#Zoltar 3.0 work in process

I just finished the software for #zoltar version 3.0. All I need to do is get the speech recognition working on a #raspberrypi 4, that’s probably hardware config. And attach a small screen to it and then build an enclosure. The latter is always my least favorite and most challenging part. But then we have an interactive Zoltar machine. A little bigger than a shoe box so I can mount it in the hallway.
I will use a Pepper’s ghost illusion. Either reflect from ten screen as demonstrated in this video or a small short throw projector. We will see…

The starfield is done in code, the warping/floating is a GLSL shader (actually two, I never knew that Raspberry Pi only supported glsl 1.00)
We are using Azure Text To Speech and Speech To Text, like in the Computer as MindReader. It's written in C# (because their TTS is easily configured in dotnet compared to C++ where you need to compile it foreach platform) and RayLib which is lightweight graphical API.

Видео #Zoltar 3.0 work in process канала CallousCoder