One of the best videos yet. More in depth on this topic would be awesome. A technical walk thru of the new PS Audio speaker in relation to the topics here would be awesome.
Excellent interview. It is really great to hear that there are people who are still innovating in this field. Agreed that THD is a terrible metric for predicting how 'good' or 'bad' a loudspeaker will sound. Years ago, I was part of a team that tried to tackle distortion using an algorithmic technique. While it was very easy to reduce THD, there was a tendency to push the high order harmonics up, which actually made the loudspeakers sound worse. In the end, we discontinued the project after concluding that a well-designed loudspeaker was more effective at reducing distortion, for the money, than would a digital correction system. BTW, MBCs, which do not correct distortion but avoid it, are useful.
Always good to hear someone cite Earl Geddes. Some of my favorite speakers are from Earl and he focused the entirety of his design based on understanding what matter in sound quality and what the human ear really didn’t care much about. I highly recommend that everyone read his various white papers on acoustics and psychoacoustics.
I agree!!! He's a prickly pear at times and I don't agree entirely with him (but what do I know). Overall though, he's right up there with Floyd Toole and where they disagree, I often find myself siding with Earl.
Great topic. Distortion matters especially for speakers. Ironically, marketing people don't want to advertise distortion figures for speakers because they are usually horrible. Any good amp will have a distortion that drowns in the speaker's distortion. I usually like to see THD plots through the frequency range at different levels as they can tell you of any spikes in THD. Still distortion in a speaker is more complex and you need to also do FFTs to see what sort of distortion it is. Ultimately your ears will judge what sounds good.
grant jonsn Well, you can’t just rely on ears. Measurements are extremely important also. But of course any compromise between options should serve what sounds best to our ears rather than what looks best on an instrument.
ThinkingBetter I agree that measurements will definitely give one access to options one could not normally have found/thought of without. Of course if a measurements-only-based setup sounds like shit then there will always be our ears.
I owned horn speakers before and they distorted audibly at near field, but at listening distance they made for a nice, warm, physical sound that would rumble my gut. I sometimes miss that flawed sound. Maybe that's how the human hearing/brain developed? Maybe we expect noisy environments so our brains don't actually mind?
Chris was awesome and complements you Paul. You are great story teller Paul and you are like a Grandfather I never had, just being calm and patient and teaching you new things all the time. Chris just ads that afterthough and hands on “from the trenches” stories. Great couple.
This is why I always said that the speakers are the weakest link in a system. So getting a decent speaker is more important than paying big bucks on an amplifier.
Very interesting insights and considerations. It's a known fact that it's difficult to establish meaningful correlations between traditional measurements (i.e., THD and IM) and listeners' sensations and preferences. Brain psychoacoustics enters the equation, so hard and fast relationships lose their objectivity and cease to be valid in many cases. Second harmonic distortion, while still being distortion of course, may be perceived subjectively as added musical richness to the instruments being reproduced. That's a kind of distortion that blends in well with the music and may in fact go unnoticed at not-so-loud volumes. This kind of distortion adds "weight" and "body" to the sound, and my take on this is that it might help even out the thin, flat tone quality character produced by many typical microphones out there during the recording phase, i.e., a thin, recessed and contained sound which lacks aural "expansion" (in subjective, not objective terms). I also guess that THD and IM measurements represent static, steady-state measurements which do not exactly translate into the way our ears hear, since these rather follow a dynamic state behavior where spectrum-related issues are especially predominant, and classical distortion measurements don't describe the way by which sound waveforms decay in time either, something quite relevant in the context discussed here.
After thinking about things said in the video and your comment, I think I have a very good idea of the things being discussed here. I build speakers as a hobby mostly to see how good I can make a cheap speaker sound by going overboard on the box. I have seen some very interesting things where the box is very easy to reduce or increase distortion just by changing the placement of supports and bracing. Thanks for giving more incite into the strangeness I was seeing.
@@99Duds The issue of cabinet resonance (and its occasional contribution to the final sound we hear from a speaker) constitutes an age-old debate as well. Today I tend to believe that a cabinet would better be either very stiff or, if allowed to resonate by design, produce good sounding harmonics, ones which blend in well and imperceptibly with the music (mainly 2nd harmonic and higher even order harmonics). Good choices for the positioning of bracings are mandatory (usually avoiding symmetry and applying bracings at 1/3 or 1/5 lengths from ends). Another good idea are baffles made up of noble solid woods allowed to mildly vibrate, supporting only the immediate area surrounding the cone, while making use of either MDF or plywood for the remainder of the box where bracings apply. In fact, resonances are not bad in themselves; but as long as they are beautiful resonances, carefully applied to enhance and embelish the main sound produced by the driver (No, I'm not speaking a heresy here!) Just think of a violin's hollow body. (And yes, speaker cones can be thought of as being musical instruments as well, since they will always produce some form of unavoidable harmonic distortion, and will never be neutral transducers in real life, regardless of what textbooks may tell you). Since harmonic distortion is always present, you better channel it in order to make it sound beautiful as you wish it to.
@@johnwhite2576 Open baffle and planar speakers are indeed interesting alternatives, but they present typical problems of their own too. Usually the problem with these is extracting good bass response; open baffles have to be carefully calculated so as to work well, and planar speakers must have large areas and be suitably positioned in rooms. (The large Magneplanar MG-20's are a good example, seeing that they are power-hungry speakers at about 400 or 500WRMS in order to work really well)
The problem with THD is the T for "Total", meaning it's a value that doesn't take into account psychoacoustics, i.e. that we're extremely sensitive to some types of higher-order distortion, but, as you mention numerous times, barely at all to 2nd order, for example.
Laurentz Duba i was just talking with my church’s head sound engineer today, we’re both hifi nuts. I play guitar there, and we were discussing how angus used a deceptively clean tone to make his guitar sound crushing. It worked!
Distortion is fine if it's intended. Then it's not distortion, it's an audio effect intended to evoke a certain emotion. Speakers should have low distortion, unless of course you enjoy a particular type of distortion signature which will appear on all your songs you listen to. Most don't like this, but a few do. The reason most hate this is vocals will always be compromised and held back.
RU-vid Curates Content, algorithm is a lie Yes, a speaker distorting is not likely to benefit a broad range of music. Usually speaker distortion is very frequency dependent. I prefer purest sound possible.
The Dutch company Philips tackled speaker distortion brilliantly many years ago with their motional feedback speakers. Their logic was that negative feedback should not end at the amplifier output.
The did a lot of things that are innovative now (that AD9710 from the 60's also had a copper sleeve over the pole) but they where horrible at marketing.
You'll never get out of this distortion dilemma as long as the traditional moving coil speaker is around ... Non linear Bl and compliance .. flux modulation are here to stay .. EV and JBL have made impressive inroads over time with there powerful PA systems .. tackling head on these distortion artifacts .. long throw cone suspensions are a no no really .. maybe we might have to go and revisit the good old days when woofers had limited excursion and still had low fs and high Qts but required large enclosures
The new Purifi woofers have solved the "long thrown no no" problem -- they have distortion numbers at least as good as best big drivers from the pro audio companies and they're small enough for bookshelf speakers.
Yep, low excursion isn't so bad. A speaker should really only play down to 60hz. Targeting 40hz and below causes a ton of problems. Clarity, distortion, etc all get worse trying get another half octave out of a tower or bookshelf. Besides, around 80hz and below is really a subwoofer's job.
@@thunderpooch great comment, and with a high pass at 80hx and above we can remove the ports and slave drivers to get even more sound quality out of the speakers.
@@sudd3660 For normal sized towers I wouldn't bother with the additional crossover unless it's dsp. But for small 2 way line arrays where you have small mid range drivers playing quite high I would recommend using a high pass filter to filter out the sub woofer range. It cleans up the mids and vocals.
Yeah, I hate that THD and impedance charts are all I have to go with, and it's hold my nose and dive in on a driver that seems promising. On the other hand, I'm SO glad that there are engineers like Chris who even acknowledge that psychoacoustics exists, it wasn't long ago when it was treated as a pseudoscience in hardware even though software engineers had long been mapping out how to fool the ear in as few bits as possible.
Nothing more beautiful in a club system... at certain moments in a song where you intend them to distort, it is not a reproduction of an acoustic sound - it IS an acoustic sound. With the right speakers it is a good trick to keep up your sleeve, not something you want to feature through a whole song...but music producers today know this trick and build it into pop records...it isn't the same timbre though..there is maybe a primal connection to acoustic sound that the human ear has had millions of years to get used to, which can get those hairs to stand up
Long coil is the cheap solution. Short coil in large magnetic gap is the most linear. Copper is not the ideal solution to eddy currents - a high mu material that is non conductive is slightly better. Thermal Compression is another huge issue. The rise in distortion with volume means the majority of speakers can’t be tested above 95 dB SPL. Distortion in very large ATCs is around 0.3% THD full range at maximum volume. Masking is indeed a factor - design a speaker with boomy underdamped bass and you won’t hear as much mid range distortion as it is masked. Tight critically damped bass will expose the mid range quality! GREAT TO SEE PS AUDIO DOING REAL ENGINEERING ON OPTIMIZING THE ENTIRE SPEAKER SYSTEM rather than put off the shelf parts in a nicely veneered box - only wish more than just 1 or 2 companies would take this approach.
@@cbrunhaverI wouldnt go over $1000 myself as the whole point becomes moot honestly. Be great to see a certain speaker done but say with different driver configurations based on low/med/high price points and their sonic differences. A "poor mans" wilson watt puppy would be absolutely awesome!
@@AllboroLCD well, you might consider kit speakers. www.troelsgravesen.dk/ has a few designs resembling a WATT/puppy but even those will cost more than $1000 in parts.
@@cbrunhaver Whoah, hey Chris! I was mildly joking, I also caught the vid with Paul talking possible price tiers on the PS speakers too. A set on offer around the $5k mark is certainly reasonable all things considered. I did NOT know about the watt puppy kit, thank you for that! Perhaps once the PSA speaker line gets some traction you folks could offer a kit, that would be very cool! Godspeed Brotha
Chris or any other experts on here, I am buying a pair of Elac unifi ub5s, 4 ohm, sens- 85 db at 2.83 v/1m, rec power 40-140. I was thinking of using my old Krell Kav 300i, 300 watts into 4 ohm. Is it too powerful for these speakers.
That's interesting I wondered years ago way sticking a washer on the pole help to my speaker sound better, I think under gap is better than over gap. I did think why didn't they split the coil have a gap between an extra pole staggered in that gap, like a braking system, I don't know if any manufacturers have gone down that route, what hes saying about staggering the pole maybe it's the same. That's why tannoy Earth the chassis of the drivers to cut down oscillation
I believe Audiovector do similar.. with their freedom grounding system which involves earthing the driver chassis ... Plus some other things such as titanium voice coil formers are also used to try and reduce distortion.
@@ronbradshaw7404 -- So, you would regard a driver that is, given a suitable small (15.5"h x 8.5"w x 11d) enclosure, down 10 dB @33 Hz and pretty nearly flat starting at 65 Hz through 1 kHz as what exactly, a penny-whistle? en.wikipedia.org/wiki/Woofer
@@ronbradshaw7404 -- I forgot to mention that it's less than 2 dB down @50 Hz in that enclosure, so by definition it's a woofer. Need deeper bass? Get yourself a sub and turn its low-pass knob way down. I live in apartment, so live concert volume is out of the question here, as are massive floor-standers with 12-15" woofers. Moreover, check out this reviewer description: "...most of the time distortion is an indicator that the volume is too loud. We hear distortion and we turn it down. With these speakers, I was often in the 95-100dB level (at 11 feet, no less) without having a clue it was that loud." If you need more than a clean 100 dB SPL to enjoy music, I guess this isn't enough of a woofer for you -- but it's still a woofer, period. :-)
And then, why even bother about reducing the THD in amplifiers? If the loudspeaker will add 1-20% distortion, why on earth is an amplifier with 0.1% bad? Why not focus on other aspects in the electronics part of the chain?
Sleeving the pole pieces in copper, So you use ALUMINUM COIL? HOW DO YOU ENGINEER SPEAKERS , OVER THE past 100YEARS the basic design hasn't changed. I guess ENGINEERING Means (Giving them an inferior product but making it seem good quality? Thank God we still have Bose. 🙏🙏🙏
Well, depending on the design, copper, copper clad aluminum or aluminum is used. Of those, Copper gives the most number of turns (and resulting force factor)for a given resistance, but has the highest mass, whereas aluminum gives fewer turns but is lower mass. Copper has a more gradual positive temperature coefficient and leads to somewhat lower power compression. Typically, we’re using copper coils in woofers. Well, cooking was invented many years ago but new, delicious recipes are being made all the time. Sometimes engineering doesn’t mean starting from first principles and innovating an entirely new thing, it’s optimizing a type of “recipe”, understanding the inherent design compromises.