Home
This is the user manual of the ORLANDOviols consort box, or in short 'ovbox'. For installation instructions, see INSTALL.md.
Before powering on the device you need to connect all components:
-
Connect the network cable. Attach one end to the network socket of the Raspberry Pi, and the other end to a LAN socket of your router or another active LAN socket in your room.
-
Connect the external sound card via USB cable. Any USB socket can be used.
-
Connect a microphone with the microphone input of your sound card.
-
If you also use an electric output of the instrument (e.g., electric viol, bass or guitar), you may connect it with the second input.
-
Connect the headphones.
Some USB powered sound cards do not startup properly when the phantom power is switched on while powering on. Therefore always switch off the phantom power before switching on the device. Also make sure that all connections are established - the software will not auto-detect the sound card when connected after starting the system.
To switch on the device, simply connect it with the power outlet. It takes about one minute to boot and connect to the service. Once the sound card is activated (the control light of the sound card is on), you may switch on the phantom power.
The correct way to power off the system depends on your device: The first generation contained a power off button to shut down the system. Shortly after pressing the button the LED of the sound card will switch off, and the green light on the Raspberry Pi will stop flashing. Now you can unplug the device. The later generation without a power button can be simply disconnected from the power outlet.
The device can be configured via the web portal box.orlandoviols.com. Currently you need to ask the author or anyone from the ensemble ORLANDOviols to get a free account (this is a completely non-commercial system, without any warranty).
Please note that currently it should be avoided to log into the same account simultaneously from two different browser windows when using multiple devices in the same user account!
If you start your device for the first time, it needs to be linked with your account. To do so, log onto the web portal. Shortly after powering on your device it should be listed as an unclaimed device. You may need to reload the web page. You can click on the device ID (MAC address) to claim the device.
When your device does not appear as unclaimed device this may have several reasons. First check that your device is properly connected to the internet, and that your router allows outgoing network connections (http). You might need to reboot your device, and to reload the web portal page. If all this does not show your device, it might have been claimed by another user. In that case you can contact the portal administrator or any ensemble member. Please provide the MAC address of your device if possible (you may find that information in the settings of your router). If you accidentally claimed a device which is not yours please click 'not my device' in the device settings dialog.
Once the device is connected with your account, you will find it in the list of your devices.
After selecting your device, you may change the device settings. In the device settings dialog you will find three sections: general settings - currently only the human-readable label of your device. This label will be visible to other users. In the 'audio settings' you can change all parameters related to the audio setup of your device, e.g., connections, reverb rendering etc. In the 'network settings' you will find parameters related to the network connection.
In the input port fields you may define to which sound card input
your microphone or instrument is connected. Typically you will have
your microphone connected to system:capture_1 (the first hardware
input) and your electric instrument to system:capture_2 (the second
hardware input). If you leave one of the fields empty, no connection
is made and no signal is transmitted for that channel. If both fields
are left empty, no signal is sent at all (see also section 'network
structure' below for information on the network load). To connect two
inputs, but send only one channel, you may set input port 1 to
system:capture_[12] (or any other regular expression for port
matching) and leave the second field empty.
The output ports define the channels where your headphone is
connected. Typically these are system:playback_1 and
system:playback_2.
The distance field is only used if you transmit two channels. This is the distance in meters between your two channels in the virtual acoustic scene.
The ego monitor gain field controls the level of your own channels in your own mix. This value does not affect the mix of the others.
The render reverb checkbox controls rendering of reverberation in the virtual acoustic scene. The reverb settings will be taken from the room you entered.
In raw mode no virtual acoustic simulation is used at all. Instead the signals are directly passed to the output ports.
Here you can control the buffer length used to compensate for your network jitter. The delay in the field sender jitter is the value added to the buffer of all other users when receiving your signal. The delay in the receiver jitter field is added to your buffer. Both values are given in Milliseconds. The total delay will be network delay (half of the ping latency), the total buffer length (sum of your receiver jitter and the peers sender jitter), and the hardware delay of the sound cards (half of your roundtrip delay and half of your peers roundtrip delay). The peer-to-peer mode checkbox toggles between peer-to-peer mode (direct connection between your device and your peers) and the server mode (connection via the room relay server).
The available rooms are listed below the device selector. To enter a room, click 'enter'. If you would like to change the position in the virtual acoustic rendering, you may click on the name of a peer to swap the places. It may take up to 15 seconds until the changes made on the web portal will be applied in your device.
Only rooms open to the public, or rooms belonging to one of your groups will be displayed. If you own a room you may also change the room acoustic parameters or its name.
The levels of your own instrument, of the others and of the reverb can be controlled in a remote mixer. This mixer is provided by your Raspberry Pi device. When your browser is in the same network as your device, you can open it by clicking on the 'open mixer' link.
Please note that settings made in the remote mixer will be lost after any changes to the room selection, after altering the network or audio settings of your device or the device of your peers in the web portal, or after restart of your device.
In the ORLANDOviols consort box system, three main components are involved: The box (your device and that of your peers), the web portal for configuration and room booking, and the room relay server. The audio connection is handled between the room relay server and your device.
In the peer-to-peer mode the room relay server is used for determining the peer's IP addresses and port numbers, and for establishing the peer-to-peer connection (mesh structure). Your device will send your audio signal(s) to all of your peers. For example, if you play with four musicians, your device will send four copies of your signal. This is approximately 1 MBit/s for each channel and peer, e.g., when sending two channels to three peers you will need an upload rate of at least 6 MBit/s. You will also receive a signal from each peer - if two are sending one channel and one other peer is sending two channels, this would result in approximately 4 MBit/s downstream bandwith.
In server mode all audio traffic is sent via the server. On the receiver side the amount of traffic is the same (just that all the traffic is received from the server and not from the peers directly). However, now the server creates the copies for the peers, and your device will send only one copy, which in the same situation as above would result in 2 MBit/s upload requirements. However, the delay might be larger because the ping time from your device to the server and from the server to the peer need to be summed. Furthermore, the peer may introduce some additional jitter which requires an increased buffer size. Therefore, typically the peer-to-peer mode should be preferred.
It is possible to have some devices in peer-to-peer mode and others in server mode within the same session.
Next to the optimised latency and audio bandwidth a main difference to network communication solutions such as Skype, jitsi or zoom is the spatial rendering of the sound: The sound of the ensemble is reproduced with virtual acoustics. You and the peers are virtually arranged in a circle with a diameter of 2.4 metres. This circle of musicians is in the middle of the room. Early and late reflections are simulated based on the room acoustic properties.