[bigshot]

Calibration is an objective measurement that doesn't take any perceptual issues into account. If you are going to start a mix on a stage in New York and finish it on one in Los Angeles, you want the frequency response of the two stages to be calibrated the same.

Fletcher Munson applies to every human being. The mixer in New York is mixing to his ears that are subject to Fletcher Munson, and the guy in LA is too. Most mixers work at a reference level that is fairly loud. That is what they mix to optimize. Fletcher Munson isn't an issue for either calibration or mixing.

Therefore, if you calibrate your response curve and listen at a fairly loud volume you are hearing what the people who made the recording intended.

If you want to playback at a low volume, you need to apply loudness compensation. You do that by turning on the loudness button. If you have a recent amp, that button works dynamically, so there is more loudness correction applied at lower volumes and less at higher ones. Just turn that on and you don't need to ever even think about it. It just does its job. Loudness compensation is applied by the listener. Not the engineers.

Is that clearer to those who are confused? (even the ones who are confused but don't realize it!)

 

[Steve999]

I read up on this a little last night and I believe it to be perfectly correct.

Thanks. This should be framed and a lot of the vitriol and misconceptions back and forth deleted. It could even go in the first post of the sticky thread, with a reference or two.

Calibration is an objective measurement that doesn't take any perceptual issues into account. If you are going to start a mix on a stage in New York and finish it on one in Los Angeles, you want the frequency response of the two stages to be calibrated the same. Therefore, if you calibrate your response curve and listen at a fairly loud volume you are hearing what the people who made the recording intended. Fletcher Munson applies to every human being. The mixer in New York is mixing to his ears that are subject to Fletcher Munson, and the guy in LA is too. Most mixers work at a reference level that is fairly loud. That is what they mix to optimize. Fletcher Munson isn't an issue for either calibration or mixing. If you want to playback at a low volume, you need to apply loudness compensation. You do that by turning on the loudness button. If you have a recent amp, that button works dynamically, so there is more loudness correction applied at lower volumes and less at higher ones. Just turn that on and you don't need to ever even think about it. It just does its job. Loudness compensation is applied by the listener. Not the engineers.

 

[bigshot]

Glad to clear it up! We should all feel free to ask for clarification if we need it.

 

[amirm]

When, according to decades of tests and measurements, the average human's hearing is most sensitive around 3kHz(+-0.5kHz)?

Harmonic distortions of the 1 kHz land in that region so it is fine from that point of view. From perceptual masking, you also want some distance between your masker (tone) and its distortions.

Finally, in a band limited channel, you want to leave plenty of room for harmonics to show up. The higher the test frequency, the less of these you will have.

 

[TheSonicTruth]

Calibration is an objective measurement that doesn't take any perceptual issues into account. If you are going to start a mix on a stage in New York and finish it on one in Los Angeles, you want the frequency response of the two stages to be calibrated the same.

Fletcher Munson applies to every human being. The mixer in New York is mixing to his ears that are subject to Fletcher Munson, and the guy in LA is too. Most mixers work at a reference level that is fairly loud. That is what they mix to optimize. Fletcher Munson isn't an issue for either calibration or mixing.

Therefore, if you calibrate your response curve and listen at a fairly loud volume you are hearing what the people who made the recording intended.

If you want to playback at a low volume, you need to apply loudness compensation. You do that by turning on the loudness button. If you have a recent amp, that button works dynamically, so there is more loudness correction applied at lower volumes and less at higher ones. Just turn that on and you don't need to ever even think about it. It just does its job. Loudness compensation is applied by the listener. Not the engineers.

Is that clearer to those who are confused? (even the ones who are confused but don't realize it!)

I created my own loudness curve for low to moderate volume listening in the Equalizer app on my mobile devices. It resembles very closely ISO 226:2003(recent update of Fletcher-Munson) in the mid-upper-mid ranges, but not nearly as much bottom or top boost. You can tell me I'm wrong to do that, if it feels better just to, but it's all water off a duck's a$$ to me.

 

[spruce music]

Getting back to the original question I think 1 khz is a fine place for a basic distortion measurement. I'd like more info, but picking one frequency the 1 khz isn't a bad choice. Why?

The 3rd harmonic is often the highest in distortion with most gear. It is an odd harmonic making it less desirable as the human ear doesn't like odd order distortion. So the 3rd harmonic of 1 khz is 3 khz which is where the ear is most sensitive and most likely to hear it.

 

[pinnahertz]

Getting back to the original question I think 1 khz is a fine place for a basic distortion measurement. I'd like more info, but picking one frequency the 1 khz isn't a bad choice. Why?

The 3rd harmonic is often the highest in distortion with most gear. It is an odd harmonic making it less desirable as the human ear doesn't like odd order distortion. So the 3rd harmonic of 1 khz is 3 khz which is where the ear is most sensitive and most likely to hear it.

While it's very true that 3kHz is where the ear is most sensitive and 3rd harmonic is often highest, the frequency of highest distortion doesn't often coincide with where it's most audible because the mechanisms that cause distortion are often frequency dependent, rising at the extremes of the audible spectrum. Well before the days when consumer audio equipment even had distortion specifications, professionals were testing and verifying distortion of audio transmission systems including broadcast transmitters, telephone lines and audio networks. Large audio transformers of the day exhibited far higher distortion at low frequencies, and the systems in general had a really hard time performing at all at 15kHz to the point that legally required testing allowed for higher distortion at the extremes of frequency.

Then, along comes consumer HiFi, and frequency-dependent distortion did not go away. I think the choice of 1kHz speaks much more to a sort of marketing approach, where choosing a nice round number that happens to be where distortion is likely lowest as the advantage in specsmanship. If you couple that with the fact that very old gear didn't even pass 50Hz or certainly 15kHz well, it seems to make sense that 1kHz was chosen for the single-spec THD frequency. And, even though none of that really applies well today, here we stand with a very old, yet very common, but technically meaningless, means of specifying distortion.

 

[Sphere 57]

When, according to decades of tests and measurements, the average human's hearing is most sensitive around 3kHz(+-10.5kHz)?

Sorry, I realise this is a long dead thread, but it is one of the top results when you search for 'equal loudness compensation'.
I believe the answer to your query is because at 1kHz the numerical value of dB is the same as the same as that for phon. So it is a useful frequency for anyone interested in both SPL and Loudness.

 

[gregorio]

I believe the answer to your query is because at 1kHz the numerical value of dB is the same as the same as that for phon.

While I’m not absolutely sure, I believe your answer to be incorrect because as far as I’m aware, the use of a 1kHz test tone predates the invention of the phon.

G