Silent Subliminals [solved?]

To download latest version (recommended):

With some help of a professional hypnotist (in the medical sense, not in the humbug sense) and Steve’s code from:

I think I have a finally working Silent Subliminal plugin producing an SSB modulated carrier like the original Lowerey machine. It works with a cheap 44.1kHz standard soundcard. I cannot find any aliasing or mirror frequencies with the Audacity “Analyse > Frequency Analysis” and everything looks pretty well.

;nyquist plug-in
;version 1
;type process
;name "Subliminal..."
;action "Subliminal..."
;control carrier "Carrier" real "Hz" 17500 14000 20000

(setf carrier (max 14000 (min carrier 20000)))

;; We have two Nyquist frequencies, carrier/2 and *sound-srate*/2.
;; The CUTOFF is the maximum allowed frequency in the modulator.
;; It must not be greater than carrier/2, but also not greater than
;; the difference between the carrier and *sound-srate*/2, because
;; otherwise the modulated carrier aliases.

(setf cutoff (min (/ carrier 2.0) (- (/ *sound-srate* 2.0) carrier)))

(defun cut (function sound frequency)
  (dotimes (ignore 10 sound)
    (setf sound (funcall function sound frequency))))

(defun subliminal (sound)
  (let ((result (mult 2 (cut #'lowpass8 (hp sound 80) cutoff)
                        (hzosc carrier))))
    (cut #'highpass8 result carrier)))

(if (< *sound-srate* 44100)
    (princ "The track sample frequency must be minimum 44100Hz.")
    (multichan-expand #'subliminal s))

The maximum modulation depth is at a carrier frequency of “track-sample-frequency / 2 / sqare-root-of-two”:

Example: 44100Hz / 2 / 1.414… = 15592Hz

Original under Silent Subliminals with Audacity - text is in German but nothing you do not already know…

  • edgar

I’ve made a note in the original topic pointing to this plug-in.
I like, and found it educational, the compact way that you’ve reused the (cut) function as both a high order low-pass and high-pass filter.

I’ve attached Edgar’s code as a file for easy installation.
See here for how to install Nyquist plug-ins:
subliminal-rft.ny (1.07 KB)

I have tried this code but get an error message “nyquist did not return audio”
I also tried the code from the German website and got a different error message "nyquist returned the value 22050.000000.
Coding’s not my strong point, but I think the issue in the first one is an undefined perimeter, but I have no idea about the second error message. Any ideas?

What Audacity version are you using? Do the errors also appear with Audacity_1.3.13-beta? Audacity_1.2.6 is probably too old for this.

Hi Edger

I am using 1.3 Beta. It is probably me doing something wrong, but I can’t see what it is

I just have tested the (obsolete) Version from above plus the (more recent) Version from the German Audacuty forum (attached below) and both versions work with Audacity_1.3.13-beta on Debian Linux with no errors.

If the attached plugin below still raises errors (“Nyquist did not return audio” or similar) then in the “Subliminal” effect window press the “Debug” button instead of the “Ok” button and copy the text from the Audacity Nyquist “Debug” Window here into the forum.
subliminal.ny (1.11 KB)

The version number has three parts, for example 1.3.13
The third part is important as early 1.3.x versions use a different version of Nyquist.

Well…Is there any plug in which gives subliminal bath effect? Did you check the output frequency of subliminal plug in effect, whether it matches with the plug in’s carrier frequency or not?

What is a “subliminal bath” effect? Listening to subliminals in a bathtub? Or do you mean a “bat” (small flying animal emitting ultrasonic sounds)?

Sorry, I do not understand what you’re talking about.

Which is a better choice, earphone or piezoelectric transducer for subliminal induction? I have been listening to the silent subliminal messages for two weeks now. still not a slight visible change in me. I used the plug in with 17500hz. How can I make the recordings 3-D and coming from different directions? Is it possible with Audacity? Regards.

This forum is about Nyquist programming and Nyquist plug-ins. Whether or not silent subliminal messages have any effect on people is really a question outside of the scope of this forum. We are only able to help with technical questions about designing Nyquist Plug-ins.

As I wrote several times in the original topic:
“I make no claims regarding the safety or effectiveness of this technique. It is only intended as a demonstration of a simple implementation of Lowry’s silent subliminal presentation system (US Patent 5,159,703 October 27, 1992).”

By using a multichannel soundcard together with more than two loudspeakers or a 3D-headphone with more than one speaker per ear.

The same way as you do with ordinary audio tracks. In the box at the beginning of an Audacity audio track, move the “panorama” slider (the lower slider below the “Mute” and “Solo” buttons) from out of the middle to the left or right, then go to the Audacity “Tracks” menu and apply “Mix and Render”.

Depends on the earphone. Cheap earphones cannot reproduce high frequency signals very well. Earphones with reasonable quality should not have any problems to reproduce signals between 17500 and 22050 Hertz. Piezo speakers if used with high volume levels can cause ear damage, even if you don’t “hear” the signal. Please be careful.

I’m not sure what you mean here. The plugin produces a single-side-band modulated signal with a suppressed carrier, so if the output frequency “matches” the suppressed carrier (the carrier itself should not appear in the output signal) then it’s a bug in the plugin. But I think that this is not what you meant.

You can test the output signal yourself by first applying the “Subliminal” effect, then select a piece ot the subliminal audio track, and then go to the Audacity “Analyze” menu and apply the “Frequency Analysis”. In the analysis window there should only appear frequencies above 17500 Hertz.

See also the text box directly below how to convert the “Subliminal” signal back to normal speech again.

The “RFT” in edgar-rft is the german abbreviation for audio/video recoding equipment electrician. I only translated the electrical description in Lowry’s patent paper into Nyquist code. I cannot promise that the “Subliminal” plugin (or any other subliminal message machine) produces any useful result.

I generally try to be unbiased in answering this question, but I think that it is mainly matters if you believe in subliminal messages or not.


1. If I get up early in the morning and think “everything is bad” then probably the whole day long everything will appear as “bad” to me, mainly because I make myself think that everything is bad, even if the day is good. The same effect can also work the other way around. Strictly spoken, I do not necessarily need subliminal messages to make me feel good, it’s mainly my own thinking that infuences my perception, often much more than I am really aware of this. Maybe subliminal messages can help me to think “good”, so I have nothing against subliminal messages, but I cannot promise that it really works.

2. There had been a predecessor “ultrasonic” plugin in the german Audacity forum, that had originally been written for a dog training school. They had a 192kHz soundcard playing ultrasonic “dog commands” over a piezo speaker system. The dogs, who can hear ultrasonic sounds, then learned to do things according to the signals, where the humans hear nothing. Some time later I was notified via some posts in the german Audacity forum that people reported success by using the “ultrasonic” plugin to produce subliminal messages. To me it was instantly clear that the “ultrasonic” plugin couldn’t work on a 44100Hz soundcard because the signals produced by the plugin were above 30000Hz, which cannot be reproduced with 44100Hz sample frequency. In some subsequent discussions I prooved several times that the “ultrasonic” plugin with 44100Hz sample frequency only produced signals containing not more “message” than contained in dust-buster noise, but people were still continuing claiming “astonishing success” and similar nonsense, where the “subliminal” messages produced by the old “ultrasonic” plugin on standard 44100Hz soundcards contained nothing but pure audio garbage and no “message” at all. This is another proof that it’s our own wishing or thinking that mainly influences our perception. But at the same time this is another reason why it is very hard to decide if subliminal messages really work or are pure “placebo” effect mainly based on simple self-suggestion.

The current “Subliminal” plugin:

With the help of Karl J. Wargan and Steve’s link to Lowry’s US-patent it was possible to “translate” the description of the original Lowry machine, that used two slightly de-tuned radio transmitters to produce the subliminal audio signal, into Nyquist code that saves the up/down frequency transformation. Lowry didn’t have digital filters at that time, so it was “easier” first to produce a SSB-modulated shortwave signal and then transform it down to the audio range, but the output signal is exactly the same. I have to admit that I do not have an original Lowry machine to compare the signals, so it’s only “mathematically” identical, but the probability is very high that both signals are also physically identical. Please complain if you find problems.

I am still interested to test the “Subliminal” plugin in some medical context.

  • edgar

Here is how to convert the “Subliminal” signal back to speech again:

1. Select the entire “Subliminal” audio track.

2. In the Audacity “Effect” menu, open the “Nyquist Prompt” and copy the following into the text field of the “Nyquist Prompt” window:

(defun ssb-demodulator (sound)
  (highpass8 (lowpass8 (s-max 0 sound) 3000) 300))

(scale 5 (multichan-expand #'ssb-demodulator s))

Then click the “OK” button below the text field.

Do expect a very distorted speech signal, because there is not much “room” for sound quality in the range between 17500 and 22050 Hertz on a soundcard with only 44100 Hertz sample frequency. It probably sounds as if you shout with a megaphone into a telephone.

  • edgar

Better quality conversion of the “Subliminal” signal back to speech again can be achieved like this:

(defun demod (s-in cf)
   (mult s-in (hzosc cf)))

(multichan-expand #'demod s 17500)

This assumes that during “encoding” a carrier frequency of 17500 Hz was used. If this is not the case, change the number as necessary.

There is a typo: #’ demod must be #‘demod (no whitespace between #’ and demod)

Question: to what number do I change the number if I do not know the number?
(the original code was written to descramble subliminals downloaded from the internet)

I’m aware that the SSB-demodulator produces poor quality, but is there a better way known to descramble “unknown” signals?

asks… - edgar

Thanks Edgar, now corrected, though strangely enough it still worked with the space.

Good question.
A fairly close guess can be made of the carrier frequency by looking at the spectrum and using the low cut-off frequency.
Other than that I think it would need to be “tuned in” like a radio.

Edgar, u really solved my some of the most uncleared questions. Thanks!
Well, would u make available that “ultrasonic” plug in u mentioned, used to dogs training?
Setting the project rate @ 96000Hz, if subliminal plugin with 17500Hz frequency used twice , will it deliver 35KHz?

If you set the Audacity track samplerate to 96000Hz and then, in the Subliminal plugin window, you type 35000 into the text box beside the frequency slider, then you’ll get a 35kHz subliminal. The “20000” (20kHz) limitation of the slider in the plugin window is only an artificial “fool proof” limit, the plugin can internally work with frequencies up to half of the Audacity track samplerate minus 3000Hz (speech signal bandwidth).

This means if you set the track sample frequency to 96000Hz before applying the Subliminal effect, then the plugin can produce sbliminals with frequencies up to (96000Hz/2)-3000Hz=45000Hz.

Note that the maximum possible carrier frequency is internally limited to (tracksamplerate/2)-3000Hz, so if you type in higher frequencies, then the carrier frequency will be internally set down to this limit and you will not get the desired output. That’s why it’s important to set the Audacity track samplerate to an appropriate value before applying the Subliminal effect. Nyquist plugins cannot change the Audacity track samplerate.

The track samplerate can be changed in the Track Drop-Down Menu via “Set Rate”.

Below is an english “ultrasonic.ny” version, modified to work with arbitrary frequencies, not only 30kHz. Be aware that the Subliminal plugin with the text-box trick from above produces much better audio quality. The “ultrasonic.ny” does not filter the speech signal before modulating the carrier, so the result will alias (produce undesired mirror frequencies) if the speech signal contains frequencies above (tracksamplerate/2)-carrier. The text-box trick from above also works with the “ultrasonic.ny” plugin, but there are no internal safety limitations. If used with “wrong” values, the “ultrasonic.ny” plugin it can produce the most unusable results ever.
ultrasonic.ny (764 Bytes)

hi everyone

Do you have any hussing, clipping sounds after modulating with plug-ins ?
what kind of sample rate you use for traditional silent subliminal
i used 44.1 hz and i have it , i use 96 khz and have some funny speedy sounds in track
what rate sample do you recommend for sound card