No way without DSP.
I read your posts/threads and waiting for this phrase
So basically the idea is like those guitar tuner boxes but for the human voice?..this is aimed for on-the-go musicians who want to do a bit of pitch practicing before they record their tracks or perform live.
Would a STM32 F4 be able to do low latency audio DSP like EQ and pitch recognition and have analog read and write? About how much power would it consume?Of course you will want to use a processor that uses a lot less power than a Pi. Perhaps an STM32 F4 or some such.
Thanks for the offer.I'll pay you to help me complete this project if you think you could do it.
You want free consultancy. Well you have had some. The fact that the advice you got isn't what you hoped for might actually mean something.fohman wrote: ↑Mon Feb 05, 2018 6:07 pm
What I feel that I really need is a way to find people who might want to help me complete this project for some negotiated amount of pay based on a finished product. How do I find such people? Sites like Upwork and Toptal are problematic because you have to pay up front for the time spent explaining the project to the would-be designer, and I don't want to do that. The reason I was digging around here in this forum was because I figured I might just have to do enough of my own research to present a more finalized solution to an hourly engineer on TopTal. Where do I find project-based designers and engineers?
That is THE question.Heater wrote: ↑Mon Feb 05, 2018 6:45 pmfohman,
Ah, I see, so you want us to advise on your scheme to make money. For free?
I'm inclined to tune out at this point but...So basically the idea is like those guitar tuner boxes but for the human voice?..this is aimed for on-the-go musicians who want to do a bit of pitch practicing before they record their tracks or perform live.
So how about forgetting the phantom power thing and just use batteries. In a stand alone box. Like those guitar tuners do?
Of course you will want to use a processor that uses a lot less power than a Pi. Perhaps an STM32 F4 or some such.
For some reason fohman wants to combine some audio processing with the vocal tuner, hence in-line AD-DA and low latency. It doesn't seem a good fit with the practice tuner aspect. Maybe there is an unstated desire to add auto-tune. There is a logic there to making it a personal mic add-on. The singer might be able to correct poor pitch without others realising they are using an auto tune (far fetched). That would have to discretely plug into the existing mic lead and run on phantom power. It would explain why it's not done in a rack mount processor.6by9 wrote: ↑Tue Feb 06, 2018 10:27 amAs I suggested earlier, T off the audio signal into your device and nick some phantom power. Do not convert to digital and back again to forward on to the desk. You'll save a bunch of power and remove all latency concerns. The function of measuring tuning accuracy can then be a much simpler job (analogue electronics possibly), and you may be looking at something like an Arduino instead of a full processor to do the final analysis and display.