Dante (network protocol)
Dante is a combination of hardware , software and a network protocol that allows multiple channels of uncompressed digital audio signals to be transmitted over a network cable with low latency .
description
Dante was developed in 2006 by the Australian company Audinate. Like most audio over Ethernet applications, Dante is primarily intended for commercial use in the professional sector. It is mostly used when multiple audio channels have to be transmitted over long distances and to different locations.
In contrast to analog transmission, digital transmission offers several advantages. The signal is comparatively insusceptible to interference even with transmissions over longer distances, and the demands on the quality of the cable are lower than with analog audio transmissions. Compared to other audio over Ethernet applications such as CobraNet or EtherSound , Dante offers integrated gigabit support, an enormous transmission capacity of over 500 high-resolution digital audio channels in each direction with a single network cable, lower latency and automatic word clock configuration.
background
Audio transmission formats via network cables differ with regard to the transmission method used, which is subdivided into layers ( OSI layer model ). The simplest transmission layer (Layer 1) only defines the transmission medium used ( e.g. copper or fiber optic cable ) in terms of its properties, including the connector ( RJ-45 ). A transmission variant based on Layer 1 is, for example, AES50. Such audio devices use network cables to transmit the audio signals, the audio transmission itself takes place between the two directly connected devices using the point-to-point method in one of the conventional digital multi-channel formats ( ADAT , MADI, etc.). The main advantage compared to transmission via ADAT, USB or Firewire is the significantly greater maximum possible cable length (max. Approx. 90 m compared to max. 10 m).
Audio transmission protocols based on Layer 2 (e.g. EtherSound , AVB , CobraNet ) are based on the Ethernet transmission protocol . They also use network cables for transmission. However, like Layer 1, the system does not offer any addressability of the devices used, so that specific routing between several devices is not possible.
In contrast to this, the transmission at Dante takes place on layer 3. In this transmission layer, the transmission no longer takes place in one of the conventional, purely PCM-based digital audio formats, but IP- based in data packets. The addressability of the data packets breaks the traditional point-to-point connection between just two audio devices that are directly connected to each other, and it becomes possible to connect any number of audio devices via a network distributor (switch) and send audio signals to any desired receiver in the network received to send as well as from any device on the network audio data (real audio equipment Ver network ung). The distance between two Dante devices or one Dante device and the next network distributor can always be up to approx. 80 m. Such a format is therefore particularly interesting for audio installations in stadiums or universities with many rooms or buildings that are relatively far apart. The same applies to television studios with central server installations. Since Dante is based on a conventional, very widespread and inexpensive network infrastructure, network cabling that is already in use for data networks can in many cases simply be used for Dante audio networks, which saves considerable costs and eliminates structural changes due to the relocation of specific audio cables.
The direct sending of Dante audio data via WLAN is currently not supported, particularly due to excessive transmission latencies and the risk of signal interruptions. The transmission of Dante via dLAN or PowerLAN currently fails due to the same and other specific restrictions . A possible workaround for the wireless digital transmission of audio information (but not in the Dante network format!) Is the XIRIUM system from Neutrik .
Applications
In a Layer 3 network, high-resolution audio data packets and other data (Internet, file transfers) can be transmitted simultaneously without any problems. In order to guarantee the interference-free transmission of the audio data, configurable gigabit network distributors are to be used, which offer the possibility of prioritizing audio data over the rest of the network data traffic ( Quality of Service , QoS). It should also be possible to switch off the energy-saving functions of the switches ( Green Ethernet ) so as not to slow down audio data. Cat 5e and Cat 6 network cables are suitable.
Numerous mixing consoles, analog-digital converters, multi-track recorders, remote-controllable preamplifiers, microphones, stage monitors as well as PA amplifiers and PA speaker systems from various manufacturers with Dante interfaces are now available. Dante devices have a mechanism that allows them to automatically recognize each other (auto discovery) as soon as they are connected to other Dante devices directly or via a network distributor. Only a few seconds after connecting to the network, all audio inputs and outputs of the newly connected device are known to all other devices and audio signals can be routed between all devices as required. In addition, the required word clock synchronization of all devices in the network is regulated automatically in most cases. Connections to devices that are connected to the network but not currently switched on are ignored, but the connections are automatically restored immediately after switching on again.
Compared to other digital transmission formats, the minimum achievable latency time in relation to the number of devices in the network is very small (depending on the size and complexity of the network, between 0.15 ms and 5 ms). This makes Dante networks suitable for real-time audio applications such as live sound reinforcement.
Dante controller
The routing of audio signals between any audio devices in the network takes place without special routing hardware. Instead, Audinate offers the free routing software Dante Controller . In addition to routing with a simple click of the mouse, it also offers some network and device status and diagnostic functions. Signals can be split almost arbitrarily to different receivers. In contrast, all analog and traditional digital audio connections require the use of hardware to route signals between more than two devices. With very little effort, Dante Controller creates virtual routings of a complexity that cannot be implemented or financed with any other means. These routing configurations can also be saved and retrieved. The physical distribution / routing of signals by means of special splitter hardware, which has been common for a long time, has been replaced by virtual, logical, purely software-based routing.
Dante Virtual Soundcard
In addition to audio devices that are factory-equipped with a Dante-enabled audio connection, the LAN interfaces that are usually available on computers can also be made Dante-enabled using a chargeable, cross-platform driver from Audinate (Dante Virtual Soundcard, DVS) . As soon as the driver is started, the computer's LAN interface functions as a Dante audio interface with up to 64 inputs and outputs. Connected to other Dante audio devices, a computer can thus be used as a multi-channel player or recorder. Since the built-in LAN interface was usually not designed and optimized in-house for such a purpose, the achievable latencies are significantly higher than with dedicated Dante devices (DVS: between 1 ms and more than 20 ms). In the case of applications in which the computer does not have to carry out any real-time processing on the audio (such as in the case of multi-channel recording), this restriction is irrelevant. The network interface of a computer can still be used for data transfers while DVS is in use.
Device remote control
Many Dante-capable audio devices use their Dante interface in parallel to transmit the audio signals in order to be remotely controlled. The parameters of preamplifiers (gain, pad, phantom power, etc.) can be remotely controlled from a control computer via the network cable connection, which allows these devices to be positioned as stage boxes on stage at live events . This eliminates the need to initially route the analogue source signals from the stage to the FOH mixer via unwieldy multicores . Instead, the analog-to-digital conversion as well as the feeding of all signals into the Dante network takes place on the stage and cheaper and much easier-to-use network cables are sufficient for forwarding. If required, all signals can be distributed very easily via a network switch to a monitor desk placed next to the stage, the FOH sound system and - for multi-channel recording - to a computer with a Dante Virtual Soundcard .
redundancy
Dante is designed so that devices with two Dante connections (primary and secondary port) can feed two parallel ( redundant ) networks to avoid signal failures . This requires double cabling and possibly double the number of network distributors. Compared to conventional redundancy methods, the effort is much lower. In the event of an interruption in one audio data stream, the system automatically and without interruption switches to the functioning data stream of the emergency network. The user can choose to do without the redundancy function of the two Dante ports and use the connections for simple chaining of Dante devices.
Flows
While the number of devices in a Dante network is in principle not limited, there is still a limitation, since the Dante boards manufactured by Audinate and sold to licensees can only transmit a limited number of flows . The data packets from 4 audio channels of a device are combined in a flow before the packets are sent. Different devices have an individually defined number of flows (max. 32). If an attempt is made to route more than the maximum possible number of flows from a certain device via the Dante controller (to too many receiving devices at the same time), no further channel can be routed from this device without further ado.
Unicast / Multicast
Audio data packets can be sent as unicast or multicast packets in a Dante network . Normally, unicast is used automatically. In unicast mode, the sending device only sends audio packets to recipients who actually requested these packets, but it also has to send multiple requested channels (from different recipients) several times, which increases network traffic. In the case of multicast, the sending device sends the data packets of the requested channels only once. In an appropriately configured network with multicast-capable switches, this infrastructure ensures that the stream arrives via the shortest route to the recipients who have requested this signal. All other network participants do not receive this stream. This means that multicast is considerably more efficient with several recipients, but requires a suitable infrastructure. In the case of a switch that is not capable of multicast or is not configured for multicast, multicast is usually treated like a broadcast, i.e. all ports are flooded. In such a network, multicast has more disadvantages than advantages over unicast.
Dante Via
Dante Via is another Audinate virtual audio driver that makes it possible to route audio between computers without the presence of a Dante device. Without Dante Via , at least one Dante device must be present in an audio device network, which in particular ensures the required word clock synchronization of all devices. The Dante Via driver can take over this function . With Dante Via it is also possible to integrate other audio devices or interfaces connected to a computer into the Dante audio network. This means that both a built-in laptop microphone and the channels of an audio device connected via USB or Firewire can be fed into the Dante network.
history
Australian developer Aidan Williams, who previously worked at Motorola , spent three years developing the foundations of Dante with a team at NICTA in Sydney . The project received funding from the government. In 2006 he founded the company Audinate and commercialized the product. He is currently (2016) the CTO of Audinate. The company is the first successful spin-off of a NICTA research project.
More than 300 companies from the audio industry have licensed Dante , and now well over 700 products are available for a wide variety of applications. The products of all manufacturers are compatible with each other, so that users can combine products with one another as desired and are not forced to use products from a single manufacturer.
Web links
Individual evidence
- ↑ Overview on the subject of Dante (Eng.)
- ↑ Multimedia campus installation Goethe University Frankfurt a. M.
- ↑ current manufacturer and product overview
- ↑ DANTE licensee . Retrieved April 16, 2016.
- ↑ https://www.audinate.com/about/news-activity/press/dante-rockets-past-300-manufacturers-and-750-enabled-products