What Is "Perfect Sound," Anyway?

[Anderton]

This is a spinoff from the thread on native vs. hardware processing. KuruPrionz said in it "Until we can eliminate or profoundly improve speakers, microphones, cables and outboard processing - 'perfect sound' (whatever that is) will never exist in this world." 

 

So...let's take a stab at defining it. To me, it's a sound quality that is identical to live music. In other words, if there's a solo classical guitar player sitting in my room, and a recording of that guitarist playing in my room, they should sound indistinguishable.

 

As he points out, the problem is tranducers. Converting one form of energy to another will never be easy, and that happens twice in the process - capturing the sound, and reproducing the sound. Yes, sound is air waves, and yes, speakers can create air waves, but they do so with physical limitations.

 

We can do anything we want between capture and playback, and throw as much technology as possible at it, but that's not where the problem is.

 

However, there is a potential (I suppose) for perfect sound. If the sound is generated without needing microphones (e.g., in the box instruments and signal generators), and if there was some way to bypass the ear and get sound directly into the part of the brain that processes what comes into the ears, then we'd be really close to the ideal - no mics, and no speakers pushing air. It probably won't happen anytime soon, but at least theoretically, that should produce an audio experience unlike live performance or traditional recordings.

[Theo Verelst]

The sources of imperfection are often misidentified. A good microphone and decent cable, reasonably accurate pre amp and good spec ADC are available, for moderate prices, but they have their own properties which much be understood. A measurement mic with all striaght and wide range and large dynamic range is possible electronically, but the DAC and the way the wave field is captured by the microphone aren't neutral by a long shot: the DAC has major signal reconstruction errors, and the microphone will not automatically incorporate the way humans listen at all, and have harmonic distortions and frequency unevenness for a lot of directions and types of sound.

 

Speakers are difficult, and headphones have their own perks when sensitive people listen to them, speakers which are full range, low distortion and not very directive are not around much, but improvements are certainly possible. My self made 5 way, 3 way multi amped high quality amplification is neutral enough with any source I use all day. Once I (with a very difficult method) use my AKG headphones with more properly digital to analog converted samples, and compensate for binoral problems, they're similar to the speakers, minus the local acoustics (big factor).

 

So, it is possible bu people misunderstand what a studio should be able to do, how DAC distort the sound (always), and what sound waves behave like.

 

Theo V

[Anderton]

The problem with speakers is that they are point sources of sound. The entire body of an acoustic instrument radiates sound. I don't think there's any simple way a loudspeaker can imitate the multiplicity of sound patterns that emanate from an acoustic instrument like a piano or nylon string guitar. (Maybe you could come closer with drums, by having a separate speaker for each drum.)

 

Surround has more potential for not being a point source, but it's still just more point sources. A piano has a near-infinite number of point sources, all spraying sound into the air. 

[KuruPrionz]

Just now, Anderton said:

The problem with speakers is that they are point sources of sound. The entire body of an acoustic instrument radiates sound. I don't think there's any simple way a loudspeaker can imitate the multiplicity of sound patterns that emanate from an acoustic instrument like a piano or nylon string guitar. (Maybe you could come closer with drums, by having a separate speaker for each drum.)

 

Surround has more potential for not being a point source, but it's still just more point sources. A piano has a near-infinite number of point sources, all spraying sound into the air. 

And the same piano will sound different in different rooms or at different distances in the same room or with a different musician playing it. 

 

Microphones are also point source devices. A completely flat response omni-directional microphone can only reproduce sound from the exact location it has been placed. 

Essentially, so are ears (which are in fact transducers). 

 

Choices are made and we try to move forward. Or, we allow the lack of perfection to halt progress. Are paint brushes perfect? No and it never stopped Rembrandt, did it?

 

Theo is correct that there is compromise in ADC and DAC, there was also compromise in tape itself and every tape recording device ever made and obviously in all vinyl formats. Should we all stop recording the music we create and wait for the perfect system?

 

At this point, we can quibble about the defects in the devices that we use to record, mix and produce music or we can just make music. 

At some point in the future perhaps these problems will be solved, I have no answers that are feasible and I do tire of even pondering the current challenges since they are truly endless and focusing on one or the other supposed "primary defect" is like pushing jelly up a hill with a rope.

[Notes_Norton]

I think perfect sound is a judgement call.

 

Perfect reproduction of sound is impossible with today's tech and probably in our lifetimes.

 

I took electronics (communication option) decades ago in college. One thing I remember my teachers telling me is that whenever you do something to the signal, it introduces some distortion. We can minimize it, but we can't eliminate it.

 

Starting with the mic, I've never seen one with a perfectly flat frequency response for the entire audible spectrum. Same for speakers, which are a bit like mics in reverse. And for both, the volume of the signal also affects the bandwidth curve.

 

Preamps and amplifiers do not have a perfectly flat frequency response either. Add FX and they add distortion.

 

When you digitize something, you introduce errors. Cutting into vinyl or transferring that to electrical impulses via a cartridge introduces distortion too.

 

So we all have to pick the distortion that appeals to us the most.

 

The pickups on my guitar don't do a perfect job of converting the vibrations of the strings into electrical impulses. The vol/tone circuit introduces more distortions. And if I put that guitar through a compressor and distortion pedal to get more sustain and put more edge on the tone, I would be thinking the FX box was improving the tone for the particular song I'm playing by distorting it.

 

If I take my voice, put it through a band-pass filter, so it sounds more like a thin, current Nashville style sounding voice, is the distortion of the filter improving my voice for that genre of music?

 

If so, is it possible that the right kind of distortion is making a perfect sound?

 

Just thinking out loud here.

 

Notes ♫

[KuruPrionz]

4 minutes ago, Notes_Norton said:

I think perfect sound is a judgement call.

 

Perfect reproduction of sound is impossible with today's tech and probably in our lifetimes.

 

I took electronics (communication option) decades ago in college. One thing I remember my teachers telling me is that whenever you do something to the signal, it introduces some distortion. We can minimize it, but we can't eliminate it.

 

Starting with the mic, I've never seen one with a perfectly flat frequency response for the entire audible spectrum. Same for speakers, which are a bit like mics in reverse. And for both, the volume of the signal also affects the bandwidth curve.

 

Preamps and amplifiers do not have a perfectly flat frequency response either. Add FX and they add distortion.

 

When you digitize something, you introduce errors. Cutting into vinyl or transferring that to electrical impulses via a cartridge introduces distortion too.

 

So we all have to pick the distortion that appeals to us the most.

 

The pickups on my guitar don't do a perfect job of converting the vibrations of the strings into electrical impulses. The vol/tone circuit introduces more distortions. And if I put that guitar through a compressor and distortion pedal to get more sustain and put more edge on the tone, I would be thinking the FX box was improving the tone for the particular song I'm playing by distorting it.

 

If I take my voice, put it through a band-pass filter, so it sounds more like a thin, current Nashville style sounding voice, is the distortion of the filter improving my voice for that genre of music?

 

If so, is it possible that the right kind of distortion is making a perfect sound?

 

Just thinking out loud here.

 

Notes ♫

I often add distortion to recorded vocals, bass and snare drum. It brings those essential elements out in the mix. 

I am a big fan of distorted electric guitar, it allows us to do things that would not otherwise be possible. 

 

It's "ally ally outs in free yo" as far as I'm concerned. 

[KuruPrionz]

2 hours ago, The Real MC said:

I don't strive for perfection.  I strive for excellence.

Perfection is very subjective.  There's no one single sound or mix that pleases everyone.

A perfect sound to musicians may not have mass appeal.  Getting both is a tall order.  Two examples who succeeded were Steely Dan and Boston.

I HUGELY agree with this. I will add 10cc to your list and my recollection of James Brown's Gravity album is that it was pretty much flawless. 

At a certain point, the Eagles achieved "perfection" but I don't like Hotel California at all. The Moody Blues stand tall as a band who made "perfect" records that I just can't bear to listen to, the lyrics are simply not excellent. 

 

On the other hand, Exile on Main Street by the Rolling Stones is a perfect album because nothing on it is perfect. Great songs, played from the gut. 

Boogie Chillun by John Lee Hooker is a perfect recording as well but it's because the song/story cannot be topped. Emmylou Harris has quite a few songs that achieve both perfection and excellence - Easy From Now On from the Quarter Moon In a Ten Cent Town comes to mind. 

[David Emm]

The brown note you aim at your next door neighbor. He's working on a car at 2 a.m. and playing Top 40 radio at 130 dB. Suddenly, BWOMP! I sold the video to A&E. It'll be on the next season of "Neighborhood Wars."

 

I'm not holding out for "perfect sound" or super-surround when my current system reproduces bells so well, I'm amazed at how my frequency range is holding up. I had expected the world to sound like I was wearing ear muffs by now. Every album by "new age" harpist Andreas Vollenweider is utterly immersive. I want THAT engineer to mix my stuff! Its stereo, but it feels like a mature form of quad that's well past its novelty phase.  

 

What's perfect for Pink Floyd is overkill for Tom Waits. Which kind of perfect are we discussing?

[Theo Verelst]

I you need surround for a decent "stereo" sound, it's not much of a pro mix, probably.

 

For people just playing around with a sequencer and some chorus and eq, t's not logical in any way to create a sound in a space, that doesn't mean there are no (partial) solutions to the problem, maybe there's just no manual.

 

T

[RABid]

There is no perfect, especially in sound. It is too subjective with the goal being to satisfy the listener. The most sought after microphones, compressors, preamps, etc... are wanted for the little peaks they add here and there that is pleasing to the listener or make the voice sound better. To even start on the perfect sound you would first have to make everyone hear the same and desire the same things.

[Anderton]

5 hours ago, KuruPrionz said:

Should we all stop recording the music we create and wait for the perfect system?

 

Order of priority for the average listener:

 

1. Vocals

2. Beat

3. Rest of the song

.

.

.

46. Sound quality

 

This is NOT a diss on the average listener. It simply means that they listen to the music, not the recording. Which I think is the correct priority, anyway.

[Anderton]

52 minutes ago, Theo Verelst said:

I you need surround for a decent "stereo" sound, it's not much of a pro mix, probably.

 

Unless the music was designed from the ground up to be a surround experience, not a surround version of physical experience.

[KuruPrionz]

5 hours ago, Anderton said:

 

Order of priority for the average listener:

 

1. Vocals

2. Beat

3. Rest of the song

.

.

.

46. Sound quality

 

This is NOT a diss on the average listener. It simply means that they listen to the music, not the recording. Which I think is the correct priority, anyway.

I know this. A great song sung loud is the best thing ever but without a beat it just shrivels down to nothing. 

I've played a TON of gigs in "cover" bands (meaning we are playing our own version entirely). If you can sing the chorus, play the beat and act cool, the dance floor will fill up. 

It doesn't hurt to sound pretty good (if not WAY TOO FRIGGIN LOUD) but I think as long as you aren't unbearably shrill people will hang out. 

 

Somebody listen to music at home, in the car, at the supermarket, etc. t really doesn't take much and Craig's list above is spot on. 

[KuruPrionz]

21 hours ago, Theo Verelst said:

A measurement mic with all striaght and wide range and large dynamic range is possible electronically, but the DAC and the way the wave field is captured by the microphone aren't neutral by a long shot: the DAC has major signal reconstruction errors, and the microphone will not automatically incorporate the way humans listen at all, and have harmonic distortions and frequency unevenness for a lot of directions and types of sound.

 

So, it is possible bu people misunderstand what a studio should be able to do, how DAC distort the sound (always), and what sound waves behave like.

 

Theo V

As with all things in digital electronics, DAC has improved over time. My "newest" interface is an SSL 2+,  somewhere on the upper end of the inexpensive interfaces. 

The output from the unit is far superior in sound to my ears. My Presonus Quantum is only a smaller increment behind but all the MOTU and Mackie interfaces I owned before those 2 cannot compete for sound quality. 

 

I'm not saying they are perfect, the term I use is "Less Worse".  The newer gear seems to have better circuits and that is historically true with electronics - things get better and more affordable. There is fierce competition in audio interfaces, it's a good market to be the "best" in and things change constantly. 

 

Theo, I don't know what interface you are using? It might be worth checking out some of the newer units, maybe friends have them. 

FWIW, I started recording when I was 16 (I'm 67) using a stereo cassette deck and an Akai reel to reel. Eventually I modified a 4 track TEAC reel to reel (it played in stereo in both directions so you didn't have to flip the reels over) so that I could record 4 tracks in one direction. It had it's problems, certainly did not sound as good as my current gear. Then the 4 track cassette revolution came and I jumped on board. Not great but much better than nothing. I never had the good mics, they were very expensive back then so I used a pair of SM58s, which are certainly better than nothing. Hi-Fi JBL speakers were better than most, we have an abundance of good studio monitors available now for reasonable prices. 

 

I did get to hang out at Maximus Studios in Fresno and hear 2 synced-up Studer 2" 24 tracks playing back at 30 ips but the speakers were 2 way 15" with horn tweeters - total garbage by todays standards in terms of fidelity - essentially PA speakers from the 70's. Yuck. Still, it sounded pretty nice but tape is expensive and so were Studers. 

 

I've learned not to go too deep in terms of interface purchases, it's better to be current with a couple of tracks than to get something more versatile that does not offer the best sound currently available at prices I can afford. This trend will continue, DAC will improve. It has gotten notably better since I bought my first MOTU 824 back in 2004. I would never use that unit now, it didn't really sound that good and the software was not intuitive. 

 

NOTHING will ever be perfect, so it goes. That's been my entire point all along, quibbling about the details of current or recently obsolete gear will change as gear changes - constantly. 

 

A great song has always been a great song and will always be a great song. I'm here for the music and use what I can to create my own. 

[David Emm]

Sorry, I missed part of that. I was busy feeling rosy because I no longer have to clean and very, very carefully demag the heads of an Otari 8-track R-to-R. The frequency response was gratifying, but I still have nightmares about its massive, glowing presence and the ominous, omnipresent cloud of ozone it emitted. I also developed big biceps from hefting the reels of tape it was fed. Hooray for flash drives.

[KuruPrionz]

2 hours ago, David Emm said:

Sorry, I missed part of that. I was busy feeling rosy because I no longer have to clean and very, very carefully demag the heads of an Otari 8-track R-to-R. The frequency response was gratifying, but I still have nightmares about its massive, glowing presence and the ominous, omnipresent cloud of ozone it emitted. I also developed big biceps from hefting the reels of tape it was fed. Hooray for flash drives.

At least those are big and you can get at the head stacks without too much torment. I was the 4 track cassette kid for a good while, the rack mounted one no less. Fancy, sort of like a Yugo with Cadillac fins on it but no hubcaps. 

LOVE my pretty little interfaces that just work, don't need tape and all you have to do is not put them in the dishwasher. Those are better...

[Notes_Norton]

On 9/24/2022 at 7:03 PM, Anderton said:

It simply means that they listen to the music, not the recording. Which I think is the correct priority, anyway.

 

I can't agree more. This should be written on the wall of every home and commercial studio.

 

Notes ♫

[dmitch57]

On 9/24/2022 at 10:12 AM, Anderton said:

To me, it's a sound quality that is identical to live music.

The problem with this approach is that since "Revolver" (or thereabouts), there is so much recorded music that can not be played live, so there is nothing for a recording to be considered "identical" to. The best they can hoped for, I think, is to match the sound that came out of the monitors when the music was originally mixed. Or mastered. (Take your pick.) In so much modern music, that is the "original" sound.

 

I've seen several performances of "Dark Side of the Moon" live. None of them came close to sounding like the record. People playing instruments on stage just can't make those same sounds. 
 

Or in a more contemporary vein...most EDM music, a lot of Hip-Hop, the latest Billie Eilish...a lot of the sounds on those records were not performed or recorded with a mic, so the reference for "perfect sound" has to be something different from "live music".

[Old No7]

On 9/24/2022 at 1:29 PM, Anderton said:

The problem with speakers is that they are point sources of sound. The entire body of an acoustic instrument radiates sound.

 

Run it through a Leslie speaker!!! 😃

 

Old No7

[Old No7]

On 9/24/2022 at 12:12 PM, Anderton said:

This is a spinoff from the thread on native vs. hardware processing. KuruPrionz said in it "Until we can eliminate or profoundly improve speakers, microphones, cables and outboard processing - 'perfect sound' (whatever that is) will never exist in this world." 

 

So...let's take a stab at defining it. To me, it's a sound quality that is identical to live music. In other words, if there's a solo classical guitar player sitting in my room, and a recording of that guitarist playing in my room, they should sound indistinguishable.

 

As he points out, the problem is tranducers. Converting one form of energy to another will never be easy, and that happens twice in the process - capturing the sound, and reproducing the sound. Yes, sound is air waves, and yes, speakers can create air waves, but they do so with physical limitations.

 

OK, time for a more serious reply to this posting...  Hope you liked the Leslie joke though!

 

Interesting discussion... I agree with the premise...

 

Yet in today's world -- with top quality gear, and a gifted sound technician -- I think we're closer than we've ever been.  Maybe the limitation isn't just the speaker -- or the pair of stereo speakers -- as ALL the sounds you hear in that room are being channeled through only 2 speakers.  While we have only L & R ears, I think we all "feel" sound too (deaf people, of which I have little experience except for what I see on TV, may agree with me). The sound from a musician or band comes at us from multiple sources (instrument, voice, reflections, etc.).  Sure, we can use tools like L-R panning, reverb and delay into tricking our ears to "hear" via a stereo mix that the guitarist is front left, the bass player is back left, the singer is front middle, drummer is in the back and the keys are on the right -- but then we play that mix back through only 2 speakers -- L & R.  Pretty close, but not perfect.

 

I envision your room with the guitar player sitting in the center on one side (with you in the middle).  Then you turn your seat around 180 degrees, and instead of TWO speakers 3 or 4 feet apart -- there are literally 2 dozen speakers of all sizes and frequency response rates, each playing back its own channel, and playing at its own relative volume to what it "heard" from the artist & instrument -- as it plays it back for you to hear and "feel".  All the speakers are positioned in a cluster, from top-to-bottom & left-to-right, to better align with what it heard on the actual recording, so the sounds being played back from that speaker -- and all the others -- more closely resembles what you heard for real.

 

That would be something like a "human & instrument shape of sound"; not mono, not just stereo -- but the "total sound", maybe similar in some ways to "surround sound", although most sound would be coming at you, not from behind you (depending on the acoustics of the room, of course). To play it back on 24 speakers, then I think it'd be best to record it with 24 individual mikes, each with their own T-B/L-R channel.

 

Would it sound any better?  Do we have the technology to do this?

 

I don't know, I'd have to give it some more thought; but this is what I thought about to overcome the "point sources" comment about the speakers.

 

Old No7

[analogika]

20 hours ago, dmitch57 said:

The problem with this approach is that since "Revolver" (or thereabouts), there is so much recorded music that can not be played live, so there is nothing for a recording to be considered "identical" to. The best they can hoped for, I think, is to match the sound that came out of the monitors when the music was originally mixed. Or mastered. (Take your pick.) In so much modern music, that is the "original" sound.

 

There is another problem, which is that anything that isn't classical music has virtually never really been about sounding like the live performance, but about allowing you to focus on what's happening as if it were a live performance, in my estimation. 

 

The marketing was "sounds as if it were live", but the engineering reality was that Peggy Lee's voice, close-miced into a U47, compressed and limited and slammed into a two-track recorder alongside a band playing behind her recorded onto the other track, sounded nothing like what you would have actually heard sitting in front of the music. But that was what engineers figured out to best capture the sex that appealed to audiences about her performance. (reference: "Black Coffee", 1953)

 

The feeling you get from Sonny Rollins' "Way Out West" of his horn being completely in your face is not because that's what the performance would sound like if you were in the room, but because it was placed there to draw your attention as it would if you were in the room. 

Even in the days of a single horn carving into wax, musicians would move around closer or farther away from the recorder as they soloed or receded into the backing band, when possible. 

 

TL;DR: "Perfect sound", IMO, is not about perfectly reproducing the sound of a live performance, but about what works best for a production.
It's the clarity of focus on what the artist/producer/engineer deems most important about a particular performance. 

 

This could be the raw energy of a band, best captured by mashing all the track together and pushing the upper mids to thrash at your ears, or the pure clarity of space on a production like "Eye in the Sky" by Alan Parsons, or the live room feel, with lots of audience response, of Donny Hathaway's "Live" album, or the super-compressed close-mic'ed danceable style of some of the late-60s Blue Note albums. 

[Notes_Norton]

1 hour ago, analogika said:

Even in the days of a single horn carving into wax, musicians would move around closer or farther away from the recorder as they soloed or receded into the backing band, when possible. 

I even do that live, as the tone of the mic changes with distance from it. Also, when singing out something much louder than the rest of the song, backing off saves the ears of the audience. (same mic control we used to use in the studio)

 

Playing with the mic as a guitarist picks from different distances from the bridge is part of an expressive performance.

 

And yes, there are plenty of songs that can't be performed live and sound like the record. So what do I do? Either don't learn them, and if I have to, reinterpret them. There are a couple that the elderly ladies want to do their line dances to, and I'll DJ them and noodle along on the guitar or wind synth.

 

And now that we have remixes of so many classic pieces of music, which one is perfect? The original? The remix?

 

What is perfect anyway?

 

Before DJs cover bands had to play the song as close to the popular record as possible. Now we have tribute bands, covering one artist. In my younger days we had to be an Elvis Tribute Band, Beatles Tribute Band, Stones Tribute Band, Aretha Franklin Tribute Band, Temptations Tribute Band, Michael Jackson Tribute Band, and so on. These new folks have it easy in the fact that they only have to cover one artist, but difficult in that before the Beatles, you could reproduce just about any recording live.

 

Notes ♫

[analogika]

Ah, but if you're covering dozens of artists, you can get away with having your own band sound and just covering the songs recognisably. 

 

I've MD'ed a Santana tribute show for several years and currently work a Rammstein tribute show (talk about job variation!). Those have been some of the toughest gigs I've worked so far, because most people who come to the show are intimately familiar with the material, and you can get away with much less than a run-of-the-mill cover band. 

And the Rammstein thing involves meticulous programming, backing tracks with pyro and flame thrower controls, theatrical stage performance, and live keyboards. It's easily the most challenging job I've ever had, and that includes several years in a Top-40 cover gig with northwards of 400 songs repertoire. (FWIW, the band actually got name-checked by the Rammstein keyboardist in his autobiography, as he mentions watching some of our stuff on Youtube when prepping for his own shows…)

[Anderton]

On 9/26/2022 at 11:00 AM, dmitch57 said:

The problem with this approach is that since "Revolver" (or thereabouts), there is so much recorded music that can not be played live, so there is nothing for a recording to be considered "identical" to. The best they can hoped for, I think, is to match the sound that came out of the monitors when the music was originally mixed. Or mastered. (Take your pick.) In so much modern music, that is the "original" sound.

 

But that's kind of my point. If a system of some kind can reproduce acoustic sounds with perfection, then it can produce anything the way it's supposed to be heard. 

[dmitch57]

On 9/27/2022 at 2:24 PM, Anderton said:

If a system of some kind can reproduce acoustic sounds with perfection, then it can produce anything the way it's supposed to be heard.

 

I see; that's a good point. But I'm not sure: can't a synth create sounds, harmonics, and so forth that no known acoustic instrument can make? If so, how do we know that a system than can reproduce any acoustic instrument perfectly can also accurately reproduce all synth sounds?

 

I'm thinking about four different scenarios with different levels of "perfection" we might hope to achieve when recording and reproducing them.

1. A single stationary acoustic instrument recorded in a reasonably non-reverberant space. I think we're pretty much able to perfectly record and reproduce this right now.
2. A large ensemble (big band, orchestra) in a large reverberant hall. I don't think we can accurately reproduce this yet, though we'll probably be able to in the future. (It'll take more than two speakers.)
3. Thunderstorms, or waves crashing on an ocean beach, or an FA-18 flying over your head. I doubt we'll ever be able to reproduce these. (No evidence here of course, just my gut reaction whenever I hear any of those things.)
4. Modern electric music, mostly made without mics. This, we can already perfectly reproduce: Just play it back through the same monitors it was mixed or mastered with.

A playback system that can reproduce #1 perfectly can't necessarily reproduce #4 perfectly, since #4 can have sounds/harmonics/etc. that no acoustic instrument can make.

 

A system that can create and reproduce #4 perfectly can't necessarily record and reproduce any of the others accurately.

 

So as usual, I think there's no one-size-fits all answer, and this stuff is not black and white.

[Notes_Norton]

We can't perfectly reproduce anything. The act of recording creates distortions.

 

On the other hand, when the distortions are so slight that the ear can't tell the difference, for all practical purposes, it's perfect.

 

Notes ♫

[dmitch57]

1 hour ago, Notes_Norton said:

We can't perfectly reproduce anything. The act of recording creates distortions.

What about sounds that were not recorded at all, but were synthesized in the box? What comes out of the speakers at mix time is the original sound. We can reproduce that perfectly. I do it every day.

[KuruPrionz]

15 hours ago, dmitch57 said:

 

I see; that's a good point. But I'm not sure: can't a synth create sounds, harmonics, and so forth that no known acoustic instrument can make? If so, how do we know that a system than can reproduce any acoustic instrument perfectly can also accurately reproduce all synth sounds?

 

And if no system can accurately reproduce all synth sounds, how would you know?

It is what it is.  🤔

[Theo Verelst]

It's easier: *if* your amplification/speakers/acoustics are "neutral", the input signal is simply transformed to audio waves with no change. It's unlikely you're in a dead room. and there are individual head transfer functions for headphones, so there's variation. Then, there's the stereo problem, which only the A grades do right at times by creating very complicated to make wave patterns (difficult subject). which unfortunately translate very lousy to digital processing steps.

 

So in short, digital is a problem because it doesn't reconstruct well, electronics can be made fairly neutral, which shouldn't be extremely expensive but people search more for an  effect box than a straight signal path.

 

Digital sound similar in synthesis, too, because the way things are computed isn't all the way the same as electrical signals (it's really not, but many dsp programmers don't know this) and *IF* you want to sound alive, low-mid mid an high distortion poor and even use digital to sound like a neutral recording, you're going to need signal preparation methods for standard DACs (any, also the latest, thus far) which aren't available in any program or VST, which are a lot of work (serious processor load) and require all kinds of givens about the signals you want to prepare for the studio or consumer DAC.

 

T

[Anderton]

17 hours ago, dmitch57 said:

A single stationary acoustic instrument recorded in a reasonably non-reverberant space. I think we're pretty much able to perfectly record and reproduce this right now.

I'm not sure I agree. I've done a lot of classical sessions for nylon-string guitar, keyboard, and voice. What I hear in the room is not what I hear coming out the speakers. It's close, but the speakers don't reproduce the soundfield that the instrument creates. I think a piano will always resonate differently in a room compared to speakers playing back that piano in a room. I guess the only way to know for sure is to have a blind test where speakers are placed as close as possible to a piano, and see if people can tell the difference between someone playing the piano vs. playing back the same performance over speakers.

 

As to synthesized tones that always existed in a purely electronic form, I would venture to say they aren't an instrument until they go through speakers. At that point, the combination of synthesized tones going through transducers becomes the instrument, because that's what you hear. Those speakers serve the same function as the piano's soundboard. At that point, you have the same issues capturing that instrument's sound as you would any other instrument.