I realize this may be a strange question, but I've been wondering it for awhile. My highest quality set of speakers is a pair of Zaph SR-71's. To me they sound perfect, but even with these speakers, there's something that your ears can 'feel' that I'm not sure how to describe objectively that lets you know it isn't a real instrument playing in front of you. Certain sounds sound very much like a live instrument, like the lower end of a guitar or a bass guitar, but a piano for example doesn't create that 'twang' in your ears that you feel when you're standing next to a real piano. I'm wondering if you guys can explain the technical reason why speakers can sound accurate but not create the 'feeling' in your ears that live instruments create? Is it just based on sound pressure or the size of the transducing material? For example, tweeter surface area being much smaller than the surface area of an instrument?
Thanks!
Thanks!
Of course you are absolutely right, there's a blaring difference anybody would easily notice ... if they ever cared to listen to real, live Music anyday.
I mean Music not passing through electronics or speakers, so the weekend visit to a local club or Church dos not qualify, sorry.
Speakers are wonders of Technology and it's incredible they perform as well as they do ... but still miles away of real sound.
FWIW at least once a year, 3 or 4 if possible, I go on a weekday (so I don't have to dress up formal) to Teatro Colón, Buenos Aires Opera House and listen to some Symphonic Music.
I close my eyes and pinpoint instrument by instrument, marvel at the transparent sound, the harmonics, the REAL "soundstage" , the lack of distortion/shrillness/mud.
NO speaker sounds like that, sorry.
If you can, try to do the same
Part of the problem is the listening room itself, no way those acoustics will even feebly imitate the real thing, even less through 2 tiny speakers placed against a wall (or a couple feet from it) .
Only speakers I ever listened to approach reasonably that kind of sound are those big panel electrostatic speakers .... but cones ... domes .... horns (ugh!!) ..... not even close.
I repeat: you must now and then "recalibrate your ears" or you'll never have a real reference.
The basic system uses 2 speakers to produce a sound field which attempts (maybe) to mimic a sound field which is produced by a number of variously spaced instruments or voices, or a single instrument in some kind of (enclosed, or not) space.
The system is imperfect.
The recording is imperfect. The mics have a non-flat frequency response, as does every individual part of the system, mixing desk, playback amplifier, speakers, the fact that an experience which is basically 3D has been compressed into 2 channels. Often there's no pretence at doing anything other than stringing instruments out in a line from left to right. It's a big ask, to expect your ear to be completely fooled. And it's not. But it comes remarkably close, and it's very difficult to improve upon, even if you throw a lot of money and technology at it.
There's also no one aspect of the system which is more responsible for non-idealities than the speakers. It's just difficult to turn a lot of electrical energy into sound energy efficiently and accurately, no matter how you go about it. Everything else can be done to a pretty high degree of accuracy.
Remember, however, that the problem of days not so very long past was to make the live performance sound sufficiently like the record, not vice versa.
The system is imperfect.
The recording is imperfect. The mics have a non-flat frequency response, as does every individual part of the system, mixing desk, playback amplifier, speakers, the fact that an experience which is basically 3D has been compressed into 2 channels. Often there's no pretence at doing anything other than stringing instruments out in a line from left to right. It's a big ask, to expect your ear to be completely fooled. And it's not. But it comes remarkably close, and it's very difficult to improve upon, even if you throw a lot of money and technology at it.
There's also no one aspect of the system which is more responsible for non-idealities than the speakers. It's just difficult to turn a lot of electrical energy into sound energy efficiently and accurately, no matter how you go about it. Everything else can be done to a pretty high degree of accuracy.
Remember, however, that the problem of days not so very long past was to make the live performance sound sufficiently like the record, not vice versa.
I think "reason" in the title should be plural, but here's my take on what I think is the main one.
It seems to me the biggest difference is dispersion/radiation patterns between loudspeakers versus various musical instruments. Loudspeakers are designed to have specific dispersion patterns, not just so they "sound good" when your ear is off axis as well as on axis (or at least that the highs drop off fairly "naturally"), but so the sounds reflected off the walls, floor, ceiling and other things in the room aren't seriously different in frequency response from the on-axis response. This of course doesn't work at all frequencies, as bass is generally omnidirectional, and the higher end rolls off more off-axis (but that's true of many musical instruments as as well as loudspeakers).
Most musical instruments DO have "seriously skewed" radiation patterns, so the sounds coming off the instrument and reflected back to you from different directions in a room have very different frequency responses. This also varies greatly between different instruments. But since all the sounds from the loudspeaker have the same dispersion, there's a "sameness" in the off-axis and reflected sounds in the room.
This is further complicated by all the instruments coming from a same point in the room (the loudspeaker), which of course never happens with an unamplified live performance of several instruments.
It seems to me the biggest difference is dispersion/radiation patterns between loudspeakers versus various musical instruments. Loudspeakers are designed to have specific dispersion patterns, not just so they "sound good" when your ear is off axis as well as on axis (or at least that the highs drop off fairly "naturally"), but so the sounds reflected off the walls, floor, ceiling and other things in the room aren't seriously different in frequency response from the on-axis response. This of course doesn't work at all frequencies, as bass is generally omnidirectional, and the higher end rolls off more off-axis (but that's true of many musical instruments as as well as loudspeakers).
Most musical instruments DO have "seriously skewed" radiation patterns, so the sounds coming off the instrument and reflected back to you from different directions in a room have very different frequency responses. This also varies greatly between different instruments. But since all the sounds from the loudspeaker have the same dispersion, there's a "sameness" in the off-axis and reflected sounds in the room.
This is further complicated by all the instruments coming from a same point in the room (the loudspeaker), which of course never happens with an unamplified live performance of several instruments.
A recording is exactly that. A copy of the real thing. Ambience being condensed into a single dimension just isn't the same. The recording is as much to blame as the speaker..
The simple, "technical" reason is that you're asking the wrong question, as nearly everyone does - the right question is "The technical reason systems sound different from live instruments?", 🙂.
Because the speaker behaviour seems so obvious, and it's the biggest thing in the room, it always has the finger pointed at it - but if one does the right sort of experiments one realises that the speakers are only one small part of the "battle", to get "right". Essentially, almost any sort of speaker can be made to sound realistic, provided the rest of the system is up to it! Very few people bother to get the latter happening, hence nearly all systems sound different from the real thing ...
Because the speaker behaviour seems so obvious, and it's the biggest thing in the room, it always has the finger pointed at it - but if one does the right sort of experiments one realises that the speakers are only one small part of the "battle", to get "right". Essentially, almost any sort of speaker can be made to sound realistic, provided the rest of the system is up to it! Very few people bother to get the latter happening, hence nearly all systems sound different from the real thing ...
I think some of what you mean is the lack of focus of the speakers. Coaxial drivers do much better at this, even in the pro monitor type speakers.
If you saw the time domain waveform coming out of most speakers, you wouldn't believe how bad it is.
This smearing is clearly audible in a single mono channel. Two channel stereo just messes things up even more, if not carefully set up.
At its best, stereo is just "about as good" as mono, except it can spread out several sound sources laterally.
If you saw the time domain waveform coming out of most speakers, you wouldn't believe how bad it is.
This smearing is clearly audible in a single mono channel. Two channel stereo just messes things up even more, if not carefully set up.
At its best, stereo is just "about as good" as mono, except it can spread out several sound sources laterally.
Last edited:
Nah.
Reason (1) best case, stereo recordings are point samples from two points in space. This does not fully describe the real world soundfield and can not.
Reason (2) best case the playback comes from two point sources (not that you own true point source speakers) which can not provide more than 1.5D of spatial information. It provides left and right, and via amplitude and phase cues, some front to back. Zero up and down, and nothing behind or to the side (save inadvertent out o phase information that appears to come from the side).
The best one can get is a fairly convincing stereo image from well recorded material and high quality playback systems.
Personally, I'd really like to see some digital multichannel format made available, and recorded appropriately to retrieve more of the "live space".
Reason (1) best case, stereo recordings are point samples from two points in space. This does not fully describe the real world soundfield and can not.
Reason (2) best case the playback comes from two point sources (not that you own true point source speakers) which can not provide more than 1.5D of spatial information. It provides left and right, and via amplitude and phase cues, some front to back. Zero up and down, and nothing behind or to the side (save inadvertent out o phase information that appears to come from the side).
The best one can get is a fairly convincing stereo image from well recorded material and high quality playback systems.
Personally, I'd really like to see some digital multichannel format made available, and recorded appropriately to retrieve more of the "live space".
Trouble is, one day - perhaps by accident - one gets quite amazing, and convincing, soundscapes being generated from the most unlikely, "poor", recordings - and the realisation sets in, that there is a lot of misunderstanding in the audio world of what is truly important ...
What will happen if the mics are placed at where your ears were, in the opera hall, with a manequinn of the the listener's size. Will the speaker listener have the same experience...
Gajanan Phadte
Gajanan Phadte
That is called binaural recording and there's a particular label -I don't remember the name ...!- that makes recordings for reproduction with headphones ( only ?)
Agree.
Binaural recording, sometimes even putting a couple of small condenser microphones inside the ears of a realistically molded latex head, and then reproducing through good headphones, overcomes most of the problems and is incredibly realistic.
Even so, there's a simple test that shows it's not the real thing in 1 second: you simply move your head from right to left and viceversa while wearing the headphones: in the live situation you would notice yourself moving and the artists staying in a fixed space; while the headphones move with you, so the "soundspace" also moves
An Argentine colleague, Hugo Zuccarelli, developed a technique which through some processing he does not disclose in detail, (comb filtering? ... selective delay processing? ... mixing parts of the audio out of phase? .... all together?) he boosts the channel separation in binaural recordings so much that when played back through speakers placed in very precise spots, he recreates those fields reasonably well .
The system is too unforgiving to be used at home, so he regularly plays it only in theaters.
You sit, lights are turned off, and he plays various music and sounds, and in a large space such as a theater it's very convincing.
The system was good enough to be used at the last Pink Floyd album, The Final Cut, go figure.
But it's not practical for general use at home.
Holophonics - Wikipedia, the free encyclopedia
http://www.acousticintegrity.com/acousticintegrity/Holophonics.html
Most speakers lack bass, so there could easily be a missing octave at the bottom. There aren't many musical notes down there, but there are lots of mechanical noises which real instruments (and performers) make.
And don't forget the whole purpose of a recording is to produce a sound you enjoy listening to. All the recording processes are there for that, not to make a lab quality warts-and-all true record of the sound.
When you buy a steak in a restaurant, and the chef sprinkles some salt on the meat before grilling it, he does that so it tastes better, not to make it taste more steakish.
The recording engineer sprinkles his own acoustic salt on the recording for the same reason. It sounds better to your ear.
When you buy a steak in a restaurant, and the chef sprinkles some salt on the meat before grilling it, he does that so it tastes better, not to make it taste more steakish.
The recording engineer sprinkles his own acoustic salt on the recording for the same reason. It sounds better to your ear.
What would happen if you recorded each instrument individually and gave each one its own speaker in its own place during playback.
Other than a divorce.
Other than a divorce.
You'd have all the same influences, but 100 times more of them.
Think about one flute or one violin, sitting in one chair out in the orchestra. The instrument makes sound in all directions, a microphone picks up only in its pattern and a speaker makes sound only out the front.
Think about one flute or one violin, sitting in one chair out in the orchestra. The instrument makes sound in all directions, a microphone picks up only in its pattern and a speaker makes sound only out the front.
Ahem, not so. At least not if realism is what we're after.
The RE does what the producer asks for - which, in 90% percent of cases IS NOT so that it sounds better. Mostly a crap load of compression post ADC and a pile of limiting/compression pre ADC.
Nuff said by others on mike placement but then we might get some reverb and panning thrown in by the RE to compensate for that.
The 10% which does sound better ON MOST SYSTEMS is to remove content which will upset said systems - bass below 30Hz. (Plus auto-tune pitch correction because the singer can't 😛 )
Still all of the above make it sound less realistic.
At the end of the day, stereo is still only as good as a view-master (think about it).
(if you want some real recordings, see Tom Danley's files mentioned in this thread
Last edited:
I don't want to get into a tussle here, but how many recordings start out with realism as the goal? Go into any music store and pick the top 50 selling recordings, or top 100. How many of them say on the cover this is REALISTIC sounding? How many people prefer the singer missing notes to pitch correction?
The ultra hifi symphonic music crowd is a VERY small portion of the music consuming public.
The ultra hifi symphonic music crowd is a VERY small portion of the music consuming public.
- Status
- Not open for further replies.