Beware Of Your Toes!

[Axiom]

A file I uploaded called "axiomcd1.mp3" contained the words "be aware of your toes and feel any tension disappear". This file was voiced by a speech synthesizer which unfortunately ran the words "Be" and "aware" together. I stayed up to the small hours completing this file and didn't notice the problem in the final proof listening but when I listen to it now it just cracks me up and spoils the induction.

I see some people said in the Forums that they use Virtual Hypnotist to voice scripts. My computer is a Linux machine and VH is not available for it so I am curious how the voice quality compares. I used the Festival speech synthesizer V1.95 with the default voice.

I looked up binaural beats on Wikipedia. Based on this minimal research I have attempted to make a binaural beat version of the file using Audacity. I used a 220 Hz background tone at 1/4 of the maximum amplitude. I increased the pitch on one stereo channel and reduced in on the other to try to get the beats. I used used 20 Hz to start (210 and 230 Hz), 10 Hz during the induction falling to 7 Hz in the actual mind-warping bit and back to 20 Hz at the end. Then I mixed the tracks and exported as mp3 at 32 kbit/s. Does this sound reasonable. Is there a better approach?

So I now have an updated standard mp3, an updated script and a binaural mp3 file. How do I upload these so that they all get combined in one entry in the Files replacing the existing version? Does the script file have to be MSDOS line-terminated? ASCII or unicode? What is the file size limit for uploads?

Axiom.

[whatthe75]

Toes can be deadly and sneak up on you!

[sarnoga]

So I now have an updated standard mp3, an updated script and a binaural mp3 file. How do I upload these so that they all get combined in one entry in the Files replacing the existing version? Does the script file have to be MSDOS line-terminated? ASCII or unicode? What is the file size limit for uploads?

Axiom.

I have never done it but on the upload page there is an option for deleting an existing file. I think you have to delete the current file and then post the new files.

Sarnoga

[Axiom]

Hi whatthe75 and sarnoga,

Thanks for responding. As a newbie I was starting to think I had done something to offend the regulars. I hope I haven't. Some of the posts I have seen in various forums look pretty strange but everyone seems friendly and helpful. A nice community if you don't mind eccentricity!

I will retry the upload as suggested.

Axiom[/quote]

[Axiom]

sarnoga,

That hint was just what needed. I got the improved file uploaded along with the script and the binaural. Thanks.

whatthe75,

There are no deadly toes in that file anymore, only the usual hazards of hypnosis files.

Axiom

[krindil]

jeeze now im paranoid that my toes are going to get me

[VeryGnawty]

jeeze now im paranoid that my toes are going to get me

That gives me a good idea for a file....

Just kidding :lol:

[homerj1620]

Slightly off topic, but what voice did you use with Festival? How's the quality? I've messed with it occasionally, but never was very happy with the voices. I can't hear the file since it's premium, but I'm curious about that. Currently I'm using Cepstral's Callie, which works very well in Linux (primarily Debian and Ubuntu). I also sometimes use "Vicki" on my PowerBook.

[Axiom]

I had to go through all the voices to find which one is the default - it's kal_diphone.
It sounds like an older US male voice I think but a little robotic; maybe that's appropriate for an evil hypnotist kind of voice. You can get used to kal and I think it helps if you have read the script and know what he's saying.

If you do get used to kal's voice reading hypnotic scripts be careful - you may be giving Stephen Hawking the power to enslave you. 8O

the command to select this voice and say something in festival is:-

festival> (voice_kal_diphone)
festival> (SayText "hello, this is festival")

ked_diphone sounds like a younger US male voice
and there is at least one female voice
festival> (voice_cmu_us_slt_arctic_hts)

I listened to a sample of Cepstral's Callie and I think it's generally better than the festival voices but festival had the great advantage that it's free. Now that my financial situation is improving maybe I will buy Callie.

You can hear several TTS speech samples, including festival using kal_diphone at http://www.student.oulu.fi/~vtatila/reviews_of_speech_synths.html

Axiom

Is that clown avatar scary or is it just me?

[homerj1620]

I had to go through all the voices to find which one is the default - it's kal_diphone.
It sounds like an older US male voice I think but a little robotic; maybe that's appropriate for an evil hypnotist kind of voice. You can get used to kal and I think it helps if you have read the script and know what he's saying.

If you do get used to kal's voice reading hypnotic scripts be careful - you may be giving Stephen Hawking the power to enslave you. 8O

That's what I never really liked about Festival, too Stephen Hawking-like. I still play with it occasionally, though. What voice do you like to use? That's more of what I'm curious about.

I listened to a sample of Cepstral's Callie and I think it's generally better than the festival voices but festival had the great advantage that it's free. Now that my financial situation is improving maybe I will buy Callie.

It's pretty nice, though if you want to use it on a website or something, it's $200 for a license. For personal use it's $30. I use it for personal scripts, though I'd prefer something Free (as in speech) or at least free (as in beer).

Is that clown avatar scary or is it just me?

It's meant to be scary. It's Captain Spaulding from Rob Zombie's films "House of a Thousand Corpses" and "The Devils Rejects".