Adding sound or a voice to your doll

center_doll · Post by **center_doll** » Wed Dec 15, 2021 2:04 am

it's a great idea but how do you make the sound interative? I mean moaning with touching or penetration?

Technician · Post by **Technician** » Wed Dec 15, 2021 8:22 am

center_doll wrote:it's a great idea but how do you make the sound interative? I mean moaning with touching or penetration?

Bubbles has a Bluetooth speaker/microphone in her head. With that I can run an AI program I wrote on a nearby computer and converse with her, but there are limitations. First, she can only hear so many words, if I try to give her a broader vocabulary to recognize she tends to hear incorrect words so it's best she hear and respond to keywords. Responding is a problem too because very high quality voices mean sending data to Google or others to process but I prefer to keep my pillow talk just between us and not on some server somewhere.
For that matter speech recognition could be better with an internet based solution too but I still have that privacy issue.
I'm working now on sensory input, ESP modules like 32 & 8266 have sensor inputs and can communicate wifi to the computer with her AI. So once I have sensors working Bubbles is getting surgery, a much bigger battery, and at least one ESP8266 module along with a bunch of sensors.

Technician · Post by **Technician** » Sun May 07, 2023 7:28 am

Well, Bubbles was (as far as I know) the first doll with ear canals back in 2020, now she's the first doll (again, as far as I know) to have nasal cavities.
A recent project prompted me to re-examine the BT voicebox mod and as I'd noticed before, the sound gets much clearer if I pull her mask away from her skull just a little.
Naturally I don't want her talking to me from her neck, so tonight on a whim I put a sharp poker up both sides of her nose, then lubed it with Vaseline and got two 1/4" pieces of vinyl tubing pressed up through her nose and into the speaker cavity. Just that little mod made a difference and now I get cleaner speech from the center of her face. I could have ported it to her mouth but didn't want anything that wasn't soft in there if you know what I mean.
The shape of her nose changed slightly and for the better, it was a bit thin and now it's much more realistic, and has open nostrils. I started a new job last year that takes a lot more of my time but yields much more, and more stable income so I'll be moving soon and my lab will be at my residence, so hopefully I can carve out some time to move forward with sensors and a computer or ESP module.

Christines Man · Post by **Christines Man** » Wed May 10, 2023 12:25 pm

Technician wrote: ↑Wed Dec 15, 2021 8:22 am
Bubbles has a Bluetooth speaker/microphone in her head. With that I can run an AI program I wrote on a nearby computer and converse with her, but there are limitations. First, she can only hear so many words, if I try to give her a broader vocabulary to recognize she tends to hear incorrect words so it's best she hear and respond to keywords. Responding is a problem too because very high quality voices mean sending data to Google or others to process but I prefer to keep my pillow talk just between us and not on some server somewhere.
For that matter speech recognition could be better with an internet based solution too but I still have that privacy issue.
I'm working now on sensory input, ESP modules like 32 & 8266 have sensor inputs and can communicate wifi to the computer with her AI. So once I have sensors working Bubbles is getting surgery, a much bigger battery, and at least one ESP8266 module along with a bunch of sensors.

I recommend trying out whisper, at https://github.com/openai/whisper/. I have this running on a computer on my lan and it's surprisingly accurate. It'll run fine on CPU but if you have a GPU it gets even better. My code is at https://github.com/ChristinesMan/Christ ... _server.py. It's python but I think there are ways to use C++, too.

BuildingFun · Post by **BuildingFun** » Sat May 13, 2023 2:33 pm

Like the text to speech stuff coming out for AI, found some on you tube that looked good 'Bark' looks very promising when added to a local run AI.

Technician · Post by **Technician** » Sat May 13, 2023 5:23 pm

I've transcribed with Whisper, it's very fast! Haven't tried the C++ API yet but it's on my to-do list. I'm still busy working through my sound commission stuff

xXmartian · Post by **xXmartian** » Sun Oct 22, 2023 9:51 pm

Kira-sama wrote: ↑Wed Jan 01, 2020 12:57 pm
manobon wrote:
Kira-sama wrote:
manobon wrote:
Kira-sama wrote: For sex, I use two different online apps with built-in accelerometer support.

1. http://plaisonar-0001.netlify.com/

2. http://www.finsexdocka.se/app/ (setting 33 for me seems to be the best for keeping in sync with the thrusting)
Thanks for linking those! For some reason, they can't read my accelerometer (Pixel 2) - any suggestions?
Glad to help out

When I got that message, I restarted my browser (FireFox) several times until it finally stopped showing that message. That's what worked for me. Maybe you can try that or uninstalling updates/clear memory/history/cache/etc...
Ah, I was using Chrome - FireFox works so far, and I'll try clearing if it stops - thanks again!
Glad I could help

Super interesting thread! I'm wondering if anyone knows of any other accelerometer based audio apps other than these 2? I'm using an iPhone 13 Pro Max and neither app sees my accelerometer regardless of the browser I use, and I've tried all of them. Lol Also tried clearing history and all of that to no avail. Thanks for the help!