Bell Lab maid a good work, but nowadaws: qualitiy is dependant of the network, this is not anymore analogic even if you have old Equipment the network earth is totaly digital (at least here !). Final quality dépends on the compression algoryhtms provided by the suppliers and the QoS your able to manage (if you do it alone). With the better scenario, I didn't find the listening quality of the digitilased phone we use today (even if you have an analogic terminal phone ) is better than yesterday : most of us are old enough to lived this transition. As ancient IT Manager and multi sites manager I worked with big company as BT & France Telecom-Orange : Ok it's cheaper, quality dépends on many factors.... but if price and TCO would not be involved I would come back to the old Erickson phone solutions... must say I'm very attentive to voice quality as an audiophile🙂... Of course todays service layers witrt digital are far better (ToIP, etc... PC, phones, smartphones, databases, mobility working together !) . But sounding quality... is moving a lot fron an infrastructure to an other with some parameters which are moving also (earth networks of operators) !
All that to say you can not anymore to benchmark the -300-3000 hz of phone bandwitch to makes yourself an efficienty benchmark (in relation to the old phone tech which was audibly better... surmise compression in digital and the last harmonics !)
All that to say you can not anymore to benchmark the -300-3000 hz of phone bandwitch to makes yourself an efficienty benchmark (in relation to the old phone tech which was audibly better... surmise compression in digital and the last harmonics !)
Last edited:
Did the test with Noise
Just ran the test with 1/3 band filtered noise.
For me bass transitions to low midrange at 200Hz. Highs start maybe 6K. Something of a subjective call, of course, but that's about where I found the transitions.
I have attached the test tones I made. They are Pink Noise that has been sharply filtered into 1/3 octave center points. 3 seconds each. The 1/3 octave bands have been level adjusted to follow the equal loudness curve, so they should all sound about the same by ear. They do on my system.
The 1st track is normal 6 seconds of Pink Noise with an RMS value of -18dB FS. Start with that fairly loud, and then the noise bands should be loud enough. The low tones will help you find rattles in your room. 🙂
Just ran the test with 1/3 band filtered noise.
For me bass transitions to low midrange at 200Hz. Highs start maybe 6K. Something of a subjective call, of course, but that's about where I found the transitions.
I have attached the test tones I made. They are Pink Noise that has been sharply filtered into 1/3 octave center points. 3 seconds each. The 1/3 octave bands have been level adjusted to follow the equal loudness curve, so they should all sound about the same by ear. They do on my system.
The 1st track is normal 6 seconds of Pink Noise with an RMS value of -18dB FS. Start with that fairly loud, and then the noise bands should be loud enough. The low tones will help you find rattles in your room. 🙂
SReten -
Bass is below ~ 300 Hz.
Mid is ~ 300Hz to ~ 1.5KHz.
Presence is ~ 1.5 KHz to ~ 6KHz.
Treble is above ~ 6KHz.
That was actually very helpful. I never thought about analyzing the frequency spectrum from this perspective.
But, while charts and graphs and discussion are helpful to a general understanding, the core question was - Where do you perceive these transitions?
Certainly that will be subjective. What I, my brain, and my ears consider bass, but not be the same as another person's perception of those same things.
Another person posted and said he heard the bass transition at 150hz. Re-running some test tones, I can see that. But I still hear a shift in the sound at around 300hz.
Perhaps I framed the question wrong. Perhaps it is not so much, where does bass become Mid, but rather where does the character of the sound change. Where do you hear a transition in the character?
Again, the results may be all over the place, or we may see patterns in the results. It may not be the pattern we expected, but there might be a pattern.
The only way to see it and not see it is to gather data.
I will run the test tones on my computer speaker again. Then I have two pair of large Stereo speakers, I will run the test on them as well. That will give us a start.
Give me a day on that.
Steve/bluewizard
Bass is below ~ 300 Hz.
Mid is ~ 300Hz to ~ 1.5KHz.
Presence is ~ 1.5 KHz to ~ 6KHz.
Treble is above ~ 6KHz.
That was actually very helpful. I never thought about analyzing the frequency spectrum from this perspective.
But, while charts and graphs and discussion are helpful to a general understanding, the core question was - Where do you perceive these transitions?
Certainly that will be subjective. What I, my brain, and my ears consider bass, but not be the same as another person's perception of those same things.
Another person posted and said he heard the bass transition at 150hz. Re-running some test tones, I can see that. But I still hear a shift in the sound at around 300hz.
Perhaps I framed the question wrong. Perhaps it is not so much, where does bass become Mid, but rather where does the character of the sound change. Where do you hear a transition in the character?
Again, the results may be all over the place, or we may see patterns in the results. It may not be the pattern we expected, but there might be a pattern.
The only way to see it and not see it is to gather data.
I will run the test tones on my computer speaker again. Then I have two pair of large Stereo speakers, I will run the test on them as well. That will give us a start.
Give me a day on that.
Steve/bluewizard
Bass is below ~ 300 Hz. : Scroeder frequency, room mods transition, close walls & XO interaction with bass driver.Change of equalisation with the ear (M-F curve)
Mid is ~ 300Hz to ~ 1.5KHz. : not all the first harmonics of acoustical instruments but most of "life" is here. And non tenor/baryton male voice first harmonics...99.99% of the interpretations with voice !
Presence is ~ 1.5 KHz to ~ 6KHz. : F-M beginning and end sensibility, focused on details and low accoustic energy for survey in the old ages?
Treble is above ~ 6KHz. Needed for inteligibility of first harmonics below and bass transcient ?
Makes sense ?
Mid is ~ 300Hz to ~ 1.5KHz. : not all the first harmonics of acoustical instruments but most of "life" is here. And non tenor/baryton male voice first harmonics...99.99% of the interpretations with voice !
Presence is ~ 1.5 KHz to ~ 6KHz. : F-M beginning and end sensibility, focused on details and low accoustic energy for survey in the old ages?
Treble is above ~ 6KHz. Needed for inteligibility of first harmonics below and bass transcient ?
Makes sense ?
Last edited:
Charts and refernces are fine. But man up - boys. Run the test and stop talking. ;
I have even supplied tones if you need them. Get to it!
I have even supplied tones if you need them. Get to it!
Manning Up
Despite believing that BlueWizard's way of looking at this problem is quite wrong, I decided to "man up". My experiment uses headphones via the headphone jack of my computer - reasonable quality sound from it+them.
Using Tone from Tolvan I find that going up in frequency the perception changes at about 300Hz, and coming down in frequency the perception changes at about 150 Hz. It takes a few attempts to do this with your eyes closed.
These frequencies are similar using sine wave, square wave and sawtooth.
Rather more fun than I expected.
Despite believing that BlueWizard's way of looking at this problem is quite wrong, I decided to "man up". My experiment uses headphones via the headphone jack of my computer - reasonable quality sound from it+them.
Using Tone from Tolvan I find that going up in frequency the perception changes at about 300Hz, and coming down in frequency the perception changes at about 150 Hz. It takes a few attempts to do this with your eyes closed.
These frequencies are similar using sine wave, square wave and sawtooth.
Rather more fun than I expected.
Last edited:
RAndyB -
1.) It never occurred to me that stepping down through the frequencies might give me a different result than stepping up through the frequencies. I always stepped upward. I may have to try this again to see if I get a different result stepping down.
Or, perhaps, that was a typo and should have been 150 and 3000 instead of 300???
2.) You gave us the Low/Mid transition points, could you perhaps also give us the Mid/High transition.
Lastly, good idea using Headphones. That eliminates a lot of room acoustic problems and resonances.
Thanks again for taking the time to help me out.
Steve/bluewizard
1.) It never occurred to me that stepping down through the frequencies might give me a different result than stepping up through the frequencies. I always stepped upward. I may have to try this again to see if I get a different result stepping down.
Or, perhaps, that was a typo and should have been 150 and 3000 instead of 300???
2.) You gave us the Low/Mid transition points, could you perhaps also give us the Mid/High transition.
Lastly, good idea using Headphones. That eliminates a lot of room acoustic problems and resonances.
Thanks again for taking the time to help me out.
Steve/bluewizard
Not a typo.
Mid/high transition point is difficult for me - my hearing is quite accute at mid and high frequencies (unusually shaped ear canals), although now limited at about 11kHz pure sine wave.
I did try for a transition mid to high and something happened, however I got the impression that the headphone/sound card was probably not reliable enough for it to be worthwhile sharing what I found.
As I know that what I hear in the concert hall, Cathedral, theatre (yes, a string quartet on a theatre stage) cannot be reproduced in my room, I have gone for the least worst option for me - a full range driver plus a 10" bass driver taking over below 100Hz. As I have said before on this forum, I can tell the difference between a flute and an oboe with this setup, which is enough to demonstrate what the composer intended. If you refer to my first post in this thread, you will realise that for me there is a definite musical transition below 150Hz, so putting a crossover here is not a serious problem. A sudden phase change at a frequency that is still in the lower harmonics of the higher instruments - violin, flute, oboe - is what makes for difficult listening.
Last edited:
Steve,
Hall Acoustics made an analog generator (the ATG-301) that puts out finite octave filtered pink noise in 1,1/2,1/3,1/5,1/10 or 1/20th of an octave, in three ranges, 20-220Hz (low) 200-2200Hz (mid) 2000-22,000 Hz (High). Using the test generator, one can do a continuous sweep, rather than the individual 1/3 octave bands as Pano provided. The set comes with a calibrated SPL meter, so one can also see the level in any particular frequency band, and make corrective EQ adjustments if desired.
My first one disappeared in a theft in 1985, a real heartbreak as they were unavailable by then, but I purchased a used one recently. Listening to the swept filtered bands literally for hours over the years I never noticed any specific "transition of perception" or "character" change going up or down in frequency, other than the obvious switch range that requires one to choose between the three decades (Low, Mid, High) of the audio bandwidth.
Regarding another discussion, I had mentioned that musical peaks throughout the audio spectra can be equal throughout the entire audio spectrum, found an example illustrating that fact. Note that in the "Mr. Brightside" song by the Killers, full amplitude peaks occur from 20-20kHz and beyond, with near continuous VLF, while the "classic rock" of Cream's "Sunshine of your love" resembles a more "orchestral" spectral distribution, with a concentration over a more limited low mid range, and much more dynamic variance. Different genres and compression schemes can have quite different spectral distribution, but it seems the trend in pop is becoming more like pink noise.
Makes me remember the answer to an old question posed to the headliner's engineer before an outdoor show: "Would you like to pink noise the system?", "That's alright, I'll just give a quick listen during Motorhead's set".
Art
Attachments
Last edited:
Thanks for that response--I didn't want to be the first to broach the subject of the elephant in the room...not to dilute the original topic but rather to attempt to focus the discussion a bit on something that I can get a handle on.
When someone says "I hear the transition from bass to midrange at XXX frequency", I'm wondering what decision criteria they are using (really, not kidding here) in addition to whether they are listening through headphones--which brings its own set of "features", or perhaps sine waves or some other type of timbre, i.e., having pure frequency or harmonics played at the same time or at differing relative loudnesses, such as a piano or other instrument. How long to you wait between changing frequencies, etc. What loudness (SPL or Phons)? What background level (SPL or Phons) while seated in an anechoic chamber/very, very quiet studio, or on the corner of a traffic intersection?
It seems to me that the original argument is somewhat reversed from that which I'd consider: "How do I perceive the different frequencies, based on my knowledge of Fletcher-Munson effects, precedence effect, different room acoustics, am I considering headphone effects, stereo or multichannel effects, and/or audibility effects of harmonic distortion levels, etc.?"
I suppose that the objective here is to propose to reallocate loudspeaker crossover frequencies based on some tacitly agreed-to criteria. That's fine. My crossover frequencies are 40 Hz (i.e., transition from extreme low frequency to bass bin) and 425 Hz (i.e., just below the interaural directionality frequency based on the distance between the eardrums). I find that those crossover frequencies make a lot more sense than typical box type, three-way direct-radiating loudspeaker crossover frequencies that the OP has referenced. But the rest of the physics of the discussion must also be considered. For instance, I'm using high quality fully horn-loaded loudspeakers throughout.
Higher in this thread, I reposted a figure from a web site URL mentioned earlier in the thread to more subtly inform that there are more things going on than perhaps determining (apparently by group acclamation) an arbitrary breakpoint between bass and midrange.
If using my criteria, I'd put the breakpoint somewhere in the region of 60 Hz to "midbass" and 30-35 Hz as the breakpoint to "infrasonic" bass, depending on SPL or Phon levels, and about 400-500 Hz for strong stereo directionality (which you can't determine nearly as well using headphones), making sure all the while that the harmonics being generated by the loudspeakers at those frequencies are very, very low--especially second harmonic. But you didn't ask for my decision criteria, so I didn't offer it... 🙂
Sorry about all this, but my experience in my profession has taught me to be more specific in order to clearly understand customer intent. Again: What is the objective? Is it to reallocate crossover frequencies for three-way direct-radiating loudspeakers based on some perceived artifact related to human hearing?
Chris
I too have noticed the up down difference around 150 to 300. But the deciding factor for me is maintaining vocal coherency. The consonant sound of the female voice is around 250Hz, so 300 wouldnt be too bad, but male consonant is around 150 so additional attention would be required in spacing ctc distance and time alignment to maintain coherency. The lower 150 point allows all to be handled by a single driver (or not 😉
Current project is crossed @142Hz and 2,100Hz. Many use below 150 as the lower cross point in F.A.S.T. systems. Have been a proponent for sub/sat, FAST setups since the early 80's.
Art, excellent example, was listening to Mr. Brightside as a spectral balance test just before reading this thread. 🙂
Current project is crossed @142Hz and 2,100Hz. Many use below 150 as the lower cross point in F.A.S.T. systems. Have been a proponent for sub/sat, FAST setups since the early 80's.
Art, excellent example, was listening to Mr. Brightside as a spectral balance test just before reading this thread. 🙂
I don't want to discount anything you said, but you answered your own question.
"I hear the transition ..."
I would be the height of absurdity for me to require people to find an Anechoic Chamber somewhere. Equally so for me to require them to gather any fancy or complex test instrumentation.
It is as simple as ..."I hear"..., in your room with your equipment with the resources you can gather.
I hear a definite transition in the sound; a definite transition in the character of that sound. What more do I really need?
Though I was perhaps wrong to frame is as when is bass not bass. So now it is framed as when do you hear a change in the character of the sound. In my previous example of ~300hz, it is not so much that bass went out, but rather that Midrange clearly came in.
While we have limited data so far, the only pattern that seems to be emerging is that what people are hearing is not that far off from the text book transition points we might expect.
I still have a couple more speakers to test in different locations, and I have some headphones I can test, though the headphones are nothing special.
Steve/bluewizard
Last edited:
I find that it's usually important to quote complete sentences. This is what I wrote:
This actually has a very different meaning than your usage of the first portion of the sentence.
What decision criteria are you using? One other person has said that he hears no "breakpoints". I agree with him.
What kind of loudspeakers or headphones you listening to? What kind of room are you listening in (if loudspeakers) and where are they located in the room?
My experience has been that when you hear breakpoints in the sound, then other factors are coming into play beside those that you think are contributory.
The criteria is - I HEAR...
The source doesn't matter as much as what you hear. We are not looking for hard empirical scientific evidence. Rather were are looking for a preliminary indication of the possibility of a pattern.
The goal is not to prove anything, but rather see if there is anything that might need proving.
You do agree that bass exists? And you do agree that Midrange exists? Then there must be a transition between the two. There must be a point where we all agree that one is not the other.
Again, I've been distracted, but I have 3 sets of speakers, and 3 set of headphones, and 3 different amps I can try this on (soon), and eventually will. If my results in these tests are even broadly consistent, then I think I've found a pattern.
There is always someone in an Internet discussion that tries to twist the subject around to what it is not. Keep this all in a reasonable context.
It is what it is, and that's all that it is.
Steve/bluewizard
Steve,
perhaps Greebster is taking us to an explanation. Does the brain assign a different importance to frequencies outside the range of the human voice? giving this effect of a change of quality in both live and reproduced music.
Could it also explain why the tiny piccolo is almost aways audible over a full orchestra in the concert hall? (not talking about multi-mic. recordings). And why the super-high sopranos are described as "other worldly", and why the huge sound of Russian ultra-low basses is described as "thrilling"?
Andy
perhaps Greebster is taking us to an explanation. Does the brain assign a different importance to frequencies outside the range of the human voice? giving this effect of a change of quality in both live and reproduced music.
Could it also explain why the tiny piccolo is almost aways audible over a full orchestra in the concert hall? (not talking about multi-mic. recordings). And why the super-high sopranos are described as "other worldly", and why the huge sound of Russian ultra-low basses is described as "thrilling"?
Andy
http://www.akutek.info/Papers/MS_Schroeder_Revisited.pdf
My listening room's Schroeder frequency is ~125 Hz - hence my response of 60 Hz for the transition to bass (stated above).
If you are in smaller room, the Schroeder frequency can be 200-300 Hz. It makes a difference.
If using headphones, the effect of closed-tube acoustics affects frequencies in the 250-300 Hz region (see slide 46).
My listening room's Schroeder frequency is ~125 Hz - hence my response of 60 Hz for the transition to bass (stated above).
If you are in smaller room, the Schroeder frequency can be 200-300 Hz. It makes a difference.
If using headphones, the effect of closed-tube acoustics affects frequencies in the 250-300 Hz region (see slide 46).
Steve,
We can agree that an equal three way split of the frequencies from 20-20 kHz results in a decade of range for each, 20-200, 200-2000, and 2000-20,000.
Whether you call them bass, mid, treble, low, mid, high or whatever convention of naming makes no difference to what I hear.
If you pick a point in the center of each three decades, we could probably agree as to there being low, mid, or high.
Art
Bass is Bass when it sounds fishy.
Bass, a spiny-finned fish found in rivers, lakes, and seas that is caught for food. Families: Centrarchidae, Percichthyidae, Serranidae
Bass, a spiny-finned fish found in rivers, lakes, and seas that is caught for food. Families: Centrarchidae, Percichthyidae, Serranidae
- Status
- Not open for further replies.