Do you use/need SMPTE LTC?

Somewhere to chat about MIDI music creation, sequencers and related hardware

Moderators: lotek_style, Mug UK, Moderator Team

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 347
Joined: Wed Feb 03, 2016 7:12 pm

Do you use/need SMPTE LTC?

Postby Foxie » Thu Mar 15, 2018 6:13 pm

A sister thread to this: viewtopic.php?f=111&t=33275&p=341484#p341484

If I'm going to implement MIDI input on my MIDI expander device, would anyone want a SMPTE LTC input too? This would then make it a complete and superior replacement for the SMP24/II.

Some extra hardware is needed to support SMPTE, but not much. The main complexity will be in the firmware. Since I have no clue how the SMP24/II conveys SMPTE information to Cubase, I'll be converting the SMPTE to MTC instead. The MTC will appear to Cubase as another MIDI input. Since Cubase has MTC freewheeling, I don't think it's necessary to implement freewheeling in the hardware. It's a bit tricky to implement the firmware.

This all depends on being able to make MIDI inputs work with Cubase. If I can't figure that out, then of course SMPTE will be impossible too.

I would think these days most people are recording audio digitally, and therefore using MTC instead of SMPTE LTC. And of course you can get converter boxes which convert SMPTE LTC into MTC.

User avatar
rEs
Retro freak
Retro freak
Posts: 11
Joined: Sun Jul 26, 2009 11:15 am
Location: Toronto

Re: Do you use/need SMPTE LTC?

Postby rEs » Thu Mar 15, 2018 7:47 pm

yes please. in my opinion, LTC is preferred over MTC due to LTC being a steady audio signal, not a serial protocol subject to traffic.
um...Atari computers did that in 1989.

User avatar
Atarieterno
Atari Super Hero
Atari Super Hero
Posts: 584
Joined: Mon Jan 18, 2016 3:40 pm
Location: Spain

Re: Do you use/need SMPTE LTC?

Postby Atarieterno » Thu Mar 15, 2018 7:59 pm

I have several devices with SMPTE, but does not bother me that a new device also includes that function.
You have to assess the time you will need to implement it and the price increase.
Greetings.
ST/fm/e, STacy, Mega ST/e, TT, Falcon, C-Lab MKX... and more music tools.

dbsys
Captain Atari
Captain Atari
Posts: 326
Joined: Fri Aug 31, 2012 6:11 am
Location: Germany

Re: Do you use/need SMPTE LTC?

Postby dbsys » Sat Mar 17, 2018 7:26 am

I think, SMPTE is an important feature!

I often use SMPTE.

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Sat Mar 17, 2018 9:13 am

you know what would really float my boat is DIN Sync

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 347
Joined: Wed Feb 03, 2016 7:12 pm

Re: Do you use/need SMPTE LTC?

Postby Foxie » Sat Mar 17, 2018 8:55 pm

ranix wrote:you know what would really float my boat is DIN Sync


This is an interesting idea! Fortunately, DIN sync uses pins 1/2/3 while MIDI uses pins 2/4/5. Therefore, it should be possible to implement DIN sync output on the MIDI output sockets.

Some devices use pins 4/5 on the DIN sync socket for restart/fill in. It might be more difficult to support such devices. I could include a DIP switch to disable pins 4/5 when a DIN sync device is connected. Or you could just use a three pin DIN cable which doesn't connect pins 4/5. Do you mind using a 3 pin cable?

I assume DIN sync input is not needed?

I think it may also be possible to convert MIDI clock to DIN sync with a simple passive cable. MIDI was carefully designed so that the MIDI clock byte is just a simple pulse. Of course you wouldn't have control over start/stop using such a cable. You'd need to manually reset the drum machine to the start whenever you stop the sequencer. If you want to try it, wire pin 2 of the MIDI output to pin 2 of the DIN sync input. Wire pin 5 of the MIDI output to pin 3 of the DIN sync input. Wire pin 4 of the MIDI output to pin 1 of the DIN sync input. You may need to put a 10k resistor between pin 4 and 5 of the MIDI output, but it's probably not necessary. You'll need to make sure you send only MIDI clocks to the MIDI output. Any other MIDI data will cause incorrect clocking. This might be a problem if the sequencer sends start/continue/stop messages - the drum machine will get slightly offset from the sequencer when it starts.

This method won't work with 48ppqn equipment unless you double the sequencer tempo. It might be difficult to support 48ppqn equipment even with a dedicated DIN sync output. I could send two pulses for every MIDI clock, but the resulting pulse train won't be even. I'm not sure if this will upset any drum machines.

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Sat Mar 17, 2018 11:37 pm

The difference between MIDI clock and DIN sync is that there's no other traffic on the DIN Sync cable that could shunt the timecode messages. Pin 4 for reset is important but pin 5 for Fill-in would be useless outside a drum machine and I don't think I'd try to support it. DIN Sync would be best on its own connector.

Realistically the reason I got the Atari was so the MIDI timecode messages would be less jittery than my Mac so it's not really that important. DIN Sync lets you do some cool stuff like clock the DAW from a modular synthesizer with no innate sync capabilities though. The sync line is just a binary high/low signal that clocks in on the rising edge.

I do have other equipment that lets me convert MIDI timecode to DIN Sync, I could clock the beatstep pro from the modular and clock the Atari off its MIDI timecode.

Miguel
Obsessive compulsive Atari behavior
Obsessive compulsive Atari behavior
Posts: 128
Joined: Mon Sep 21, 2015 2:59 pm

Re: Do you use/need SMPTE LTC?

Postby Miguel » Sun Mar 18, 2018 12:01 am

Unlike the old Roland MIDI cables that I have most MIDI cables sold these days tend to have all 5 pins wired up which is not recommended for use with the Atari's non standard MIDI out/thru connector so anyone rocking an Atari should be using a MIDI cable with just the 3 pins wired unless of course they have a splitter to separate the Atari's MIDI out and thru ports.

Ideally Sync24 should have it's own socket rather than use up a MIDI socket and any decent MIDI-to-Sync24 box should have both MIDI in and out (or thru) I use the Sync24 output of a Novation Drumstation myself. Perhaps you should contemplate using panel mounted sockets rather than ones soldered to the PCB if space is an issue Foxie...after all PCB mounted sockets are rubbish anyway.

I read somewhere that there is a Sync24 cable for use with Maxymiser which plugs into the printer port, but then that's a waste of a perfectly good printer port which could otherwise be used for printing off dot matrix pictures of Samantha Fox.

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Sun Mar 18, 2018 1:07 am

+1 on pcb mounted sockets being rubbish, the best sockets are panel mount with flying wires that connect to 100pitch headers

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 347
Joined: Wed Feb 03, 2016 7:12 pm

Re: Do you use/need SMPTE LTC?

Postby Foxie » Sun Mar 18, 2018 1:33 am

ranix wrote:The difference between MIDI clock and DIN sync is that there's no other traffic on the DIN Sync cable that could shunt the timecode messages. Pin 4 for reset is important but pin 5 for Fill-in would be useless outside a drum machine and I don't think I'd try to support it. DIN Sync would be best on its own connector.


I can't find much information about pin 4 at all. It seems it's only supported by a minority of devices. Some sources claim it's used to continue/pause, others claim it's used to reset. But I'm not sure what it resets, because pin 1 is supposed to reset to the beginning.

It seems a lot of MIDI clock to DIN sync converters only drive pins 1 and 3. One of them also implemented the continue message by pausing the clock output - without resetting with pin 1. I found one which drove pin 4, but it doesn't say how.

Do you know if pin 4 is edge-triggered, or level-triggered? I would imagine being a TTL input it must be edge-triggered. If you used a 3 pin DIN sync cable, it would float high.

If you put MIDI traffic on the same output that's used for MIDI clock, there will be a jitter of 0.3ms. It's never more than that, because MIDI always prioritises MIDI clock over ordinary traffic. Of course it would be nice to do better than 0.3ms, but Cubase itself has lousy performance and I doubt it can do better anyway.

I'm hoping to get more accurate timing than Cubase in any software I write. Cutting MROS out of the picture should help a lot.


Miguel wrote:Unlike the old Roland MIDI cables that I have most MIDI cables sold these days tend to have all 5 pins wired up which is not recommended for use with the Atari's non standard MIDI out/thru connector so anyone rocking an Atari should be using a MIDI cable with just the 3 pins wired unless of course they have a splitter to separate the Atari's MIDI out and thru ports.


It should normally be safe to use a 5 conductor MIDI cable, providing the receiving device is MIDI compliant. Where you could run into trouble is if the receiving device deviates from the spec and uses pin 1 and 3 for something. This seems to be very rare, but I have come across one example over the years. I think it was an effects unit remote control that applied power to those pins. Not a good idea! Fortunately, most MIDI outputs (at least those which aren't made by Steinberg and C-Lab) are pretty robust.

In general, there are more compatibility problems between MIDI devices due to flagrant violation of the voltage/current specs. It doesn't help that the spec is badly written and contains contradictions, leading to manufacturers creating incompatible devices. The Atari ST is pretty close to spec, but they have used a 4.7k pullup on the PC900 which is strange.

Miguel wrote:Ideally Sync24 should have it's own socket rather than use up a MIDI socket


This might be worth considering if there's enough demand for DIN sync. I would think it's pretty niche though. You might want to drive multiple DIN devices, in which case you'd need multiple DIN sync outputs. You could use a custom cable assembly to split it, but then you may as well just build a custom cable that splits off the DIN sync from the MIDI out - allowing you to use both.

I would think with 12 MIDI outs, there would be enough to sacrifice one or two for DIN sync even if you didn't use a splitter cable. 8 would be more restrictive, but there's no reason to limit the number of outs to 8 any more.


Miguel wrote:Perhaps you should contemplate using panel mounted sockets rather than ones soldered to the PCB if space is an issue Foxie...after all PCB mounted sockets are rubbish anyway.


I'm not a fan of PCB sockets either. I'm not sure if there's much to gain by ruggedizing the MIDI expander considering almost all MIDI devices also have PCB sockets. It's possible to cram panel mount sockets a little more densely than PCB sockets, but the big problem is how much it adds to assembly costs. It's very labour-intensive to solder all those panel sockets, and the case will be more expensive due to all the circular cutouts.


Miguel wrote:I read somewhere that there is a Sync24 cable for use with Maxymiser which plugs into the printer port, but then that's a waste of a perfectly good printer port which could otherwise be used for printing off dot matrix pictures of Samantha Fox.


It should be pretty easy to write a Cubase driver to support such a cable. But considering my MIDI expander also plugs into the printer port, that's not really ideal. The printer port's interrupt input makes it perfect for MIDI I/O, better than the cartridge port. I suppose the modem port could be used with a zener diode to clamp the voltage. You could use the two handshaking lines as pin 1 and pin 3, and the TX line for an extra MIDI out.

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 347
Joined: Wed Feb 03, 2016 7:12 pm

Re: Do you use/need SMPTE LTC?

Postby Foxie » Mon Mar 19, 2018 12:10 am

ranix wrote:Pin 4 for reset is important but pin 5 for Fill-in would be useless outside a drum machine and I don't think I'd try to support it


Do you own any DIN sync devices which support the pin 5 continue function? It seems that the later Roland devices used a short pulse on pin 5 - overlapping the rising edge of start - to indicate the device should continue where it left off rather than start from the beginning. An example is the MC-202.

I don't know exactly how those machines behave. If you don't pulse pin 5, will it ignore any attempts to set the start position via the front panel? If so, it could be quite annoying to have no support for pin 5 continue. You'd never be able to set the machine to a specific bar and then start the sequencer without resetting it.

There is a possible conflict if I do connect pin 5. Some older machines (like the 808) used pin 5 for fill-in or other undocumented functions. The only way around this is to have two drivers - one of which disables the continue function. Or I could have two DIN sync outputs, one of which doesn't have pin 5 connected.

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Mon Mar 19, 2018 7:18 am

Take all this with a grain of salt because I have never used DIN sync properly - I use breakout cables to connect the signals to weird poo. This post represents the entirety of my DIN Sync knowledge.

I don't think DIN Sync has a concept of bar count, it just clocks pulses out on pin 3 on the rising edge. The edge may go high 48, 24, or 1 time per quarter note (pulses per quarter note, ppq or ppqn).

When you reset the sequence to the first bar on a master device, pin 4 asserts high. When pin 4 is asserted on the slave device, the slave device resets to the first count. I have never heard of pin 4 being used for any other purpose.

I assume pin 5 is for a signal that should tell a drum machine to execute a fill. I've never used this functionality myself and do not own a device that supports it. Nor any device that uses pin 5 for anything else.

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Mon Mar 19, 2018 7:22 am

I was getting pins 1 and 4 confused in my last post, I'm also not clear on the function of pin 4. Start/stop as described above is pin 1. The cool stuff I like is just start/stop and clock (pins 1 and 3)

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 347
Joined: Wed Feb 03, 2016 7:12 pm

Re: Do you use/need SMPTE LTC?

Postby Foxie » Mon Mar 19, 2018 6:30 pm

ranix wrote:I was getting pins 1 and 4 confused in my last post, I'm also not clear on the function of pin 4. Start/stop as described above is pin 1. The cool stuff I like is just start/stop and clock (pins 1 and 3)


I could still implement continue by stopping the clock pulses but leaving pin 1 high. Then when a MIDI continue is received, the pulses can continue without lowering pin 1. If a MIDI start is received, it would lower pin 1 briefly to reset. But would this even be useful? I would think resuming where you stopped isn't useful without some method to locate to a particular position.

It could allow you to manually set the position in a pattern before issuing a continue. But then it's also possible some devices won't let you adjust the transport while pin 1 is high. Not having had much experience with DIN sync, I don't know what range of behaviours to expect. On a device which supports sequencing multiple patterns, does lowering pin 1 reset right back to the beginning of the sequence? Or just the beginning of the current pattern?

What does your MIDI to DIN sync converter do when presented with a MIDI stop and continue (not start) sequence?

Pin 4 seems to vary even between Roland devices. The 808 uses it for "reset and start," but I have no idea what that actually means. The 303 uses it for tap, it's wired directly to the tap switch. I imagine holding it high will cause a problem.

Although I'm not against including a separate DIN sync socket, it appears there is a way to safely combine DIN sync and MIDI on one socket. The MIDI output would drive pin 4 high instead of pin 5 low. Providing you don't send any MIDI data at all to the MIDI out, it will function like a normal DIN sync output. I would only do this for a couple of the MIDI outs. That way, you can still use non-compliant MIDI devices which need power on pin 4 on the remaining MIDI outs.

Korg gear is a little trickier to support with its 48ppqn. I could have a special Korg version of the driver that sends MIDI clocks in pairs to simulate 48ppqn. The clocks will be bunched together, degrading the perceived resolution to 24ppqn. But because of the need to meet Roland-compatible timings (4ms high/low time for pin 3), it will limit the maximum BPM to about 150 on Korg gear. I could add a special mode triggered by the driver which reduces the timings to 2ms. I'm not sure if it's necessary. Weren't there only about 3 Korg models that ever supported DIN sync?

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Tue Mar 20, 2018 8:56 am

oh the normal behavior is that a pause would just stop clocking pin 3 where it is (either high or low, whatever, you can transition it low if you want) and when you resume it just starts clocking again. So rather than a "stop and continue", it just ceases to get clocks and pauses right where it is indefinitely not advancing its internal clock. When the clock resumes it happily continues like nothing ever happened (because really nothing did from its point of view).

pin 1 is only asserted when you want to reset to 0

equipment you'd have attached to DIN sync is free-running and you interface with it more like a DJ, it is its own interface and it's not meant to fully integrate with the DAW. Like those Korg Volca things

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 347
Joined: Wed Feb 03, 2016 7:12 pm

Re: Do you use/need SMPTE LTC?

Postby Foxie » Tue Mar 20, 2018 7:37 pm

ranix wrote:oh the normal behavior is that a pause would just stop clocking pin 3 where it is (either high or low, whatever, you can transition it low if you want) and when you resume it just starts clocking again. So rather than a "stop and continue", it just ceases to get clocks and pauses right where it is indefinitely not advancing its internal clock. When the clock resumes it happily continues like nothing ever happened (because really nothing did from its point of view).


Thanks, this is useful to know! So I should implement it as follows?

MIDI start received: Bring pin 1 high (run), start clocking pin 3.

MIDI stop received: Leave pin 1 high (run), stop clocking pin 3.

MIDI continue received: Leave pin 1 high (run), start clocking pin 3 again.

MIDI stop received: Leave pin 1 high (run), stop clocking pin 3.

MIDI start received: Bring pin 1 low (stop) to reset, then high (run). Then start clocking pin 3.

Or should I always handle start and continue as the same, by resetting back to 0 each time?

----------------

Concerning SMPTE, there's a potential problem. MIDI time code requires encoding the frame rate in the messages, while SMPTE LTC does not. I could wait for a whole second before initially outputting timecode to count the number of frames. But this introduces an undesirable delay. Alternatively I could attempt to measure the actual frame rate. I don't know how accurate that measurement will be, if a response time of one frame needs to be met. Another solution is to always mark the timecode as 30fps. If using a different frame rate, you would use a different version of the driver which marks it as another framerate before sending to Cubase.

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Wed Mar 21, 2018 4:59 am

Ah, as long as pin 1 is asserted high, the slaved equipment might hold itself in position 1. So I'd say:

MIDI start received: Bring pin 1 high (assert), bring pin 1 low (release), start clocking pin 3

MIDI stop received: Leave pin 1 low, stop clocking pin 3.

MIDI continue received: Leave pin 1 low (release), start clocking pin 3 again.

MIDI stop received: Leave pin 1 low (release), stop clocking pin 3.

MIDI start received: Bring pin 1 high (assert) to reset, then low (release). Then start clocking pin 3.

There is also something that happens on my MIDI devices when I use cubase where when I press STOP twice in cubase the sequence resets to position 1. Cubase goes back to the first beat, and so do all my MIDI sequencers. When that happens, pin 1 should also go low/high/low

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Wed Mar 21, 2018 5:19 am

Foxie wrote:Although I'm not against including a separate DIN sync socket, it appears there is a way to safely combine DIN sync and MIDI on one socket. The MIDI output would drive pin 4 high instead of pin 5 low. Providing you don't send any MIDI data at all to the MIDI out, it will function like a normal DIN sync output. I would only do this for a couple of the MIDI outs. That way, you can still use non-compliant MIDI devices which need power on pin 4 on the remaining MIDI outs.


My thoughts:

I think the DIN Sync bit wouldn't worry about pins 4 or 5 and would leave those unused.
I'd tie pin 5 to ground in hardware if I were you.
I'd only support DIN Sync on one of the MIDI ports.
There's a conflict on pin 2 - you might need to provide a hardware jumper to switch pin 2 from being +5V to ground.
MIDI power is hairy, I might actually consider not supporting MIDI power if I were you and leaving pins 2 and 4 floating in MIDI mode. I would slightly prefer to have it but it's added complexity for very little gain. The only device I have ever seen that uses MIDI power is the Yamaha MD-BT01. Which I do happen to have, but I'm probably the only one around here who does

Foxie wrote:Korg gear is a little trickier to support with its 48ppqn. I could have a special Korg version of the driver that sends MIDI clocks in pairs to simulate 48ppqn. The clocks will be bunched together, degrading the perceived resolution to 24ppqn. But because of the need to meet Roland-compatible timings (4ms high/low time for pin 3), it will limit the maximum BPM to about 150 on Korg gear. I could add a special mode triggered by the driver which reduces the timings to 2ms. I'm not sure if it's necessary. Weren't there only about 3 Korg models that ever supported DIN sync?


I don't care about 48ppqn personally and won't use it, I use 24ppqn. The Korg behavior would just be that it runs more slowly at 24ppqn. If it is easy to provide configurable ppqn, 48 24 and 1 are the big values to support. But if it doesn't fit the design to provide thusly configurable options, 24ppqn is the one to shoot for :cheers:

Miguel
Obsessive compulsive Atari behavior
Obsessive compulsive Atari behavior
Posts: 128
Joined: Mon Sep 21, 2015 2:59 pm

Re: Do you use/need SMPTE LTC?

Postby Miguel » Wed Mar 21, 2018 6:07 am

The Kenton Pro Solo II has a MIDI out port which can be switched over to Sync24 or to provide additional outputs although I think these are just 2 trigger outputs rather than gate and CV, oh yeah and speaking of CV have you seen the plans for an Atari CV interface in RA Penfold's book Musical Applications of the Atari ST?

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Wed Mar 21, 2018 9:13 am

that's right up my alley so I just ordered a copy off Amazon

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 347
Joined: Wed Feb 03, 2016 7:12 pm

Re: Do you use/need SMPTE LTC?

Postby Foxie » Wed Mar 21, 2018 11:00 pm

ranix wrote:I think the DIN Sync bit wouldn't worry about pins 4 or 5 and would leave those unused.
I'd tie pin 5 to ground in hardware if I were you.
I'd only support DIN Sync on one of the MIDI ports.
There's a conflict on pin 2 - you might need to provide a hardware jumper to switch pin 2 from being +5V to ground.


I don't think there's a conflict, because the pinouts are as follows:

MIDI:
Pin 2: ground
Pin 4: high-side drive
Pin 5: low-side drive

DIN sync:
Pin 2: ground
Pin 1: start/stop, stop=low
Pin 3: clock
Pin 4 and 5: model-specific or unused

So I could tie pin 5 to ground and drive pin 4 high to transmit MIDI data. When no data is being transmitted, pin 4 is low - the DIN sync device will see those pins as disconnected.

On the one or two MIDI outputs which support DIN sync, I would filter out MIDI clock/start/stop/continue/SPP messages. Instead, when you send MIDI clock/start/stop/continue to those outputs it will drive the DIN sync pins. If you want to connect a MIDI device that uses the MIDI clock, you would use one of the other MIDI outs.

There is one weird thing. Most of the Roland gear connects pin 2 to ground through a 22 ohm resistor. Then, pin 2 is also connected to one of the input pins of the MCU. Under most conditions, the 22 ohm resistor will be enough to provide a direct short to ground. So why are they monitoring the voltage on pin 2? If you wanted to drive pin 2 high, you'd have to provide over 200mA! This would burn out the resistor for sure. I can understand the resistor being there to help break ground loops, but why they need to monitor the voltage I don't know. It's not a balanced input, if a high enough voltage develops across that resistor then it will be unable to read the DIN sync signal correctly.

Not all gear monitors pin 2 this way. The MPU-401 connects it directly to ground. I think Korg gear does too.


ranix wrote:MIDI power is hairy, I might actually consider not supporting MIDI power if I were you and leaving pins 2 and 4 floating in MIDI mode. I would slightly prefer to have it but it's added complexity for very little gain. The only device I have ever seen that uses MIDI power is the Yamaha MD-BT01. Which I do happen to have, but I'm probably the only one around here who does


When I've got my MIDI benchmark program up and running, would you be interested in benchmarking the MD-BT01? By enabling MIDI thru on an iOS sequencer and connecting the MIDI ports to the ST, it should get a reading of the latency and jitter introduced by the bluetooth link and iOS. I would have thought there would be significant latency and jitter on bluetooth, but it might still be better than USB.

ranix wrote:There is also something that happens on my MIDI devices when I use cubase where when I press STOP twice in cubase the sequence resets to position 1. Cubase goes back to the first beat, and so do all my MIDI sequencers. When that happens, pin 1 should also go low/high/low


I'm not sure what sort of data Cubase transmits when you do this, I'd need to monitor it. It might just be transmitting start/stop in quick succession. Or it might be transmitting a zero song position pointer. I don't know if Cubase ever transmits a MIDI start, perhaps it does when you start from the beginning. Cubase might really send a double stop when you press stop twice. I wouldn't have thought a MIDI device would respond to a double stop by resetting.

It's possible some devices could behave differently with pin 1. Some gear might be held at position 0 when pin 1 is in reset. But some gear might be held at the current position when pin 1 is in run - because it thinks it's running and won't let you touch the transport. I don't know if there's anything I can do about this. I could always bring pin 1 back to reset when stopping - in the hope the device will then let you manually select a start position before pin 1 goes back into run mode. Or if the user wants to change the transport position, they could press stop in Cubase twice to put pin 1 back to reset mode.


Miguel wrote:The Kenton Pro Solo II has a MIDI out port which can be switched over to Sync24 or to provide additional outputs although I think these are just 2 trigger outputs rather than gate and CV, oh yeah and speaking of CV have you seen the plans for an Atari CV interface in RA Penfold's book Musical Applications of the Atari ST?


I'll have to have a look at that. I don't currently own any CV-only gear, but I would like to in the future.

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Thu Mar 22, 2018 1:27 am

ah ok, you're right. I was looking at the incorrectly labeled colored midi pinout on a google image search. Pin 2 is indeed ground for both systems. Sure, I will test the MD-BT01

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 347
Joined: Wed Feb 03, 2016 7:12 pm

Re: Do you use/need SMPTE LTC?

Postby Foxie » Thu Mar 22, 2018 6:47 pm

ranix wrote:ah ok, you're right. I was looking at the incorrectly labeled colored midi pinout on a google image search. Pin 2 is indeed ground for both systems. Sure, I will test the MD-BT01


There's some truly horrendous MIDI designs on the internet. It's got a million times worse since the open source maker types started trying to design MIDI gear. What's worse is that many try to sell their faulty and sometimes dangerous designs, and without CE certification no less. I don't claim to be any expert on MIDI hardware design, but I try to take the time to read the standards docs and think about all the worst case scenarios. Any designer thinking of selling products should be doing that! It's just plain lazy when they don't, and everyone suffers.

Here's a really stinky example: https://thefork.github.io/uMIDI/midi_pinout.html

MIDI was very carefully designed to be completely idiot-proof, you can plug outputs into outputs and DIN sync into MIDI with no possibility of damage.

I did some testing with Cubase and MIDI clock. It transmits a start message whenever it starts at bar 1, and a continue message at other times. However, I can't get it to transmit anything when I click stop twice. Are your sequencers resetting back to 0 the instant you press stop twice, or do they reset when you press play following a double stop?

There is an option in the MIDI setup which allows you to transmit resets whenever it stops. I have that switched off, but if it's turned on it transmits quite a lot of data. It might be transmitting something to reset back to 0 there.

The other thing is I couldn't figure out how to get Cubase to transmit a song position pointer. Does it even support that?

There is an apparent limitation in Cubase in that it only lets you transmit MIDI clock on one MIDI out. This means if someone wanted to use DIN sync simultaneously with MIDI sync, they'd be out of luck. I can think of two possible ways around that. One would be to have a dedicated socket just for DIN sync. It would listen to MIDI clocks on the highest MIDI out, but the MIDI clocks also get forwarded to the MIDI out. The other solution would be to have a special version of the driver which duplicates MIDI clocks to all MIDI outs.

I'm thinking it might be more compatible to implement MIDI stop by bringing pin 1 to reset - rather than trying to hold it in run and stop clocking pin 3. Some Roland machines do use pin 5 to implement a continue, which I could support if I had a dedicated socket for DIN sync. For those machines, they don't reset to 0 when you bring pin 1 to reset. Instead, they reset to 0 when you bring pin 1 to run again - unless continue is also high. It might be annoying to use with an 808, because whenever you try to continue it will trigger a fill-in.

ranix
Atari freak
Atari freak
Posts: 71
Joined: Sun Jan 14, 2018 8:01 pm

Re: Do you use/need SMPTE LTC?

Postby ranix » Fri Mar 23, 2018 3:18 am

lol 12v. I'll test the behavior on Cubase stop, it's definitely possible I mischaracterized the behavior. My sequencers are a ways away from the Atari and I can't look at both simultaneously.

I don't think it will send a song position pointer over MIDI - at least I have never used that functionality. Doesn't LTC need to track song position though? I also have never used my LTC in my Unitor

User avatar
Foxie
Captain Atari
Captain Atari
Posts: 347
Joined: Wed Feb 03, 2016 7:12 pm

Re: Do you use/need SMPTE LTC?

Postby Foxie » Fri Mar 23, 2018 4:25 am

ranix wrote:I don't think it will send a song position pointer over MIDI - at least I have never used that functionality. Doesn't LTC need to track song position though? I also have never used my LTC in my Unitor


I think Cubase can both receive LTC, and receive and send MTC. I don't know if any standalone sequencers and drum machines can use it though. LTC is just wall-clock time, rather than based on tempo - so it's probably not terribly useful unless the sequencer has its own tempo map. I'm not really sure why Cubase can't send song position pointer, considering it can send MTC.

Any idea if the Unitor supports LTC under Cubase? I also wonder what would happen if you tried to connect a Midex and Unitor at the same time. They might use conflicting addresses.


Return to “MIDI Software and Hardware”

Who is online

Users browsing this forum: No registered users and 2 guests