can you understand if its a single note or a chord playing?

i know how to find the fundamental note its the first big peak but what else can you identify looking at that?

can you also understand if its a single note or chord playing?

 

Too many possibilities to guess.
But why would you want to figure it out from a picture instead of stuffing it into your ears.
At least your ears tell you instantly whether it's a chord or an oversaturated sound, etc.

But if my life depended on guessing... I'd guess there's a chord in there somewhere and it's probably Ab(m)maj7
OK I'm dead.

 

the video by bassdude the first one i'm watching now is very good

actually the one i provided is a female voice sample so its just a note

 

OK so I'm dead in double quick timeunless she's a very talented alien.
Any chance you could post the sample.
It would be interesting to see what overtones Melodyne revealed.

 

Not at all. Voxengo Span is an interesting beast. It presents itself as an analyser with the frequencies on the bottom and even RMS peaks on the right.
However, it has the facility to detect chords and individual notes. You approached it perfectly normally. If I did not know a little bit about it, my response would be almost identical. It used to be free and I think it still is?

 

At which point overtones are strong enough to count as separate notes? Is it a semantic trick to say every sound that isn't a sine wave (or inharmonic noise) is a chord?

 

Agreed, it's a nice ambiguity to ponder.
In most contexts, our ears don't have too many problems imposing a preferred choice and deciding which is which.
But I wonder how the algorithms in something like Melodyne go about choosing?
Maybe if the ascending frequencies emerge in a similar order to the harmonic series it calls them overtones,
whereas if they emerge NOT in that order then more likely to classify them as notes from chords.
Maybe also clues/hints from relative magnitudes
Only wild amateur guesses - those algorithms will be insanely complex

 

The best A.I. detectors available still have flaws and I agree the mathematical number crunching it would have to do matching frequency to pitch would be no small feat. Unfortunately, where it falls down is in the human aspects. It is great if you have a rigid undynamic computer generated part. If you have an acoustic guitarist who has new strings doing hammer-ons, string slides and picking noise, it tends to take the frequencies of those noises and turn them into notes. You can sometimes spend as much time discarding those as isolating what is actually a note or chord. As you indicated, your ears are your best friend there

 

here it is

https://vocaroo.com/18N4ELQgUH07

its the first singing note actually



its actually this one below but i made it acapella using an online site....btw i can hear a small leakage note from the piano in the acapella so maybe not the best example

 

Last edited by a moderator: Nov 9, 2021

I think it was Dan Worrel? ....I could be wrong, but someone did an analysis on reverbs and found they can generate additional pitches. if you think about it that makes perfect sense. Basically, a pitch detection of any make or model seems to work best with a dry signal

 

i don't know who Dan Worrel is,

its
Ava Max "Into Your Arms" (Acoustic Cover)

yes it would be slightly different without the reverb and i can also hear a piano note,
acapella extraction wasn't perfect

 

Technically he is exceptional with his technical knowledge of frequency, loudness, EQ and all the technical aspects most people using a DAW or plugins might want to know. He digs right into the deep end on why some things work better than others and the uses of them.
Worth a look. I am sure you will find at least one video that is useful. It also helps if I spelt his name correctly, sorry about that. he has a video on Voxengo Span mixing settings in there somewhere.
https://www.youtube.com/c/DanWorrall

 

seems legit i trust ppl not showing their face

hehehe must be good will bookmark it thanks!

 

So this is what I got from Melodyne.

- Upper half is asking Melodyne to reveal as much as possible about overtones.
- Lower half is asking Melodyne to ignore the overtones.

Fascinating overtone mess at the beginning. Intriguing how Melodyne doesn't just ignore that mess
but seems to actually replace it with the F# that it has more confidence about later.

Side note: I don't feel quite so bad now about my guessed Abm(maj7) chord.
That initial mess of overtones at the beginning can accommodate that chord (at a stretch )

Thanks very much @BaSsDuDe for showing me Voxengo Span
I hadn't used it before but definitely enjoyed exploring it as part of looking at this little puzzle.

 

You shouldn't. This also is not helped by the reverb which has (hidden) additional almost unheard notes.
And you are most welcome

P.S. I am kicking myself that I never saved something I read ages ago. There is a super clever guy out there who created a harmonic chart for every instrument and their range notes in the orchestra, saxophones, guitars and voices.
That is the harmonics you normally do not hear but are present across most of the frequency range. The chart is massive and I never saved it. I'd almost go as far to make an educated guess you are hearing the maj7 due to a harmonic being made more present by the verb.

 

Last edited: Nov 9, 2021

@Ad Heesive - The thing that was great about that chart is say for example you want to harmonise Contrabasses in their lower registers which you avoid other than perfect intervals, is that by having the chart, you can see where the harmonics clash and they blend. This means rules you would not normally break because they sound like mud, with that chart, you could potentially.

Theoretically, SPAN can do that I suppose. I guess I will have to put more time into it.

 

Definitely two notes. Look at phase cancellation between 2 and 3 kHz and then 6 and 8 kHz.
Also probably a sampled instrument.

 

There's a misunderstanding here - my guess was based only on a visual scrutiny of that Voxengo screenshot - I hadn't heard anything at that point. I didn't know whether the screenshot was showing a chord or a sound saturated with overtones - that was the OP's challenge for us to figure out. I guessed (incorrectly) that it was a chord. But in fact it was just a vocalist.
Having since heard it - to my ears now it's just a wobbly vocalist trying to sing F#. I don't hear a chord at all. And there's no way in a million years I'd ever be able to accurately identify subtle overtones by ear.
But the scruffy overtones revealed by Melodyne did offer some support for my visual interpretation of that Voxengo screenshot.

 

So curiously enough (at least for me), that's not the case. Wonder where the phase cancellation comes from then.