[Joe Bloggs]

My argument is that what is perceived as "bad" treble is often just an uneven distribution of quantity of different treble frequencies, eg too much at 8000Hz and too little at 2000Hz.
In other words, the human hearing is much more sensitive to frequency domain imperfections than time domain imperfections.
 
Some examples:
http://cnslab.ss.uci.edu/hearing/index.html
A negative finding for human hearing sensitivity to "edges" in sound
 
http://en.wikipedia.org/wiki/Audio_system_measurements
"The human ear is largely insensitive to phase distortion, though it is exquisitely sensitive to relative phase relationships within heard sounds."
 
http://www.ocf.berkeley.edu/~ashon/audio/phase/phaseaud2.htm
"Ohm's Phase Law in the 1800s claimed that the phase of a waveform has no effect on how the ear perceives it, and in typical circumstances, relative phase differences are difficult to perceive. One theory suggests that phase is altered by ordinary reverberant environments, being position-dependent, so our ears don't pay much attention to phase.
Despite these early beliefs, studies have been conducted demonstrating that phase distortion is audible, however subtle and specific to certain circumstances. However, many people claim that previous research shows phase distortion is not audible. They simply not read the more current research. Many loudspeaker designers are guilty in this regard. My assertion is:
Depends - Yes, phase distortion is audible under the right circumstances to certain people, but the concensus is it is rather subtle at best, especially in relation to other forms of distortion. Phase is chaotic in reverberant environments, yes, but the direct sound is not affected by reverberation. In some situations such as choral music in a cathedral from the back of the audience, phase is totally messed up, but in most other cases it still matters!"

 

[awinn17]

Ladies and gentlemen, I shall henceforth sit back, talk less and read more. Where is the bowing-to-sound-gods emoticon?! 

 

[ac500]

I should also elaborate on something I said:
 
> Like I said, this is actually not so hard to prove mathematically as long as we define mathematically what it means to be detailed and what it means to have good treble. My conjecture is that once we define "quality treble" and "detailed sound reproduction", it will be relatively easy to show that both mathematical descriptions are of the same exact thing.
 
This is not to say that treble is the only thing that matters, but rather that it's treble quality that defines this thing we call a "detailed sound".

There is absolutely a variation of qualities of true bass for example, however if you hear a good deep bass tone, you won't really call it "detailed". This is more likely described as "smooth" or "resolving" or clearly separated bass. To produce quality bass, the headphone diaphragm needs to be able to accurately extend long distances from equilibrium for long durations in such a way that produces very wide compressions of air. This won't really make a detailed bass sound, but rather a very precise and smooth and impactful bass that doesn't distort the rest of the frequency spectrum.
 
At most, bad mass could perhaps cause distortion or attenuation in the upper frequencies and therefore impacting the perception of detail, but again this would be due to an indirect effect on higher frequencies.
 
So to summarize: Detail is not everything, just as treble is not everything. The SRH940 IMO is an excellent example of a headphone with stunning detail and resolution but suffers some issues with bass. The bass still sounds quite detailed, but it doesn't sound nearly as natural as an HD650's deep bass.

 

[50an6xy06r6n]

I'm with the OP on this one. It seems like the treble affecting bass issue has been generally agreed upon, but here's my take:
 
1) "Detailed" bass really just describes textured bass. This in turn is caused by the shape of the wave signal being put into the headphone, which in turn is the sum of all the harmonics and stuff that make up the sound (this is shown in the audio sample given by Armaegis). Therefore, those higher frequency harmonics and sounds are indeed part of the sound that we perceive as the bass, even if we don't hear it directly.
 
2) The ability of your headphone to reproduce these sounds is dependent on how fast it can change its direction to match the wave input. (I.e., if you put in a square wave, its ability to match the sharp corners and produce the shape of that wave.) This corresponds to both the headphone's transient response, its "speed," since the ability to reproduce a lot of different sounds at once is also dependent on the driver's ability to match the input wave as closely as possible, and the input wave is the Fourier sum of all the different sounds and instruments on the recording, and how "good" its treble is, since treble frequencies are very high, and therefore are dependent on how fast the driver can change directions.
 
Theoretically, I guess, as long as your driver is fast enough to accurately reproduce the wave, you should be able to equalize the sound to your tastes. However, going back to the OP, if the driver cannot keep up with the input wave, the only way you can simulate more detail is to maybe increase the treble, which could emphasize the little peaks, but not really reproduce them, giving the illusion of more detail. However, it's important to note that "good" treble =/= a lot of treble.
 
Sorry if that was a bit long-winded, and if you disagree with my assessment, please feel free to try to correct me.
 
EDIT: Also, it seems to me that the good transient response results in the good treble, not the other way around. Basically, this all boils down to the speed and control of the driver IMO.

 

[ac500]

Every word you said aligns exactly with my mental picture I've been trying to describe here ^. So I think we're both pretty much in full agreement

 

[50an6xy06r6n]

Quote:

Every word you said aligns exactly with my mental picture I've been trying to describe here ^. So I think we're both pretty much in full agreement

Glad to know I've understood the argument.

 

[Tilpo]

The question then is whether the harmonics of a base tone reach all the way up to treble.

Say we have a 100Hz tone. Let's say this tone has 16 harmonics (which is a lot, btw). It would still only go up to 1.6KHz, which is mid-range and not treble. As a result I think good mid-range is more important for bass than good treble.
Additionally the 16th harmonic is going to be of very low volume, and I wonder whether the accurate reproduction of this harmonic has such a large effect on the overall timbre of the tone.

Other than that, I completely agree with you, ac500.

 

[50an6xy06r6n]

Quote:

The question then is whether the harmonics of a base tone reach all the way up to treble.
Say we have a 100Hz tone. Let's say this tone has 16 harmonics (which is a lot, btw). It would still only go up to 1.6KHz, which is mid-range and not treble. As a result I think good mid-range is more important for bass than good treble.
Additionally the 16th harmonic is going to be of very low volume, and I wonder whether the accurate reproduction of this harmonic has such a large effect on the overall timbre of the tone.
Other than that, I completely agree with you, ac500.

That's true, I guess the treble probably has a lesser effect on lower bass notes than a good midrange does, but I feel like the main point here is more that the speed of the driver, and consequently the good treble that comes along with that is what gives you a lot of that nice texturing and microdetail. 
 

 

[ac500]

Yeah, the bass point was just to make a point that frequencies higher than the one you "hear" are important. It's the most extreme example and obviously not the easiest to argue for the case, but by showing that bass does have some higher frequencies vital to its sound, it proves the whole theory in general I hope.
 
Anyway the point that this detail = transient response = quality treble = texturing = microdetail etc.

 

[Edoardo]

Quote:

If any of you are familiar with frequency domain transformations, this will probably make sense (I think too many people have the misconception that "treble" only effects high pitch sounds, and don't realize that every frequency contributes to the sound across the entire spectrum.)
 
By definition, a "detailed" headphone exposes the fine details, that is, the subtle quick curves and features of the sound waveform. By definition, this means in the frequency domain, the higher frequencies are better / not suppressed.
 
If you know what I'm referring to about frequency domain, consider the DCT of JPEG compression. As you lower the quality of the image, the higher frequency components of the (8x8 I think) freq-domain blocks are dropped off, resulting in a blurrier photo in general. Of course it's a bit more complex than that with color transformations, quantization etc., but you get the idea.
 
Lack of high frequency features, from a mathematical and practical perspective as far as I have ever seen = lack of detail.
 
I'm curious if there's something I'm missing. How can anything but treble have an effect on detail? And if nothing, why do people say "Be careful - good treble creates a false impression of detail!"
 
My impression is that when people refer to "detailed" headphones which may have recessed treble, they are referring to waveform amplitude precision / low quantization, and high quality Thighs, even if they're quiet.

 
 
 
 
 
 

The fact is that bass frequencies are omnidirectional, which means they are so big it seems they expand in every direction. That's why when you position a subwoofer, you care more about room reflections rather than putting it symmetrically against your ears as you would do with the other sound sources.
 
Our brains use high frequencies to tell us which direction the sound comes from. So... More treble, more direction information, more illusions of a better recording.
 
Another big issue are the instruments harmonics.
 
Sorry for quoting myself, but copy-and-paste is much faster
 
 
The thread I'm quoting was more about dynamics and loudness stuff, but there I tried to explain the harmonics/treble/soundstage/recording quality issue also. These four things are strictly linked as:

1. larger dynamic range allows you to get relatively lower-volumed signals
2. instruments harmonics are relatively lower-volumed and higher-pitched than their fundamentals

1. most of the higher frequencies we hear are relatively low-volumed

1. higher frequencies give our brain information about sound direction and the position of the sound source
    

Quote:

[...]
 
Every note produced by a musical instrument is composed of various, and theoretically infinite, harmonics.
 
In very poor words, the harmonics are the elements in which a sound can be divided, and of which a sound is composed.
When we play an "A" - which has a fundamental frequency of 440Hz - on a pitched musical instrument, we do not produce just a 440Hz sound wave, but a more or less complex sound composed also by a series of sound waves at 880 Hz, 1320Hz, 1760Hz, etc. depending on the instrument we are using.
 
They become increasingly less audible as their amplitude decreases as their frequency increases
 
The distributions, decreases and volumes of the harmonics in playing a note, give every instrument or voice their peculiar timbres, or colors.
 
Let me borrow a pic from the internet, showing the frequency range of the fundamentals and of the harmonics which some musical some musical instruments are able to produce.
 

 
Don't be misleaded by the way the graph is drawn down here, though.
Any musical instrument produces harmonics at any height. What's colored lighter here are just those frequencies that the instruments do not reach as a fundamental frequency, but only through their harmonics.

 
Of course, as you go upwards in the frequency spectrum, the harmonics play at increasingly lower volumes than their fundamentals. But yet, they are there. 
And it's them to give each instrument its peculiar sound, or timbre, or color if you prefer.
 
[...]
 
 
Dynamic compression, through a digital or an analog compressor, does not necessarily imply loss of information.
What does it do? It boosts lower-volumed signals, and/or flattens the peaks. For instance, electric guitars dynamic compressors boost the harmonics to get a fatter sound.
 
Sometimes audio compression is used to fool audiophiles, in the sense that boosting a tinkle, or a drum or a slap, their harmonics and/or their duration can make the illusion of the recording to sound better. In fact, in blind tests most people find compressed stuff more pleasant, just because wow, they hear every little detail so well, even if they don't realize what they're hearing is unnatural.
 
Nowadays, majors' discographic producers abuse of dynamic compression both for putting at high volume those details you want to hear, so that you don't to have to play with the volume knob in the car, or to hide details.
 
[...]
 
So... Testing anything, actually, it's just about listening to what's happening at the lower volumes.
 
What's there? Instruments playing in the background, their upper harmonics, environment reflections, the sound of those strings picked one or two seconds before that keep vibrating...
Mind that higher frequencies give our brains the direction from which the sound is coming, and therefore, as the harmonics expand upwards, the smaller or bigger richness of the harmonics can be "translated", by our brains or put into words as worse or better soundstage also.
 
 

 
 
It's not just about harmonics of course, but they play a very big role. I own a CD test  in which the same singer is recorded with a dozen microphones in front of her. Whereas there is the illusion that the recording environment changes from track to track, actually the engineer did not change the recording room, but just switched the microphones, which get more or less room reflections and/or different harmonic patterns...!

 

[ac500]

I should add a little bit of "anecdotal" evidence for this I don't think I mentioned before.
 
Due to scarring in my ear from sickness as a child, I have high range hearing "loss" in one of my ears. Well it's not bad... I can hear up to 16-17khz in my left ear, and up to 19khz in my right ear. The sensitivity of both ears is not really affected except for these higher frequencies.
 
With both ears, my brain uses my right ear to fill in the loss in my left ear, so that works out well. However, if I plug my left ear so I hear up to 19khz, everything sounds extremely detailed i.e. I can hear very very fine quiet details across an entire room (I've always had very sensitive hearing). If I plug my right ear, the sound is considerably more blurry. I can still hear it, but it's not as detailed.
 
This theory makes perfect sense not only scientifically but in experience: this blurry sound is the result of the lack of quality of these very high frequencies (15+khz typically) even though 15+khz is usually way beyond the usual range of frequencies you think of in music. But it matters a LOT... in detail reproduction, and that's what this thread is about.
 
In theory and in practice, I see everything pointing to the equivalence of "quality" treble to a detailed sound. The problem then is just how you define quality treble, but in any case, I've not seen anything that contradicts this theory so far and all mathematical proofs point in this direction as well.