Some stupid synth theory questions (timbre)

[Oscius]

OK, so I would like to know where I could find out the harmonic spectrum of diffrent instruments? (meaning the places of the harmonics and their relative volumes).

 

And a other question, why doesnt percussion instruments have pich?

 

And probably the stupidest question why does every synth have filters when a 7 band EQ can make any filter type adjustment that you can think of (well apart from some very wierd synth stuff I think all of the "normal" instruments have maxium of 7 harmonics and all of them have them in same multiples too... right?)

[Megakazbek]

 

And probably the stupidest question why does every synth have filters when a 7 band EQ can make any filter type adjustment that you can think of (well apart from some very wierd synth stuff I think all of the "normal" instruments have maxium of 7 harmonics and all of them have them in same multiples too... right?)

 

The synths have filters because they are the most basic tone shaping elements. An equaliser is actually a bunch of filters running together.

And again, the spectrums of almost all instruments are much more complex than 7 harmonics. Even a synthesized saw wave at 100 Hz has at least 200 harmonics in hearable range, and spectrums of natural instruments are pretty much imposible to describe in a simple comprehensible way. So the filters are there not for precise harmonic adjustment, but to give sounds general variation in timbre over time and in response to different controls like velocity.

[Antman261]

Okay, for a start you need to understand what a harmonic is.

 

Every note we hear has a fundamental frequency, say, A = 440Hz. Well the harmonics is are made up of multiples of that frequency. So, F1 (440Hz, F2 (880Hz). F3 (1220Hz) and so on.

 

Now not every frequency will be present, and in different proportions, and so on. Even the fundamental frequency doesn't have to be present, our brain can identify pitch from the frequencies we can hear, and also the apparent multiples between the frequencies. In fact, a pitch consisting of a single frequency is sort of ambiguous.

 

Anyway, this means that a chart of harmonics like you said is impossible, because they shift with pitch. Also, I imagine you're thinking about different components of the sound, say like, the attack of a piano, which probably couldn't be said to be at any particular pitch or made up of any certain frequencies for all instruments, because it depends on the make up of the instrument!

 

And generally, percussion instruments don't have pitch for a couple of reasons..

 

1. it is physically impossible to tune a drum head to a perfect pitch. You can get it close, but it simply cannot achieve perfect pitch.

 

2. They may not be perceived to have a pitch due to other pitches or frequencies being present that contradict any pattern of harmonics that might be formed from the fundamental frequency. It's simply too random and unrelated.

 

Sometimes this means the pitch is just ambiguous, sometimes it means it really doesn't have a pitch at all.

[Khazul]

 

And generally, percussion instruments don't have pitch for a couple of reasons.. 1. it is physically impossible to tune a drum head to a perfect pitch. You can get it close, but it simply cannot achieve perfect pitch. 2. They may not be perceived to have a pitch due to other pitches or frequencies being present that contradict any pattern of harmonics that might be formed from the fundamental frequency. It's simply too random and unrelated. Sometimes this means the pitch is just ambiguous, sometimes it means it really doesn't have a pitch at all.

 

 

You can tune any tunable/butcherable percusion intrument to ensure a desired pitch is prominent amongst those it produces. Not all percussion instruments consistently decay to a single fundamental however as you suggest. Most single skinned instruments however do and tuning those is certainly desirable, and even double skinned instrument are well worth tuning such that the frequencies of the skins are related (for eg a 3rd/5th etc out).

 

Have you (for example) never noticed when the pitch of the toms and kick (and even the snare sometimes) just clases horribly whatever key the rest of the band is playing in. Typical worst offender is when kick is tuned badly relative to close notes being played on a bass.

 

I frequently record and re-pitch drum and percusion loops to make them sit better with the music and actually treat them as notes rather than just noises, or pre-tune the sounds on my roland handsonic (which is what I mostly record percussion from).

[Antman261]

Oh, when I said the pitch could be ambiguous I meant to imply that you could still tune drums to be perceived as having pitch. Just look at Terry Bozzio! Or Bob Gatzen, he tunes his whole drum kit mostly in 4ths, 5ths and octaves.

[Oscius]

You can run samples of different instruments through a spectrum analyzer and measure it yourself. There is not much point to it though. With rare exceptions, spectrums of natural instruments are insanely complex and consist of thousands of harmonics constantly changing over time. That's why it's impossible to synthesize natural instruments on an analog or even additive synthesizer. For the instrument to have pitch, its spectrum must have a fundamental harmonic which means that the most powerful harmonics in spectrum should have frequencies that are multiples of fundamental frequency (which actually IS the pitch). Most percussion instruments have harmonics with pretty much random frequencies, not related to each other, so they don't have a fundamental harmonic and therefore pitch. There are pitched percussion instruments though, like marimba or xylophone. The synths have filters because they are the most basic tone shaping elements. An equaliser is actually a bunch of filters running together. And again, the spectrums of almost all instruments are much more complex than 7 harmonics. Even a synthesized saw wave at 100 Hz has at least 200 harmonics in hearable range, and spectrums of natural instruments are pretty much imposible to describe in a simple comprehensible way. So the filters are there not for precise harmonic adjustment, but to give sounds general variation in timbre over time and in response to different controls like velocity.

 

 

Thanks for good answers

 

But now I am a bit confused I thought that you cant have more than 7 harmonics running at the same time when you play for example a stringed instrument

 

And a other question, when you play high notes diffrent tones should start to sound alike when sustained due to the harmonics disappearing because we can only hear to 20Khz (meaning there is only one timbre for 20Khz?). This also means that the EQs/filters that have low frequency dont do a thing when you play higher note. Is this right?

[Antman261]

Most instruments we hear have fundamental frequencies from 30hz to maybe 1.5kHz.

 

However, you're right in the sense that as you go higher there would be less harmonics within our audio perception window. While a note at 40hz might have 500 harmonics within our audible range, a note at 400hz will only have 50.

 

And yes, as you change what note you're playing, it's relationship with the EQ will change. But we get around that problem by using wide enough bandwidths. You might be able to use a pretty tight bandwidth on something like a kick drum, but not so much on say a cello.

[Yoozer]

 

But now I am a bit confused I thought that you cant have more than 7 harmonics running at the same time when you play for example a stringed instrument

 

No, that's just your EQ's limitations. Perhaps you should take a look at additive synthesizers.

 

 

And a other question, when you play high notes diffrent tones should start to sound alike when sustained due to the harmonics disappearing because we can only hear to 20Khz (meaning there is only one timbre for 20Khz?).

 

Our ears don't stop at 20kHz, they simply turn into lowpass filters. What's left when you filter a signal nearly completely? Exactly - the lowest harmonic, which is a sinewave. The reason they sound alike is because there's less of a jump between 400 and 500 hz than there is between 15 khz and 16khz due to the logarithmic scale.

[mildbill]

 

OK, so I would like to know where I could find out the harmonic spectrum of diffrent instruments? (meaning the places of the harmonics and their relative volumes). ...

 

 

 

 

http://www.phys.unsw.edu.au/music/

 

 

Scroll down a bit to 'recent projects'. Then look under 'Introduction to the acoustics of musical instruments'.

 

There's some info and graphs that may be useful for various instruments.

[ChromaLord]

Dude, respectively you have a lot of misconception about audio-related physics.

Answers to harmonics can be found just about anywhere on line- start with wickepedia(spelling?) Hint: there are more than 7! Also everything in nature has a pitch- EVERYTHING.

[shaft9000]

percussion - excluding mallets - has pitch , but it's usually going from high energy to low energy. it's about frequency and amplitude.

 

most other instruments be they winds, strings etc have a much more constant energy applied to the note creating sustained pitch. all of the enregy applied to a percussion note is concentrated in the attack alone - no sustaining energy needed so the vibrations (frequencies) quickly decrease.

[xmlguy]

 

Dude, respectively you have a lot of misconception about audio-related physics. Answers to harmonics can be found just about anywhere on line- start with wickepedia(spelling?) Hint: there are more than 7! Also everything in nature has a pitch- EVERYTHING.

 

 

Even when in a vacuum, on the surface of moon, if you're deaf, if cooled to absolute zero, or if beyond your event horizon?

[veracohr]

Thanks for good answers But now I am a bit confused I thought that you cant have more than 7 harmonics running at the same time when you play for example a stringed instrument

 

Mathematically, any non-sinewave waveform can be described by an infinite series of sine waves in very specific ratios to each other. Look into Fourier series. This means that to get, for instance, a "true" square wave, you need an infinite series of harmonically-related sine waves. You can, however, approximate a square wave with a limited number of sine waves.

 

A synth will do it's best to make a perfect waveform. Whether it's analog or digital, there are physical limitations that prohibit a truly infinite frequency response, so even if you had a synth producing an unfiltered saw wave, you would not get a true, perfect saw wave at the output. And that's not to mention that most synths you can't actually bypass the filter, you can just raise the cutoff to where you don't notice it's effect. If the cutoff is above your hearing range, it seems like it's not there, but it is still filtering the signal (along with the physical limitations of electronic circuits.)

 

But if you were trying to produce a 100Hz square wave and were limited to only 7 harmonics, you wouldn't have a very good square wave. Harmonics are not something that are generated, they just are. Any non-sine wave waveform will have harmonics, which are limited electronically by the frequency limitations of the circuit and acoustically by the limitations of our hearing.

 

And a other question, when you play high notes diffrent tones should start to sound alike when sustained due to the harmonics disappearing because we can only hear to 20Khz (meaning there is only one timbre for 20Khz?). This also means that the EQs/filters that have low frequency dont do a thing when you play higher note. Is this right?

 

Our hearing has nothing to do with the harmonics present in a signal. Just because we can't hear them doesn't mean they aren't there.

 

Going back to the Fourier series I mentioned above, the idea of a harmonic series is sort of an abstract invention. If you have a 1kHz saw wave, you don't actually have an infinite series of sine waves harmonically related to a 1kHz sine wave, you just have a 1kHz saw wave. But it's a useful wave of describing a 1kHz saw wave that makes it easier to explain why the sound changes when you put a low-pass filter on the signal at 5kHz (to reduce the Fourier series concept to one particular application, audio).

 

If you read up on Fourier series and do some small experiments on a graphing calculator (ie. enter the formula for the first few harmonics of a square or saw wave and see how the graph turns out) you might understand better.

 

Let me put it a different way: say you have two different synths and you want both to play a square wave at 440Hz (A4). The first synth is a subtractive synth, and you set it to produce a square wave with the filter wide open and play A4. The second synth is an additive synth capable of producing 4 harmonics, and you play A4. The two different synths will sound different, because the oscillator in the subtractive synth will produce the best approximation of a square wave it can within its physical limitations, whereas the additive synth will produce only 4 individual sine waves, which combine in the proportions dictated by the square wave formula. But since this is only the first 4 harmonics, it won't be a very good square wave, and the signal from the subtractive synth will be closer to a "true" square wave.

[Tusks]

http://www.phys.unsw.edu.au/music/ Scroll down a bit to 'recent projects'. Then look under 'Introduction to the acoustics of musical instruments'. There's some info and graphs that may be useful for various instruments.

 

 

Thanks MB.