Bonjour,
This is a nice article , testimony to the nice work you guys have done. However I felt compelled to write to the editors (sos.feedback@soundonsound.com) about some clarification on audio interfaces.
"()...
But an important caveat is that for an audio device, a mixer might be a better tool than an audio interface, specifically if you are going to contribute many audio sources, for example voice and electric guitar, because Jamulus supports only one stereo (L-R) or mono input per participant. And we want the cleanest, shortest, most efficient path possible (processor and internet -wise) between us and the band to reduce latency as much as possible.
Note that, whatever your instrument, voice is important as you need to talk to your bandmates; an external mic into your Jamulus contribution would be a better option than using the telephone/cell or another piece of software eating at valuable processor and network bandwidth.
Looking at audio devices that have a USB/Firewire/Thunderbolt connexion to a computer, I know of no audio interface that can output mixed down channels in addition to their individual channels. On the mixer side, most of them output the mixed down channels only (good for Jamulus), although a few (ex. ZOOM L-12) can output individual channels in addition to mixed down channels (good for a DAW). So for audio interfaces to Jamulus, you would need to pipe the interface output (individual channels) to a DAW or software mixer before piping a combined stereo or mono signal into Jamulus. This would add to the latency that we are trying to reduce.
In some cases, it might be possible to use two mono chanels on an audio interface and bridge them to one stereo channel, panned hard left and right, then use this output chanel as input to Jamulus, thus bypassing a mixing software in between.
()..."
This is along the lines that I posted a few months ago and that was incorporated in the "Hardware setup" section of the site. Please feel free to add clarification , as i am no engineer.
Merci
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I must confess I am not able to follow what you want to tell, however I (and also the keyboard player) have a USB mixer, I output voice on one channel only, bass on both channels (so the others hear it at the center). When playing, I normally switch off the microphone to avoid extra noises, however a lot of times I forget to do it, so when mixing the recordings I can just delete the mic channel.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I realise that my above post might have been better understood if I had taken the original post example. Lets say you have 3 or more sources contributing to Jamulus. No problem if you have a mixer (all channels mixed to stereo L, R before Jamulus), but if you have an audio interfaces (all channels independent) you have to choose which 2 will go into Jamulus.
Hope this clarifies ...
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm gratified you liked the article in Sound on Sound since I wrote it. I have a real liking for Jamulus as it's kept my band rehearsing all through this difficult time. I also think the new change permitting recording from the server is very neat and I'll be writing on the SoS forums about that.
I think the answer to your issue is: it depends. Specifically it depends on the facilities offered by your audio interface and its accompanying software. For instance my UAD Apollo 8 is controlled by a piece of software called Console which, as its name suggests, emulates a mixer. It has 8 analogue inputs and using Console all these can be mixed down however you want to the monitor or other outputs from where the Jamulus client can pick up the audio. This meets your requirement. So far as I'm aware the majority of audio interfaces with more than two inputs come with similar software. Those that don't you'll need either to connect a mixer to your interface as you suggest or get a mixer with an inbuilt audio interface.
Hope that helps.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Thanks for pointing out that an audio interface's software (on the computer) might mix the multiple sources to provide 2 inputs to Jamulus.
My point in "...we want the cleanest, shortest, most efficient path possible (processor and internet -wise) between us and the band to reduce latency as much as possible..." is that it would include reducing/alleviating anything else eating at your computing resources (hence time) between your audio device output(s) and Jamulus's 2 inputs, even if saved CPU and system time (bandwidth) might be a lot less than internet bandwidth.
BTW I find it strange that people would host a Jamulus client + a video communication (ex. ZOOM) session on the same computer/same network as the video software is eating system resources and network bandwidth that could be used by Jamulus. I would think that one might find reduced latency and jitter using only the Jamulus client and adding a microphone for communicating ....
Lastly, I may digress somewhat here but I'd like to suggest your next article include some information about the possibility to host the Jamulus server in a cloud (application) service. From different discussion threads here (ex. "cloud server vs local server") , it may sometimes reduce latency and more importantly make the latency somewhat more equal between players than if one participant hosts the Jamulus server. I will look into this myself. However for a session with percussion / drummer, I am not sure whether having the server local at the "time/rhythm keeper" is not preferable to having it in a cloud, to help with musical timing ...
Merci.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I doubt if I'll be writing any more articles on Jamulus specifically for the magazine though I have mentioned the possibility of putting the server on a cloud service in the SoS forums. It's a very good tool but there's lots of very good stuff out there to be written about and only one magazine to put it all in. Maybe in a year or two when Jamulus has been further developed and there'd be more new to say.
As it happens we regularly run Zoom with the audio muted at the same time as Jamulus when we're rehearsing. There's no lip sync, of course, but the visual aspect helps in many other ways. Whatever Zoom's other shortcomings it's pretty efficient on bandwidth and we have noticed no ill effects on latency from doing this. My general answer to your points from my experience of working with Jamulus is yes, we want to reduce latency, hence the no Wi-Fi rule and the rest, but the end of the day I wouldn't obsess about it. Set up a reasonable system and get on and use it for making music.
Cheers,
Chris
Last edit: ConcertinaChap 2020-11-04
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I also saw no disadvantages running Jamulus and Zoom on the same computer at the same time. I was cable connected and have an excellent ISP service so it may be problematic for some. In any case Jamulus' delay times were not effected at all and I had no dropouts or other disturbances.
In the mean time we no longer run Zoom, however, as we have found that there is no real advantage running video that is not in the least synced to the sound.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Bonjour,
This is a nice article , testimony to the nice work you guys have done. However I felt compelled to write to the editors (sos.feedback@soundonsound.com) about some clarification on audio interfaces.
"()...
But an important caveat is that for an audio device, a mixer might be a better tool than an audio interface, specifically if you are going to contribute many audio sources, for example voice and electric guitar, because Jamulus supports only one stereo (L-R) or mono input per participant. And we want the cleanest, shortest, most efficient path possible (processor and internet -wise) between us and the band to reduce latency as much as possible.
Note that, whatever your instrument, voice is important as you need to talk to your bandmates; an external mic into your Jamulus contribution would be a better option than using the telephone/cell or another piece of software eating at valuable processor and network bandwidth.
Looking at audio devices that have a USB/Firewire/Thunderbolt connexion to a computer, I know of no audio interface that can output mixed down channels in addition to their individual channels. On the mixer side, most of them output the mixed down channels only (good for Jamulus), although a few (ex. ZOOM L-12) can output individual channels in addition to mixed down channels (good for a DAW). So for audio interfaces to Jamulus, you would need to pipe the interface output (individual channels) to a DAW or software mixer before piping a combined stereo or mono signal into Jamulus. This would add to the latency that we are trying to reduce.
In some cases, it might be possible to use two mono chanels on an audio interface and bridge them to one stereo channel, panned hard left and right, then use this output chanel as input to Jamulus, thus bypassing a mixing software in between.
()..."
This is along the lines that I posted a few months ago and that was incorporated in the "Hardware setup" section of the site. Please feel free to add clarification , as i am no engineer.
Merci
I do not understand exactly what you mean, but I think a Behringer XR18 can do the job.
Yep. it's a mixer ! A digital one with lots of inputs/ouputs ! As long as you fit all your sources to the 2 channels of Jamulus, it'll work !
I must confess I am not able to follow what you want to tell, however I (and also the keyboard player) have a USB mixer, I output voice on one channel only, bass on both channels (so the others hear it at the center). When playing, I normally switch off the microphone to avoid extra noises, however a lot of times I forget to do it, so when mixing the recordings I can just delete the mic channel.
Yep. it's a mixer ....
I realise that my above post might have been better understood if I had taken the original post example. Lets say you have 3 or more sources contributing to Jamulus. No problem if you have a mixer (all channels mixed to stereo L, R before Jamulus), but if you have an audio interfaces (all channels independent) you have to choose which 2 will go into Jamulus.
Hope this clarifies ...
Hi,
I'm gratified you liked the article in Sound on Sound since I wrote it. I have a real liking for Jamulus as it's kept my band rehearsing all through this difficult time. I also think the new change permitting recording from the server is very neat and I'll be writing on the SoS forums about that.
I think the answer to your issue is: it depends. Specifically it depends on the facilities offered by your audio interface and its accompanying software. For instance my UAD Apollo 8 is controlled by a piece of software called Console which, as its name suggests, emulates a mixer. It has 8 analogue inputs and using Console all these can be mixed down however you want to the monitor or other outputs from where the Jamulus client can pick up the audio. This meets your requirement. So far as I'm aware the majority of audio interfaces with more than two inputs come with similar software. Those that don't you'll need either to connect a mixer to your interface as you suggest or get a mixer with an inbuilt audio interface.
Hope that helps.
Bonjour Chris,
Thanks for pointing out that an audio interface's software (on the computer) might mix the multiple sources to provide 2 inputs to Jamulus.
My point in "...we want the cleanest, shortest, most efficient path possible (processor and internet -wise) between us and the band to reduce latency as much as possible..." is that it would include reducing/alleviating anything else eating at your computing resources (hence time) between your audio device output(s) and Jamulus's 2 inputs, even if saved CPU and system time (bandwidth) might be a lot less than internet bandwidth.
Lastly, I may digress somewhat here but I'd like to suggest your next article include some information about the possibility to host the Jamulus server in a cloud (application) service. From different discussion threads here (ex. "cloud server vs local server") , it may sometimes reduce latency and more importantly make the latency somewhat more equal between players than if one participant hosts the Jamulus server. I will look into this myself. However for a session with percussion / drummer, I am not sure whether having the server local at the "time/rhythm keeper" is not preferable to having it in a cloud, to help with musical timing ...
Merci.
I doubt if I'll be writing any more articles on Jamulus specifically for the magazine though I have mentioned the possibility of putting the server on a cloud service in the SoS forums. It's a very good tool but there's lots of very good stuff out there to be written about and only one magazine to put it all in. Maybe in a year or two when Jamulus has been further developed and there'd be more new to say.
As it happens we regularly run Zoom with the audio muted at the same time as Jamulus when we're rehearsing. There's no lip sync, of course, but the visual aspect helps in many other ways. Whatever Zoom's other shortcomings it's pretty efficient on bandwidth and we have noticed no ill effects on latency from doing this. My general answer to your points from my experience of working with Jamulus is yes, we want to reduce latency, hence the no Wi-Fi rule and the rest, but the end of the day I wouldn't obsess about it. Set up a reasonable system and get on and use it for making music.
Cheers,
Chris
Last edit: ConcertinaChap 2020-11-04
I also saw no disadvantages running Jamulus and Zoom on the same computer at the same time. I was cable connected and have an excellent ISP service so it may be problematic for some. In any case Jamulus' delay times were not effected at all and I had no dropouts or other disturbances.
In the mean time we no longer run Zoom, however, as we have found that there is no real advantage running video that is not in the least synced to the sound.