Steganography and “Silent Subliminals”

Steganography” from the Greek steganós (concealed) and graphia (writing), is the practice of hiding information within another message or physical object. It could be described as a form of “hiding in plain sight”.

There are several different stenanogrphic techniques relating to audio. This article, and the associated plug-in uses one specific technique called Amplitude Modulation (AM) encoding. This is the same process as is used by AM radio, except that rather than using radio waves to carry the audio, very high frequency audio is used instead.

AM Encoding is the most popular technique for creating “Silent Subliminal” recordings, which advocates claim can bring about positive shifts in one’s mindset. The technique was largely based on the 1989 US Patent US5159703A by Oliver M. Lowery for a “Silent subliminal presentation system”.

O. M. Lowery’s invention was based on the belief that speech messages that are AM encoded with a high audio frequency carrier, may be “decoded” by the listener’s subconscious mind, while remaining inaudible and indecipherable to the conscious mind. Despite the lack of empirical evidence to support his claims, the technique has become very popular in New-Age Wellness circles.


The author of this article and associate plug-in makes no claims whatsoever regarding the effectiveness or safety of “Silent Subliminals”.

The purpose of this article and the associated plug-in is to provide some background information about the steganographic technique of high frequency audio encoding of speech messages, and to provide an Audacity plug-in that allows experimentation with the technique.

How it works:

The idea is to hide a recorded speech message via AM encoding with a very high audio frequency carrier wave. This shifts the frequency range above 16 kHz which is largely inaudible to most adults. (People with very good high frequency hearing may still be able to detect a pulsating high frequency whistle, though it will be incomprehensible).

Sound Quality Limitations:

The format for audio CDs uses a sample rate of 44.1 kHz, which has an audio bandwidth up to approximately 20 kHz. This was considered to be the minimum requirement for very high quality audio. In order for the encoded message to be essentially silent, frequencies below about 16 kHz are unusable for our purposes. Also, given that audio equipment is usually designed for frequencies below about 20 kHz, our available bandwidth is limited to a 4 kHz range between 16 to 20 kHz. Due to practical / technical issues, the usable range becomes limited further to a usable range of around 3 kHz.

The effect of a 3 kHz bandwidth is that the sound quality is limited to be about the same as an old-school landline telephone.


This plug-in is entirely free for personal use in any manner, including commercial use, but is subject to the terms and conditions of GPL version 2 (or later).

If you wish to share this plug-in with others, please link to this post. This will ensure that everyone has access to the latest version.

Please DO NOT hotlink directly to the plug-in, as doing so is effectively stealing my (paid for) Internet bandwidth.

If you must share a downloaded versions of this plug-in, ensure that you include this copyright notice in full:

The plug-in:

Installation Instructions:

This plug-in may be used in Audacity version 2.3 or later.

Install into Audacity in the same way as any other Nyquist Plug-in. Instructions for the current version of Audacity can be found HERE.

Usage Instructions:

  • The speech to be encoded may be on a mono or stereo track.
  • The track sample rate MUST be 44100 Hz or higher.
  • The speech recording should be clear and not distorted. Ideally it should have a peak level around -3 dB.


  1. Select the audio to be encoded.
  2. Open the “Subliminal” effect from the Effect menu.
  3. Ensure that Encode is selected.
  4. Apply.


  1. Select the audio to be decoded.
  2. Open the “Subliminal” effect from the Effect menu.
  3. Ensure that Decode is selected.
  4. Apply.

45 thoughts on “Steganography and “Silent Subliminals””

  1. When listening to a subliminal for more than 30 minutes produced with this plug-in, I feel discomfort in my ears, it is as if the audio is hurting my ears, I believe it is because of the frequency at 16000 hz. Is there any way to resolve this?

    1. I’d suggest that you either listen at a lower volume, or don’t listen at all.

      I’m not an audiologist, but I’d advise against listening to anything at very high volume levels.

    2. I managed to solve it, the problem was on my PC. I used a lite version of windows 10, and this ”lite” version was without the correct drivers, this absence of drivers can distort the subliminal, hurting the ear. I formatted windows, now I installed the full version of windows, imported the WAV audio to audacity, ran the plug-in, and now it doesn’t hurt my ears anymore. Thanks!

  2. Thanks for this. Do you think uploading to Spotify or YouTube removes the subliminal characteristics at all?

    1. I generally write the term “subliminal” in quote because it refers to a belief that certain kinds of inaudible sound have causative influence on the subconscious mind.
      Considering that advocates of homoeopathy believe that “homoeopathic remedies” are causative of medical benefits even when the solution is diluted to the point that none of the final product is chemically indistinguishable from the diluent, then advocates of “silent subliminals” may believe that they still work even if they are encoded in such a way that the “message” is no longer physically present.
      Regarding the physical encoded message, it is likely that the encoding process when uploading to YouTube or Spotify, will remove, or at least substantially reduce the encoded message, but I don’t think it’s my place to tell people what to believe or not believe.
      You can test if an uploaded “silent subliminal” still contains the “message” by downloading (or recording) from YouTube / Spotify and running it through the plug-in with the “decoder” option. I would expect the decoded message to be very weak or totally absent, but please do try it for yourself.

      1. Can I download a subliminal track from a reputable creator and decode it to confirm the affirmations are “intact”?

        1. If the track is AM encoded then yes you should be able to decode it, but you will need to figure out (guess) what “carrier frequency” they used (usually somewhere around 16 / 17 kHz).

          The plug-in posted on this page uses a carrier frequency of 16000 Hz (16 kHz). You can change the carrier frequency by modifying this line in the plug-in:

          (psetq *cf* 16000 ; Carrier frequency

  3. Hi Steve,

    Got a couple questions on these plugins:

    1) How does this differ to the subliminal.ny file that edgar-rft posted on the audacity forum back in 2011?

    2) If i play a subliminal audio at say 16,500Hz i can very faintly notice the high pitch frequency , but if i play it at 17,500hz i notice nothing. I’ve been reading that as we age our ears deteriorate and at my age (28) the ear stops picking up frequencies at the 17,500 range on average. Therefore, does this mean my subconscious is not receiving the audio and i should therefore choose a frequency that i can only just hear faintly? I.e. 16,500hz instead? There is very little info on this out there and no good scientific studies i could find.

    Much appreciated

    1. > 1) How does this differ to the subliminal.ny file that edgar-rft posted on the audacity forum back in 2011?

      The code is a bit different, but basically it is the same effect.
      The main difference is that this plug-in also includes a decoder so that the “message” can be converted back to an audible signal.

      > and no good scientific studies i could find.
      I’ll leave it up to you to draw your own conclusions from that. It is certainly not for a lack of trying because many people have been trying for years to produce verifiable evidence, just as many people have been trying for years to prove the existence of the Loch Ness Monster and Bigfoot.

    1. To prevent spam, comments do not appear until I approve them. I’ve been away for a few days for the funeral of a relative, so I’ve not been available to approve comments until now. I do try to respond to genuine comments as quickly as possible 😉

  4. I am having an issue that every time I import something, the project rate is 24000. When I change it to 44100, it does not work. How do I fix this issue for all future projects?

  5. Hi, is it okay if I listen via my smartphone’s speaker? Or is it better with small speakers? Thanks in advance.

    1. It doesn’t really matter what you listen with as the point of the encoding is to make the encoded content inaudible.

  6. Love the plugin Steve. If I look at the Plot Spectrum and compare your output to, I notice they provide more on the lower end. Is this an adjustment you are willing to make?

    1. Thanks for your question Davie.

      The zenmix “subliminal-maker” has a carrier frequency of 17.5 kHz, but it does not have a high-pass filter on the output. Rather than “Single side-band” AM (as in Lowrey’s patent), the zenmix version is dual side-band, which allows the encoded message to seep into the easily audible range.

      The idea of my plug-in is that the encoded message is almost silent, which is why my version uses single side-band AM encoding.

      Personally I find the chirping sound of the zenmix version to be rather irritating to listen to.

      In short, no I wouldn’t like to copy their version, because it is far from “silent”.

      1. Hi Steve thank you for this scrip
        Can i ask if i have to normalize to -1 the audio,then applying the script and finally normalize to -23?
        Thank you very much for your time

        1. The audio to be encoded should ideally be a good quality recording with low background noise and a reasonably high signal level but below 0 dB (above 0 dB causes clipping distortion). The effect will still function with a poor quality low level recording, but the old saying “rubbish in: rubbish out” applies. If you encode a bad quality low level signal, then the plug-in produces “AM encoded bad quality low level audio”, which is probably not what you want.

          Amplifying / Normalizing the result is optional, but to avoid clipping the audio must be below 0 dB peak before exporting as an audio file.

          Basically, the same rules apply as when processing any kind of audio.

  7. hi Steve, there is a subliminal channel called @subliminalchamp2082 that makes really effective subliminal and he says he uses a technique similar to the original lowry patent that you also mentioned, however your subliminal sound a little different than his for example when it comes to his subliminal i cannot hear one sound but when it comes to yours i can hear the frequency what do you think his formula is just wondering. thank you for sharing your findings

    1. I don’t see any evidence of “subliminals” in those videos. It may be that they are fake, or it may be that the Youtube encoding has destroyed too much high frequency content for the encoding to survive.

      The “silent subliminals” are either silent, or very nearly silent because there is nothing (or nearly nothing) there. Those with music appear to just be music.

      Please note that I am a musician / artist and engineer. Unlike Lowrey and the person that created that Youtube channel, I make no claims about subconscious effects of AM encoded audio. I am not the right person to discuss such claims with as I have no evidence to support such claims.

  8. Hey so I’m not supposed to get a kind of chirping sound from this right? Tried to use TTS (dspeech from Dimio’s Software; Win7 software) but after using the encoder the resulting audio was kind of chirping and still *very* audible. Would you recommend I just use my own voice instead if I’m not willing to use any other TTS software?

  9. > Would you recommend I just use my own voice instead if I’m not willing to use any other TTS software?


  10. Hi Steve, thank you for your work on this. I was wondering if you had any suggestions for text to MP3 sites to use for making the audio affirmation files. I have been searching for free ones, but I realize it’s probable that not all are of the same quality.

    I am new to this whole field. Is there a certain speaker quality or classification needed to effectively play these files after encoding? Thank you!

  11. Hi I have a 1 hour 15 minute audio that Im wanting to change to subliminal does this plug in not working with longer audios as it doesn’t seem to want to work. When I hit the 2nd apply it just stops converting to subliminal half way through can you help at all?

    1. The maximum length may depend on which version of Audacity you are using, but I think that for all versions you should be OK for up to 45 minutes for a mono track at a sample rate of 44100, or half that for stereo.

      For longer files, process the track in sections.

  12. Hi Steve!

    I bought the book “Subconscious – Sleep Learning with Ultrasound” by Heike Bonin.

    This book contains the text of the Subliminal plugin, but it is in Russian. Can you translate the text of the Subliminal plugin into the required language, so that I can then copy it into a text file and place it in the Audacity plugin folder with the ny extension?

  13. When applying the plug in, I notice that it distorts the audio recording (It can be heard when everything is turned up to the max, not doing that with headphones of course) , does that have any effect on the way subliminals are supposed to work, or is it ok as long as the audio is clear prior to the plug-in being applied? Hope my question makes sense. TIA.

    1. Lowry’s theory regarding “silent subliminals” sounds highly dubious to me. I think he should have stuck to biochemistry.

  14. Thank you so much for your help. I am having so much fun playing around with this. I never knew there was so much to learn about sound. One more question. Do you suggest using normalize loudness or the amplify effect when choosing to lower the decibel or overall volume of the subliminal?

    1. You can use Amplify or Normalize. Changing the volume with either does the same thing – they scale the size of the waveform, thus changing the amplitude.

      1. Great thank you! My husband and I have been playing around with this. Our latest method is creating a vocal audio file with text to mp3, import audio, resample to 44100, copy the sound, drag one track to the right 100% and encode with subliminal ny.file then lower decibel max to -23. Then we select the other track, drag to the left 100% and lower the decibel max to -120. We then export audio as .wav and stereo. Our thought is this lower decibel in 44100 might help the unconscious mind decifer the high frequency being played parallel. Technically speaking, does this sound like a good method to create what we are trying to?

        1. As you may have seen in my previous comments, I’m not into “subliminals” myself, so I can’t comment on psychological / spiritual benefits that you may be wanting to achieve. From a technical point of view, it sounds like your method will effectively hide the message from conscious perception.

          1. Thank you so much for the replies. There is so much to learn about sound. I know I speak for many when I say thank you for your helpful insights. I was wondering if you would mind explaining the difference between the 17500hz and 432hz “subliminals”? When someone is claiming to have 432hz, is that possibly spoken affirmations, similar to what the Nyquist plug-in produces, but a different frequency?

  15. The “432 Hz” idea has nothing to do with “subliminals”. It is a theory, popularised by the convicted fraudster Lyndon Hermyle LaRouche Jr. that 432 cycles per second is a magical frequency. There is no evidence whatsoever to support his claims, but there is a ton of evidence that much of his “evidence” is factually incorrect. He started of as a far-left political activist, changing in later life to become a far-right conspiracy theorist. I wouldn’t go so far as to describe him as “evil”, though many of the people that he scammed may disagree.

    1. So please excuse my ignorance but I have a question. That (432) would be referring to tempo? And 17500hz would be frequency? So could anything, even a encoded subliminal be switched to 432 if one so desired?

      1. No, the “432” refers to the audio frequency 432 Hz. Specifically it relates to how instruments are tuned.

        Before the 18th century, there was no standard for how instruments were tuned, because there was no good way to define a standard. In the early 18th century, the “tuning fork” was invented. Tuning forks have a fixed pitch, and reliably retain a precise frequency, changing very little with temperature, humidity or age.

        The absence of standard tuning was a problem for musicians that played in groups, because the music would sound horribly out of tune unless all of the instruments were tuned in the same way. Over the years, many standards came and went. One standard that became quite popular was the French “diapason normal” tuning (also known as “Continental Tuning”), which was standardised in France in the mid 19th c.

        It wasn’t until 1939 that the modern “Concert Pitch” was standardised as ISO-16 by the International Organization for Standardization. While “Concert Pitch” is generally adopted by musicians that play “Western” music, orchestral works that involve a Church Organ (which are often tuned differently, and play lower in warm weather) require that the orchestral instruments are tuned to match the organ (Church organs cannot easily be tuned).

        It is also common for solo musicians to be tuned in non-standard ways, since agreeing to one reference frequency is only an issue when there are multiple instruments playing.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.