[bigshot]

Flutes are that quiet, when?  (Actually, they are at G4, I'll give you that.)  But 30 dB down is under the noise floor when close mic'ed, in the studio, or even for a solo flute recording, or...?  Recordings are never getting better than 30 dB SNR capturing flute?

You can always safely assume that I am talking about playing music on your stereo in your living room. Normal listening volume for a flute would be well under 45dB. Subtract 30dB from that and you're down in the typical noise floor of a suburban living room. Assuming other instruments are playing along with the flute, and masking by lower octaves, you would never be able to hear an upper harmonic down that far.

In practice, the first three levels of harmonics, or about two octaves, are the most important. Beyond that, unless you're talking about cymbals or triangles or bells, the sound is generally at a very low volume level and can't really be heard in music at normal listening levels. Some engineers figure three octaves rather than two. That's fine too. This stuff is important when you're doing a multimiked, multichannel mix and you're EQing each instrument separately.

 

[mikeaj]

45 dB doesn't really sound like flute to me, but maybe I'm used to being in the same room, usually very close.  Maybe way out in the audience when they're not playing loudly, in the bottom octave, do you get 45 dB and lower.  But are people listening at home on their stereo setups to a solo flute recording at 45 dB?
 
What about other instruments with much higher harmonic content, like the trumpet example posted earlier?  The 9th harmonic was at around 10 dB down, and that's over 3 octaves above the fundamental.
 
 
Rules of thumb are good for general practice but not suitable for serious discussion and sound science, I would think.

 

[jaddie]

Quote:

This is ~0.5% THD on average from a few seconds of a simple whistle, which is basically like noise passed through a narrow bandpass filter:

 
 
As a THD figure, that's interesting but a little misleading.  Past the fundamental that's mostly second (even order) harmonic content, which is far less audible than odd-order.  That's if we're talking about harmonic content as a function of non-linear distortion, which isn't the same as when talking about spectrum of specific sounds. You can really refer to spectral content as THD.

 

[stv014]

Sorry, you are on my block list. By the way, my post was trying to show a simple "real" sound with very limited harmonic content, so including a THD figure (compared to a pure sine wave) made some sense.

 

[stv014]

Quote:

Generally, yes. However, square waves and sawtooth waves have extremely loud harmonics, with their 9th harmonic having the same amplitude as the 3rd harmonic of a triangle wave. Because synthesizers often make heavy use of square and sawtooth waves, this may change things a bit when it comes to EDM.

 
Of course, these waveforms usually do not appear in actual music in as pure square or sawtooth waves with no processing, they are normally filtered and manipulated in other ways to create complex sounds. This will obviously also change harmonic levels. But resonant filtering used in subtractive synthesis can also increase the level of high order harmonics significantly.

 

[jaddie]

Quote:

...Normal listening volume for a flute would be well under 45dB. Subtract 30dB from that and you're down in the typical noise floor of a suburban living room. 

 
A flute at 45dB would be deeply background music.  Normal face-to-face conversation is 65-70dB SPL, I would expect a flute at 1 meter to be above that easily.  If we're talking about noise floor of rooms, we should probably be using NC figures, not unweighted SPL.  That's the only way a living room could ever hit 15 (that would be NC15), LF noise would keep living rooms well above that.  But that would be an unusually quiet living room.  The average (27 room average - Fiedler/Cohen) is just a shade over NC20, at least until the HVAC blower turns on.  
 
I couldn't find a reference to a measurement of a flute, but a guess would be at least a few dB SPL over conversation level, and of course, distance dependent, but lets say 75dB at 1 meter.  That would but your harmonic at 75 - 30dB, 45dB SPL, not exactly buried in the room noise.

 

[bigshot]

Normal listening level for music is between 60dB and 70 dB. The background noise of an average living room is about 35 dB to 40 dB. A flute is probably not going to be the loudest instrument in music. Subtract 30 dB from it and you're right around that noise floor.

 

[bigshot]

Quote:

Rules of thumb are good for general practice but not suitable for serious discussion and sound science, I would think.

 
You can always safely assume that whenever I speak, I am talking about conditions similar to those of a regular human being listening to his stereo in his living room. I have no interest in pure theory- only theory as it applies to getting better sound out of my stereo.

 

[jaddie]

Quote:

Normal listening level for music is between 60dB and 70 dB. The background noise of an average living room is about 35 dB to 40 dB. A flute is probably not going to be the loudest instrument in music. Subtract 30 dB from it and you're right around that noise floor.

So, you're listening to a 25 - 30dB S/N ratio?  Wow.  A cassette tape without noise reduction beats that by 10dB.

 
Might it be that somebody's figures are a little bit off?

 

[AKG240mkII]

Quote:

Engineers generally concern themselves with the first three levels of harmonics. That's two octaves above the fundamental. Above that most harmonics are at such a low level, they're inaudible. Only percussion instruments have harmonics up high.

Imagine you convert from a 'lossless' to a 'lossy' file-format  : Guess what is going to be deemed 'irrelevant' and 'compressed' away, to give you that nice small file ?.
 
I've done the experiment with Mozart's Don Giovanni - Conducted by Otto Klemperer, Cologne 1955 .
It's a recording that I've listened to for more than 30 years, on LP and CD, now on computer .
The CD's sound nothing like the LP's and the mp3's sound nothing like either ...

 

[jaddie]

Quote:

Imagine you convert from a 'lossless' to a 'lossy' file-format  : Guess what is going to be deemed 'irrelevant' and 'compressed' away, to give you that nice small file ?.
 
I've done the experiment with Mozart's Don Giovanni - Conducted by Otto Klemperer, Cologne 1955 .
It's a recording that I've listened to for more than 30 years, on LP and CD, now on computer .
The CD's sound nothing like the LP's and the mp3's sound nothing like either ...

Comparing lossless to lossy of any kind without specifying the bit rate of the lossy file is, indeed, irrelevant. 
 
When comparing CDs and LPs, you're not comparing the end medium only, you're comparing an entire audio chain, which can be very different. 
 
But, what's your point?

 

[AKG240mkII]

I disagree ..
'Lossless' and 'Lossy' are incompatible, Irregardless of the bit-rate ..
 
 
Quote:

When comparing CDs and LPs, you're not comparing the end medium only, you're comparing an entire audio chain, which can be very different.

Sorry, but my 'audio-chain' doesn't sound of anything !

 

[jaddie]

Quote:

I disagree ..
'Lossless' and 'Lossy' are incompatible, Irregardless of the bit-rate ..
 
 
Sorry, but my 'audio-chain' doesn't sound of anything !

You are welcome to disagree. Those that have tried comparing high-rate AAC to Lossless would be on the other side of that argument.
 
I wasn't talking about your audio chain, but rather the entire path the audio takes just to get from a master to CD or vinyl.  The paths can be completely different with deliberate and accidental equalization applied to either.  If you just listen to a vinyl recording then its CD release, you aren't comparing the vinyl to the CD, you're comparing the entire production path, which you wouldn't necessarily know about. 
 
But this is sort of pointless to discuss, we're both off topic.

 

[bigshot]

 

[bigshot]

Quote:

I've done the experiment with Mozart's Don Giovanni - Conducted by Otto Klemperer, Cologne 1955 .
It's a recording that I've listened to for more than 30 years, on LP and CD, now on computer .
The CD's sound nothing like the LP's and the mp3's sound nothing like either ...

 
Take your LP and using a good capture card, transfer it to digital audio.
Burn a CD of it.
Bring the CD into iTunes and convert it to AAC 256 VBR.
Set up a switchbox and preamp where you can balance all three line levels (LP, CD, AAC) and switch between them.
Let me know what you find out.
 
I guarantee you it will surprise you.