Latency with drone

[Jesse Johannesen]

Hello,

I just purchase The NDLR and love it, but I'm noticing the done is out of time from the other parts. 

Instead of starting on the "1" downbeat it always triggers on "&".  I've tried to muck around with the Drone Trigger options, but nothing seems to fix the issue. I figure that with the five dots and four completely filled boxes, I should get the drone to trigger on each beat. Instead, it is syncopated on the "&". I record four parts into a DAW and was able to confirm that motifs 1 & 2 and the pad all trigger properly and it is the drone that's out.

Also, I understand if only the right half of the boxes are filled in, the done should syncopate. I can confirm that I'm seeing the syncopated behavior with the boxed completely filled.

Do you have any advice on how to correct this issue?

Mark

Hey Mark, Let's see if we can figure out what's going on and see if we can sort this out for you. I just took a look at mine and can confirm that when I have 5 dots and 4 filled solid I am getting drones on the 1,2,3 and 4 beats of the bar, as expected so let's see if we can get your's to do so. First you mentioned recording in the DAW can you confirm that you were recording the midi, or was it just audio? If it was just audio, then do me a favor and just confirm for me that there's not some kind of long attack to the sound, so we can rule that out, and try recording the midi, just so we can confirm the issue visually. 

Provided we are still dealing with an issue after the above tests,
Try:
*power cycling the device and see if the problem persists. 
*update the firmware. (link here with instructions: https://conductivelabs.com/download/ )
Let me know what you find and whether any of these solve the issue and I'll see if Steve has any other ideas. 
Cheers,
Jesse Johannesen

[maltin]

Hello,

I just purchase The NDLR and love it, but I'm noticing the done is out of time from the other parts. 

Instead of starting on the "1" downbeat it always triggers on "&".  I've tried to muck around with the Drone Trigger options, but nothing seems to fix the issue. I figure that with the five dots and four completely filled boxes, I should get the drone to trigger on each beat. Instead, it is syncopated on the "&". I record four parts into a DAW and was able to confirm that motifs 1 & 2 and the pad all trigger properly and it is the drone that's out.

Also, I understand if only the right half of the boxes are filled in, the done should syncopate. I can confirm that I'm seeing the syncopated behavior with the boxed completely filled.

Do you have any advice on how to correct this issue?

Mark

Hey Mark, Let's see if we can figure out what's going on and see if we can sort this out for you. I just took a look at mine and can confirm that when I have 5 dots and 4 filled solid I am getting drones on the 1,2,3 and 4 beats of the bar, as expected so let's see if we can get your's to do so. First you mentioned recording in the DAW can you confirm that you were recording the midi, or was it just audio? If it was just audio, then do me a favor and just confirm for me that there's not some kind of long attack to the sound, so we can rule that out, and try recording the midi, just so we can confirm the issue visually. 

Provided we are still dealing with an issue after the above tests,
Try:
*power cycling the device and see if the problem persists. 
*update the firmware. (link here with instructions: https://conductivelabs.com/download/ )
Let me know what you find and whether any of these solve the issue and I'll see if Steve has any other ideas. 
Cheers,
Jesse Johannesen

Hi Jesse,
Thanks for confirming on your end. I recorded the midi. I was trying to figure out if there was a problem with a synth, so I wanted to analysis the data directly. I'll double check my firmware to see if it is the latest (I think it is). One thing that comes to mind is that I am currently connecting it via DIN to my iconnectivity mioXM interface. I don't see how this would only impact one of the four parts, but it is a variable that should be removed. I'll try to connect the NDLR directly to the computer via USB to see if I get the same behavior and report back. 

Thank you so much for your help.

Mark

[Jesse Johannesen]

Ok, that's great info, I agree that the midi interface is not a likely culprit, even if it is the latest firmware it may be worth a reflashing as it may just be a setting that accidentally got switched or something I can't think of that just fixes itself with a reinstall. a couple more things to try first:
For both Motifs set them to a short even numbered pattern length, then choose a rhythm that has a trigger every beat and just to keep things under control, shorten that down and make it an even number of beats. set clk division to x4. In the menu turn the setting for pad quantisation off. Try playing that and see if that lines things up for you.

Turn down the tempo to something low, and listen to the drone while watching the beat indicator, if the on screen tempo indicator is lit (red on the one and then the next three are a different color) that indicates a down beat and with the 5 dots and 4 filled squares, you should be hearing the drone trigger in time. This test can help us figure out if it's actually everything else that's been off.

Let me know if any of that returns good info!
Jesse

[maltin]

Ok, that's great info, I agree that the midi interface is not a likely culprit, even if it is the latest firmware it may be worth a reflashing as it may just be a setting that accidentally got switched or something I can't think of that just fixes itself with a reinstall.  a couple more things to try first:
For both Motifs set them to a short even numbered pattern length, then choose a rhythm that has a trigger every beat and just to keep things under control, shorten that down and make it an even number of beats. set clk division to x4. In the menu turn the setting for pad quantisation off. Try playing that and see if that lines things up for you.

Turn down the tempo to something low, and listen to the drone while watching the beat indicator, if the on screen tempo indicator is lit (red on the one and then the next three are a different color) that indicates a down beat and with the 5 dots and 4 filled squares, you should be hearing the drone trigger in time. This test can help us figure out if it's actually everything else that's been off.

Let me know if any of that returns good info!
Jesse

Hey Jesse, so I spent the weekend testing and playing around. I've localized the issue to clock sync on the midi inputs. Here was my test:

The NDLR: Powered by connecting to the PC via USB. Midi In and Out are connected to the iconnectivity MioXM (for whatever reason the MioXM USB host port doesn't see the NDLR, so I needed to use MIDI - this is so I don't have to rewire my studio when I practice performances in "Dawless" mode).

The Deluge, Korg XD, DSI Rev2, Pyramid, and Neutron are all connected to the MioXM via usb host. Finally have I have a zoia connected via midi to the mioXM.

The MioXM is connected to the PC via USB DAW port.
Both Reaper and Live are configured to send out midi clock to the Deluge, Korg Minilogue XD, Rev2, Neutron, zioa, pyramid, and NDLR (via USB and Midi).

If I program a sequence on the Pyramid and Deluge, then hit play on the DAW. They both play locked perfectly. When the NDLR is locked via USB is to rock-steady as well. I can trigger all of the different parts and they stay locked regardless of when I trigger them. (the sequence on programmed on the pyramid and the deluge matches the pattern matches the drone patter of the one solid block followed by the divided block, followed the two solid blocks, so I can see if it locked to the phase of the pattern).
When I switch to MIDI A for sync, however, it was all over the place. It doesn't drift, as far as I can tell, but the "phase" is dependent on when I push the trigger buttons. It doesn't seen to correlate to exactly when I press the trigger button. Sometimes it will be as much as a beat and a half out of sync. Pressing close to the downbeat tends to keep the sync in phase. (again this is different than the behavior i see with the sync set to USB where phase is matched regardless of where in the bar I trigger the sequence).

[Jesse Johannesen]

Ok, this is an interesting issue. Let's see if we can do some tests to see where in the chain the problem lies. Lets try to get the simplest signal chain into the NDLR MIDI A possible:
one external clock source => midi cable=>NDLR MIDI A and see it's still messing up
if yes, then replace the midi cable.
if still yes, try one more external clock source.
if no the problem went away, start putting things back in the chain until we see what is causing the hubbub.

One thing I noticed you mention is that the MioXM isn't recognizing the NDLR via USB host, and I was thinking it may not be providing quite enough power. You may be able to get the host to recognize NDLR by using a powered USB Hub, and though I've had mixed results with those in the past, it may be worth a try.

I'll get Steve and Darryl to take a look at this and see if they've got any suggestions as well.
Jesse

[Steve]

Ok, this is an interesting issue. Let's see if we can do some tests to see where in the chain the problem lies. Lets try to get the simplest signal chain into the NDLR MIDI A possible:
one external clock source => midi cable=>NDLR MIDI A and see it's still messing up
if yes, then replace the midi cable.
if still yes, try one more external clock source.
if no the problem went away, start putting things back in the chain until we see what is causing the hubbub.

One thing I noticed you mention is that the MioXM isn't recognizing the NDLR via USB host, and I was thinking it may not be providing quite enough power. You may be able to get the host to recognize NDLR by using a powered USB Hub, and though I've had mixed results with those in the past, it may be worth a try.

I'll get Steve and Darryl to take a look at this and see if they've got any suggestions as well.
Jesse

[Steve]

Hi Maltin,

Jesse and I have been discussing your issue and I think we have a solution. 
Can you give this a try…  use MIDI 5 Pin B, not A.

The NDLR only responds to start/stop/continue messages on MIDI 5 Pin B and USB 1. I believe this is the reason why you see it behaving differently based on the port used. When you have it plugged into the USB port 1, where #1 (of 4) is the virtual cable number over the USB physical cable, both the clock messages and the stop/start/continue messages are being received. So it works as expected. 

When you have it plugged into the MIDI 5 Pin A, The NDLR does NOT respond to start/stop/continue messages. So you’re seeing the phase issue, even though it is still in sync, because the midi clock messages are being received in MIDI 5 Pin A but NOT the start/stop/continue messages.

A side note:  Often when you press “stop” or “pause” on a MIDI device that is sending clock and stop/start/continue messages, the stop message gets sent and it no longer sends clock tick messages. So even though the NDLR does not receive stop/start/continue messages on MIDI 5 Pin A, it will appear to respond to the “stop” message because no clock tick messages are being received. And the NDLR is, for the moment, “frozen in time”.  When you then press “play” or “continue” the stop/start/continue messages are still not received on the MIDI 5 Pin A, but the clock tick messages start again and are received by The NDLR, so it continues where it left off, no longer “frozen in time”. :-) 

BTW we did this because sometimes you may not want the stop/start/continue messages to actually stop The NDLR.

Please let us know if this resolves your problem. Happy noodling!

Steve

[Jesse Johannesen]

Well I may have been able to actually recreate the issue on my unit. My findings are that regardless of which 5 pin din input I use, if I am clocking from Ableton out through in my case, a tascam audio interface with midi, I get a similar issue, where occasionally the rhythm of the of the NLDR is inconsistent (I didn't initially have audio monitored on the daw, but it seems possible there would be phase issues), but if I changed sources to literally anything besides Ableton, the issue went away. I tried to figure out a way to record the clock signal out of ableton and analyze the timing but the result was so far off of what I was expecting that I am having doubts about it's validity, (clock time signitures between 3 times/ms - 30 times/ms).

Regardless I planned to try and monitor the results via my Ipad today, but after retrying the initial test, I am no longer getting the weird clock variance. My supposition is that this is the result of a bad clock signal from the DAW but as my results are steady now, I don't have a bad signal to measure. I will try this a few more times over the next couple days and if I can get it to fail again, I'll see if I can record the results again for a second look.

Maltin, if you can please try to clock the NDLR from another source via MIDI A(preferably both through mio and direct) and help me to confirm this is related to the DAW clock. The tests we ran:
ableton > midi A = Wonky
ipad midi app > midi A = Steady
MPC Live > midi A = Steady
Midi Ox from pc > midi A = Steady
and finally today:
ableton > midi A = Steady
then:
restarted ableton > midi A = Steady

Let me know if any of this works better for you and I'll keep you posted if I can get it to mess up again.
Jesse Johannesen

[maltin]

Hi Jesse,

Thank you for everything! Your advice fixed it! Pluging clocking from MIDI DIN B locked everything to the downbeat!

Related but slightly off topic, is there any way to mute the clock out on the NDLR?

It seems it always outputs clock and as a result any device that is not the master or the NDLR doubles in BPM. As such, I fixed the issue by making the NDLR clock, or filter out the clock messages with the mioXM.

Filtering out the messages ins't a problem, but it would be nice to have the option to turn off the clock in cases where I'm using a less sophisticated midi router.


Lastly, I also found it odd that the NDLR wasn't found be the mioXM. I did try a powered hub. The mioXM found other devices connected to the hub, but not the NDLR. I was a little bummed at first, but now that I got the clock worked out, I'm set! (other than trying to find a way to turn off the clock output on the NDLR).

[Jesse Johannesen]

Hi Jesse,

Thank you for everything! Your advice fixed it! Pluging clocking from MIDI DIN B locked everything to the downbeat!

Related but slightly off topic, is there any way to mute the clock out on the NDLR?

It seems it always outputs clock and as a result any device that is not the master or the NDLR doubles in BPM. As such, I fixed the issue by making the NDLR clock, or filter out the clock messages with the mioXM.

Filtering out the messages ins't a problem, but it would be nice to have the option to turn off the clock in cases where I'm using a less sophisticated midi router.


Lastly, I also found it odd that the NDLR wasn't found be the mioXM. I did try a powered hub. The mioXM found other devices connected to the hub, but not the NDLR. I was a little bummed at first, but now that I got the clock worked out, I'm set! (other than trying to find a way to turn off the clock output on the NDLR).

Oh hot dang! Great to see we found a solution!

Turning off midi clock out is located in the global settings menu. To enter press and hold [shift + menu] and powercycle the NDLR it will load in the global settings menu. press motif 1 to toggle the clock out setting, then press panic/motif1/2 toggle at the upper center-right of the unit.

As it turns out the mioXM's usb won't recognize the NDLR for some reason. Steve told me yesterday but I forgot to mention it. Also, I was informed that the NDLR only draws around 90ma of current, so it's unlikely to be under supplied unless for some reason the usb isn't wired for power, which has been known to occur, but apparently is very unusual. 
Jesse