Harmonics in the "electrical" domain

[hollowman]

There is some controversy as to why the spectral content of vinyl records contain frequency "information" above what the analog tape recorder was capable of capturing.

A thought experiment:

If I have a piano that has a key which plays 15khz (primary freq. of that key's piano string), what happens to the harmonics as they:

(1) go thru the analog tape deck (with 18khz top freq.)?
(2) go thru the record cutting lathe (yes: the cutting lathe is downstream of the analog tape deck)?

For example, note the spectrogram here:
http://www.channld.com/vinylanalysis1.html

The author claims: (Above) Spectrogram (Loudness vs. Frequency vs. Time). Bell percussion, showing harmonic overtones extending to 96 kHz, which is the ADC Nyquist Limit at a 192 kHz Sample Rate

Bottom line: if the original tape recorder could only record up to, e.g., 20khz, why does the spectrogram show "harmonic overtones"?

Some audiophiles claim that "harmonic overtones" is mechanical distortion produced (ringing) from the vinyl cutting lathe and./or the playback system (i.e., turntable/cartridge/stylus).

Others claim that harmonic overtones propagate electrically. If a tape recorder plays that 15khz piano key (see Thought Experiment above), it will still ring in the "electrical domain" (e.g., in amplifiers and circuits of the cutting lathe) and produce higher-order harmonics (as seen the spectrogram above).

Thoughts?

 

[FactChecker]

Any physical component involved in the process can have a frequency response that influences the output. I do not know what the characteristics of a cutting lathe are, but it certainly might add harmonics.

 

[tech99]

Any physical component involved in the process can have a frequency response that influences the output. I do not know what the characteristics of a cutting lathe are, but it certainly might add harmonics.

Neither the tape playback nor cutter have low pass filtering.
The harmonics might even occur during tape recording, as a large amount of high frequency boost is required. I notice the higher harmonics, above 18kHz, are something like 70dB down, which is only 0.03% of the fundamental amplitude, in a system having about 10% distortion.
For professional tape recording I would guess at least 18kHz was obtained, due to the use of high tape speeds. In mono broadcasting days we had 18kHz from FM stations and recorders could handle it at 15 inches per second.

 

[hollowman]

I posted on a somewhat related issue WRT "image frequencies", here:
https://www.physicsforums.com/threads/image-frequencies.931256/

When creating digital pulses, note that images (err... "harmonics"?) are created purely in the "electrical domain".

For the topic in this thread [Harmonics in the "electrical" domain], note that overtones in the spectrogram "track" the freq content of the music signal. So as the bell percussion used in the song "Dreamer" decays, so does its ultrasonic content.

 

[Baluncore]

(1) go thru the analog tape deck (with 18khz top freq.)?

That quoted 18kHz is probably the -3dB point.

The magnetic recording head gap has a finite width and a sharp edge, so there will be many higher frequencies with lower amplitudes present in the recording.

The sonogram range you show starts at -30dB of full scale and goes to -100dB. The -3dB of fundamental is only the white part of that image. You would get a better idea of the transfer function and harmonic generation if a frequency sweep was used to characterise the system rather than a percussive impact with a microphone.

It is not possible to separate the recording and replay distortion from the original sound quality. Even if the higher frequencies were of sufficient amplitude, your ears cannot hear the harmonics you are concerned about. It is doubtful that you could sense any real harmonic content after the percussive impact abuse on your cochlea during the previous second of time.

 

[Baluncore]

In electronic systems, higher harmonics are generated by distortion of the fundamental sinewave. Where the distortion of positive signals is different to that of negative signals, it will generate even numbered harmonics. If the distortion affects the signal symmetrically, it will generate only odd harmonics. Audio processing equipment is linear and has very low distortion.
Sub-harmonics, at fractional frequencies below the fundamental, or multiples of the fundamental period, are only generated when energy or information storage is available.

Sampling a signal can be seen as a multiplication which is a non-linear process. It can generate alias frequencies if filters are not used to block some parts of the spectrum.
Image frequencies are generated in non-linear systems. The generation of an image frequency requires that two different signals be multiplied together to generate sum and difference components of the input frequencies. One of those may be called the image of the other.

Thought experiments based on false assumptions will lead you astray if you do not first understand the fundamentals. If something does not make sense you need to study the fundamental terminology of electronics, signal processing and Fourier analysis, before you extend your paralogical model into a heap of irreality.

 

[anorlunda]

Where the distortion of positive signals is different to that of negative signals, it will generate even numbered harmonics. If the distortion affects the signal symmetrically, it will generate only odd harmonics.

Thank you. I never heard of that, but now that you point it out, I see it must be true.

PF offers new things for me to learn every day.

 

[hollowman]

I'm not sure this was pointed out ... but I've dwelled on the topic a bit and the explanation may be this simple:

The tape recorder/player is band-limited. Let's say, 18khz is the highest freq. it can record/playback.

We record a very high-freq sound ... for the sake of argument, let's invent a piano key that's centered at 16khz. When this key is hit, its string will produce 16k and overtones (harmonics). The tape recorder will attenuate any signal over 18khz, so most of the overtones will be "cut off" in the recording.

When it's time to cut the record, the tape deck (with that 16k piano key audio signal) is connected to the lathe and we begin cutting. Although the lathe is also bandlimited (say 18k again), this is an electronic filter (in the lathe's preamp); the cutting stylus acts like a tuning fork and can vibrate to its mechanical limit (so it may very well vibrate at overtones well past the electronic filter cutoff).

Ditto case at the playback end (at the record player): when the stylus meets the groove with the 18k signal, the stylus may very well (naturally) ring at overtones. And this is reproduced in the generated electrical signal.
This "phenomenon" is seen in a lot of vinyl record spectrograms, as in the case of the prev. noted:

 

[hollowman]

On a related note...

Watch the following YouTube video (it's about spectral analysis of the 1985 digitally-recorded Beach Boys album, which was released on vinyl):



The vlogger concluded that post-22khz content is "inaccuracies in vinyl" or "cartridge mis-tracking".
But I'm not sure I buy that. For one, he did not repeat the experiment with another turntable/cart. Or he didn't attempt to re-align the cart. Etc. ,etc.
What could be some reasons for the post-22k ultrasonic content shown in the video? HINT: See my last post ;)!

 

[Tom.G]

That spectrum looks an awful lot like (very) minor overload distortion that could be anywhere, or cumulative from everywhere, in the signal chain from the digital-to-analog (D/A) output of the original digital recording to the A/D convertor in your computer. That level of 60db to 80db below signal is 0.1% to 0.01% of signal. Distortion of 0.1% might be detectable by ear in an A-B test of a pure sinewave... but don't bet on it in your local bar!

Here is a rather high-end 16 bit A/D from 2005 that guarantees "No missing codes to14 bits".
https://www.analog.com/media/en/technical-documentation/data-sheets/ADADC71.pdf

 

[anorlunda]

I think the OP questions have been adequately answered.

Thread closed.