Welcome to
Vox Protect
The software that protects the voice of witnesses.
A true innovation
VOX PROTECT is a robust software for witness protection at any level.
Maximum security
At level 2 (high protection level), the effect created is irreversible, making speech recognition impossible.
Maintain authenticity
Preserves the original intonation of the witness, leaving the emotion of the testimony intact.
VOX PROTECT
A unique technology
Today with VOX PROTECT, SWEALINK has developed an innovative and irreversible solution.
It includes two non-reversible linear transformations in the software and three levels of random modulations making the effect irreversible, even by a speech recognition analysis.
F.A.Q.
Frequently Asked Questions
Why are the witnesses' voices not protected by classic software?
Until now, all the proposed solutions do not protect the identification of witnesses.
For example, the most used solution, the “pitch” effect, is reversible. Moreover, this use has become obsolete and dangerous.
The same is true for all software that incorporates the effect reversal function in the digital era.
Do other people talk about the problems of witness voice protection?
Yes, this is a recurring topic that has been gaining importance lately with the issue of data privacy.
We invite you to follow these links:
Lately, the scientific community has become aware of this security flaw:
https://llgeek.github.io/pdf/VoiceMask.pdf
Some experts speak about it :
https://blog.govnet.co.uk/do-voice-modification-technologies-truly-protect-victims-
identities
https://www.pro-tools-expert.com/home-page/2016/1/13/help-please-how-to-disguise-
an-interviewees-voice-to-protect-their-identity
paragraphe : « les limites du pitch »
https://theconversation.com/ils-excellent-dans-la-reconnaissance-des-visages-et-des-
voix-ce-sont-les-super-reconnaisseurs-157955
https://www.pactenovation.fr/index.php/2021/11/22/la-reconnaissance-vocale-de-
locuteurs-comment-et-pourquoi/
https://app.livestorm.co/hub-france-ia/cycle-voix-theme-3
Are witness voices protected ?
No : the software (plug-in) used to fake voices is entirely reversible.
Only VOX PROTECT offers a solution dedicated to this protection in plug-in format.
Why does VOX PROTECT protect more than the other effects used ?
VOX PROTECT has been created for this specific use.
It required 4 years of reflection, research and development.
Therefore, It is the only software (plug-in), to date, dedicated solely to the protection of witness’ voices.
Its design protects the witness’ identity, the user of the software, and the software itself.
For more details go to “Product”
Why is it a problem if a witness' voice is not well protected ?
In the digital age, the process of reversing a witness’ voice has become accessible to everyone.
-It is pretty easy to find all the interviews of witnesses broadcasted on TV, radio, and all media in digital quality online.
-It is easy to get a pitch on the internet, under a free license, and ready to use. The digital quality allows keeping an identical quality to the original by applying the reverse process and thus finding the voice of the witness in a perfect way.
With the digital air, the issue of voice recognition is essential for many services. Therefore, It has become elementary to identify a person from an audio recording.
If you don’t apply for good witness voice protection, you may be putting your witness at risk.
Is there an obligation to protect the voice of witnesses ?
It is a fact that today voice recognition has become a significant issue.
For example:
In France, the famous case of the Jérôme Cahuzac affair recognized by the BATVOX software clearly shows the critical issue of voice protection.
We specify here that several legal obligations in the world cover the protection of witnesses as an indication and are not exhaustive:
-In the United States, in the 1970s, the anonymization of witnesses became mandatory as a result of the program to dismantle mafia-type criminal organizations
https://www.unodc.org/documents/organized-crime/09-80620_F_ebook.pdf
-In 2001, France strengthened witness protection with the law N° 2001-539, article 706-61.
http://prison.eu.org/livret-4-titre-21-de-la-protection
The full set of global recommendations can be found in the 2009 UN report, pages 8-16:
https://www.unodc.org/documents/organized-crime/09-80620_F_ebook.pdf
How does VOX PROTECT work ?
We have defined two levels of protection in the software (plug-in):
-The first level of protection, “Low level of protection” corresponds to weak security.
This protection is more reliable than a common effect such as pitch. Indeed, the result found cannot be reversed by the current market effects. However, it remains reversible and can be analyzed by more “specialized” services.
We advise you to use this level of protection for witnesses, for example, who testify on facts without significant risk of recognition or for the actors’ voice replacing the voice of the witness.
-The second level of protection, “High level of protection” corresponds to a strong
level of protection that protects, for example, a witness who denounces a drug trafficking or a terrorism case.
This level will protect in all situations, including voice recognition. The downside is that your voice will become less understandable and require subtitles in some cases.
Does the use of VOX PROTECT require an authorization key ?
Yes, we have an agreement with PACE ANTI-PIRACY with their I-loK key system.
It would be best to have an I-lok key to load the VOX PROTECT authorization onto it.
If you do not have this key system. You must acquire it on the site:
https://www.ilok.com
by creating a user account specifically for the use of this authorization.
Without this key, you will not be able to use VOX PROTECT.
Do you record and transmit the audio of interviews and witnesses ?
Now, the VOX PROTECT plug-in, working on your sound or video editor, cannot send and transmit the original audio information automatically. In addition, the software does not send any information about the protection criteria chosen.
Therefore, the responsibility for preserving sensitive data remains with the software user.
What are the recommended "buffer" preferences ?
We recommend a computer with a minimum of 32 GB RAM.
We recommend the following preference criteria:
-For Protools, Set Up/Playback Engine/HW Buffer Size :1024 samples /Cache Size: Normal
-For Audition, Preferences/Audio Hardware/SW Buffer Size and Available Ram Memory for AU 20 GB.
-For Première Pro, Preferences/Audio Material/S.E. Buffer Size and Available Ram for AU 20 GB.
-For Davinci Resolve, Preferences/Video and audio I/0 : auto or 1024 minimum.
You want to ask us something ?
Click Here to contact us
This program has received the support of the CNC, RIAM
and the BPI.
