What are the benefits of using ASIO (Audio Interface)?

Question

I just discovered ASIO (thanks to an answer here on SD).

I've read a bit about it and want to make sure I understand the benefits.

It looks like the chief benefits are:

1. Reduced Latency
2. Ability to combine several sound cards or external USB Mics to a single program (e.g., Audacity, Goldwave, etc.) . (So if I was recording audio from two such mics :maybe vocals on one and instruments on another, I suppose.

Any other benefits?

What is the benefit of the reduced latency for recording audio (other than reducing potential syncing problems between the multiple mics)?

Answer 1

ASIO is a protocol for communicating with a sound card.

Normally the operating system handles the audio path, which results in passing through many layers before actually hitting the soundcard. Each layer adds latency.

ASIO allows a software application to talk directly to the soundcard, avoiding all intermediate layers. This is how you achieve reduced latency.

Also, by accessing the hardware directly. The audio does not pass through the Operating System's audio mixer, potentially avoiding any sample/bit rate conversion losses. This is the higher fidelity gains.

Just to note. Using an ASIO driver essentially allows the software to utilize the hardware to the maximum of its potential. If the actual soundcard has poor AD/DA conversion, or only supports lower bit depth (16bit only?), one cannot exceed this physical limitation.

Hence, using something like asio4all will allow you to cut down on latency. It will not improve the conversion taking place on the hardware.

Answer 2

ASIO's main advantage is quicker, easier routing of audio inputs and outputs. It allows you to easily wire inputs and outputs directly in to your audio recording software and eliminates a lot of the overhead normally associated with consumer sound cards.

The benefit of reduced latency comes when dealing with multiple channels of audio. Even if you only have a single input and output, if you tried to do any layering, you need your input and output to match up. If it took 100ms from when your computer started trying to play audio to when your speakers made it and another 100ms for your input to be processed and recorded, if you were to listen to say, an audio track and sing to it, your singing would be recorded 200ms after the computer tried to play it. This would result in your vocals not being timed correctly in your track compared to the music.