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 5:11 pm

James Perrett wrote:
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.

Yes, I get avoiding daisy-chains of MIDI Thru, but a single Thru-box (4- / 6-way / more), driven by a single output port, should be OK.
BillB
Frequent Poster
Posts: 917
Joined: Mon Apr 07, 2003 12:00 am
Location: East Yorkshire

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

Postby Folderol » Wed Oct 23, 2019 6:45 pm

BillB wrote:
James Perrett wrote:
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.

Yes, I get avoiding daisy-chains of MIDI Thru, but a single Thru-box (4- / 6-way / more), driven by a single output port, should be OK.
In a word, Congestion
P.S. Note I said especially with kit that has active sensing.
User avatar
Folderol
Jedi Poster
Posts: 10590
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.
Onwards and... err... sideways!

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

Postby BillB » Wed Oct 23, 2019 11:43 pm

BillB wrote: 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...

OK, I think this is sorted, at least for the MPC.

With identical notes on the first beat, copied to every channel, where track 1 = channel 1 etc, the order of channelised notes arriving at MIDI-OX is
1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16.

However, if Track 1 is set to channel 10, the order of channelised notes is
10,1,2,3,4,5,6 etc.

Same holds true whatever channel is set to whatever track. The notes are output in track order, irrespective of channel. So if I want Channel 10 to have top timing priority, I set it to track 1. If I am understanding MIDI-OX’s timestamp correctly, it seems to take about 17ms to send the full set of 16 note-on commands.

So the simple answer for me is a) stick with Ch10 for drum tracks and b) place them on the first sequencer track(s).

Job’s a good’n, thank you all for your help. :thumbup:
BillB
Frequent Poster
Posts: 917
Joined: Mon Apr 07, 2003 12:00 am
Location: East Yorkshire

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

Postby desmond » Wed Oct 23, 2019 11:56 pm

Good findings. It's about what I'd expect on a hardware sequencer of this vintage - the code is probably fairly simple, and loops through the tracks from track 1 onwards outputting events as necessary.

Now you have a plan! :thumbup:
User avatar
desmond
Jedi Poster
Posts: 9601
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 resistorman » Thu Oct 24, 2019 1:00 am

Unless you want the drums a bit behind the beat :D
User avatar
resistorman
Frequent Poster
Posts: 842
Joined: Sun Nov 22, 2015 1:00 am
Location: Asheville NC

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

Postby Mixedup » Thu Oct 24, 2019 8:46 am

desmond wrote:
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.

This *was* before GM!
User avatar
Mixedup
Jedi Poster
Posts: 4410
Joined: Wed Sep 03, 2003 12:00 am
Location: Cambridgeshire, UK

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

Postby desmond » Thu Oct 24, 2019 9:10 am

Did you have an MT32 by any chance? That was iirc the first thing that had a convention of channel 10 for drums, largely because it had 8 parts, but ignored channel 1, so it’s 8 parts were on channels 2-9, with the drums therefore landing on 10.

As the MT32 was popular in 1987, a fair few software sequencers directly supported it, and drums started to be expected on channel 10 to maintain compatibility. So that’s where the channel 10 thing really started, and continued in Roland modules/workstations, before becoming officially part of the GM spec.
User avatar
desmond
Jedi Poster
Posts: 9601
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 Mixedup » Thu Oct 24, 2019 3:03 pm

No, not familiar with that one. I tended to use Yamaha stuff... much of it loaned from the music dept of the school my dad worked at!
User avatar
Mixedup
Jedi Poster
Posts: 4410
Joined: Wed Sep 03, 2003 12:00 am
Location: Cambridgeshire, UK

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

Postby BigRedX » Thu Oct 24, 2019 4:39 pm

I've never used channel 10 for drums and always thought it a bit strange (for the reasons outlined in previous posts) that it was done that way.

In all my programming, as far as possible, drums have always been on MIDI channel 1 and when using multi-port interfaces on Channel 1 of port one. And they've always been on the first track(s) on my Logic arrange page. The only exception to this was if I was using a MIDI device that didn't normally do drum sounds for an additional percussion part in an arrangement.

I think the first MIDI device I came across which defaulted to MIDI channel 10 for drums was the Roland TR707.
User avatar
BigRedX
Frequent Poster
Posts: 795
Joined: Fri Sep 03, 2004 12:00 am
RockinRollin' VampireMan

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

Postby desmond » Thu Oct 24, 2019 4:52 pm

BigRedX wrote:I think the first MIDI device I came across which defaulted to MIDI channel 10 for drums was the Roland TR707.

Hmm, interesting. The MIDI implementation chart just says it can be set to 1-16 and can be changed to whatever you want and memorized, but it doesn't say in the manual (that I can see) it defaults to 10 when shipped.

If that's true, that's interesting, and suggests that Roland internally tending to think of MIDI channel 10 as for drums, even before the MT32. I wonder if their other e-drum stuff of that era, like the DDR30*, also defaulted to channel 10..?

*Edit: the manual for the DD3 regarding MIDI channels etc is basically the same as the TR707, no mention of the default either.
The TD7 defaults to channels 10-13.

I suspect Roland wanted to keep the drum channels sufficiently far away from the lower end so that people buying DX7's wouldn't hear drum sounds when playing notes etc, which would be confusing, and they settled on 10 as a sensible choice...
User avatar
desmond
Jedi Poster
Posts: 9601
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 desmond » Tue Oct 29, 2019 6:17 pm

Some more detailed info on how Creator/Notator prioritises MIDI transmission, from Nov 91's Sound On Sound "Software Support", C-Lab section:-
http://www.muzines.co.uk/articles/software-support/7536

DATA PRIORITY

We confirm Paul's reassurances regarding the transmission of MIDI data taking precedence over, say, screen updates. In principle, C-Lab programs always 'prioritise' MIDI notes, with controllers second, followed by everything else. The computer screen comes last (though needless to say, the C-Lab screen-refreshing routines are optimised for speed). While on the subject of screen-refreshing, the most extreme example of Paul's "molasses" occurs when a large amount of Sys Ex data is transmitted: the screen may well freeze for a few seconds while the Atari grunts away at transmitting the bytes at all possible speed (once the header has gone, the main body of Sys Ex data follows on as quickly as the MIDI baud rate allows, with no regard to pulses, clocks etc).

There's another interesting twist: where more than one note occurs on the same pulse (eg. where the music has been quantised), C-Lab programs transmit the lowest-pitched note first, followed by the other pitches in ascending order (you can follow this hierarchy in the event list) — it's done for psychoacoustic reasons, because the ear/brain extracts information on relative timing from low-pitched sounds.

And because MIDI is a serial killer (sorry, couldn't resist that one!), and Intensely hierarchical in nature, you can always predict which events will be sent in which order, given some background information:

• In a chord, lowest note first.

• In a pattern, track 1 first, followed by 2, 3 etc.

• In an arrangement, chain 'a' first, then 'b', 'c', 'd'.

• Between notes of the same pitch, but on different MIDI Channels, on the same pulse in the same track: dictated by the order in the event list.

• Between notes and MIDI controllers on the same pulse: notes first, thanks to the 'Play Algorithm' function in the Flags menu; if this is disabled, then notes and controllers share equal importance in the hierarchy, a situation that is occasionally required by some devices such as mixing consoles.

• Between MIDI Out ports A, B/C/D, E and F (for example, in one pattern tracks 1,2, 3 and 4 each contain a C3 at the same time position; each track is assigned to a different port, say A, B, E and F). Here, the speed at which the notes are transmitted via their respective ports is appreciably higher than if the same port were used by all four notes, since there is no queueing at the output buss. Using this information, you should spread your MIDI data load between your available MIDI Out ports to optimise the speed of transmission.

Using more than one of Export's Outputs gives you more Channels, but without the corresponding speed enhancements of the other ports; only use the extra two parts for non data-intensive, 'special case' devices (old, dubious synths, mono-mode devices etc).

So here at least, there doesn't appear to be any *MIDI Channel* event prioritisation - track order yes, but not MIDI channel order - these events are transmitted in the order they appear in the event list.
User avatar
desmond
Jedi Poster
Posts: 9601
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 Eddy Deegan » Tue Oct 29, 2019 6:53 pm

I've encountered the 'default to channel 10 for drums' thing a few times, though I tend to override that nowadays. I don't know where I first saw it myself but I have a strong sense of it being a Roland thing, although that could just be co-incidence.

Looking through a few of my old track notes regarding devices, channels, patches etc. I see that I did use channel 10 for drums quite a lot in the early 2000s. On a whim, I just checked the manual for the Roland D10 which was released in the late 1980s (I had one in the early 90s for a year or two) and I see that there too, they denote channel 10 for drums:

Image

Incidentally, the Roland D10 manual is a great example of, shall we say, somewhat rough translation in places. One section is headed "Mamory backup" Image
User avatar
Eddy Deegan
Moderator
Posts: 4072
Joined: Wed Sep 01, 2004 12:00 am
Location: Brighton & Hove, UK
Some of my works.
Please consider supporting the SOS Forum Album project.
 

Previous