, 2001 12:00 am
Location: The wilds of Hampshire
JRP Music - Audio Mastering and Restoration. JRP Music Facebook Page

PreviousNext