Page 1 of 4 1234 LastLast
Results 1 to 20 of 77

Thread: Trends in speaker design. What to listen out for?

  1. #1
    Join Date
    Sep 2007
    Location
    England
    Posts
    235

    Default Trends in speaker design. What to listen out for?

    Is it possible that loudspeakers along with clothing and cars go through fashion cycles? Are these technical, cosmetic or sonic?

    What should the listener/buyer watch out for?

  2. #2
    Join Date
    Dec 2007
    Location
    Suffolk, UK
    Posts
    293

    Default Stagnation?

    I think speaker design has stagneted for some time myself. I genuinely get the feeling that the conventional "two-way drivers with passively-driven crossovers in an oblong wooden box" rulebook was written decades ago and only in recent years, have advances in materials and technology, coupled with a deeper understanding? of the way and how we suspend disbelief in audio reproduction made any real improvements possible - and even then, many of these improvements have been evolutionary in my opinion, rather than groundbreaking.

    For the future, separates HiFi as we know it will almost certainly fade away with "our" generation, although interest in legacy vinyl formats may see this survive CD. My hope for "loudpeaker future" is another look at fully "active" operation, where, at the very least, the crossover dividing the drive units can be done in the electrical domain, feeding dedicated amps then directly to the drive units, thus freeing the driving amps from having to drive the effects of inductors and other components often in the signal path.

    Advances in digital technology may well bring in full in-room equalisation, since the listening room is often the weakest link after the speakers themselves. My small experience of early versions of this technology were not very enjoyable, but so much has happened in recent years along these lines, I await developments with much interest.. In the meantime, the simpler the load as seen by the driving amp(s), the better in my opinion :)

  3. #3
    Join Date
    Jan 2006
    Location
    South of England, UK
    Posts
    4,016

    Default

    In the last few months, I have been in conversation with folks who have far more opportunity to listen to generally available loudspeakers than I do. They have, completely independently and with no knowledge of each other, casually commented that 'today's speakers seem to be extremely toppy'. I have been vaguely suspicious of this for some time, and have a general recollection that the frequency response curves that occasionally I see in hi-fi magazines do seem to hint at this.

    Now, to be fair, I'm not sure if we can have full confidence in any loudspeaker measurement that is not using calibrated analysis equipment, positioned carefully and in a measurement environment that is known to contribute as few reflections as possible: there are lots of variables there. Can we assume that published curves are reasonably reliable and that it is easier to accurately measure high frequencies than mid or low ones, where the room has a substantial influence?

    After further discussion I have prepared what I believe to be three design trends that some, not all, modern contemporary loudspeakers evidence - see attached plot. If you are not familiar with interpreting frequency response curves, they are not at all alarming and they certainly do not need great skill to read - or I wouldn't be able to use them. First, the X (horizontal) axis: this is graduated from 10Hz (subsonic) to 40kHz, but the audio band is generally about 40Hz to 20kHz so it more than covers what we need in audio. The Y (vertical) axis is scaled in decibels: and each little vertical fine-graduation is one decibel, 1dB. I've hand drawn three curves onto the one chart, to save paper. That really is all we need to know.

    If we say that the crossover region for most good speakers is no lower than about 3kHz (you can't see that from the curve, this is just for background information) then Curve A is the equivalent of stepping-up the tweeter's entire contribution by about 4dB across the tweeter's entire operating band. Just one resistor adjustment in the crossover could produce that effect, by accident or design.

    Curve B exhibits a different design choice. Here the bottom end of the tweeter's rage from, say, 2-5kHz (remember the tweeter is fully operational at about 3kHz) is in line with the general bass/midrange output, but the mid-high region of the tweeter is elevated by about 8-9dB. As audiology textbooks say that 6dB is a doubling in sound pressure, and 10dB is considered to be a doubling in loudness we could expect this region to be markedly more prominent replaying music that has a significant energy in the mid-tweeter range.

    Curve C exhibits a marked boost at the upper edge of the audio band. As most of us in our 50s have significant reduction in acuity above 10kHz or so simply due to age, it could be argued that if a speaker was targeted at an older customer that a boost to the extreme-top would somewhat compensate for age-loss in the extreme high frequencies and make the speaker sound 'right' to ears that were significantly 'wrong' through age or life on the rock and roll circuit. However, one would have to caution the designer that if a younger person or one who had not abused his hearing happened across such a speaker, he/she may be aware of the tilt. So now we have the overall shapes that we may experience or read about. Comfortable with this idea so far? OK, now let's take a recording that is 'flat', and process to to give it the sound of A, B or C as you would experience if you listened on-axis to speakers exhibiting these traits.

    What prompted me to write this today is that this afternoon when driving, I heard a live concert from the BBC Symphony Orchestra of Bernstein's Symphonic Dances from West Side Story here. I thought this would give us a good clip to manipulate.

    NOTES ON HOW TO PROCEED WITH THIS COMPARISON

    What I think you'll find interesting is this: play Ref. clip Q and familiarise yourself with it entirely then replay it to about 1/3 and pause. Then play 1/3 of A and pause, then 1/3 of B and pause then 1/3 of C1 and pause so that Q, A, B and C1 are all stopped at about the same point in the music. Then play a little more of A and stop, then B and stop then C1 and stop (but not Ref Q). Inch A, B and C1 forward by playing/stopping in that way. After a few start/stops like that then play Ref Q. Quite a shock isn't it: it sound so dull and lacking in top. But we know that it is the reference, exactly what left the studio. How can this be? How can the reference sound so wrong? If we were comparing speakers in a hifi store, and we started with model Q, listened to A, B and C1 how many listeners would recognise the correct balance of Q having acclimatised to the others with the boosted top? One in a hundred? Key point: It is to be expected that the average consumer will be attracted to a brighter, more 'revealing', 'crisper' sound even under quasi A-B comparisons like this and the marketing awareness of average consumer preference drives the design process that sets the overall speaker balance.

    Loading the player ...
    Reference clip Q. This is what it should sound like. This is what was sent from the studio.


    The HF characteristics that are not uncommon in modern speakers can be heard here:


    Loading the player ...
    Example A (see attached curves)

    Loading the player ...
    Example B

    Loading the player ...
    Example C (ignore this one - go to C1)


    UPDATE
    --------

    Silly me. I forgot that the BBC cut-off everything above 15kHz at the studio before it is distributed. So the boost curve of C is boosting nothing above 15kHz. So here is Curve C1 where I'm moving the boost of C left a little to give more prominence to the 10-15kHz region. I'll re-do this with a full bandwidth CD shortly.


    Loading the player ...
    Example C1


    AFTER THE TEST

    Stop all audio playing. Just sit in silence and reset your ears for a few minutes. Do you feel that your ears have 'pricked-up' and have now heightened sensitivity in the upper region? Do they whistle? Are the ear muscles tingling? Is that a pleasant sensation or not?

    >
    Attached Files Attached Files
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

  4. #4
    Join Date
    May 2009
    Location
    England
    Posts
    825

    Default Screaming top end v. Harbeth

    The trend in speakers, for some years now, is for slim, floor standers, the smaller the better. Bass appears to be an optional extra.

    Sonically, modern speakers appear to be thinner, brighter, sucked out mid and with a screaming top end for extra 'detail'. Not all but many.

    Harbeth are bucking these trends, which is why they sell well in audition. IMO of course.

  5. #5
    Join Date
    Sep 2009
    Location
    Malaysia
    Posts
    508

    Default

    Please note the clip was wrongly labelled as R.

    It is taking forever to download those clips. Meanwhile, the Reference clip Q compared to my TER ORBIS CD sounds a bit flat. Now, which one suppose to be the correct sound? My CD was recorded DDD at Abbey Road Studios. i would go for the other clips but still unable to download them fully. Anyone having problem with the loudness level of the clip(s)?

    ST

  6. #6
    Join Date
    Jan 2006
    Location
    South of England, UK
    Posts
    4,016

    Default Proper comparisons need a 'control' ...

    Quote Originally Posted by STHLS5 View Post
    Please note the clip was wrongly labelled as R.

    It is taking forever to download those clips. Meanwhile, the Reference clip Q compared to my TER ORBIS CD sounds a bit flat. Now, which one suppose to be the correct sound? My CD was recorded DDD at Abbey Road Studios. i would go for the other clips but still unable to download them fully. Anyone having problem with the loudness level of the clip(s)?

    ST
    Clip re-labled, thanks.

    What are you suggesting by introducing an unquantified new variable into the experiment, namely your CD version of the music? If I say that clip Q is the reference, that's a fixed variable whether or not you like the sound. Reference means reference. Comparisons need a control. My control is clip Q which is the source sound for the other clips. Q is clearly and unambiguously labelled in red as the reference. Like it or not, that is the sound that was sent from the studio to the listening audience. It is a 'broadcast balance'* - what you would hear at the hall, (not a hotted-up CD balance) and would be familiar to a concert-goer or BBC listener even if the audiophile didn't like or recognise it. The BBC balance is right technically, artistically and acoustically. It is not necessarily a 'commercial' sound; it is not designed as a substitute for purchasing a commercial CD. It exists for an entirely different (probably non-audiophile) audience familiar with live sound and I picked it with some care.

    The BBC sound will be available for a few days here.

    As you mention Abbey Road, are you aware that a significant part of their income is from 'sweetening' or mastering recordings for commercial release? To quote the above link:

    (Abbey Road Mastering) ... involves polishing the mixed sound with equalisation, compression and other outboard processing and then producing the final master.
    And only yesterday I gave up my morning to demonstrate what mastering can/does do to natural sound.

    * This is a whole new subject which we need to look at together.
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

  7. #7
    Join Date
    Sep 2009
    Location
    Malaysia
    Posts
    508

    Default False reference etc.

    I am not introducing a third variable and regret that it is seen that way. I am addressing what was said earlier, i.e., ” Key point: It is to be expected that the average consumer will be attracted to a brighter, more 'revealing', 'crisper' sound even under quasi A-B comparisons like this and the marketing awareness of average consumer preference drives the design process that sets the overall speaker balance. “


    So the point is to an average consumer I am unwittingly got attracted to the sweeten sound. There may be 100s of CD in my collection that may qualify as a perfect recording to my ears, but in reality, it would not meet the standards of correct recording of BBC. The same applies to the speakers as you are trying to guide us in this thread.

    As most people here in HUG or others are in the middle or last leg of our life, we move forward with quick decisions. Now, the examples that you picked with great care failed to drive into my head the very important point you were trying to make, which was, the correct sound that a speaker should represent. I was just pointing out the danger that for whatever reasons that my preference in the above clips were based on ignorance. The same ignorance that made by many in choosing excessively bright speakers and the examples you gave me shows the modified clips which were representing today’s bright speakers fit into my liking.

    It is simply difficult to cross the first hurdle, when my heart is telling me, the examples of the excessively high on treble speakers sounds better in the illustration above, and my head is telling if that’s what I like, then that’s right for me.

    The questions that I am asking myself now are - “Did I audition the Harbeth SHLS5 with my reference CD, which are excessively bright?” and “ Did Harbeth make the sweetened CDs sound perfect to my ears?" If the answers were affirmative to both questions, then Reference Q will never ever be my liking listening over the SHL5. And if I were to search for ideal speakers with Reference Q clip as my reference track then Harbeth SHLS5, no matter how perfect they are, would have never met my flawed standards as I would probably choose a speaker that would make Reference Q to sound a little brighter. This was the danger that I tried to highlight.

    The problem here are the consumers not the perfect speaker.

    As a designer, you have years of experience and knowledge about sound but for poor souls like myself, I have only one reference, i.e. the recordings that I liked very very much that I want to hear over and over again which may technically be flawed but deeply appreciated by myself with no means to convey how my ears hear and my brain interprets them and my search for good speakers was based on how well those speakers capable of making this technically flawed recording to sound best to my ears. As a designer there is no way for you to go please every individuals preference and you have to make your own reference.

    End of the day, it is not how good the speaker must be but how well the combination of the reference CD ( which may not be technically correct) that I used and the speakers sound. That’s what consumers want. That is the danger that this thread seem to leading me to.

    ST

  8. #8
    Join Date
    Jan 2006
    Location
    South of England, UK
    Posts
    4,016

    Default

    Oh dear. You are beating yourself up needlessly again as you have jumped over all the incremental knowledge-building steps I'd planned, and are now at the very end of the mental journey, standing alone in unfamiliar territory without a route map and understandably anxious. But the genie is out of the bottle now and I just don't have the time to retrace back to base camp.

    1. Not only have you introduced the variable of an alternative recording/performance, you have now introduced hi-fi speakers to your listening
    2. You have now drawn all sorts of confidence shattering conclusions based about your entire CD collection based on just the few audio clips I've presented you with ...
    3. You have unwittingly positioned yourself in a highly vulnerable mental position such that, presented with 'solutions' to your self-imposed anxiety by a marketeer, could you resist paying-out for a gadget to reduce your anxiety?!
    4. I didn't actually ask for a preference between A, B and C1. That's up to you and your ears. What I asked for was a report on "Do you feel that your ears have 'pricked-up' and have now heightened sensitivity in the upper region? Do they whistle? Are the ear muscles tingling? Is that a pleasant sensation or not?".

    Could you be thinking about this too deeply, and too quickly. Please don't. You'll make yourself ill. If you don't mind, leave the route planning to me. I'm an expert tour guide and I can lead you through the wilderness and back to safety. If you run ahead without me, you will be vulnerable. You will be confused. You will put yourself in danger. You will take wrong turnings. You will stumble about in circles chasing your own tail. You will eventually be sucked down into a mental quagmire. Can you trust me to reveal, step by step the way out of the audiophile mess? You don't have to believe me - why should you? - but at least I'll show you where we are on the map and you can make your own decisions; you can always re-read the map and reconsider where you've been and how you got there, even plan another exit route. I am nothing more than a compass. You can follow me or not: it's your choice. But to rush ahead is deadly.


    Moving on ....

    I'm certain that I've said before that as I don't have a hi-fi system at home, all audio clips I present here are made/monitored on my Logitec LS11 active speakers. Cost about $20. Pretty good actually for plastic boxes. That's all you need for listening (or headphones). You do not need to turn on your hi-fi. You do not need Harbeth speakers. You could probably even just about get along with the speakers built into some PC monitors. But you have, unexpectedly, taken a huge earth-to-moon leap and conclude that your entire CD collection, which has given you years of pleasure is now, today, 'wrong' based on audio clips made on $20 speakers. That's a massive over-reaction don't you agree? We're not here to drive an emotional wedge between you and your music. I'm solely illustrating a point about speaker trends, and the audio clips are intended to be interpreted literally as A v. B within the confines of the thread step and not indiscriminately applied to the big, outside world. To apply to the bigger picture, we have to complete this journey first and not get ahead of ourselves.

    Your observations about preference suggest one thing to me. You (in common with most audiophiles I meet) are not exposed to live music. Regrettably then, your own internal frame of reference about how instruments and halls sound is significantly different to those of us who have the chance to attend concerts. That is not a criticism - it's a fact. And yes, it must come as a shock to be presented with a sound balance that is closer to what a concert-goer (as opposed to a CD buyer) would recognise*. And yes, that must indeed be confusing and make you anxious. We've covered that before here. But, as I said in post #3, "this afternoon when driving, I heard a live concert from the BBC Symphony Orchestra of Bernstein's Symphonic Dances ... I thought this would give us a good clip to manipulate". I didn't say that I heard this broadcast sitting in the sweet spot between my M40.1s did I! I said that I recognised its merits just listening over my manufacturer's fitted, standard car audio system on FM radio with lots of crackles whilst my wife and her mother were chatting. It was blindingly obvious to me that it was the sort of balance that I am familiar with in the hall.

    We need to explore recording balance in more detail, and I just do not have the time now. I have to set off for The X Factor studios in London.

    * The recorded balance of a (classical) CD is not necessarily intended to convey the sound of the instruments in a concert hall with an audience and by an audience. A BBC 'broadcast balance' must convey what the audience would hear. The recording approach is almost opposite.

    [51 mins.]
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

  9. #9
    Join Date
    Oct 2011
    Location
    United States
    Posts
    14

    Default Thanks!

    I want to thank you Alan (again) for taking the time to educate on aspects of recorded and reproduced sound. It is appreciated!

  10. #10
    Join Date
    Jan 2006
    Location
    South of England, UK
    Posts
    4,016

    Default

    I much appreciate your feedback. I could move faster and with more enthusiasm if I could keep subjects strictly in focus and not have to spend time pulling them back into track. It must be hugely confusing for those reading. Explanations, like chapters of a book, need to flow in a logical, incremental way or the reader gives-up in confusion.
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

  11. #11
    Join Date
    Sep 2009
    Location
    Malaysia
    Posts
    508

    Default Very fatiguing

    Quote Originally Posted by A.S. View Post
    ...What I asked for was a report on "Do you feel that your ears have 'pricked-up' and have now heightened sensitivity in the upper region? Do they whistle? Are the ear muscles tingling? Is that a pleasant sensation or not?"
    Ok, I just finished spending listening to all the clips and after 35 minutes it is very fatiguing. The ears feel like the rain has just stopped.

    ST

  12. #12
    Join Date
    Jan 2006
    Location
    South of England, UK
    Posts
    4,016

    Default Extreme fatigue

    Quote Originally Posted by STHLS5 View Post
    Ok, I just finished spending listening to all the clips and after 35 minutes it is very fatiguing. The ears feel like the rain has just stopped. ST
    Fantastic! That is exactly what I hoped you would say: bang on track.

    That's just what I experienced - extreme listening fatigue.*

    Ok, great .... as a matter of interest, which of the 'enhanced' clips did you find the most fatiguing? I found one especially fatiguing and it was always a pleasure to revert back to Q, which as I warned, in comparison, did sound dull on the switch-over. But as I said, Q was and is correct even though one could be forgiven for being attracted to one of the enhanced clips due to inexperience. We'll put that right. I picked this music because of the prominent brass instrument content. That was deliberate.

    *Even reference clip Q, which I think you previously dismissed as too dull would become tiring (but hopefully not fatiguing: not the same thing at all) after long exposure I'd expect.
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

  13. #13
    Join Date
    Sep 2009
    Location
    Malaysia
    Posts
    508

    Default Fatigue results

    Quote Originally Posted by A.S. View Post
    Ok, great .... as a matter of interest, which of the 'enhanced' clips did you find the most fatiguing? I found one especially fatiguing and it was always a pleasure to revert back to Q, which as I warned, in comparison did sound dull. But as I said, it was and is correct.

    I picked this sonic example to make a point.
    I find Clip A to be more fatiguing but it is difficult to say because at a different passage of the music. I find Clip B to be even more fatiguing.

    Clip C1 is not that bad and even though you asked to skip C, I find it smooth much better than B.

    B is very fatiguing.

    ST

  14. #14
    Join Date
    Jan 2006
    Location
    South of England, UK
    Posts
    4,016

    Default

    Quote Originally Posted by STHLS5 View Post
    I find Clip A to be more fatiguing but it is difficult to say because at a different passage of the music I find Clip B to be even more fatiguing. Clip C1 is not that bad and even though you asked to skip C, I find it smooth much better than B. ST
    Excellent. That's a very astute observation.

    Ok let's have a think about C or C1 first. You can see from my frequency graphs showing the frequency bands that I have boosted, that C and C1 are only enhancing the very top end of the audio band - the 'extreme top'. Yet, you (and I) could tolerate what is a massive amount of boost in that region without distress (at least, on this short clip: we cannot necessarily say that hour after hour of exposure would be so benign.) How can that be? How can I have cranked up the extreme top without it becoming obvious or irritating to the critical listener?

    Possibilities come to mind .....

    1. Whatever you listened to the clips on (plastic speakers, nasty headphones ...) didn't actually reproduce much HF (due to cheap design, poor parts) so even when the signal was massively boosted, no sound was generated at you ear ... or perhaps

    2. There is so little energy in music at the extreme high frequencies as a proportion of the entire audio band that even when boosted it's an insignificant part of the energy spectrum.

    What do you think? Probable?
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

  15. #15
    Join Date
    Sep 2009
    Location
    Malaysia
    Posts
    508

    Default 10k limited

    Quote Originally Posted by A.S. View Post
    ..
    What do you think? Probable?
    In this case, looks like nothing much above 10kHz.

    ST

  16. #16
    Join Date
    Jan 2006
    Location
    South of England, UK
    Posts
    4,016

    Default Wide bandwidth and sonic character

    Quote Originally Posted by STHLS5 View Post
    In this case, looks like nothing much above 10kHz [in the audio itself].ST
    That's possible.

    How could we prove this in our usual quick, low-cost, low-hassle way?

    I saved the frequency-shaping envelopes A,B, C and C1 in the audio editor when I created them. So I can recall them and apply them to any other audio at will. So, how about if I apply the same preset effects to ...

    1. Wide bandwidth music that we pre-check has components extending up to at least 20kHz (to be sure that there are actually tones present to boost) and/or
    2. Non-musical hiss or noise which we also pre-check for truly extended bandwidth to eliminate music as a variable

    and place them here as clips then see if we can hear the C/C1 effect on the extended bandwidth source. If we can then hear C/C1, that validates our hearing acuity. Agree?

    There is also one aspect I want to check with you. Would you feel comfortable with the concept that whichever component in the audio chain exhibited the sonic character A, B, C/C1 (from the microphone right through to the speakers at home) would impose that character on all subsequent component steps right down to the listener? For example, if the microphone had sonic character B (refer to clip), regardless of whether the home listener's CD, CD transport, DAC, amplifier, stands, interconnects, room treatment, biwire links or speakers were perfectly flat in themselves (like ref. clip Q), audio character B would be evident to the listener?

    In simple language, 'once the damage is done it cannot be undone' regardless of how much time or money the audiophile throws at 'remedying' the undesirable tonality issue. It would imply that the sound quality perceived by the home listener can never, ever improve upon that of the recording itself.

    That would suggest, if true, that the most crucial 'components' in the hi-fi chain are the sonic characteristics of the recording microphones and hall acoustics. And we know that they are far from flat.

    Comfortable with that idea?
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

  17. #17
    Join Date
    Jun 2011
    Location
    Belgium
    Posts
    115

    Default Is hi-fi about music? Maybe not.

    A bit unrelated perhaps, but reading some of the responses here made me wonder;

    Why do I get the feeling, that people who own a hi-fi installation that is probably of better quality than that of 99% of the population, seem to enjoy music less than those 99%. Ironic, no?

  18. #18
    Join Date
    Nov 2010
    Location
    UK
    Posts
    184

    Default

    Great thread. My responses to the clips was exactly the same as ST's.


    Quote Originally Posted by A.S. View Post
    ...and place them here as clips then see if we can hear the C/C1 effect on the extended bandwidth source. If we can then hear C/C1, that validates our hearing acuity. Agree?
    Yes. (Although if we can't hear C/C1, that doesn't validate our hearing non-acuity as it were, if our PC speakers at home fail to dish out much of anything over 10 kHz).

    Quote Originally Posted by A.S. View Post
    For example, if the microphone had sonic character B (refer to clip), regardless of whether the home listener's CD, CD transport, DAC, amplifier, stands, interconnects, room treatment, biwire links or speakers were perfectly flat in themselves (like ref. clip Q), audio character B would be evident to the listener?
    Yes definitely.

    Quote Originally Posted by A.S. View Post
    In simple language, 'once the damage is done it cannot be undone' regardless of how much time or money the audiophile throws at 'remedying' the undesirable tonality issue. It would imply that the sound quality perceived by the home listener can never, ever improve upon that of the recording itself.
    Not sure about that... if the recording microphone boosted a certain frequency range, could not the home listener with a good equalising deck bring back the response to something like flat?

    Quote Originally Posted by A.S. View Post
    That would suggest, if true, that the most crucial 'components' in the hi-fi chain are the sonic characteristics of the recording microphones and hall acoustics. And we know that they are far from flat.

    Comfortable with that idea?
    Yes, definitely happy with that.
    Ben from UK. Harbeth P3ESR owner.

  19. #19
    Join Date
    Jan 2006
    Location
    South of England, UK
    Posts
    4,016

    Default

    Quote Originally Posted by BAS-H View Post
    ....Not sure about that... if the recording microphone boosted a certain frequency range, could not the home listener with a good equalising deck bring back the response to something like flat? ......
    This is not actually possible in the complete sense you imply. It is not possible to remove every molecule of your presence from the scene of the crime and it is not possible to completely reverse whatever undesireable sonic corruption proceeding stages in the audio chain have imposed. Just consider the effect that the frequency boost has on harmonics of the boosted frequency band .... how do you remove boosted harmonics even if you can cancel the boosted fundamentals?

    Furthermore, as stated in the matrix here, tone controls are anathema to audiophiles. They represent an unspeakable crime against audio - a truly barmy position. Those that can get along without the occasional touch of the tone control, are surely putting themselves in a position where they have absolutely no ability to correct for technical or artistic blemishes in the audio chain. Their position is the same as removing the colour control from the domestic TV: the viewer/listener must then take whatever is dished-up to them without complaint, like it or lump it. A very curious state of affairs.

    Can we assume that only a smattering of audiophiles actually have tone corrections controls on their equipment? Can we also assume that those controls offer only a very broad-brush approach that gives only a generalised boost/cut across quite wide audio bands like 'bass' and 'treble'? Can we anticipate that unless the preceding corruption is of exactly the same frequency-shaped profile as the generalised fit-all tone controls, that the effect of applying a generalised solution to a specific problem may to actually make the resulting sound even less attractive?

    Furthermore, you have now touched on one of the most serious issues in audio - the limitations of microphones which compared to every other component in the chain (and second only to the speakers and pickup cartridges) are very challenging devices, as you would expect for any electo-mechanical system.

    I was back in the studio today, and asked three top class sound engineer (without priming them) for their opinion on which audio component they believed had the most profound influence on the audio chain right through to the listener. The unanimous opinion was, obviously, that it was the selection of the microphone, where it was placed and its condition and serviceability. In their view, the second most important factor is the treatment of the listener's room, and they referred me to the SOS DIY diffuser article which is their 'bible'; I'm sure we've posted here before (anyone?).

    As a matter of interest, if you were selecting a microphone for a pro recording, what sort of sonic characteristics would you shortlist? What unwelcome characterisitic (limitations) would you be very much aware of? (Your answer to this very much relates to your belief above about being able to cancel-out preceeding sonic aberations.)

    We'll return to analysing the clips shortly.
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

  20. #20
    Join Date
    Jan 2006
    Location
    South of England, UK
    Posts
    4,016

    Default The usefulness of pink noise

    Back to why we couldn't seem to hear any effect for eq settings C or C1. It's possible that there just wasn't any or enough energy in the music reference Q to boost to audibility. Or it could have been the speakers/headphones on which you were listening. Or it could have been that at our age, we just can't hear much above 10kHz .... or maybe a little of all three. So we need to do some elimination.

    Let's forget about finding a suitable wide-bandwidth music clip and jump straight to the quick solution from experience. We'll use pink noise. Pink noise is usually generated as the bright, sharp 'white noise' hiss you hear when an FM radio is off tune, but then it's processed to have a particular sonic character. You can see from the attached thumbnail, that pink noise has a diminishing energy gradient from low to high frequencies - and as you can hear it has much less 'top' than the white noise of an off-tune FM radio. Pink noise is a very useful test tool for quick and accurate analysis of audio related issues. Although pink noise sounds distinctly bass heavy (or lacking in high frequencies which is really the same subjective effect) in fact, there is far more high frequency energy in pink noise than western classical music. Hard to believe? How can it sound so dull and yet contain more HF energy than classical* music? It's all about total energy, not instantaneous energy.

    Have you ever taken a long exposure photo where the camera shutter is open for many minutes? Perhaps of the night sky? What will be recorded is every event between the opening and shutting of the lens. Events which were slow moving and/or especially bright (or dark) will have made more of an impression on the image sensor (film in the old days); fleeting events such as a meteor flashing past will have little influence on the scene, because they just don't have enough luminance to really effect the average brightness. So what we will see when we examine the image is that the scene will be weighted or biased away from fleeting events towards more persistent events. In audio terms, the persistent events will be the almost continuous low frequencies (cello, double bass etc.) and the fleeting events will be cymbal crashes. So .... if we set the audio analyser to take the audio equivalent of a long exposure picture of the audio over an entire performance (say, the complete 23 minutes of Symphonic Dances) we can determine how much of each frequency is present across the whole piece. And as a generalisation, most western classical music sound about the same, and unsurprisingly has about the same spectral content. Using pink noise may sometimes saves us having to critically listen to music over a long period: just a few seconds of pink noise can tell us as much, or possibly more, than having to carefully absorb the entire performance. Every seconds worth of pink noise is alike every other seconds (unlike music).

    Here is computer-generated pink noise: have a listen and see if you find it seems to be lacking in high frequency sparkle compared with music we like ....

    Loading the player ...
    Clip P: stereo pink noise

    It really doesn't have the characteristic clarity and brightness of tone that we associate with good music on a good system does it.

    OK, now let's open the shutter in the audio analyser and make a long-exposure capture of pink noise playing. (Refer to pinkshape-sc.jpg). Ignoring the minor wiggles we can see that there is a straight-line function, and that gradient precisely defines 'pink noise' to acoustic people. Pink noise is not just any old noise: it is scientific noise.

    Next I played into the audio analyser the entire 23 minutes of Symphonic Dances. Ignoring the difference in vertical offset between Dances and pink noise (I didn't calibrate the analyser as we're working in dBs so no need to) I've plotted Dances on the same graph as the pink noise (ref: dances-pink-sc.jpg). You can see that below about 150Hz the pink noise has much more bass (the green trace is off the top of the image) than Dances averaged over 23 minutes, and perhaps surprisingly, above about 2kHz, the pink noise has much more high frequencies - six times more (20-5dB) at 10kHz than the averaged musical performance. Surprised just how little total high frequency energy there is in western classical* music? So our initial impression playing the pink noise clip P above that it was lacking in HF was in fact, wildly incorrect. It has roughly six times more.

    Are you Ok with the idea that we can substitute listening to a short blast of pink noise for listening to many minutes of music when we are trying to compare general response shape characteristics? There is certainly more than enough HF in pink noise, and as we can see, its bandwidth extends right out to beyond 20kHz without the droop we see with classical music. It's going to allow us to nail either our hearing or speakers/headphones as the reason why we can't hear the effect of clip C or C1.

    For my own curiosity, I applied a filter having the overall shape (the three orange lines) of the Symphonic Dances profile onto the pink noise (as per clip P above) and made it have the same sonic spectrum. This now audibly exposes how little HF total energy there is in classical music. Compare with the pink noise clip above. As I have said for many years, the crucially important part of music is the midrange - the speech band - where almost all of the energy is concentrated. The 'tweet' bit is a non-essential luxury, as listening to AM radio confirms.

    Loading the player ...
    Clip S: stereo pink noise shaped to have the same energy profile as 23 minutes of classical music


    * Pop music has an entirely different spectra with much more top end. If you were presented with a plot of audio long-exposure of pop or classical music, you could identify them with certainty by the gradient of the curves.

    >SUBJECT CONTINUES ON POST #29

    [2.6 hrs.]
    Attached Images Attached Images
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

Page 1 of 4 1234 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •