[VNandor]

Vowel type sounds are much lower, the highest voice is the female Soprano, which typically extends no higher than “top A” in choral music (~880Hz) or “top C” in operatic music (~1046Hz) and even the highest note on a piccolo flute isn’t as high as 4kHz.

The harmonics vary greatly between different wovels. I'm sure you know that for example the "o" vowel singed at 200Hz will have a different spectrum than an "i" wovel singed at the same pitch. I don't know if he is right about which vowels accentuate which part of the spectrum though and it can depend on how exactly the vowel is pronounced anyways.

 

[KinGensai]

But that’s the problem or rather, the two problems: 1. Can the “abrasiveness“ be referenced by a FR reading? You state ”abrasiveness” commonly refers to the 4-8kHz region but what potential property of a cable could affect specifically that region? For example, given certain conditions, say an impedance mismatch along with too small a gauge, a cable could roll-off the HF and potentially affect freqs even as far down as into the mid freq region but that would affect the whole region above where the freq roll-off starts, for example 4k to 22kHz and the greatest loss would be in the HF, the loss would not be limited to only 8kHz. And 2. Are you sure “abrasiveness” refers to FR, maybe just a lightly lower volume could result in the perception of less “abrasiveness” or maybe a difference in transient response or phase, although it’s hard to see how correctly functioning cables would cause either of these.

I don't know, I don't own the Mellianus or know whether Ryokan matched the gain on both his cables (I suspect not because he doesn't have a coupler). We have some measurements on the IEM from RAA, so by looking at that graph, there's a slight drop in the bass and treble after normalizing at 1k that could explain what Ryokan is experiencing IF his cable has a significant amount of resistance. From the looks of the impedance response on that graph it's quite linear, so I guess that possibility can be discounted. I tend to believe what you're saying though because that cable has to have an unusually high amount of resistance to cause that change.

Even in the rare example above (too small a cable gauge) of an actual audible difference, still “perception is not telling the truth” because the result would not be limited to the 4-8kHz region. In practice of course, pretty much every time an audiophile makes claims about audible FR differences caused by cables, typically using perceptual descriptions such as “brightness”, “harshness”, etc., the actual FR reading/measurement does not substantiate their claims.

Yeah, gotta have some measurements here, I agree.

I’m not quite sure how that trick helped, those ranges are not correct. “Sss” is a broadband sound made by the human voice, usually starting somewhere below 2kHz and extending to 10-12kHz (although it can vary considerably). Vowel type sounds are much lower, the highest voice is the female Soprano, which typically extends no higher than “top A” in choral music (~880Hz) or “top C” in operatic music (~1046Hz) and even the highest note on a piccolo flute isn’t as high as 4kHz. Men’s voices are of course much lower than women’s, the famous tenor “top C” is an octave lower than the soprano’s (so around 500Hz) and most of the time people making those sounds aren’t straining at the top of their vocal range, so generally All those sounds will be lower than 500Hz (or a lot lower). There are many online resources for listening skills training or free audio software containing a signal generator if you just want to learn what different freqs sound like.

G

As the trick was explained to me, the letter sounds are approximations of what elevated sound in those areas are in pink noise, so I use that as a point of reference to direct me to the relevant frequency range for EQing. It's not exact, but it's helpful for quickly figuring out how I want to EQ a new IEM.

 

[71 dB]

A trick that helped me was using letter sounds to index particular frequency ranges. Uu for 250Hz, oo for 500Hz, aa for 1k, eh for 2k, ee for 4k, sss for 8k, etc.

63 Hz "abdomen"
125 Hz "chest"
250 Hz "oo"
500 Hz "o"
1000 Hz "ah"
2000 Hz "a"
4000 Hz "ee"
8000 Hz "s"
16000 Hz "ts"

I’m not quite sure how that trick helped, those ranges are not correct.

G

It is not about the fundamental frequencies or overall bandwidth of the sounds, but the characteristic frequency ranges.

 

[bigshot]

I don't know how useful it is to assign a vocal sound to a frequency outside the range of vocals. It's easier to just play around with an equalizer and find the frequencies in music. Then you can get a feeling for where musical instruments sit.

 

[james444]

^ or consult this interactive frequency chart:

https://alexiy.nl/eq_chart/

 

[castleofargh]

^ or consult this interactive frequency chart:

https://alexiy.nl/eq_chart/

I'm so glad anytime I see that one still working(as the original source went down). I think I still have it somewhere saved on a hard drive as a web page, but I never uploaded it so it could be shared. Good guy Alexy did it, whoever he is.

 

[gregorio]

The harmonics vary greatly between different wovels. I'm sure you know that for example the "o" vowel singed at 200Hz will have a different spectrum than an "i" wovel singed at the same pitch.

Indeed but as I read it, the context of the statement was “listening”, a trick to help identify freqs when listening. With pitched sounds we tend not to directly hear the harmonics, we tend to hear the pitch of the fundamental and perceive the higher harmonics as timbral information rather than as pitches. A “oo” vowel at 200Hz will sound roughly the same pitch as say an “ee” sound, probably within 100Hz or so, you won’t perceive it as a 4kHz pitch. If you want to train your listening to identify pitches/ranges of pitches, then using musical notes or a sine wave generator is far more effective/accurate.

I don't know if he is right about which vowels accentuate which part of the spectrum though and it can depend on how exactly the vowel is pronounced anyways.

That’s an additional problem, different pronunciations/vocalisations (saying the vowel quietly or loudly for example) will change the spectrum of higher harmonics and different voices will also change it. Plus, some of the other sounds referenced, for example the “sss” sound, doesn’t have a defined fundamental and harmonics, it’s far more comparable to band limited noise which also varies considerably from person to person and according to how it’s vocalised.

63 Hz "abdomen"
125 Hz "chest"
250 Hz "oo"
500 Hz "o"
1000 Hz "ah"
2000 Hz "a"
4000 Hz "ee"
8000 Hz "s"
16000 Hz "ts"
It is not about the fundamental frequencies or overall bandwidth of the sounds, but the characteristic frequency ranges.

No, it’s not. As far as listening is concerned, it IS about the fundamental freqs, where a fundamental freq is clearly defined, such as with vowels. I’m not sure where you got that list from but it appears to be nonsense and not only due to the facts already mentioned! For example, the word “chest” obviously ends with the “s” and “t” sounds, the “s” sound is a band of noise (probably around 2-8kHz), the “t” sound is also a noise band but probably starting lower (around 1kHz) but extending to 12kHz or so and the “ch” is also a band, typically starting in the mid or high hundreds and extending to around 6kHz or so, so where does the 125Hz come from? ”Abdomen” is even worse because it is actually 3 different syllables, each with different freqs/harmonics AND none of them have any content at 63Hz! In spoken language there is virtually never any content below 80Hz (which is why mics sometimes have an optional 80Hz roll-off). There are some exceptionally low male voices but it’s still very rare for there to be content as low as 63Hz. With a typical adult male voice there probably would be anything below about 100Hz when pronouncing the word “abdomen”, with the potential exception of a “plosive” on the “b”. In fact, the first two items in your list are so absurd I wonder if you mean them to be different to the other items in the list, for example where you might physically sense those freqs (when using speakers at a high level) rather than vocal sounds which produce those freqs?

All the freqs I’ve given above are extremely vague, they vary significantly from person to person, also according to dialect and even vary very significantly from the same person. The same person may pronounce the letter “t” similarly to the action of spitting, which would produce content up to 16kHz or more, or on another occasion more like a “duh” type of sound which might not have much content above 5kHz or so. With all the vagaries and potential for such large variations, using vocal sounds as a reference for frequency/pitch training is arguably amongst the very worst content one could use IMHO, a sine wave generator or MIDI instrument would be far better, both of which can be obtained with free audio software or even more simply on websites, there are numerous online virtual pianos and sine wave generators available for example.

G

 

[71 dB]

No, it’s not. As far as listening is concerned, it IS about the fundamental freqs, where a fundamental freq is clearly defined, such as with vowels. I’m not sure where you got that list from but it appears to be nonsense and not only due to the facts already mentioned!

G

I won't argue with you because I get nothing out of it (I have realized I'm a "happier" person when I use my time on other things than arguing online. That's why I haven't been very active here lately), but I can tell you where I got that list from:



Maybe you should comment on that video telling everybody the videos on that channel are nonsense. I don't care.

 

[bigshot]

“I give myself such very good advice, but I very rarely follow it.” -Alice in Wonderland

 

[KinGensai]

@gregorio
I don't know how to explain this besides the trick working for saving me time when identifying freqs by ear. The audio university video is indeed where I learned it, and as you can see in the video the actual sound is pink noise with EQ applied. It's a form of mnemonics and it helps me when critically listening for the purpose of judging tuning and EQing to preference.

I had to do EQ sweeps at the start to figure out what freqs were causing problems. Other charts helped a bit, but that video helped tie everything together and land me very close to problem spots the first try vs hunting for it.

 

[gregorio]

Maybe you should comment on that video telling everybody the videos on that channel are nonsense.

Why? That video isn’t nonsense, what was nonsense was the interpretation of it!

Firstly, I draw your attention to the statement to which I was responding: “A trick that helped me was using letter sounds to index particular frequency ranges. Uu for 250Hz, oo for 500Hz, aa for 1k, eh for 2k, ee for 4k, sss for 8k, etc.” - What the video demonstrates is the use of a sort of mnemonics to very roughly “label” octave ranges (when boosting freqs within noise), not the actual “letter sounds” themselves, which in reality not only contain a far larger range of freqs but which in most cases have perceptually dominant freq content far outside the range they are being used to “label”, as already explained. Secondly, the video confirms my guess that the first two items are not in fact sounds/mnemonics as are the other listed items, but descriptions of where the sound might be physically felt with a full range speaker system. Additionally, the freqs listed are arbitrary, 63Hz (rounded from 62.5Hz) and the subsequent freqs are used simply because they result in easy to write octave multiples but in reality the 63Hz listing (for example) could be almost anything from around 20Hz to 80Hz or so, a two octave range.

It's a form of mnemonics and it helps me when critically listening for the purpose of judging tuning and EQing to preference.

Yes, I can see how, as “a form of mnemonics” in the context of boosting freqs (within broadband noise/sound using a default Q value), it could be a useful training trick to help identify very rough octave ranges, as a beginner and if other references aren’t available (or known). Personally, I never favoured employing this trick with students because of the potential for confusion between the actual sound and the “mnemonic”, a potential for confusion that only worsens as students progress, for example when considering “formants”. I (and my education establishment) therefore much preferred using analogous descriptive terms, such as “honk”, etc., along with references to aspects of instrument sounds, say the boom of a kick drum (or the impact), the wash of a splash cymbal, the snare portion of a snare drum (or it’s “body“), the impact of a closed hi-hat, etc. However, our students would already know these instrument references and we had easy access to demonstrate them. I realise that’s not necessarily going to be the case when trying to learn from scratch informally on your own, in which case the method described in the video could be useful, as long as the distinction is made between the actual sound and it’s use as “a form of mnemonic”.

G

 

[71 dB]

Why? That video isn’t nonsense, what was nonsense was the interpretation of it!

Firstly, I draw your attention to the statement to which I was responding: “A trick that helped me was using letter sounds to index particular frequency ranges. Uu for 250Hz, oo for 500Hz, aa for 1k, eh for 2k, ee for 4k, sss for 8k, etc.” - What the video demonstrates is the use of a sort of mnemonics to very roughly “label” octave ranges (when boosting freqs within noise), not the actual “letter sounds” themselves, which in reality not only contain a far larger range of freqs but which in most cases have perceptually dominant freq content far outside the range they are being used to “label”, as already explained. Secondly, the video confirms my guess that the first two items are not in fact sounds/mnemonics as are the other listed items, but descriptions of where the sound might be physically felt with a full range speaker system. Additionally, the freqs listed are arbitrary, 63Hz (rounded from 62.5Hz) and the subsequent freqs are used simply because they result in easy to write octave multiples but in reality the 63Hz listing (for example) could be almost anything from around 20Hz to 80Hz or so, a two octave range.

G

This illustrates why I avoid arguing online nowadays. You straw-man me telling me how I supposedly interpreted something and then you tell me I interpreted that something wrongly. How do you even know how I interpreted the video? All you have to work with is a list I copied from the video. I didn't give much interpretation of my own of it. You have the tendency to assume people mean different things if they don't use the exact same vocabulary you do. As I have told several times, I have done my university studies predominantly in Finnish (some material such as textbooks where in English and two math books were in Swedish* ) and even if I did them in an English speaking country, vocabulary is not always exactly the same everywhere. I used the expression "characteristic sound" and you use "a sort of mnemonics", but we both mean the same thing, and we both know, educated as we are, how the video should be interpreted! In fact, in my opinion you are the one here who has had the most struggle interpreting things (it shouldn't take much intellectual effort to realise the words "abdomen" and "chest" mean frequencies you "feel" in those parts of your body). You are welcome to suggest me better English vocabulary since that's not my strength as someone whose first language is not English, but please don't always assume the way I express myself in a language not my own indicates I don't know what I am talking about.

Hopefully this makes you understand why I am less active on this forum. I am just tired of this game of yours. BTW, 63 Hz is a nominal octave/third octave band center frequency. Sure, calculated frequency is 62.5 Hz (125 Hz/2), but nominal frequencies are used. As for "abdomen" being anything between 20 Hz and 80 Hz, I am not sure about that. Perhaps for the fattest people in the World it can be 20 Hz, but for most people it should be around 63 Hz. The question is how fat is the fattest sound engineer in the World?

* Arne Persson: Analys i en variabel, Arne Persson/Lars-Christer Böiers: Analys i flera variabel.

 

[bigshot]

This illustrates why I avoid arguing online nowadays.

 

[KinGensai]

@71 dB
Well, it was actually me that used the term mnemonics here, I used my original verbage as a colloquialism for mnemonics, I didn't anticipate that it would be taken literally as the sound of the spoken letter itself.

Gregorio has a tendency to interpret the content of an argument/statement very literally, which is comical at times, but understandably frustrating if English is not your primary language. I find it amusing how foreign English speakers show such uncertainty about their English when they end up using English better than many native english speakers.

 

[bigshot]

At the high end of that scale, the only thing to relate to is the sound of the letters spoken, and it doesn’t represent the sound at all. Ts doesn’t sound like the high frequency squeal of upper frequencies. I really don’t know the difference between two os and three. I think there are much better ways to learn what frequency bands sound like. Just sitting down with an equalizer and dialing frequencies in and out will vividly illustrate what the various bands sound like… and hearing the changes in music will show you the relative importance of the various frequencies and will help you identify imbalances when you run across them.