Damn, I really wanted to make some brief comments and I’ve written another frickin chapter of a book. Anyhow, thanks much for everyones posts, and if I’m taking up too much bandwidth here I apologize. Here’s my comment:
—————————
The subject of this thread is evidence based speaker design. My question is, what evidence?
I take it on reputation that the Harmon studies are well done, and they have identified characteristics of speaker design that can be objectively measured that correlate with reported listener preferences. For brevity let’s call them “accurate” speakers.
There are two levels of “evidence” here. One is the evidence of psychoacoustic testing, and the other is the measurements.
On a practical level, a speaker designer could take the social science aspect as done, and then use measurements to help design the speakers in accordance with those standards.
If you are a speaker manufacturer, then this could very well be the gold standard for audio performance. But I’m sure no speaker manufacturers can ignore the other factors of perceived value in speakers!
I think it is unlikely that visual aspects of audio gear designs can be dismissed as important elements in perceived value. We can consider those elements as having objective value, if the results are repeatable, in an individual or group study.
Especially for speakers, which are the sound source, I would be very surprised if visual/design elements didn’t have consistent effects on perceived quality.
While I doubt the effect is as significant as smell to taste, our ears, proprioceptors, and visual sensors are tightly coupled. These senses are all integrated in the brain through a complex, evolved, culturally informed processing system that gives us the end experience of “listening to music.”
In particular, I find speaker designs where the cabinet “disappears” to be unsatisfying. I do think this is in part because I favor rock music, meant to be loud, but I don’t like to listen loud. The cabinet resonance provides psychoacoustic cues that help create the illusion of high energy content.
I also think one of the biggest problems in music, and getting worse with the almost exclusive use of digitally generated sounds in pop music is that most music is multi-tracked and therefore must be integrated into a whole by the mixer. This is the crafting of an illusion that these sounds really were made together.
Sounds in space affect each other, (as do electric signals in analog gear). A big part of this is the reverberations, and artificial ambiance is often used to “glue” or “mix” the sound together. There are complex effects at play.
Digital audio has some problems as a recording medium, and these are especially exasperated with the current practice of digital mixing. Essentially (and um, this is just my own pet theory, so take it as informed speculation) sounds do not “mix” in a digital mixer. They are simply superimposed. You can completely remove a sound from a straight digital mix by inverting the phase and adding it back in.
This is just not how sound behaves in the real world. Digital recording also has very high dynamic range it can represent, so when signals from different sources are mixed straight, they mix poorly because the summer behavior does not mimic how acoustic and mechanical energy sums in the physical world. The result is that the relative levels are uncorrelated, and the sounds repeatedly mask each other. (This is another place where visual cues have an effect on auditory perception. If you can watch people playing, what I think of as the perceptual dynamic range is increased, increasing the ability to perceive the quieter sound under the load).
Mixers apply a lot of processing, and often distortion, to overcome this limitation and integrate the sound.
My pet theory is that this process happens in the speaker as well. But the higher the resolution of the speaker, the more ability it has to only present the signal and add nothing else, the less able it is to integrate the sound.
On mixes that are not well integrated, sonically, musically, this leads to an “uncanny” or “canned” quality. Fake.
This is less of a concern with more acoustically based recordings, captured in a real space, as this integration is done by the musician, playing in the room. While it’s tricky, a good recording can capture this, and good speakers can present this information in a way that makes you feel like “you’re there!”
Mr Toole outlined an ideal music reproduction environment, where studio producer types work on neutral, standardized monitors, crafting a signal, a work product, that is then presented to the listener on neutral speakers, so the actual work product alone is passed to the listener to experience.
While I really get this ideal, I think it way over simplified and does represent how the cultural production and consumption of audio products really works.
While there has been a move towards some standardization on the high end sound-for-picture world, the music world is following its own twisted path, and has escaped the construct he outlines completely.
One of the thing that surprises me about these studies of listener preferences in the lab is the behavior of so many folks in the “real world” as I experience them (anecdotally). The actual listening environments people choose and create for themselves seem wildly departed from any type of accurate playback. Bass and/or treble is cranked. Speakers are placed asymmetrically, and the sub is cranking out wherever space can be found.
The overwhelming trend is away from stereo reproduction. Car systems are crazy complex multi speaker environments. A lot of the little blue tooth speakers emanate sound from multiple drivers at a single site, projecting outward. I’m fascinated that some of these systems can sound pretty good, but I have no idea how the stereo signal is handled.
A lot of younger folks like to crank music in their cars, and in the case of hip hip, the music is mixed to rely on a specific, very loud, playback environment to be fully expressed.
I’m the one that goes around dialing back people’s systems to a semblance of accurate reproduction!
I don’t know, could just be me
Anyhow, research on any these effects could also be evidence! I’m not sure that would be fruitful or interesting. It kind of comes down to the intended target market, price point, tech trends, cultural trends.
Toole did make a comment that consumers are unhappy when speakers deviate from these aspects of accurate reproduction, so maybe Hartman has done this kind of research.
But core of my skepticism on this subject is that my personal preferences deviates from “accurate” speakers for fun listening. In the other hand, relatively accurate speakers are a must for audio work.
I mentioned the Genelecs we use at our studio in my other comment. Not only are these good speakers to work on (with the sub) they also do have a kind of pleasant sound. We’ve gotten far more comments about how good things were sounding on these monitors than others, and we’ve tried a lot. I’m pretty confident these would win the approval of many non-pro listeners. They fill up the room relatively evenly, and have a wide sweet spot, which is less important for the engineer, but more important for a person doing critical listening around the room. But I can’t bear them to listen to for fun. I would rather listen to a little blue tooth speaker, even with the limitations!
I’m highly confident that there is an objective reality to my subjective experience. For me that counts as evidence. I just can’t be imagining this. This pattern has held consistent for about the last 15 years. But I could very well be totally unique here, and am at best a market of one
I don’t consider this issue with speakers as being in the same category as the audio attributes of other audio gear, like DACs or amplifiers.
My subjective experience is that different amps have significant differences in audio quality. But I am convinced this is more in the realm of “imagination” than reality. For one thing, my perceptions will change. I think this amp sounds better...wait! On these other speakers my crap amplifier sounds great. Or I’ve thought I was listening to one amp, grumbling to my self about its deficiencies, and I find out someone had switched the cables to the amp right below it in the rack.
As far as I am concerned, it
has been proven that well made amps sound undistinguishable under blinded conditions, carefully level matched operating under clipping levels.
One of the weird things about “placebo” though it is not just a figment of the imagination. If someone compares their lamp cord speaker cable with their newly purchased gazillion dollar per foot speaker cable, and they hear big improvements, they really do hear them.
Blind testing can tease out whether there is actually an audio element to this perception.
I remain open minded, and am actively looking for good speakers. If anybody has some suggestions for good sounding, good measuring speakers that don’t cost big dollars, I’m all ears, hah hah
I would like to find some speakers I love for home listening. Right now I have a handful that I like, so in that sense, life is good. My current faves are ADS L1290, but it requires a fair amount of detailed EQing to set them up how I like.
This is definitely a downside to the vintage speakers I know, they do have annoying resonances. But if the overall sound is pleasing, then EQ can do wonders.
I also have a listening space where I’ve been trying out bookshelves on stands. There a handful of speakers that i have won extensive stays here: ADS L400, B&W DM12, Advent Baby II.
I like these of bookshelves a lot, but they are all idiosyncratic, need helping EQ, and fussing about with placement. They are fun speakers though. I find myself wishing I could combine some of their attributes.
These were all entry mid-fi speakers back in the day, so as an experiment I decided to try out a modern iteration of the market category, and picked up a pair of KEF Q150, mainly because they had a bunch of hype, and were marked down. (Maybe that should have been a clue.)
In any case, I was flabbergasted. These are simply unlistenable. I’ve heard cheap computer-media speakers that were much more enjoyable to listen to.
My main gripe was a kind “harsh” quality in the highs, and an overly resonant bass sound, I assume from the attempt to squeeze more bass from the port. And a lack of focus, for lack of a better term. It was hard to put my finger on, I could hear everything, but it just seemed kind of lifeless or dull. Lacking immediacy.
They did have a pretty uniform frequency response compared to my vintage standbys, and project a significant amount of sound into the space.
I’m not sure if these measure well. But I did expect much more based on the reputation of KEF, and rave reviews online, Amazon, Reddit, etc, and was hoping for some magic with the concentric drivers.
I know these are not really high fi speakers, but I just wanted to get a feel for what people are liking in this category. I need to find a better prospect for further research.
I wonder if anyone else has noticed this issue with newer/older designs?