Eliminating loud sound of a recording

Hallo, I’m completely new with this kind of audio thing, and I hope someone here can help me to find out whether it is possible to erase/eliminate loud sound on a recording while keep very low one with Audacity? I’m using Win2.1.0 version.

I’m trying to decode a subliminal recording. I guess if I can eliminate such loud masking sound, I can amplify the rest very low sound to listen the hidden message on it.

Any help very appreciated.

The short answer is no. The software doesn’t know what content is. It just knows sound or noise. We can’t separate two different people speaking at the same time for the same reason. Your ears can tell there are two people speaking, but the software has no idea. You have the additional problem that you don’t know what the hidden work actually is.


That’s mostly done in reverse. So try opening the file by importing it in to audacity then use the arrow and left click on the area to the left of the song wave under the sliders for the song. This will select the whole song. Then left click on the word effects to drop a menu of tools slide the menu looking for the reverse one left click it. Now find the play button and left click it if there is back masking on the song you should hear it now as it will play in forward direction.
Then if you don’t care about the song file you cand start boosting the eq settings for mid range. The eq is on the same menu as the reverse is. Boost the middle range frequencys for the lowest ones you get the best sound of the male voice. Boost the treble ones on the right for the female voices. If you find immorality in the back masking delete the song contact the record company to ask if they knew it was there.

looking for the reverse one left click it.

Are you expecting the hidden message to be speaking backwards?

Thats what back masking is. The message sounds like some strange words when playing the song normally. But reversing it makes them sound normal.

Someone put a bad backmasking on the Eagles’s song " hotel California." When it was released. Then the public heard it found it and started protesting the group and record company. I’ve not heard of any more back masking but they might have been doing it with underground studios.

Bad news for me :frowning:

I just tried your suggestion to reverse it but seems didn’t work. My guess is he stated the subliminal sentence in very low voice, so what we can hear is just the loud sound that masking it.

And for the eq function, could you tell me what exactly should I do? I see two pointers on the left, the top one set to 20db and the bottom set to -17db, how should I set them?
Btw there is no human voice can be heard on my recording. I just can hear sound of train machine.

Omg, I’ve serious problem with this recording, I hope there is a way to solve this.

Back masking of a train? I’m going to say its not made to be anything but might be a power hum. The eq settings have a hum removeal one try that. You change the settings by clicking on the option at the bottom called unnamed. If you change to graphic view you see sliders to move up and down for all the frequencys the bass ones are on the left so if the power hum doesn’t remove it try the different sliders to turn down some of the bass frequencys to remove it. If you are not feeling too good about going that way. Then try the noise remover tool. Find a silent space at start or end of the song select that part of the wave. Then open the noise remover left click the get noise sample button. Now re select the whole song wave go open the noise remover tool again this time left click the remove noise button on the bottom.

I have to say that I impressed of how we can sort each frequencies with that function. I moved down all of the sliders to the bottom, and then tried to listen to each those 30 frequencies by moving it up one by one alternately. Unfortunately I didn’t find any human voice on any frequency.

Noise reduction tool also came with no result :frowning:

Noise Reduction only works on sounds that don’t change over time like refrigerator or air conditioning or computer fan noise. It doesn’t work on moving or changing content like speaking, singing or street noises.

It also only works if you have a sample of the noise without the subliminal content to “train” the tool. If you can’t tell anything about the subliminal content, then I don’t see anything we can do.


What some claim is intentional-backmasking is unintentional via pareidolia.


Haha hahaha yep that makes more sense then what they were said to have done.

I find a decoding method by twisting high freq and low freq with other editor on this link: http://www.subliminalrich.com/english/all/16-how-to-subliminal/39-decoding-a-subliminal-audio-to-check-the-sentences.html
Can we do the same with Audacity?

There are many techniques used for embedding subliminal messages in audio files. In order to “decode” the message you need to know what method has been used to encode it. “Amplitude modulation” is one such technique and there is information about decoding that type of “silent subliminal” in this topic: https://forum.audacityteam.org/t/silent-subliminals-solved/20100/1

Also probably worth mentioning that there is little scientific evidence that subliminal messages “work”, and no credible scientific evidence supporting “AM”, “FM” or “frequency shift” methods of encoding subliminal messages.

And then there’s the variation between the versions you are expected to derive intelligence by decoding, and the one that supposed to help you give up smoking while you listen. Again, if you don’t know what you have, you may never know what’s there.


Kozikowski and steve opinion of the impossibility of my case sound reasonable. But I still have a bit hope since rovingcowboy idea of frequencies hacking on eq seems make it somehow near to possible.

I think I prefer someone who expert to do this for me. So if any of you guys on this forum can help me for this decoding thing, please drop an email to gokspay {at} gmail com. I will bought you a doughnut and a cup of coffee for this :wink:

Thank you for all of your thought kozikowski, rovingcowboy, steve. I want to stop here. So if mods gonna close this thread please to do so. Thank you