wes480
New member
ya...so that makes skippy and my physics teacher that disagree with prorec...or, its a confusion of terms as you said slack. time to read the cave article.
skippy
New member
http://hyperphysics.phy-astr.gsu.edu/hbase/sound/db.html#c4
The human auditory apparatus is nonlinear as hell. It is very sensitive to signals in the human speech range (around 1kHz), and gets progressively less sensitive at higher and lower frequencies. It also has very different responses at high and low levles: the quieter the sound, the more profound the differences. That's why I pointed out the Fletcher-Munson equal loudness curves-there's a link to them on that page above.
Anyway, the just-noticeable difference for Joe and Mary Sixpack is usually regarded as 1dB SPL, and the "doubling of intensity" is usually regarded as 10dB SPL. That doesn't mean that you will always notice a 1.1dB change, or will never notice a 0.9dB change, or that a 10dB change will sound "twice as loud" to you. When you're talking about "percieved loudness", it is very subjective: trained listeners will usually have significantly different ideas about the JND when compared to the man-on-the-street. Those are the industry-standard rules of thumb, and they are at best a convenient approximation.
I can easily detect 0.5dB changes in the 1kHz range at 60dB SPL. I'm hard-pressed to detect 3dB changes at 30Hz at 60dB SPL. That's the nature of the hearing mechanism...
6dB is too small an increment to sound "twice as loud". However, expressed in voltages (not SPL!) a 6dB hotter signal _will_ have twice the _amplitude_ of a reference signal. How we perceive it, and how it looks on a voltmeter, are two very different things. I think that the author of the ProRec article is stating things in his own engineering space: that 6dB does give twice the amplitude, all right. But it doesn't _sound_ twice as loud. If he were to sit down with a good audiologist and have a hearing test done, his "twice as loud" (with respect to his _perception_) would vary dramatically across frequency and intensity: but it would _never_ be 6dB, I'd bet. He's thinking about this as a meter-reader just like me: but Joe and Mary Sixpack won't say that that 6dB hotter signal was "twice as loud".
That's why the question "how many dB sounds twice as loud?" is such a loaded deal: the answer is "it varies all to hellandgone". Once you insert some listener's highly subjective auditory apparatus in the path, it's all approximation and judgement calls.
It's much easier to look at the meters and make your judgement based on that- but that 6dB value isn't correct, for a percieved doubling in intensity, for the average listener. Make sense?
And here's that old thread in the Cave:
https://homerecording.com/bbs/showthread.php?s=&threadid=42960
And now I have to post this before the damned laptop dies and eats it again...
The human auditory apparatus is nonlinear as hell. It is very sensitive to signals in the human speech range (around 1kHz), and gets progressively less sensitive at higher and lower frequencies. It also has very different responses at high and low levles: the quieter the sound, the more profound the differences. That's why I pointed out the Fletcher-Munson equal loudness curves-there's a link to them on that page above.
Anyway, the just-noticeable difference for Joe and Mary Sixpack is usually regarded as 1dB SPL, and the "doubling of intensity" is usually regarded as 10dB SPL. That doesn't mean that you will always notice a 1.1dB change, or will never notice a 0.9dB change, or that a 10dB change will sound "twice as loud" to you. When you're talking about "percieved loudness", it is very subjective: trained listeners will usually have significantly different ideas about the JND when compared to the man-on-the-street. Those are the industry-standard rules of thumb, and they are at best a convenient approximation.
I can easily detect 0.5dB changes in the 1kHz range at 60dB SPL. I'm hard-pressed to detect 3dB changes at 30Hz at 60dB SPL. That's the nature of the hearing mechanism...
6dB is too small an increment to sound "twice as loud". However, expressed in voltages (not SPL!) a 6dB hotter signal _will_ have twice the _amplitude_ of a reference signal. How we perceive it, and how it looks on a voltmeter, are two very different things. I think that the author of the ProRec article is stating things in his own engineering space: that 6dB does give twice the amplitude, all right. But it doesn't _sound_ twice as loud. If he were to sit down with a good audiologist and have a hearing test done, his "twice as loud" (with respect to his _perception_) would vary dramatically across frequency and intensity: but it would _never_ be 6dB, I'd bet. He's thinking about this as a meter-reader just like me: but Joe and Mary Sixpack won't say that that 6dB hotter signal was "twice as loud".
That's why the question "how many dB sounds twice as loud?" is such a loaded deal: the answer is "it varies all to hellandgone". Once you insert some listener's highly subjective auditory apparatus in the path, it's all approximation and judgement calls.
It's much easier to look at the meters and make your judgement based on that- but that 6dB value isn't correct, for a percieved doubling in intensity, for the average listener. Make sense?
And here's that old thread in the Cave:
https://homerecording.com/bbs/showthread.php?s=&threadid=42960
And now I have to post this before the damned laptop dies and eats it again...
Last edited:
H
Harvey Gerst
New member
Slack has it right, but there has always been some confusion between a doubling in power, voltage, and loudness.
When you go from a 10 watt amplifier to a 20 watt amplifer, the power is doubled, but it doesn't sound twice as loud to the ear. Why?
The ear is capable of a great range of sound levels and the closest explanation of how the ear hears would be that it hears logarithmically.
If your talking about watts, you'd hafta go from a 10 watt amp to about 100 watts before you said that sounds twice as loud. That's about a 10dB increase in wattage. Every time you double the wattage , you go up 3dB in power. That 10dB (10X) increase only sounds to your ear like you've doubled the signal level.
Voltage is calculated differently and the signal increases 6dB when you double the voltage. But people don't hear in terms of voltage increases - they hear in terms of loudness (how loud does it sound) and that's also usually measured in watts or dB when you're talking about amplifiers or the ear.
If you consider 3dB as being double, you'll be pretty safe, unless you're talking about voltage (which most people don't). But if you're talking about how loud something is, relative to something else, a 10dB increase in power sounds about twice as loud as before. A 3dB increase doesn't sound very much louder.
When you go from a 10 watt amplifier to a 20 watt amplifer, the power is doubled, but it doesn't sound twice as loud to the ear. Why?
The ear is capable of a great range of sound levels and the closest explanation of how the ear hears would be that it hears logarithmically.
If your talking about watts, you'd hafta go from a 10 watt amp to about 100 watts before you said that sounds twice as loud. That's about a 10dB increase in wattage. Every time you double the wattage , you go up 3dB in power. That 10dB (10X) increase only sounds to your ear like you've doubled the signal level.
Voltage is calculated differently and the signal increases 6dB when you double the voltage. But people don't hear in terms of voltage increases - they hear in terms of loudness (how loud does it sound) and that's also usually measured in watts or dB when you're talking about amplifiers or the ear.
If you consider 3dB as being double, you'll be pretty safe, unless you're talking about voltage (which most people don't). But if you're talking about how loud something is, relative to something else, a 10dB increase in power sounds about twice as loud as before. A 3dB increase doesn't sound very much louder.
Slackmaster2K
Gone
Ok, then, Harvey & Skippy....I know you guys are probably busy, but could you take a look at that article that wes posted a link to at prorec?
He is talking in terms of dbSPL here. Something he said caught my eye, and now that both Skippy and Harvey have posted, I feel safer saying that I think he's using the term "loud" improperly. He explicitly says, "an increase of 6db of a sound means twice as loud", and that "If you are boosting a signal by 6 db, that means you're doubling the amplitude."
He sure used a lot of math and sounded very right. But it appears to me that he is in error in this case. I do see why it would be beneficial as an engineer to know that +6db equates to twice the amplitude, but I don't believe that amplitude should be refered to using the word "loud", which seems to me to be a subjective unit of measure. In fact, from an engineer's perspective, the term "loud" probably doesn't mean all that much when raising a fader. I can't imagine somebody saying "turn that guitar up so it's 1.2 times louder".
Slackmaster 2000
He is talking in terms of dbSPL here. Something he said caught my eye, and now that both Skippy and Harvey have posted, I feel safer saying that I think he's using the term "loud" improperly. He explicitly says, "an increase of 6db of a sound means twice as loud", and that "If you are boosting a signal by 6 db, that means you're doubling the amplitude."
He sure used a lot of math and sounded very right. But it appears to me that he is in error in this case. I do see why it would be beneficial as an engineer to know that +6db equates to twice the amplitude, but I don't believe that amplitude should be refered to using the word "loud", which seems to me to be a subjective unit of measure. In fact, from an engineer's perspective, the term "loud" probably doesn't mean all that much when raising a fader. I can't imagine somebody saying "turn that guitar up so it's 1.2 times louder".
Slackmaster 2000
skippy
New member
Oh, I did- it's a very well-written article, and I'm going to keep that link around to use in explanations in the future. It'll save me a lot of typing!
That "6dB is twice as loud" error was the only error that stuck its head up to me. And I understand why the author made it: like I said, he's a knobtwister/meterreader like me. He's not the first person to mistake "level" for "loudness", and he won't be the last: the problem is with his choice of words to describe what he's doing. He's just exactly correct in the "level" department, but he describes it as if it is correct for *perceived loudness*. Not so, regrettably. In this case, the physics prof and I (and a lot of psychoacousticians) win.
Other than that, I have no stones to throw at him: he attacked a nasty subject with humor and a good theoretical background. Fat-fingering one part of it is no suprise: there are textbooks that do _much_ worse...
As recording engineers, we spend a lot of time worrying about levels: we very quickly come to make the voltage levels that we read with our VU meters be the basis for our frame of reference. We mix with our ears, but meter with our eyes- and learn to largely ignore the differences, or just incorporate them in our working styles. So to us, 6dB being twice as big makes sense: one more bit in linear PCM, twice the amplitude. We learn to ignore the fact that a 6dB move on the master fader does a lot less than "doubling" what we hear. It's just too damned subjective, and we _know_ that that meant the amplitude doubled- so we scale what we hear with that knowledge.
There's one other nasty thing about trying to translate meter levels into dB SPL: the speakers (even the very best monitor speakers) are, if anything, even *more* nonlinear than our ears. Here's an example: you're mixing the Pure Sine Wave Orchestra's latest CD on your Tannoy Reds, and you look at your Rat Shack SPL meter- and it says that you're monitoring the tune "Constant Drone" at 95dB(c). You look at the master VUs, and you notice that in your boredom, you've let the levels drop- you're way down at -6dBFS. Well, hell yes, give 'em what they want even if it is boring: so you push up 6dB on the master faders.
It doesn't sound much louder, so you look at your Rat Shack meter, and you find that you're now monitoring at 97dB(c). What? How the heck did _that_ happen?
Your 6dB *level* shift only resulted in a 2dBSPL increase in the room: because your speakers are getting out of their linear region monitoring that loud, and are gonna need a _lot_ more power to give you 6dB more SPL. Their poor little cones are just a blur.
Your perception of the sound, the metering of the sound, and the transducer producing the sound all have their nightwierdies... The VU meters win, for us, once we have 'em calibrated: that's what tells us if we're using all the dynamic range on the gear, and what's printing on the tape or the disk. But what the meters say, and what we _hear_, will forever be related only by hours of practice and good intentions, not by any set of numbers. Does that make sense?
That "6dB is twice as loud" error was the only error that stuck its head up to me. And I understand why the author made it: like I said, he's a knobtwister/meterreader like me. He's not the first person to mistake "level" for "loudness", and he won't be the last: the problem is with his choice of words to describe what he's doing. He's just exactly correct in the "level" department, but he describes it as if it is correct for *perceived loudness*. Not so, regrettably. In this case, the physics prof and I (and a lot of psychoacousticians) win.
Other than that, I have no stones to throw at him: he attacked a nasty subject with humor and a good theoretical background. Fat-fingering one part of it is no suprise: there are textbooks that do _much_ worse...
As recording engineers, we spend a lot of time worrying about levels: we very quickly come to make the voltage levels that we read with our VU meters be the basis for our frame of reference. We mix with our ears, but meter with our eyes- and learn to largely ignore the differences, or just incorporate them in our working styles. So to us, 6dB being twice as big makes sense: one more bit in linear PCM, twice the amplitude. We learn to ignore the fact that a 6dB move on the master fader does a lot less than "doubling" what we hear. It's just too damned subjective, and we _know_ that that meant the amplitude doubled- so we scale what we hear with that knowledge.
There's one other nasty thing about trying to translate meter levels into dB SPL: the speakers (even the very best monitor speakers) are, if anything, even *more* nonlinear than our ears. Here's an example: you're mixing the Pure Sine Wave Orchestra's latest CD on your Tannoy Reds, and you look at your Rat Shack SPL meter- and it says that you're monitoring the tune "Constant Drone" at 95dB(c). You look at the master VUs, and you notice that in your boredom, you've let the levels drop- you're way down at -6dBFS. Well, hell yes, give 'em what they want even if it is boring: so you push up 6dB on the master faders.
It doesn't sound much louder, so you look at your Rat Shack meter, and you find that you're now monitoring at 97dB(c). What? How the heck did _that_ happen?
Your 6dB *level* shift only resulted in a 2dBSPL increase in the room: because your speakers are getting out of their linear region monitoring that loud, and are gonna need a _lot_ more power to give you 6dB more SPL. Their poor little cones are just a blur.
Your perception of the sound, the metering of the sound, and the transducer producing the sound all have their nightwierdies... The VU meters win, for us, once we have 'em calibrated: that's what tells us if we're using all the dynamic range on the gear, and what's printing on the tape or the disk. But what the meters say, and what we _hear_, will forever be related only by hours of practice and good intentions, not by any set of numbers. Does that make sense?
wes480
New member
skippy said:
HAHAHA. you guys will not believe this - but, you know...me..who started this thread questioning what my professor said in class?
Yeah....thats my profressors web site/article. Heh, it's his pride and joy. TONS of info on there.
Small world. Want the proof...GA State...see me...from Atlanta? Woo!
skippy
New member
Well, there you go. His web site is nicely done, and it agrees with my copy of Tremaine's "Audio Cyclopedia" and the texts I had to buy during my misspent youth...
Or, at least I think it does. Those reference materials are on my bookshelf in Colorado, and I'm in California for the week, so I'm working from a rapidly-aging and _very_ fallible memory. So I could be completely braindead wrong here too- but I don't think so.
But for other references, try this Google search: http://www.google.com/search?hl=en&q="Fletcher-Munson"+phons+sones&btnG=Google+Search
Such fun...
Or, at least I think it does. Those reference materials are on my bookshelf in Colorado, and I'm in California for the week, so I'm working from a rapidly-aging and _very_ fallible memory. So I could be completely braindead wrong here too- but I don't think so.
But for other references, try this Google search: http://www.google.com/search?hl=en&q="Fletcher-Munson"+phons+sones&btnG=Google+Search
Such fun...
Similar threads
- K-dub
- MP3 Mixing Clinic