How Fast Can Audio Matching Identify Audio Content? The Variables Explained

by Intrasonics Team

Speaker concept picture id1069348184

In a time before smartphones and high-speed internet, this scenario will likely be something you are familiar with; you are out and about, maybe at the pub or club or at the gym, and you hear a song come on through the sound system, but are not sure what it is called.

The only way you could find out back then would be to ask a friend or someone that worked wherever you were. Or you had to suffer without knowing.

That was then. But it’s now a thing of the past, thanks to high-speed internet and smartphone technology and a process called audio matching.

Using a technique called audio watermarking also achieves a similar result.

At Intrasonics, we provide bespoke solutions for both methods of matching and watermarking audio.

How Does Audio Matching and Watermarking Identify Audio?

Audio matching uses an audio file database to match sounds.

While watermarking involves encoding a unique identifier into audio, helping companies and owners of the audio identify the signal and the source that shared it.

Although it is known more as a form of copyright control, audio watermarking also has the side benefit of identifying audio.

You can read more here about how audio watermarking and matching works.

How Audio Matching is Used

One of the uses for audio matching is in the scenario we outlined at the start.

Besides its obvious use for the above, audio matching is also used to track content exposure for the research and advertising industry.

The technology allows companies to receive data on what consumers are listening to, whether on TV, radio or the internet.

This information is captured via the users mobile phone, which can hear the audio being listened.

With audio matching software like Shazam and other apps, you can quickly find out exactly what the name of the audio is that you are listening to in that bar, restaurant, or gym.

Even if you haven’t personally used that kind of software before, you will know someone who has.

Contact our team today for a free demo and find out how our technology can help your business.

How Quickly Can Audio Matching Technology Identify Content?

Back to the original question of how long does it take for audio matching to identify audio content?

The short answer is anywhere from a few seconds to around 20 seconds.

The longer answer is it depends on the type of audio matching software you use. The environment and background noise also affects audio watermark detection

Let’s explore the different options. To give you an idea of how quickly they can get to work and provide you with results, we will look at a few of the popular ones.

Smart speaker with a smartphone picture id1200139132

AWT2 Watermark Listener

An android mobile phone app, AWT2 Watermark Listener, is an accompaniment to AWT2 or Audio Watermarking Tools 2.

This app is designed to help you detect any inaudible digital codes or watermarks in content using the AWT2 watermarking tech.

The best way to use this application is to hold your smartphone as close to the device you are listening to the sound through, whether it’s your TV, music system or external speakers.

The search process takes between 15 and 20 seconds, and if it finds a match, it will display the information on your smartphone screen through the app.

Shazam

Shazam is often cited as one of the first audio matching and audio recognition applications.

To find an accurate match for the audio you are listening to or want to identify, you need to capture 20 seconds of the song.

It can be just about any part of the song, whether it’s the chorus, verse or intro, and Shazam will create a fingerprint of the sample.

It will use that sample to quickly look through its extensive database music recognition algorithm to find an exact match.

It usually does this in seconds.

SoundHound

An interesting alternative to the above is SoundHound.

SoundHound is popular because it is more than just an audio recognition app; it also offers speech recognition capabilities.

How this differs from Shazam is that it can actually recognise spoken and sung lyrics and even humming.

Great for those of us who can never remember lyrics, only tunes!

Again, this takes only a few seconds to get results.

Intrasonics

At Intrasonics, we provide businesses with a bespoke audio matching solution to find the specific audio content you’re looking for.

In terms of detection speed, we’re among the fastest since our audio matching doesn’t require an internet connection to identify your audio.

Here are some other differences vs the above audio matching solutions:

  • Can provide bespoke solutions to search for exactly and only the content you’re looking for (e.g. Shazam and Soundhound are helpful only to end-users but not to businesses).
  • We can provide on-device audio matching, i.e. matching that doesn’t require internet and therefore is entirely privacy-centric and fast.
  • Our audio matching naturally combines with our watermarking to provide a solution that covers all bases. This is optional.
  • The very low false-positive ratio for professional solutions: an audio match is a guarantee and not a guess

Some Issues to Bear in Mind

On the subject of how quickly it takes audio matching to work, you need to consider what can prevent audio matching from working or slow down the process.

As is noted for the programs mentioned above and others like them:

  • You must hold your device close to the sound source.
  • Any interference is likely to cause issues and prevent it from working efficiently enough. It could even result in inaccurate search results.
  • There needs to be an existing file somewhere out there in a database for it to work correctly.

Most audio matching algorithms rely on a database of existing sound files, for example, Shazam and other similar software.

In this case, if you wrote a song and had never recorded it or made it available on any streaming service and there was no physical copy out there, using audio matching is unlikely to find any accurate results.

Intrasonics’ audio matching algorithm’s benefit is that it’s a bespoke solution for business and only searches for the specific content you want to find.

Summary

So, now you know how fast audio matching algorithm’s can identify content. Anywhere from a few seconds to around 20 seconds.

But, you now also know the variables that can affect speed of audio detection.

At Intrasonics, we pride ourselves on our fast and robust audio matching and watermarking solutions.

Contact our team today for a free demo and find out how our technology can help your business.


< Previous Page
Intrasonics Ltd

Bateman House
82-88 Hills Road
Cambridge
United Kingdom
CB2 1LQ

Get In Touch

+44 (0)1223 927 070
Make an enquiry