Audio Recording - is it ever realistic?

[Timcognito]

Of course there are realistic recordings and recording is where 95% of the accuracy comes from. Inaccurate and poorly placed speakers, and the room can rob realistic sound but garbage in is garbage out. It is ironic that some posters here will argue that SINAD of 90 is distinguishable over 80 and over 100 represents an improvement and yet say that there are little realism in the source. Do some searching on binaural recordings, those guys are all about real sounding playback which must start at the microphone(s) and and the venue. Also there are recording engineers and labels known for eerily real recordings especially small acoustic settings.

The Search for Roy DuNann

Roy DuNann

jazzprofiles.blogspot.com

 

[Robin L]

No matter what, a recording of music first requires microphones. Microphones are transducers, they are subject to the distortions/nonlinearities that are inevitable with transducers. Anyone who has made recordings has heard how profoundly the sound is altered by microphones compared to the sound of the live performance without. I'm speaking of recordings of acoustic instruments and/or voices. With amplified instruments there's no "there" there, there's no correlation to reality. And most recordings are studio artifacts with even less correlation to reality. Is a recording "real"? Yes, it's a real recording, but it cannot be the same as "live" for the reasons just cited.

 

[Timcognito]

No matter what, a recording of music first requires microphones. Microphones are transducers, they are subject to the distortions/nonlinearities that are inevitable with transducers. Anyone who has made recordings has heard how profoundly the sound is altered by microphones compared to the sound of the live performance without. I'm speaking of recordings of acoustic instruments and/or voices. With amplified instruments there's no "there" there, there's no correlation to reality. And most recordings are studio artifacts with even less correlation to reality. Is a recording "real"? Yes, it's a real recording, but it cannot be the same as "live" for the reasons just cited.

Generally agree but there are exceptions that there are recordings so close to real as to be indispensable unless you heard them played back in the same room within minutes of each other. Do I think its common, no, but best is very close so we can go back to arguing about SINAD 80 versus 90.

 

[Barrelhouse Solly]

I've been playing acoustic guitar for over 60 years. I've heard myself from behind the guitar a few times and heard many others from in front of the guitar. I've never heard a recording that sounded like an acoustic guitar sounds to me. Some are closer than others but I'm still waiting for one to one correspondence. I agree with other posters that recordings are generally made to suit preferences of the producer and musicians which are not necessarily biased toward "photorealism." I've fooled around with home recording and mastering and have an idea of what is going on on commercial recordings.

As to why and how, I have a lot of conjectures and no proof as to why we don't hear the "real" sound. I suspect it's a combination of production and equipment. Microphones don't hear the way that human ears do. Speakers are limited. I'm hopeful that the grail is out there to be found.

 

[watchnerd]

I just read, in Stereophile, a history of the first DDD CD…”Brothers in Arms”. It is worth reading for both the historic content and the recording techniques utilized. The piece is detailed, including microphone types used, positions and engineering. I consider that recording as a tour de force in every way - except what we hear is not “real”, but a creation of its own.

In my brief alterego as a self styled recordist, I used an excellent R2R Technics with RCA Ribbon microphones in the Blumlein configuration. In the 1970s I recorded the Oberlin College Choir in an Episcopal Church in Palm Beach, Florida. By happenstance, many of the choir and its Director were at my home, listening to the recoding on my Fulton Premier Speakers powered by Audio Research amps.

The Director was so impressed by the recording, he borrowed my tapes to be used on their vinyl recording. The vinyl was produced, crediting me as a recoding engineer. The album sound came nowhere near the tape, and sadly, the tapes were destroyed in return shipment!

My comment now on real… the tape, playing back on my system, came close, at least to my and the conductor’s ears, to approximating the performance in the church. Even so, having been in the venue for the concert, it was still an approximation, therefore - not real.

That said, having thousands of vinyl and CD/SACD recordings and now, access to high quality streaming music, approximation is pretty damn good!! Being 84 years old and somewhat disabled, live performances are no longer within my reach and, without my likely ultimate system, I would not be able to enjoy the experience!

We stand on the shoulders of giants to be able to enjoy the vast catalog of performances of all genre of music - kudos to all of them… and the approximation of reality!

Yes, it's an approximation.

Microphones alone have their own voicing, patterns, independent of choices made about placement, let alone what happens in mixing and mastering thereafter.

 

[Timcognito]

I've been playing acoustic guitar for over 60 years. I've heard myself from behind the guitar a few times and heard many others from in front of the guitar. I've never heard a recording that sounded like an acoustic guitar sounds to me. Some are closer than others but I'm still waiting for one to one correspondence. I agree with other posters that recordings are generally made to suit preferences of the producer and musicians which are not necessarily biased toward "photorealism." I've fooled around with home recording and mastering and have an idea of what is going on on commercial recordings.

As to why and how, I have a lot of conjectures and no proof as to why we don't hear the "real" sound. I suspect it's a combination of production and equipment. Microphones don't hear the way that human ears do. Speakers are limited. I'm hopeful that the grail is out there to be found.

Again I agree that it takes a trained ear to distinguish but it all starts at the recording end and there are many places in the chain that can take it away, the transducers and room at the other end are the main issue. The fact so few threads at ASR are focused on recording authenticity and instead on tiniest nuances on playback authenticity has always puzzled me, when it is the former that has by far the biggest impact.

 

[Robin L]

Generally agree but there are exceptions that there are recordings so close to real as to be indispensable unless you heard them played back in the same room within minutes of each other. Do I think its common, no, but best is very close so we can go back to arguing about SINAD 80 versus 90.

I've made lots of recordings that would be more likely to be "real" than the vast majority of recordings; small ensembles or solo instruments in friendly acoustic settings using first-class microphones. Because of that, I've heard lots of recordings where I'd hear the recording and "the real thing" within moments of each other. Of course, the recording sounded the same as what was coming off the mixer (most of my recordings were made on DAT recorders), so I could easily judge any adjustments I'd need to make without having to rewind the tape. I've heard plenty of recordings where I'm happy with the sound, but recordings where I've been involved never sound exactly the same as what I would hear standing in front of the musicians. The bigger the ensemble (orchestral vs. chamber), the less likely the two would sound the same or close to the same, including sessions where I was only an assistant and the engineer had a wide selection of microphones. Of course, recordings involving a lot of microphones introduce a number of problems one would not encounter with a simple X/Y or ORTF array.

My "Holy Grail" of microphones was the Neumann 84. I foolishly passed up the chance to get a pair for a reasonable price when they were being phased out, being replaced by the KM 140. I got the 140s (and a couple of 130s). Very similar to the KM 84 but not quite as transparent. However, it was more practical for a variety of reasons. I managed to rent a pair of Klaus Heyne modified KM 84 microphones, they had a lower noise floor than the standard KM 84 microphones, probably the best microphones I've used. Schoeps Collettes are excellent in a similar way, a first choice for recording harpsichords.

 

[MattHooper]

It should be noted that a close-miked multi track recording can never sound like the real live performance despite how well it's mixed. It typically turns out sounding hyper realistic, much like an HDR photo plus tons of Photoshop treatment looks hyper realistic. Having isolated instruments and voices allows you to bring out the details of each track and combine them in such way that they don't step on each other, which would not be possible with just a neutral capture. Strings might end up sounding thin or otherwise strange but that's often because they would "cloud" or "step on" other more important tracks if they were not thinned out. This is where the most important part of making a record comes in, namely arranging and getting "good" sounds as the source. Some producers are working with a solid vision of the final product and are making crucial choices ahead of time. That makes their final mixes stand out. Others will capture a ton of material and attempt to "fix it in the mix", which sometimes works out and sometimes does not. It's a big field and making generalizations is typically a mistake.

I play a fair amount of orchestral music on my systems. I understand lots of audio files have traditionally liked “ naturally mic’d” orchestral recordings, the capture with a feel to be a natural balance of the hall sound to the orchestra, rather than a spotlight sound on the sections. I enjoy those recordings to.

But I absolutely love orchestra based soundtrack recordings that are more dramatic and creative. Plenty of Bernard Herrmann’s soundtracks feature close micing of the instruments, especially strings, which achieve a real textual Technicolour presence. I absolutely love it. And it can sound very “ real” in its own way because of that vivid presence.

 

[Robin L]

I play a fair amount of orchestral music on my systems. I understand lots of audio files have traditionally liked “ naturally mic’d” orchestral recordings, the capture with a feel to be a natural balance of the hall sound to the orchestra, rather than a spotlight sound on the sections. I enjoy those recordings to.

But I absolutely love orchestra based soundtrack recordings that are more dramatic and creative. Plenty of Bernard Herrmann’s soundtracks feature close micing of the instruments, especially strings, which achieve a real textual Technicolour presence. I absolutely love it. And it can sound very “ real” in its own way because of that vivid presence.

There's a number of early Decca orchestral recordings that have more presence than one would encounter in any seat in a concert hall. The perspective allotted would be slightly above and slightly behind the conductor. Great for hearing into the orchestra, but not actually possible in the real world.

 

[MattHooper]

I've been playing acoustic guitar for over 60 years. I've heard myself from behind the guitar a few times and heard many others from in front of the guitar. I've never heard a recording that sounded like an acoustic guitar sounds to me. Some are closer than others but I'm still waiting for one to one correspondence. I agree with other posters that recordings are generally made to suit preferences of the producer and musicians which are not necessarily biased toward "photorealism." I've fooled around with home recording and mastering and have an idea of what is going on on commercial recordings.

As to why and how, I have a lot of conjectures and no proof as to why we don't hear the "real" sound. I suspect it's a combination of production and equipment. Microphones don't hear the way that human ears do. Speakers are limited. I'm hopeful that the grail is out there to be found.

I grew up with both my brother and I playing acoustic guitar and I continue to play (very modestly), and the majority of my friends were musicians who also played acoustic guitar as well as electric.

So acoustic guitar for me has always been one of my first go-to sounds to test a system.
If I don’t get the impression of that woody timbre and the (steel) strings with rich sparkling harmonics, then I will never fully enjoy the system. Some systems can make an acoustic guitar sound like it’s made of plastic or carbon fiber.

That’s why I used to do some live versus reproduced tests of recordings I made of my own acoustic guitar versus the live guitar being played through various speakers.

Some speakers sounded more like the real thing than others, especially timbre.

The speakers I had that sounded the closest were Hales Transcendence 5s, Thiel 3.7s and MBL 121 omnis. All of which had sounded “ right” to my ears on their own with instruments like acoustic guitar which is why I bought them.

The Thiel and MBL could both do fairly startling recreations of my acoustic guitar in my living room.

 

[bluefuzz]

Oddly it’s the studio manufactured productions that may be the most “real” type of recordings.

Exactly. The recording is the performance. Not a second class imitation of a separate event.

when artists do try to perform those pieces live they are only approximations of the recordings, live imitations of the actual artwork.

I'm not sure I would say that. Typically, in a rock or pop setting, songs will be composed to be performed live and then 'produced' for an album release. Alternatively they may be composed in a studio setting and then arranged to be performed live. I don't regard either as better or worse than the other,. Simply different. What puzzles me is when 'audiophiles' expect their rig to both transport them to the live event and put them in the producer's chair in the recording studio while at the same time adding some kind of euphonious fairy dust.

If I want to experience reality I can go for a walk in the country. But that doesn't stop me enjoying the Turner or Constable hanging on my wall (I wish) ...

 

[MattHooper]

What puzzles me is when 'audiophiles' expect their rig to both transport them to the live event and put them in the producer's chair in the recording studio while at the same time adding some kind of euphonious fairy dust.

I think that’s me you might’ve just described

My ideal for a system is that it can sound lifelike, but also very revealing about the nature of any recording because I really love the differences in recording and production techniques, and I’m happy to sprinkle a little of my own flavouring if it helps me get there.
So sort of ideal would be “ transported into the recording where it feels like I’m hearing live musicians, but in the acoustic context in which they’ve been placed whether it’s real or artificial.”

So I sort of view my system, and recordings themselves as a sort of “ Star Trek holo-deck” experience.

 

[Timcognito]

I think most would agree that key is that what and how much is missing is real issue or is it significantly changed or augmented in any way. If a soprano were to sing her solos with same accompaniment in three different rooms mic'ed the the same we would want to hear subtle differences between each brought about by the venue. Just by sitting in different places in those rooms would naturally change the sound. As Matt just said "sound lifelike, but also very revealing"

 

[Robin L]

I think most would agree that key is that what and how much is missing is real issue or is it significantly changed or augmented in any way. If a soprano were to sing her solos with same accompaniment in three different rooms mic'ed the the same we would want to hear subtle differences between each brought about by the venue. Just by sitting in different places in those rooms would naturally change the sound. As Matt just said "sound lifelike, but also very revealing"

I've had lots of experiences like that. I recorded a string quartet + guitar in two venues on the same day. One was a big church with high ceilings and a long reverb tail. Made for a lovely recording - one could easily balance the direct and hall sound. The other recording, made later that day, was in a small room with 10' ceilings. A room small enough that temperature would rise during the recordings, requiring that the performers had to tune between movements. Sound was predictably nasty. These recordings were made for the group in advance of making commercial recordings for Harmonia Mundi. Those turned out just lovely:

 

[Kal Rubinson]

The other recording, made later that day, was in a small room with 10' ceilings. A room small enough that temperature would rise during the recordings, requiring that the performers had to tune between movements. Sound was predictably nasty.

Ah. Something like a typical listening room.

 

[Multicore]

I've been playing acoustic guitar for over 60 years. I've heard myself from behind the guitar a few times and heard many others from in front of the guitar. I've never heard a recording that sounded like an acoustic guitar sounds to me. Some are closer than others but I'm still waiting for one to one correspondence.

Mics are one dimensional - what comes out is a scalar function of time. Acoustic guitars have size and each part emits sound waves and each part emits differently in different directions producing a complex field and then the room reflects those waves back to us too. I think that this is part of what makes playing acoustic guitar so interesting. Recordings of acoustic guitars can sound great and I love them but it's not the same thing.

Electric guitars are also one dimensional, producing a scalar function. So perhaps this is why on electric I use lots of effects, synths and stuff and with acoustic I'll have none of that.

 

[Robin L]

Just for the record:Telarc made digital recordings at least since 1979 (probably earlier), they were published on vinyl back then as DDA, of course.

I've got a CD from Denon - DDD. Beethoven: "Archduke Trio", performed by the Suk Trio. Session recordings were from 1975; the CD came out in 1987. Denon was experimenting with digital recording as far back as 1969. Denon's early digital recordings used a sampling rate of 47.25 kHz and 13 bits of resolution. Their first commercial digital recording was Nippon Columbia NCB-7003, "Something" by Steve Marcus, released January 1971. My recording of the Suk Trio sounds just fine, FWIW.

https://www.aes.org/aeshc/pdf/fine_dawn-of-digital.pdf

 

[tmtomh]

A recording, when played back in the right environment, can trigger the feeling of realism in the listener. I would argue strenuously, though, that this experience of realism does not necessarily correlate with the actual reality of the performance. It very well might approximate it - but it does not necessarily have much of anything to do with it. As has been noted already, the majority of modern recordings are in no way "real" - they are studio creations of multitracks of instruments and voices not recorded at the same time, or even in the same space, treated with different individual kinds of post-processing, and then mixed into an artificial soundstage because there was no actual soundstage to begin with since the band was not recorded live in the studio, all playing at once, arranged in a concert-like array of players on a stage.

Even with that, there are studio creations like this that trigger a feeling of "realism" - the timbre of an instrument, the presence of a vocal, something about the mix that triggers a mental image of a band playing right in front of us, and so on. Conversely, there are minimal-mic'd live orchestral recordings that don't necessarily sound "real" because of huge mismatches between the size of the recording venue and our listening space, or because we're used to listening from the middle of the hall and the mics make it sound like you're very close to the stage, or whatever.

My point is that when elements of realism come through, we should not simply assume that we're getting a glimpse of actual reality, of what the performance(s) really sounded like. We might be - but just as often, and probably more often, we're not.

I think this confusion causes our discussions to go around in circles sometimes.

 

[Kal Rubinson]

I've got a CD from Denon - DDD. Beethoven: "Archduke Trio", performed by the Suk Trio. Session recordings were from 1975;

I had that recording on LP.

 

[Multicore]

A recording, when played back in the right environment, can trigger the feeling of realism in the listener. I would argue strenuously, though, that this experience of realism does not necessarily correlate with the actual reality of the performance.

The experience of realism may not have much to do with fidelity in either production or reproduction and the experience is produced by the brain. So the possibilities for artistry in making recordings that produce feelings of realism can be a whole aesthetic specialism.