Home Office – SBRI Digital Forensics Challenge – MADCAT WHISKERS wins both Phase 1 and 2 funding


In August 2015, the Home Office challenged UK businesses and academia to come up with ways to improve the speed, efficiency, and effectiveness of recovering and analysing data from digital devices seized from suspects under investigation. This was conducted through the SBRI (Small Business Research Initiative) competition run by Innovate UK on behalf of the UK Home Office.

Oxford Wave Research participated in this highly competitive call and was one of the eleven businesses and universities selected for Phase 1 funding. OWR successfully completed Phase 1 and was one of the five companies from the Phase 1 participants that was further selected for Phase 2 of the competition.

The funding for this challenge helped us to enhance our existing MADCAT algorithms for extracting and comparing fingerprints in audio or video, to using GPU-acceleration to perform millions of comparisons in seconds and hundreds of millions in tens of minutes. We further developed this into a client and GPU-server based architecture capable of sharing and comparing audio fingerprint data for high speed cross-case comparisons.

Our R&D team attended the SBRI Digital Forensics Showcase Event in London on Friday, 3rd November 2017, to showcase MADCAT WHISKERS, our high-speed media fingerprint comparison solution that was developed as part of this challenge.

The showcase event provided an excellent opportunity to demonstrate how we used the Home Office funding and input to develop and improve a product which is now being used in certain UK law enforcement agencies with tremendous results.

To find out more about our software please contact us directly at info@oxfordwaveresearch.com

TAPAS Kick Off

Oscar and Nikki are excited to be attending the kick off of the TAPAS project in Switzerland. OWR are privileged to be industry partners on this exciting new project which brings together some of the best academic and industry leaders in Europe.


SpectrumView and SpectrumView + release V2.2 

As SpectrumView approaches 400,000 downloads, we continue to be intrigued and amazed by the diverse applications our users have found for this real-time frequency analyser app, such as detecting ultrasonic high frequency noises, school physics experiments, detecting bats, recording wildlife and birdsong, to mention just a few.

We have had some great feedback and many feature requests which have been prioritised and we are pleased to roll out with the release of SpectrumView V2.2.

This version is fully compatible with iOS11 and includes the following new features;

  • Data-snapshot capability – ability to save the spectrogram or spectrum analyser data you are exporting and view this in Excel, Matlab, or any other software that supports CSV formatted files.
  • Over the air downloads – ability to download recordings, data snapshots and images from SpectrumView to other devices using a web-interface over WiFi
  • Simultaneous Playback and visualisation – ability to play, pause and scrub through recordings while viewing the spectrograms
  • Share files with other apps – share your SpectrumView recordings with other apps on your device such as ‘Whatsapp’, ‘Facebook Messenger’ etc or ‘Airdrop’ them to your friends.

You’ll be pleased to know that the most used and favourite features from SpectrumView 2.0 are still available and have been enhanced further. These include;

  • Saving location info of the recording – Never forget where a clip was recorded
  • Allow Peak hold – Enables you to freeze and hold the peaks of the frequency envelope (three holds in three colours) in Spectrum Analyser mode
  • Logarithmic scale – Enables toggling between logarithmic and linear scale modes in Spectrogram and Spectrum Analyser view
  • Auto lock disabled during playback and recordings – Prevents the device from automatically locking (if enabled), for uninterrupted recording/playback
  • Open from other applications – Allows audio clips to be opened in SpectrumView from other applications (e.g. Mail, Facebook, and iCloud Drive)
  • In-App Purchases – Only purchase the features you need from SpectrumView Free

‘SMART Subtract’ – New Product Announcement!


We are proud to release an innovative new audio enhancement product called ‘SMART Subtract’ that can be used to easily reduce or remove interfering music, television or even speech from recordings using reference recordings of these interferences. This product combines the capabilities of three leading audio processing companies namely Salient Sciences (previously Digital Audio Corporation),  Acon Digital (creators of Acoustica) and Oxford Wave Research.

The SMARTSubtract Software can be used to quickly and easily subtract music, television or speech from an audio recording if you have a reference song, TV show, or speech recording that you want to remove from the recording and preserves the underlying speech, making it more intelligible.

SMARTSubtract  uses  MADCAT(TM) audio fingerprinting technology from Oxford Wave Research to precisely align the file with the interfering noise in it, and the reference audio file containing the interfering sounds. It then performs high quality ‘reference cancellation’ using one of the two powerful VST plugins supplied by leading audio processing software companies.  The resulting file can be easily export into WAV format.

For more information please contact us – sales@oxfordwaveresearch.com

Or visit the webpage http://oxfordwaveresearch.com/products/smartsubtract/

'We're delighted that our time-domain Reference Canceller filter that is part of the CARDINAL MiniLab Suite can now be used within SMARTSubtract, making use of MADCAT's audio alignment technology to provide a practical reference noise cancellation tool for our audio and video forensics users.'

Donald Tunstall, General Manager, Salient Sciences

'The Acon Digital Reference Canceller combined with SmartSubtract's audio synchronisation capabilities allows for high quality frequency domain reference cancellation of interfering noises that is both powerful easy to use.'

Stian Aagedal, CEO, Acon Digital

How fast can a fidget spinner video about frequency analysis go?

Last Friday, when we were just winding up for the end of the week we started getting a large number of messages on our website chat app, and also a a huge spike in the number of hits on our website (1283.05%), and in particular from our audio frequency spectrum analyser called Spectrumview.

Had we been hacked? Had some rivettingly interestingly pictures of the intimate details of audio analysis been unwittingly released on our webpage? Thankfully not. Our app had been used by the extremely talented stand-up comedian and maths communicator Matt Parker (@standupmathsto measure how fast he could get a fidget spinner to go. This video had hundreds of thousands of views each day, and just under 300,000 at the time of writing.

This is a brilliant video that shows how to use Spectrumview to calculate the frequency and thereby the speed of the tips of the fidget spinner. We are delighted to see such a weird and wonderful use for our little app.


Before you ask, we don’t have an Android version. There are no plans to have one just yet, but we may be persuaded. If you ask nicely.

The Linguistic Data Consortium and Oxford Wave Research announce a new collaboration

The Linguistic Data Consortium (LDC), USA and Oxford Wave Research (UK) are proud to announce a new collaboration. Oxford Wave Research (OWR) is an audio and speech R&D company based in Oxford, UK that works on audio processing and speaker diarization and recognition. This collaboration encompasses the use of LDC’s speech corpora and OWR’s audio fingerprinting, speaker diarization and recognition software.

In particular, LDC will use OWR’s audio fingerprinting technology as part of the MADCAT software (Multimedia Audio Duplication & Content Analysis Tool) to find repeated content in broadcast data including audio signatures that mark program boundaries.

The OWR Research Director, Dr Anil Alexander says,

“The OWR team is really looking forward to working with LDC on this exciting real-world application of audio fingerprinting of short utterances.”


LDC Executive Director Chris Cieri says

“The Consortium continually looks for new ways to integrate speech technology into data collection and annotation processes to improve speed, scale and quality while avoiding bias. We are excited by the increased capability that OWR tools offer.”