Choosing an Audio Interface - Project Requirements
The implications of different types of audio and video project for your choice of audio computer interface.
The computer audio interface (AKA 'sound card') is a vital component of the audio path, and arguably the single most important hardware choice when configuring an audio digitisation workstation.
Its primary role is to convert an analogue audio signal into its digital counterpart, and to present the computer with a stream of data which a software codec can then encode as a digital audio file. Similarly, it decodes the digital output of your workstation and converts this into an analogue waveform, playable through speakers and headphones, and so is equally critical to the monitoring process.
In simple practical terms, this means that you plug your audio source(s) into its inputs for recording, and your headphones and speakers into its outputs to listen to the output of your digital audio workstation (DAW).
In addition to this core role, the audio interface may also be called upon to
- Amplify the signals from your microphones and other input devices
- Split the output(s) for monitoring and/or routing to other devices
- Allow you to monitor the input signal you are recording in real time
- Offer additional DSP processing capability and plug-in effects
- Convert analogue signal(s) into different digital formats (S/PDIF, AES, ADAT etc)
Most standard PCs include a soundcard of some description, but unless designed with audio/video production in mind, these interfaces are rarely of a sufficient quality for critical tasks. Many external audio interfaces are available which connect by USB or Firewire to your computer, and here we examine which features you should look for to ensure their suitability for your project(s). A few internally mounted PCI solutions are still available, but external interfaces offer superior portability between systems, and their performance is comparable so they will be our focus.
Rather than try to cover the various features of the huge range of interfaces available today, we will examine the core functions of the interface and explain the terms used to differentiate their various capabilities. We also offer a comparison table - Choosing an Audio Interface - Unit Specifications - which lists the key features of a few notable models, but this is by no means exhaustive.
There are also some proprietary systems such as ProTools and Sadie, which require specific own-brand hardware/software combinations. These systems may not suit an education environment where transparency and long-term accessibility are required, due to their reliance on the policy and longevity of a small group of companies. The proprietary nature and limited compatibility of the data they produce - despite its often high quality - cannot be considered to conform to open standards-based practice in even a limited sense.
They do, however, offer outstanding system integration, quality and reliability, and for these reasons are used widely in commercial media facilities, including the film and broadcast industries. If your aims include the training of staff and students for work in these environments then you may want to look at the de facto preferences of your target market, and factor these into your decision.
Your project objectives will determine which primary qualities you should be looking for in an interface. If you are recording educational materials on location for VLE distribution, you will have different needs from an archivist digitising a library of legacy analogue material. Though some hardware will be ideal for particular projects, good quality equipment can potentially suit many uses, and interfaces are by no means limited to specific tasks.
The digitisation and archiving of legacy audio materials places stringent demands on an AD (Analogue to Digital) converter, as the intention should be to create as honest and uncoloured a copy of the original as possible. Any artefacts introduced in the conversion process are undesirable, and all dynamic and harmonic content of the original should - as far as possible - be entirely and honestly preserved in its digital replica.
Additional features such as microphone preamplification are not required - though some archival standard units, such as the Prism Orpheus and RME Fireface range, offer similarly high specification mic amps. The prime motivator in your choice should be conversion quality, and additional features considered a bonus.
The most important job of the archiving interface is analogue to digital (AD) conversion. For use in archiving, a converter should offer at least a 96kHz sampling rate (use of higher rates than this shows marginal subjective improvement, and is discretionary), and 24-bit conversion. 16-bit conversion (the standard for CDs) is not considered sufficient for archiving purposes, as its maximum possible dynamic range of 96dB is too limited to capture dynamic nuances in sufficient detail.
To produce archive audio objects suitable for restoration, remastering etc, the first stage of the process must be the capture of a reference-quality 'warts and all' digital copy. This first generation copy should be archived at full quality, in the event that superior restoration tools to those currently available are developed in future, or that the true nature and character of the analogue original (imperfect as it may be) are of interest to researchers. Restoration and processing should be applied to digital surrogate copies only.
For further guidelines and recommendations for audio interfaces suitable for archiving, see our companion guide to Choosing an Audio Interface - Technical Considerations.
[Suggested units: RME Fireface800/400, Prism Orpheus]
Recording spoken word educational materials - Oral History, Podcasting, etc
While the quality of AD/DA conversion is still important when making recordings of speakers and other 'live' events, it is not the sine qua non that it is to archivists. For the job of live recording your interface will be called on to amplify microphones - an important process - and provide performers and engineers with mixes to monitor the signal(s) as they are recorded.
If you wish to use studio condenser microphones (recommended where possible for spoken word recording) then your interface should have a number of phantom powered microphone inputs corresponding to the maximum number you envisage using simultaneously when recording. Signal-to-noise ratio (SNR) will also be of importance, given the significant levels of amplification these mics require, and though no 'generic' desired minimum value can be given, a comparison of SNR performance can give an idea of the relative performance of different units.
You should have enough headphone outputs (or line level outputs plus external headphone amplifiers) as you intend to provide separate monitor mixes, and direct monitoring (see below) will enable performers to hear themselves without any system latency, and choose a desired mix between their own voice level and that of any backing track(s) over which they are being dubbed.
[Suggested units: Apogee ONE, Duet or Ensemble, Tascam US-100, Edirol UA-101 or UA-25EX]
[Note - The built-in audio connections on current Macintosh computers are of sufficient quality and latency for podcasting use (operating at up to a respectable 24-bit 96kHz), though direct monitoring, phantom powered mic preamps etc are not available.]
Working with video
When digitising archive moving image resources, all audio capture will be handled by the video capture card. Similarly, when working with live video (lectures, interviews etc) audio may be captured on the camera to avoid any later issues with synchronisation between picture and audio (though this is not generally recommended).
In either scenario a dedicated audio interface is therefore not a requirement, though one may be necessary to monitor video files' audio content accurately, especially if surround sound is to be implemented - a topic covered in detail below.
An interface will of course be required if voice-over is to be dubbed onto the final video. In this scenario the same guidelines apply as for spoken-word recording, as above.
Music recording and production
Recording, editing, mixing and producing music brings yet another set of requirements to the digital audio system. Music recording is often 'multitrack', where several performers or microphones are recorded simultaneously onto separate tracks on the computer, and/or in successive passes, or 'takes', and then edited and mixed together to a stereo or surround master.
Again when choosing an interface for musical work you should ensure that you have sufficient inputs and outputs to handle your envisaged multitracking requirements (though you may be content to submix signals before digitisation - see below). Consider how many microphones and other sources you will want to connect simultaneously, and also how many monitor mixes and headphone outputs you may be required to supply (as well as your own!).
The clinical accuracy of an archiving system is not necessarily required (though as stated, units suitable for archiving are usually very high specification, and suitable for most audio tasks) and some degree of colouration can be acceptable or even occasionally desirable. For example, interfaces by Apogee are sometimes considered to add some subtle 'warmth' to the digital signal. This is not to say that colouration is desirable - far from it - but it is worth researching user experiences of any interface you are considering.
Signal-to-noise performance and microphone amplification quality are of great significance, if no additional external microphone pre-amplification is to be used. Again, the subjective qualities of their 'sound' are best judged by researching user feedback and reading reviews. Many suppliers and manufacturers (e.g. Apogee UK - Sonic Distribution) offer loan services to educators to assess in situ the performance of different units.
The other key role of your audio interface - as well as digitisation - is the inverse process of converting digital files (presented as a binary data stream) into analogue sound, for listening and critical monitoring purposes.
When applying any form of processing or edit, it is vital to have accurate representation of its effect in your monitor signal. Also, when making decisions about file types and compression techniques the sometimes subtle differences between the alternatives will be more honestly portrayed by a reference monitoring system attached to a high-quality DA converter.
So, even if it is not being used as the primary interface for digitisation but merely for monitoring duties, the interface in your edit suite should still offer sample and bit rates at least as high as the highest quality files with which you work, and if you are using digital restoration tools a minimum of 24bit 96kHz DA conversion should be demanded, combined with reference monitor playback.
Expanding the interface's capabilities with a mixer
If you have an existing interface of sufficient quality for your project, but which does not offer the connectivity and routing options you require (e.g. Macintosh built-in audio connections), you may wish to add these connections with a small analogue mixer.
Even if you don't have an existing interface, many new interfaces are now available which are built into a small mixing desk format. These usually then offers microphone and instrument inputs, and allows various routing and monitoring options, so can be a good solution for podcast or simple music production, though consider that if you wish to upgrade either the mixer or the audio interface, the whole unit will need replacing.
Two Alesis mixer interfaces - Multimix 4 USB and Multimix 8 Firewire
Using a mixer to submix signals to a stereo pair before recording will of course mean that you will not be able to alter the balance between them later, as you would with a multitrack recording, so be aware of this limitation when deciding on this type of system. Simple 2-track separation is however easily achieved by panning inputs hard left or hard right and feeding these to the respective inputs of your interface. Signals can then be placed within the stereo field later, when mixing.
Another significant factor in your choice may be system portability. Though many systems will be fixed facilities built around a desktop computer, increasingly professional quality systems can now be built around laptops, and offer the ability to record on location without the need for a dedicated field recorder.
The same guidelines apply as before when weighing up the sound quality and audio connectivity of the various interfaces - including having the relevant number of mic inputs, headphone out etc - but if use on location with a laptop is also desired then ensure that the unit you choose can be powered by the laptop if required (USB or Firewire bus powering), and of course sports the relevant connector to interface with your laptop. Some laptops now only offer USB2 or Firewire 800, so read system requirements carefully.
Further details on the technical specifications relating to audio interfaces are included in our companion advice documents Choosing an Audio Interface - Technical Considerations and Choosing an Audio Interface - Unit Specifications