Keep getting weird tones while recording

Hey all...

I have been really frustrated with my theremin recording lately... I hope one can guide me on what I need to do! I keep getting this weird tones in my recordings. Here is a link to an example

https://soundcloud.com/#amethyste/theremin-recording-woes

It sounds like other frequencies in there? I tried through an amp and it sound horrible. THis recording is through my M-Audio External souncard right into my recording software...

What could it be? :(

Due to lazy output filtering in the theremin, there is most times a RF component remaining with the audio signal. This is normally not a problem since each amplifier and the human ears have a low-pass filter characteristic.

Things become different when this signal mix is sampled (digitalized) for digital recording: Then the RF component interferes strongly with the sampling frequency of the sound card AND its harmonics (google for Mr. Nyquist who described the problem mathematically).

Let's take an example: You play an 1kHz tone on your Subscope theremin. The fixed oscillator is at let's say 482kHz and the variable at 481kHz. Thus you'll find all three frequencies (1kHz, 481kHz, 482kHz) in the output signal.

Now the sound card samples this mix with a 96kHz square wave signal. Thus you'll get
- your 1kHz plus 95kHz (96-1) and 97kHz (96+1)
- your 481kHz plus 577khz (481+96) and 385kHz(481-96)
- your 482kHz plus 578khz (482+96) and 386kHz(482-96)

And with the 5th harmonic of the 96kHz square (=480kHz) wave:

- your 1kHz plus 479kHz (480-1) and 481kHz (480+1)
- your 481kHz plus 1khz (481-480) and 961kHz(481+480)
- your 482kHz plus 2kHz (482-480) and 962kHz(482+480)

I've drawn in green your original signal, ghost tones in orange and a "new" RF signal which risk to stray back into the theremin and trouble the pitch oscillator in red.

And that's only a little simplified excerpt from all combinations of sampling harmonics and oscillator signals. And there is still the volume oscillator which may stray other things in...

Conclusion: Digital recording of a theremin needs perfect low-pass filtering before sampling AND perfect grounding and shielding to avoid stray interference - a thing which is very difficult to realize!

i appreciate the explanation, but I am having a hard time understanding what I need to do to not have these tones in my recording.

another sounds card? Maybe I need to run my theremin on batteries to avoid grounding problems?

"Conclusion: Digital recording of a theremin needs perfect low-pass filtering before sampling AND perfect grounding and shielding to avoid stray interference - a thing which is very difficult to realize!" - Thierry

I agree with everything you say Thierry -

But I dont think its quite as hopeless as you (unintentionally, I believe) seem to convey. "perfect low-pass filtering" and "perfect grounding and shielding" are impossible to obtain - but IMO "perfection" is not required - "Extremely good" is usually good enough.

I believe that your statement "Due to lazy output filtering in the theremin" is the primary source of the problem.

Alas, care and attention to detail is seriously lacking in most theremin designs when it comes to interfacing the output to studio equipment - they are designed to drive an analogue amplifier, with the fact that technology has changed since Lev's days, and that one should now expect that users will want to interface their theremin to digital (sampling and class D amplifiers for example) being overlooked or ignored.

If the final output stage of a theremin incorperates a good filter, and if the connection between this stage and the output socket is correctly shielded, then in my expierience almost all Nyquist related issues drop to acceptable levels.

Other things which greatly reduce the problems is high quality screened lead from the theremin, and ensuring that the primary ground connection between the theremin and equipment is not via the screen (in fact, I find that best results are obtained if the screen is disconnected from the plug at either the theremin OR the amplifier end (NOT BOTH!), and a seperate ground wire is run from the theremin to the equipment - this way there is no conduction path through the screen so no unwanted inductively induced ground signal coupling to the signal wire can occur)

With the above, if the theremin is kept a reasonable distance from the external equipment (so that radiated signals are not picked up strongly by this equipment) one can get a clean digital recording.

But it all starts at the theremin output - If the signal here is full of HF, nothing one does will have a major impact.. But even so, good screened cable, seperate ground wire (and perhaps removing screen connection to plug at amplifier OR theremin end) and distancing the theremin from the equipment, can improve things somewhat.

Fred.

Regarding the earth connection -

If one does not have easy access to an earthing point on the theremin or equipment, one can take this off the plugs, and run the seperate earth wire along the lead.. I will draw a diagram and post it here ASAP.

It is actually better to run the ground wire as a seperate wire away from the screened cable, and certainly DO NOT wrap this wire 'round the screened cable.. Coupling will be greatly reduced because the screen is between the ground wire and the signal wire, but this will not completely eliminate any inductive coupling (only the fact that the distance between the signal wire and any ground current flow has been increased, will reduce inductive coupling).

Also - Good plugs with metal shells NOT plastic shells! - It really is worth spending money on the lead!

What Thierry and Fred said.
This would be my luddite short term solution: output from theremin to a guitar (type) amp and record the output from the amp speaker with a good quality mic to PC soundcard.

Ok, I know that this is a work around but at least in the short term it would get you there.

Apologies if I haven’t answered in the spirit of the correct solution!

R

"output from theremin to a guitar (type) amp and record the output from the amp speaker with a good quality mic to PC soundcard." - RoyP

In many cases, this is the only solution..

"another sounds card? Maybe I need to run my theremin on batteries to avoid grounding problems?" - Amey

You could play about with sound-card settings - with different sample frequency and a lot of luck, the tones might be shifted out of hearing.

Battery operation is not likely to help - there is no way* to avoid ground when interfacing to grounded equipment.

*not absolutely true - optical interfaces have no ground coupling... But they convert signal to digital form, so wont solve the problem.

Fred.

SOME SOUND CARDS / INTERFACES ARE BETTER THAN OTHERS ...

some high-end soundcards and audio-digital interfaces have better input filters than others - Good filtering is expensive, so cost is cut by minimizing these filters. It is really the responsibility of the equipment providing the audio signal, to provide AUDIO, not RF!

NOTE TO ALL THEREMIN DEVELOPERS!

Digital is here, users will want to interface their theremins to digital equipment! It is time we all faced this reality and put good filtering on the output stage of our theremins!

Hello Amey,

Thierry's explanation is excellent. I use to have this issue and eliminated it (I think) by using "several" tricks of the trade. If you are not using one, you will need it sometime in the future. Get a "ground loop hum filter" which uses a lower frequency audio transformer for isolation. If it does not eliminate your current issue it will help with other things in the future.

I use one from Walmart Online $15, One side RED between the theremin and my EQ-preAmp Input and the other side of the filter White from the EQ-preAmp Output to the computer sound card. Theremin sound is very simple so it does not create any noticeable change to the sound. I use my grandson for these experiments as he has youthful hearing.

Christopher

Edit: You will need the theremin to be on its own direct earth ground connection as it won't get it from the amplifier. One day I will post a webpage how I do this mixing of the Walwart DC and Ground connection in a little black box at the power source. Then I use a single 1/8" plug on a 6' cord to my theremin.

Edit2: The way I analyze sound visually on the computer your harmonic balance is ideal and your signal to noise is excellent. If that sample is bad you make me worry everything I do is in vain. LOL

That does not sound like an interfering power supply, batteries i guess won´t be a solution like Fred already mentioned. I just can tell about the way it works for me with recording (semi-) directly to the computer (older MacBook Pro): i play through a Boss DD7 delay/looper as a signal splitter. It has two outs, one is routed to the guitar amp (and serves the grounding), the other one directly to the input of the computer. I never tried to record to the computer without that device because this way it is easy to get the grounding, the monitor and the effect if i like to. Thus i have not compared and cannot secure if this might be a solution in general. But well possible that the output stage of the Boss DD7 might be the more sophisticated one compared to the subscopes in digitally recording cases.

"(older MacBook Pro):" - Dominik

My understanding is that the older MacBook had a really good audio interface - Macs in general were the choice for audio work..

But - The fact is that it cannot be the Mac which is sorting the problem out, as you are going via the DD7 which itself is A/D D/A input/output.. If there were any Nyquist related issues at the DD7, the audio sidebands would appear at the Mac input, and all the HF filtering in the world would not eliminate them!

So it seems that the DD7 is "behaving" and passing a clean audio signal to your Mac..

The question is, why !? ..

And here we have a range of possibilities.. DD7 has good input filters and/or DD7 sampling frequency is not conflicting with HF signals from the theremin (as in, not producing sidebands in the audio) or something else - grounding, proximity, screening, whatever - is making your results better than what Amey is getting.

For that matter, perhaps we are wrong in assuming the trouble is via the audio lead at all - It may be that the sound card is actually picking up radiated signal - it is quite easy for RF to de detected by some non-linearity at a component, and this component could even be down-stream of whatever filter is fitted.

Moving the theremin well away from the PC should test this possibility.

Fred.

Hi there,

I have used my best tools which would make any musician laugh, but my conclusion is what Amey hears is coming from her method of play back and not burned into the recording. There is one person I know here that has super sonic hearing....hello, a little input please. lol