hey there,
i'm new to this forum and i hope this is the right place for my question.
with software it is possible to shift sounds up and down in quite extreme ranges. can this be done with hardware too and what is the minimal approach?
my special case requires me to shift a sound at 3400Hz – 6500Hz down to 300Hz –*3400Hz.
i already found some diy projects about pitch/frequency shifting. one of them is using a PIC24, but i think it can't handle the extreme range i would need. 8http://hackaday.com/2011/03/17/pitch-shifter-makes-your-band-sing-higher/). i'm wondering if there are some dedicated ICs that could do the job? or any other smaller scaled hardware approaches.
just for reference: the adjacent musical notes in my case would be: D4 (293.66Hz) and G#7 (3322.44Hz)
thanks in advance,
hans
i'm new to this forum and i hope this is the right place for my question.
with software it is possible to shift sounds up and down in quite extreme ranges. can this be done with hardware too and what is the minimal approach?
my special case requires me to shift a sound at 3400Hz – 6500Hz down to 300Hz –*3400Hz.
i already found some diy projects about pitch/frequency shifting. one of them is using a PIC24, but i think it can't handle the extreme range i would need. 8http://hackaday.com/2011/03/17/pitch-shifter-makes-your-band-sing-higher/). i'm wondering if there are some dedicated ICs that could do the job? or any other smaller scaled hardware approaches.
just for reference: the adjacent musical notes in my case would be: D4 (293.66Hz) and G#7 (3322.44Hz)
thanks in advance,
hans
Have you TRIED commercial DSP stuff with pitch shifters? Nothing new or rare about pitch shifting in the music industry.
There are plenty of DSP chips out there, though I am not aware of any designed for your specific task.
Your example is 3400Hz down to 300Hz, roughtly 10/1 but the other end of the scale is 6500Hz down to 3400Hz, only 2/1. SO your example isn't a linear effect already. 3400 to 300 is not far from a four octave drop. 6500 to 3400 is approximately a one octave drop.
There are plenty of DSP chips out there, though I am not aware of any designed for your specific task.
Your example is 3400Hz down to 300Hz, roughtly 10/1 but the other end of the scale is 6500Hz down to 3400Hz, only 2/1. SO your example isn't a linear effect already. 3400 to 300 is not far from a four octave drop. 6500 to 3400 is approximately a one octave drop.
hmm… you're right about the non-linear effect. maybe i'm searching for somthing different, but i could prototype the setup with abletons live and built in "Frequency Shifter" effect unit.
I want to layer a phone call (typical transmission bandwith: 300 –*3.4k Hz) with an ambient sound (frequency shifted out of the phone call range). a small piece of hardware should filter the sounds, and shift the ambience back (down) to the phone calls frequency range.
I want to layer a phone call (typical transmission bandwith: 300 –*3.4k Hz) with an ambient sound (frequency shifted out of the phone call range). a small piece of hardware should filter the sounds, and shift the ambience back (down) to the phone calls frequency range.
some graphics to make this more visible:
(orange: my voice; blue: the sound i want to layer on top of it; blue sound gets linear shift out of the orange spectrum)
this is what i can do with software already (sreenshot shows "blue sound processing"):
(eq the sound to fit 300-3100Hz, frequency shift up)
i'd like to build a piece of hardware that can divide this single sound stream by:
- voice/orange channel: low-pass filter, 3.4k
- layer/blue channel: high-pass filter 3.4k -> linear shift down

(orange: my voice; blue: the sound i want to layer on top of it; blue sound gets linear shift out of the orange spectrum)
this is what i can do with software already (sreenshot shows "blue sound processing"):

(eq the sound to fit 300-3100Hz, frequency shift up)
i'd like to build a piece of hardware that can divide this single sound stream by:
- voice/orange channel: low-pass filter, 3.4k
- layer/blue channel: high-pass filter 3.4k -> linear shift down
This looks like SSB modulation and can be done with a ring modulator.
At least it is what I gather of what you are trying to do.
http://en.wikipedia.org/wiki/Single-sideband_modulation
jer 🙂
At least it is what I gather of what you are trying to do.
http://en.wikipedia.org/wiki/Single-sideband_modulation
jer 🙂
Yes, SSB. Used for telephone trunk lines at one time. Two ways to achieve it: filtering or phasing. (There is a third method, but it is really a variant of phasing). Also can be thought of as an image-rejecting mixer.
Thank you so much for bringing up that amplitude modulation topic.
I built myself a quick Max-patch for the transmitting computer, and now i'm about to set-up a piece of hardware to demodulate the signal. (For a first test i'm happy with transmitting one modulated signal and demodulate it.)
In a small book by Forrest Mims i found a basic AM radio circuit.
Because my transmission is cable based i connected the line-ins ground to the other end of the coil. But how to dimension coil, cap and diode to demodulate carrier frequency under 20 kHz?
Thanks in advance –*You already helped me a lot!
I built myself a quick Max-patch for the transmitting computer, and now i'm about to set-up a piece of hardware to demodulate the signal. (For a first test i'm happy with transmitting one modulated signal and demodulate it.)

In a small book by Forrest Mims i found a basic AM radio circuit.

Because my transmission is cable based i connected the line-ins ground to the other end of the coil. But how to dimension coil, cap and diode to demodulate carrier frequency under 20 kHz?
Thanks in advance –*You already helped me a lot!
Last edited:
An envelope detector cannot detect SSB; for that you need a product detector (unless the incoming signal includes the carrier - SSB usually suppresses that).
- Status
- Not open for further replies.
- Home
- Design & Build
- Construction Tips
- extreme frequency shift-down