Hi,
I still currently use Movie Edit Pro 2018 Premium, and was wondering whether or not there is a way to change your voice? For example, if I wanted to change my voice to make it sound like a woman or a child, is that possible?
Hi Jak.
Yes you can alter the pitch but not the timbre so would not really sound like a female without some work from the performer. It may also add some artifacts into the recording. MEP and VPX is not that as smooth in it's abilities compared to a dedicated music DAW in that regard. whether an individual can hear the difference or not is a different matter. I can hear it if the pitch is taken beyond a certain point but then I'm sensitive to things like auto-tune.
Right click an audio track and select Timestretch / resample.
Ray.
Hi Jak.
Yes you can alter the pitch but not the timbre so would not really sound like a female without some work from the performer. It may also add some artifacts into the recording. MEP and VPX is not that as smooth in it's abilities compared to a dedicated music DAW in that regard. whether an individual can hear the difference or not is a different matter. I can hear it if the pitch is taken beyond a certain point but then I'm sensitive to things like auto-tune.
Right click an audio track and select Timestretch / resample.
Ray.
Hi Ray,
I actually do my own voice overs for my videos. I’m quite good at doing female voices, however I am trying to find a way to make each female voice sound slightly different so that it doesn’t become obvious that it’s the same person doing all of the voices of that makes sense?
Hi Jak.
One of the most complex sounds to alter is the human voice. It is not easily sculpted of synthesized. The best way to alter a voice is to use the human voice. You can alter it's pitch and increase or decrease frequencies bands but you can't alter it's timbrel or harmonic structure convincingly by artificial means. Most AI text to speech starts by recording various actors performances building a library of sounds of hundreds of thousands of recordings. So if you are looking for a sort of plugin that could do that then there are specialist companies like Sonantic that produce products for game developers and the like but their products are not for the consumer.
Beyond pitch correction, I can't think of any effect or plug-in that would be remotely useful for this.
Maybe someone else has come across something I haven't seen. I would be interested in looking at something that could do that myself but I doubt it would be cheap.
Ray.
Hi Jak.
One of the most complex sounds to alter is the human voice. It is not easily sculpted of synthesized. The best way to alter a voice is to use the human voice. You can alter it's pitch and increase or decrease frequencies bands but you can't alter it's timbrel or harmonic structure convincingly by artificial means. Most AI text to speech starts by recording various actors performances building a library of sounds of hundreds of thousands of recordings. So if you are looking for a sort of plugin that could do that then there are specialist companies like Sonantic that produce products for game developers and the like but their products are not for the consumer.
Beyond pitch correction, I can't think of any effect or plug-in that would be remotely useful for this.
Maybe someone else has come across something I haven't seen. I would be interested in looking at something that could do that myself but I doubt it would be cheap.
Ray.
There are a couple of programs I briefly looked at called VoiceMod & Voxal Voice Changer, but it doesn’t look they do what I want. Like they’re more cartoony/silly voice effects, which I don’t want.
Yes, those types of programs add a ring modulator module (Think Moog synthesizer module) that rapidly fades in and out the samples on a waveform (Sine, square or saw tooth) to a pitch changer /re-sampler which also does not work that well as it literally slices the waveform at the sampling frequency and repeats the slice or takes some of it away or speeds it up or slows it down depending on whether you select to resample or stretch the audio. Pitch changers / re-samplers that can work at higher sampling rates can do a better job if the recording is made at a higher sampling rate but even then they are not ideal.
That is the limit of what can be done at current technology levels without the use of artificial intelligence and whole huge bank of real life samples to work with. Vocoders just add extra synthesis modules to get the sort of 'Sparky the magic piano effect'.
Ray.
What you need is a vocal processor VST that allows you to change the formats.
Meldaproduction Mtransformer is an example of a plugin that has this capability. You can try the 14 day demo version to see of this fits your needs.
https://www.meldaproduction.com/MTransformer
Thank you, I’ll check it out. Have you ever used it yourself?
Yes I have this plugin and I have used it myself (albeit in a DAW) to change the timbre of a singers voice in a music track.
Hi Jak.
If you are still interested have a look at Watson speech to text program.
Ray.
Thanks Ray, I’ll have a look at it.
Hi Jak.
That was meant to read text to speech. 😆
It's supposed to be one of the better ones. To give more convincing gaps there needs to be added commas and full stops where you need pauses. There should be a selection of voices to choose from that could also be pitch shifted for additional variation. Some are more convincing than others. There is also Google's offering. I have found others but some need commercial licensing for use on publicly viewable material. Wideo is another free program that creates an MP3 file to download. I found quite a few in the end of varying quality. Most though are not that natural. Watson so far has been the best I think I found.
Also be aware it does not always work well. 'I am having a row with bill'. Will come out as in 'I saw the two people row a boat'. So Row will be pronounced roe. Other words also get mispronounced.
Ray.