Let's both be clear that no one is saying that one can eliminate > 7 kHz with no audible change in quality, its just that eliminating > 7 kHz has a small effect when compared to eliminating 700 - 7 kHz.
One has to learn in audio that everything is not equally important. There are vast differences in the audibility of different aspects of the problem. Learning to balance these aspects to yield an optimum given a set of constraints is what it is all about.
What I find often happens is that we get fixated on one aspect of the problem, focus in on that, both mentally and audibly, thus raising its level of importance. This is natural, and necessary to improve that aspect of the problem. But we must always remember to come back down to the "big picture" where our last fixation may only be a very small part.
Last edited:
That might have been in Houston, but not at headquarters in DC. The Manned Spaceflight building was across the street from the NASA Headquarters building, which was shared with the Department of Health, Education, and Welfare (they had the bottom 4 floors, we had the top 3).
We had a lot of funky ex-military gear, nothing too fancy. The two projectors were ex-Navy 16mm projectors modified for brighter xenon-arc lamps (closed). The three Vu-Graph machines were the usual high-school style with a reflex mirror and a simple projections lens. The rear-projection screens had 3 sections of frosted glass with a small slit between them. From my side, sitting in a darkened room with dark-painted walls, all I could see was a white glow from the screens. What they saw were one to three high-contrast screen images that had good contrast despite all of the fluorescent ceiling lights being on (they needed to see their notes during the meetings). I was told to stay away from the screens during a presentation, because they would see a ghostly face (mine) appear on their screens if I got within a foot or so. So I scampered around the room, flicking the Vu-Graphs from the IN pile to OUT pile, watching audio levels, listening for requests for out-of-order Vu-Graphs, playing a movie when requested (sometimes complete with changeover from one projector to another), etc. The voice-controlled analog version of a three-screen PowerPoint presentation today.
I never went to the facilities at Houston, but I gathered that's where all the techno razzle-dazzle was. We did have those little Bell System speakers through all the rooms where we worked, and listened to the live, uncensored feed from Houston during the Moon landings. Weirdly enough, I was offered a free ride to Woodstock during that summer, but I would much rather stay working at NASA. Besides, roughing it in the mud and crowd of a rock festival is not my thing, so I was perfectly happy to put in my time zipping those Vu-Graphs across the projectors while we listened to the live feed from Houston.
Last edited:
My method is not very scientific; I use inverse Fletcher-Munson weighting to assess the severity of spectral peaking, narrowband diaphragm resonances, diffraction problems, and inter-driver phase spread in the most critical 1~5 kHz band. In other words, problems in these four areas are much more critical around 1~5 kHz then they are a decade lower. From my perspective (which is not that common in the high-end industry), problems in 1~5 kHz region need the most attention, since they really jump out at you, and can sound the most unnatural and artificial. Room colorations are dominant a decade lower, and have a way of mixing in with cabinet coloration or recording-balance issues.
If the 700 Hz to 7 kHz region is smooth, flat, free of diffraction and time-domain energy storage problems, you are well on your way to a good loudspeaker.
Last edited:
Interesting. Doing a crossover at 1K or near there is something I find frustrating. Any tiny little change seems audible. Like no room for error, no fudging it, no slack. It's a tricky area.
Oh yeah, that's kind of the other side of me. I was in Hong Kong during all the run-up in Europe to color television, and followed with great interest the political and technical contest between NTSC, PAL, and SECAM. All we had in Hong Kong at the time was 405-line monochrome distributed via twisted-pair Rediffusion cable-TV. Two channels, one Mandarin Chinese, the other English.
Also an avid photographer; my first "real" camera was a Pentax Spotmatic, which I got a a couple months after it debuted in Hong Kong (which was a world center of photographic distribution at the time). Shot Tri-X at ASA (ISO) 320 and developed in Rodinal for maximum acutance, and printed on either Du Pont or Ilford paper. Kept that until I got a Nikon FE2 around 1982 or so. It feels a little weird to see a book in a photo store describing the Nikon FE2 as a "classic" camera. If my 2nd good camera is now a classic, what am I?
Also an avid photographer; my first "real" camera was a Pentax Spotmatic, which I got a a couple months after it debuted in Hong Kong (which was a world center of photographic distribution at the time). Shot Tri-X at ASA (ISO) 320 and developed in Rodinal for maximum acutance, and printed on either Du Pont or Ilford paper. Kept that until I got a Nikon FE2 around 1982 or so. It feels a little weird to see a book in a photo store describing the Nikon FE2 as a "classic" camera. If my 2nd good camera is now a classic, what am I?
The big trick is inter-driver phase angle. This is extremely audible. You can hear it across the room as a weird, non-tonal coloration that comes and goes, depending on program material. Pink noise is the most reliable subjective test material for assessing problems with inter-driver phase angle; most music so dynamic it's hopeless trying to figure out what you're hearing.
It's awkward to directly measure inter-driver phase angle, so I just use a summation method of temporarily reverse-phasing one driver and trimming the crossover slopes for maximum null depth. I aim for at least 15 to 20 dB of null at the listening position, and reasonable symmetry laterally and vertically. Then I return the phase to the "normal" position and make sure that a broad rise has not developed. I'd rather have the nominal crossover frequency be just a little down (maybe by a dB) than elevated, where it can draw attention to itself.
If one of the drivers is acting up, with uncontrolled resonances, that really messes up the inter-driver phase angle. This is the number-one crossover problem in old-school crossovers from the Fifties, along with gross driver colorations. Smooth out the driver, using notch filters if necessary, and the whole crossover region behaves better.
If you're doing the job right, you can walk right up to the speaker, while playing pink-noise at a moderate volume, and you'll hear what sounds like one single driver, no matter where you move your head. In my experience, that takes an inter-driver phase angle no greater than 10 to 15 degrees, with no narrowband deviations in the crossover octave.
Interesting discussion about 700 Hz - 7 kHz. If you've ever heard a horn playing just 700 Hz and above, it is fascinating to see how much music is below that range. Almost all of it. Intelligibility may be completely coming from the 700 Hz - 7 kHz range, but most of the music fundamentals are below 700 Hz. My point is, the range below 700 Hz is of extreme importance and is not secondary to the range above it.
I agree, I think what happens is folks that haven't lived with a good horn in that frequency range (most people ) discount the entire range as secondary and suppose a direct radiator is enough. The 700-7000 range is ultra important too but without the realism down to 100 cycles or lower I feel it's a big yawn. It's like me driving my Miata feeling it's all I need without ever driving that new Porsche. I guess this was worth repeating?
I agree, I think what happens is folks that haven't lived with a good horn in that frequency range (most people ) discount the entire range as secondary and suppose a direct radiator is enough. The 700-7000 range is ultra important too but without the realism down to 100 cycles or lower I feel it's a big yawn. It's like me driving my Miata feeling it's all I need without ever driving that new Porsche. I guess this was worth repeating?
Proper mid bass is the guts and the glory; keeping the range of 700-7k intact, is the yummy icing on the cake
Nikon FE2 an FM. I shots 1000s of photos with those and loved them. Had them up to about 4 years ago, when they just fell apart. Sold the lenses.
Lynn, if you remember 8x10 cameras, you might find it interesting that I used to shoot with one, outdoors. That will get you some attention for people on the street! It's a slow, deliberate process, and you don't shoot a lot of film. I never had an enlarger for 8x10 negs, so only made contact prints. Gorgeous things they were. Real high fidelity.
Lynn, if you remember 8x10 cameras, you might find it interesting that I used to shoot with one, outdoors. That will get you some attention for people on the street! It's a slow, deliberate process, and you don't shoot a lot of film. I never had an enlarger for 8x10 negs, so only made contact prints. Gorgeous things they were. Real high fidelity.
I never shot (outdoors) with 8x10, but I was struck by the rendering of textures by large-format negatives and the cinematic equivalent, 70mm. I worked at the outer bound of 35mm thanks to careful technique, and superb lenses like the Nikon Micro-Nikkor 55mm and the Schneider Componon enlarging lens. It was sharp, no problem there, but the rendering of something like sand, or the texture of wood bark, just didn't happen. Large and medium format are about textures, not sharpness. As a result of more natural rendering of textures, you could see the grainless airy openness of the sky and clouds contrasted against bark of a tree in the foreground.
I feel the same applies to artisan audio (not really high-end per se). A properly running direct-heated triode system is striking in the rendering of textures, an almost tactile quality to the sound ... and it can be heard even at very quiet levels. This is not quite the same as the ultra-resolution of electrostats or good headphones, which tell you a lot about the details of the mix, but more a feeling of connection to the music. Karna calls this quality the "essence" of the music, and is more felt than heard. This quality does not lend itself to the usual audiophile lingo, or the even more ridiculous checklists that some reviewers use. It's an emotional feeling that is evoked. Audiophiles who are hyper-analytical may block these interior feelings that arise, and can miss the subtler aspects of the listening experience.
Speaking only for Karna and myself here, this experience of "essence" is rare with transistor amplifiers. Adding in a bunch of harmonics with a software or hardware gizmo does not result in a simulation of a tube amp. It'll warm up the sound, and maybe rescue a harsh mix, but it doesn't have the effortless natural quality of a good triode setup. I'm not sure what's going on, but it's pretty consistent, and both Karna and I hear it in a matter of seconds.
I feel the same applies to artisan audio (not really high-end per se). A properly running direct-heated triode system is striking in the rendering of textures, an almost tactile quality to the sound ... and it can be heard even at very quiet levels. This is not quite the same as the ultra-resolution of electrostats or good headphones, which tell you a lot about the details of the mix, but more a feeling of connection to the music. Karna calls this quality the "essence" of the music, and is more felt than heard. This quality does not lend itself to the usual audiophile lingo, or the even more ridiculous checklists that some reviewers use. It's an emotional feeling that is evoked. Audiophiles who are hyper-analytical may block these interior feelings that arise, and can miss the subtler aspects of the listening experience.
Speaking only for Karna and myself here, this experience of "essence" is rare with transistor amplifiers. Adding in a bunch of harmonics with a software or hardware gizmo does not result in a simulation of a tube amp. It'll warm up the sound, and maybe rescue a harsh mix, but it doesn't have the effortless natural quality of a good triode setup. I'm not sure what's going on, but it's pretty consistent, and both Karna and I hear it in a matter of seconds.
Last edited:
Well, your idea about audiophiles baffles me. I consider myself an audiophile, yet, the major factor I'm looking for in a sound set-up is the closest possible feeling of recreating the feeling of live music, or of music being played in the room. All other attributes are secondary.
It looks like according to your definition, or terminology, I'm not an audiophile. How would you label me?
Joshua,
My approach is simple: visit live concerts as much as you can.
I am in the happy circumstance that I can visit live (classical) concerts about twice a month.
The live sound of individual instruments, voices and orchestras is settled in my brains.
In my younger years I played the French horn in orchestras and chamber music ensembles.
This active and passive reference, when listening to reproduced music at home, helps me enormously to recreate some of the live experience, but, I admit, you will never get quite there.
Even direct heated triodes or horn loudspeakers (I don't like them because all horns IMO and hearing more or less color the sound) will not get you there.
Accepting this will make it easier to enjoy reproduced music.
I know you visit concerts (too many "audiophiles don't); consider yourself a happy camper, don't feel to belong to some class of audiophiles....
My approach is simple: visit live concerts as much as you can.
I am in the happy circumstance that I can visit live (classical) concerts about twice a month.
The live sound of individual instruments, voices and orchestras is settled in my brains.
In my younger years I played the French horn in orchestras and chamber music ensembles.
This active and passive reference, when listening to reproduced music at home, helps me enormously to recreate some of the live experience, but, I admit, you will never get quite there.
Even direct heated triodes or horn loudspeakers (I don't like them because all horns IMO and hearing more or less color the sound) will not get you there.
Accepting this will make it easier to enjoy reproduced music.
I know you visit concerts (too many "audiophiles don't); consider yourself a happy camper, don't feel to belong to some class of audiophiles....
Last edited:
I agree 100%!
You can get a good visualization of inter-driver phase matching in HolmImpulse by measuring the two drivers' impulse responses (IRs) with the same fixed "time zero", then subjecting the two IRs to ex-post band-pass filtering centred on the crossover frequency, and looking at whether they overlap.
Marco
Last edited:
... and too large as well (the blankett): here guys are also working about horns'dirctivity, waveguides... to reduce colorations of the rooms ! Not sure some distorsion can be superior to rooms coloration !
Choose your poison !
PS : an audiophile who "hears" the reccording at 90 dB max spl with poor dynamics is not a music lover but an idiotdiophile : classical music has its requirements : yes go to the concerts to understand it and sold your little three ways !
Choose your poison !
PS : an audiophile who "hears" the reccording at 90 dB max spl with poor dynamics is not a music lover but an idiotdiophile : classical music has its requirements : yes go to the concerts to understand it and sold your little three ways !
Last edited:
Last edited: