Latency

[legobeats]

Here we go:
- Superior Drummer 3 standalone
- audio buffer set to 64 (that's the minimum within SD3, not sure why I can't set it to 32)
- Mac Mini 2018, macOS 10.14.6, 32GB
- Roland PDX-12 Snare Pad
- scan time 2ms
- Motu M4
- all tests done with eDRUMin 10 and Roland TD-17

Total roundtrip latency from hitting the snare pad to hearing the sound out of the monitors:
- 44,1 kHz sample rate: 8ms
- 88,2 kHz sample rate: 7ms
- 176,4 kHz sample rate: 6ms

This time I couldn't find any differences between the TD-17 and the aDUMIN 10. Well done, Rob!

Disclaimer: Please keep in mind that there may be some inaccuracies in my measurements. But the overall tendency is clear and shows in all the tests. Playing with something around 6-7ms feels really good and now I know how to get there.

[Dadwrshpdrum]

Maybe this is simple answer, but I don’t know. Would there be any positive or negative impact to latency if the eDRUMin 10 or TD-17 is connected to the MOTU M4 via a MIDI cable? So the the only USB connection is the MOTU M4?

[mprinz]

That is an interesting questions.
But I think, the MIDI data transfer times are that low - only mircroseconds, if the you read Rob post of "internal" latency - that it doesn't matter.
I mean, the incoming MIDI over USB events are even closer to the CPU than the MOTU MIDI jack ... aren't they?

As I get the this all right it's mainly an ASIO driver issue feeding the CPU fast enough and THEN a matter of the CPU creating data and feeding again the ASIO driver + DAC. Just playing WAV samples should not be a big deal, but with VST insert FX it becomes more difficult for the CPU.
Now I am thinking of - SD3 uses many FX in there mixer. Hmmm...
But ok, I learned something like, if there aren't complicated convoluted reverb FX or other CPU intensive FX it should be fine.

Due to the Jamulus software for online sessions I have to go to 48 kHz SR and I remained at 64 samples. Now, when having Cubase, and Bidule EZDrummer2 and Firefox etc. running I got crackles and drop outs. And I have a i7 with 16GB RAM from 2013.

But I have a NVIDIA graphics card, does anyone know if that thing can be useful in software drumming.
Do some of you use the GPU?

Greetings

[MisterE]

Maybe this is simple answer, but I don’t know. Would there be any positive or negative impact to latency if the eDRUMin 10 or TD-17 is connected to the MOTU M4 via a MIDI cable? So the the only USB connection is the MOTU M4?

I'm not sure what your question actually is intended to ask. If you are asking whether MIDI has latency, then the answer is "yes". In general, a MIDI signal adds no more than about 1 msec, but that's not zero.

However, when most people are talking about latency, we're talking about audio latency, not MIDI.

If you were to connect an eDRUMin or TD-17 is connected to a MOTU M4 via MIDI, the MOTU M4 (or any audio device for that matter) will not create an audio signal. MIDI only carries digital performance data, not audio.

What would your goal be by connecting the eDRUMin to the MOTU M4 via MIDI?

[Dadwrshpdrum]

My only goal would be to reduce the number of USB connections to the laptop. My 2013 MacBook Pro has two USB ports, and currently I have the eDRUMin 10 using one port and my Alesis Strike Multipad using the other port (I use the multipad as a simple audio interface in addition to its other features). But when I filmed my first drum cover for an audition a couple weeks back I needed a third USB cable to connect my iPhone so I could frame the shot. Anyways caused me to pull out a USB hub and complain to myself about so many cables and the mess. I know I could plug my MIDI cable from the eDRUMin out port to the multipad, but I tried that with my TD-17 module before I received my eDRUMin, and the results were not very good. The multipad seemed to do a terrible job sending MIDI in back out the USB connection. Missed notes and other weird artifacts. I assume an audio interface, such as the MOTU M4, would not have the same issues. But I didn’t know if there would be a noticeable improvement or degradation in latency.

I was just wondering in the test setup, would you have had a positive or negative impact if you used a MIDI cable for the eDRUMin and TD-17 instead of the unit’s USB cable. Sounds like one would expect an additional 1ms using the MIDI cable from the eDRUMin vs the USB port.

I’m not an expert on USB connections, drivers and such. So I didn’t know if the MOTU M4, or other quality audio interface, could transmit the MIDI “faster” to the DAW compared to the eDRUMin or TD-17 USB connection. Based on this feedback I think the answer is no because the MIDI cable from the eDRUMin or TD-17 adds approximately 1ms.

[mprinz]

It depends on which answer...
MIDI is a serial protocol. So, I would state, there is no relevant latency, way less then 1ms. BUT having several MIDI instruments connected sending a lot of data that needs to be transferred you will have timing issues.

But with edrums you have max. 4 notes simultaneously plus one CC04. And I guess, that wont cause any delay issues.

Please check also Rob post on internal latency. He is talking about microseconds.

[MisterE]

My only goal would be to reduce the number of USB connections to the laptop. My 2013 MacBook Pro has two USB ports, and currently I have the eDRUMin 10 using one port and my Alesis Strike Multipad using the other port (I use the multipad as a simple audio interface in addition to its other features). But when I filmed my first drum cover for an audition a couple weeks back I needed a third USB cable to connect my iPhone so I could frame the shot. Anyways caused me to pull out a USB hub and complain to myself about so many cables and the mess. I know I could plug my MIDI cable from the eDRUMin out port to the multipad, but I tried that with my TD-17 module before I received my eDRUMin, and the results were not very good. The multipad seemed to do a terrible job sending MIDI in back out the USB connection. Missed notes and other weird artifacts. I assume an audio interface, such as the MOTU M4, would not have the same issues. But I didn’t know if there would be a noticeable improvement or degradation in latency.

I was just wondering in the test setup, would you have had a positive or negative impact if you used a MIDI cable for the eDRUMin and TD-17 instead of the unit’s USB cable. Sounds like one would expect an additional 1ms using the MIDI cable from the eDRUMin vs the USB port.

I’m not an expert on USB connections, drivers and such. So I didn’t know if the MOTU M4, or other quality audio interface, could transmit the MIDI “faster” to the DAW compared to the eDRUMin or TD-17 USB connection. Based on this feedback I think the answer is no because the MIDI cable from the eDRUMin or TD-17 adds approximately 1ms.

Aha, I see the point. I'll keep it simple. When it comes to MIDI, assume the latency is negligible. I said it was less than 1 msec. It's far less than that, and 1msec is imperceptible to any human. Simply do not factor in MIDI latency. That applies to either 5-PIN standard connections or USB MIDI. Having said that, 5-PIN MIDI is considered more reliable. No one ever complains about the stability or reliability of 5-PIN MIDI connections, but occasionally USB MIDI issues pop up. Again, we're not talking about latency issues with MIDI.

When it comes to latency, the typical culprit is the audio interface. There are some good options to minimize audio latency; I have the MOTU M4 and it has maybe 4 msec which is plenty low for me.

So, go right ahead and use 5-PIN MIDI connections if you are aiming to save your USB connections for other purposes. I am using one from my eDRUMin10 to my Mimic Pro and it works just great.

[mprinz]

Hi guys,

yesterday I was still searching for some information about USB MIDI as I found that playlist
https://youtube.com/playlist?list=PLgWv ... hLC4ZagyJB

Then I ask him about the best way to connect MIDI devices for the lowest latency and he provided that video:

https://youtu.be/7QtTfR9VSa8

And I want to share this for all guys who are actually thinking about that instead of playing the MIDI instrument ..

Edit: like me, of course!

[Alx_d]

Using a cymbal with a splitter to connect to the Roland and eDRUMin at the same time, I used MIDI-OX to monitor the Roland TD20X and eDRUMin. The Roland was always 5-6 msec behind eDRUMin 10. I then looked at the MIDI data in Studio One with SD3 and saw the same difference, Roland behind 5-6 msec.

[Alx_d]

Using a cymbal with a splitter to connect to the Roland and eDRUMin at the same time, I used MIDI-OX to monitor the Roland TD20X and eDRUMin. The Roland was always 5-6 msec behind eDRUMin 10. I then looked at the MIDI data in Studio One with SD3 and saw the same difference, Roland behind 5-6 msec.