Article: Realism vs Accuracy For Audiophiles | Part 1: Soundstage

[fas42]

32 minutes ago, bluesman said:

 

 

 

Pretty obvious, actually - just on laptop speakers ... thanks for that.

 

The Steinway is more 'stolid'; the Fazioli has a 'bounce' to the attack; a liveliness to the tone.

 

[ARQuint]

Along the same lines, I highly recommend a program on the Channel Classics label (a hybrid multichannel SACD, no less) with Paolo Giacometti playing an identical all-Ravel program on two pianos, a modern Steinway and a restored 1888 Érard - the brand that Ravel used at home when composing. The differences in the performer's approach to the same music on the two instruments—tempo, pedaling, dynamics, etc—is very interesting.

 

[The title of the 2-SACD set is Compared. The catalog number is CCS SA 31612]

[firedog]

4 hours ago, ARQuint said:

Along the same lines, I highly recommend a program on the Channel Classics label (a hybrid multichannel SACD, no less) with Paolo Giacometti playing an identical all-Ravel program on two pianos, a modern Steinway and a restored 1888 Érard - the brand that Ravel used at home when composing. The differences in the performer's approach to the same music on the two instruments—tempo, pedaling, dynamics, etc—is very interesting.

 

[The title of the 2-SACD set is Compared. The catalog number is CCS SA 31612]

Nice recommendation, Andrew, thanks. The performances do sound really different. Not hard to hear it at all.

[bluesman]

9 hours ago, firedog said:

Nice recommendation, Andrew, thanks. The performances do sound really different. Not hard to hear it at all.

Indeed!  I also recommend that you go to the Channel Classics website and read Giacometti's elegant introduction to this project.  He says quite eloquently what I've been saying crudely.  You can also download the liner notes for these recordings HERE.  I'm a liner notes junkie, and these are a fantastic accompaniment to the music.  I think this is the meat of the matter:

 

"The different tonal worlds of the two instruments result primarily from the parallel stringing and largely wooden frame of the Érard, and the cross-stringing and steel frame (to accommodate the higher tension of the strings and case) of the Steinway. The sound of the Érard is thus more stringy and dies away differently, while the various registers are more individual. The Steinway glories in the resonance of the strings and case, and the mélange of the notes. Since relationships between the immense richness of colour and timbre in Ravel’s music are therefore different, this is naturally of influence on the interpretative choices made by the performer. The pedalling, dynamic transitions, balance between the parts, harmonic colours, build-up of tension, tempos – all these must be approached differently depending on the instrument used."

 

It's of great interest to me that Érard was Lizst's piano of choice.  One of the things that distinguishes one piano from another is the action, which is the mechanism that transfers a fingerstrike on a key into a hammer strike on strings.  Lizst was one of the finest technical pianists in history.  But his choice of the Érard enabled him to develop his mastery further than he might have on a less responsive piano.  He could not have played as fast or as expessively on a piano with heavy or sluggish action.  So he would not have written the technically demanding music for which he was acclaimed if he didn’t have a piano on which it could be played.  Liszt was a bit of a show off. 😏

 

I was interested to read in those notes that Giacometti says that "...the Steinway grand has hardly changed since the late nineteenth century".  Steinway has made many major changes in their pianos over the years, the most important to the player being in the action (the mechanism that transfers a finger strike on a key into a hammer strike on strings).  This page on the Steinway website has a wonderful illustrated history of Steinway actions from 1720 to today.  This quote from Steinway's literature says it all:

 

"One day in 1932, Josef Hofmann, the American pianist, composer, and inventor who had already spoken of the STEINWAY’s 'extraordinary perfection of action,' came to STEINWAY HALL and said, 'It isn’t quick enough. Can’t you make it still more sensitive, still more responsive?' "

 

Once you start hearing these subtle and underappreciated characteristics of music of any kind, you both understand and use your ears, your mind, and your systems muich better.  And you begin to understand the deeper essential meaning of accuracy in reproduction.   It's not just about frequency response, tight bass, and precise imaging - it's about delivering the soul and the essence of music to you as the composer and performer conspired to shape them.  It's about hearing and feeling how Richter's Chopin differed from Ashkenazy's Chopin (Richter was often said to have made all other pianists sound clumsy). 

 

This is not limited to classical music. Art Tatum was a jazz pianist with an uncanny sense of timing that let him play seemingly a tempi through an entire verse or chorus of a song, only to emerge right on the beat long after most listeners have lost it.  And if you pay careful attention to his left hand while his right apears to be taking a vacation from counting, his comping for himself is as solid as a rock, despite being very very subtle.  And now that we've opened this line of communication and education, I hope all AS members and participants will have discovered a new dimension in their listening pleasure. 

[WAM]

17 hours ago, fas42 said:

The Steinway is more 'stolid'; the Fazioli has a 'bounce' to the attack; a liveliness to the tone.

 

I agree. Thanks Bluesman for your explanation!

 

Another interesting album is Andras Schiff's take on the Diabelli Variations: he plays them on a 1921 Bechstein and on a 1820 Brodmann fortepiano. 

 

 

[Claude]

Wonderful text and information. Thank you. Looking forward to the next.

[audiobomber]

“SOUND STAGE” AND “IMAGING”

 

According to J Gordon Holt's Audio Glossary, https://www.stereophile.com/reference/50/index.html

 

soundstaging, soundstage presentation The accuracy with which a reproducing system conveys audible information about the size, shape, and acoustical characteristics of the original recording space and the placement of the performers within it.

 

imaging The measure of a system's ability to float stable and specific phantom images, reproducing the original sizes and locations of the instruments across the soundstage.

 

There is another component of imaging that Holt's description does not capture. I've heard it called "thereness", which is a clumsy made-up word but does convey the meaning. The other term for this is palpability. From Holt again:

 

palpable Describes reproduction that is so realistic you feel you could reach out and touch the instruments or singers.

 

Palpability should be included in imaging IMO, to further differentiate imaging and soundstaging. Instrument placement is a fairly easy trick. Sound that can make you believe the artist is in the room is much harder.

[Temporal_Dissident]

Great article. Raises a number of interesting questions and pushes on a few sacred cows. 
 

I certainly question the value of pursuing some platonic ideal of “accuracy” in soundstage. We can fret over whether the guitar appears Left-Center-Left as it did the moment it was recorded, but the reality is the musicians didn’t really obsess over their seating arrangement and half-way through the song, the bass player moved to the other side of the stage, so where does that leave you? 

I definitely take the point that with un-amplified live music, there is a single point source for most seats in the house. 
 

The more important thing to me is that the sound is clear and mixed in a way that best serves the music and the song,...not in a way that accurately reproduces some arbitrary spatial soundstage, which brings me to this question:

 

Does stereo serve another function (other than reapers in an “accurate” soundstage)? By mixing various instruments so they are distributed among different channels reduces “congestion” at any given frequency and makes it easier for our speakers to reproduce all sounds more clearly. Isn’t THAT the more valid justification for stereo, multi-channel mixing? If an artist wants to create a spatial mix that serves their music, who cares if it accurately reflects where musicians were sitting during recording?

[The Computer Audiophile]

26 minutes ago, Temporal_Dissident said:

half-way through the song, the bass player moved to the other side of the stage, so where does that leave you? 

My favorite comment of the month 🙂

[bluesman]

4 hours ago, Temporal_Dissident said:

half-way through the song, the bass player moved to the other side of the stage

I must be missing that.  Are you talking about the Great Guitars video?  If so, I see the bass player move toward stage center a few times.  But his amplifier stays where it is, and that’s what making the sound.

 

Stage position does matter to the musicians.  We position ourselves to hear and communicate with each other as well as possible, unless we all have monitors (which is rare in the world of small clubs, dive bars, and parties in which most jazz, blues, and commercial musicians live).

 

FWIW, I just noticed that the Great Guitars video I linked is not the same one I used when writing the piece. Nothing’s different except Charlie Byrd’s guitar, which was not his touring Ovation.

 

[Brian A]

Fascinating read.  Thank you!!!

 

The article certainly moots criticism of the likes of Gilles Martin rejiggering the sonic image (did I use that term correctly?) when remastering the Beatles stuff. 

[bluesman]

2 hours ago, Brian A said:

Fascinating read.  Thank you!!!

 

The article certainly moots criticism of the likes of Gilles Martin rejiggering the sonic image (did I use that term correctly?) when remastering the Beatles stuff. 

You’re very welcome! 
 

The first two Beatles albums were made on 2 track tape machines - so the sonic image was both captured (rather than engineered) and fairly accurate.  “I Want to Hold Your Hand” was the first Beatles music recorded on a 4 track machine, and they used “sound on sound” to overdub multiple parts.  So from that point on, their sonic image was engineered.  “Dear Prudence” was probably their first recording on an 8 track machine, and from there the tracks multiplied year by year.

 

Martin’s remixes and remasters are pretty good, in my opinion.  I can’t quibble with his concepts - they’re as valid as the originals were.  Once you start recording with separate mics on separate channels for individual instruments and voices, there is no image until the engineer creates it.

[AudioDoctor]

regarding realism... Do you think the final version of this album sounds like this video shows it would if you were in the room with the performers when it was recorded?

 

edit: except for rare cases, realism is a fools errand when it comes to playing back recorded music. You want realism, go to a live in person performance.
 

 

[fas42]

18 minutes ago, AudioDoctor said:

regarding realism... Do you think the final version of this album sounds like this video shows it would if you were in the room with the performers when it was recorded?

 

 

No ... how it works if you were in the room with the performers is that you would be aware of a single acoustic, where the sounds blend, with part direct and part reflected sound.

 

In such a recording, what you hear are separate, acoustic layers - each performer is in their own space; which are all on top of each other, in front of you. This sounds like it would be messy; but the human hearing system is very good at untangling this - given enough clues ... on a very transparent system, each layer exists with full integrity, and individuality - it's easy to focus on each layer, in turn, and hear exactly what's going on, with that particular sound source - be it voice, or the drums, say.

 

Is it realistic? No ... but it has its own qualities which are perfectly satisfying to experience. The recording is realistic in that you can hear exactly what happened with the mic'ing, etc - but it doesn't get in the way of enjoying what you're hearing.

[bluesman]

1 hour ago, AudioDoctor said:

Do you think the final version of this album sounds like this video shows it would if you were in the room with the performers when it was recorded?

I assume you realize that they’re not in the same space.  There are clear walls separating and sonically isolating each performer from the others, with separate microphone(s) on each.  So you couldn’t be in the room with them because there is no room - and there is no image until the engineer assembles it.

 

That’s not to say that the final product is not realistic - the individual instruments can be captured with great fidelity and reproduced with great realism by capable systems.  The article at the head of this thread is addressing the difference between realism and accuracy - and any judgment about imaging or soundstage would be inaccurate because neither existed so neither could be captured and reproduced.  Both were created de novo, but both can be realistic because they can sound like the music is being created in your space.

 

If a fine portrait artist imagined a woman’s face and painted her purely from his mind, the portrait might look quite realistic - but it couldn’t be accurate because she never existed.

[AudioDoctor]

39 minutes ago, bluesman said:

I assume you realize that they’re not in the same space.  There are clear walls separating and sonically isolating each performer from the others, with separate microphone(s) on each.  So you couldn’t be in the room with them because there is no room - and there is no image until the engineer assembles it.

 

Yes... that's the entire reason I chose it.

[Brian A]

I wonder how many artists take an interest in the sonic image.  Certainly Pink Floyd did with all kinds of sounds flying around all over the place.

 

On the flip side (har, har), I sometimes wonder what the sound engineer is thinking.  One of many examples that jumps to mind is "My Dear Companion" on the "Trio" album.  Most of the song is fine with Emmylou Harris crisply placed on the left, Dolly Parton in the middle and Linda Ronstadt on the right.  However, toward the end, Harris and Ronstadt abruptly float to a combined position kinda sorta a bit to the left of Parton.  It drives me nuts.  I wonder why it was done?

[Temporal_Dissident]

19 hours ago, bluesman said:

I must be missing that.  Are you talking about the Great Guitars video?

No, sorry. I was just referring to a hypothetical situation. Musicians do move; stages are different sizes, etc. I understand that stage position matters to musicians (I play drums), but after reading your article it occurs to me that we listeners may imbue “soundstage” with more importance than do the musicians themselves. I thought the point about the stereo image collapsing as the you move farther from the stage was especially interesting. 
 

The point I was trying to make is that it does seem arbitrary and unimportant that stereo or multi-channel mixing generate a soundstage that “accurately” reflects the placement of musicians at recording. What is more important (to me) is that the mix serve the music and how it will sound coming out of two speakers in my living room (or car, or headphones, etc). Will more stereo separation help me hear each instrument more clearly? Will it better serve the song? That is important; not where the bassist sat in relation to the pianist in a studio when it was recorded originally. 
 

The “accuracy vs realism” frame is interesting, but my guess is that many producers are going for neither accuracy nor realism, but rather for what makes the recording sound most pleasing and engaging through speakers. After reading your piece, I am more at peace with that. 

[bluesman]

47 minutes ago, Temporal_Dissident said:

my guess is that many producers are going for neither accuracy nor realism, but rather for what makes the recording sound most pleasing and engaging through speakers.

Many are. Unfortunately, many are also going for what will sell and many more are tailoring their final mixes and masters to a target medium.  Compression, EQ, delay etc are often used to punch up the sound for mobile device listeners using earbuds, for example.

 

I think the drive to have an accurate reproduction in playback of the 3 dimensional relationships among performers is fueled by audiophiles who believe that there was a soundstage to be captured.  That drive would be better served by providing the pan settings for each performer, so we could at least know how well our systems are reproducing the emperor’s new soundstage.

 

Here’s a very interesting piece on binaural vs stereo recording and some of the engineering that goes into tailoring a recording for different uses, users, and devices.

[firedog]

19 hours ago, bluesman said:

You’re very welcome! 
 

The first two Beatles albums were made on 2 track tape machines - so the sonic image was both captured (rather than engineered) and fairly accurate.  “I Want to Hold Your Hand” was the first Beatles music recorded on a 4 track machine, and they used “sound on sound” to overdub multiple parts.  So from that point on, their sonic image was engineered.  “Dear Prudence” was probably their first recording on an 8 track machine, and from there the tracks multiplied year by year.

 

Martin’s remixes and remasters are pretty good, in my opinion.  I can’t quibble with his concepts - they’re as valid as the originals were.  Once you start recording with separate mics on separate channels for individual instruments and voices, there is no image until the engineer creates it.

On Sgt. Pepper, his professed goal was to take the sound of the original mono, and somehow translate that sound to modern stereo. I think he did a good job. It's clearly stereo and modern, yet does have some of the spatial concepts and feel of the original mono. 

[bluesman]

6 minutes ago, firedog said:

On Sgt. Pepper, his professed goal was to take the sound of the original mono, and somehow translate that sound to modern stereo. I think he did a good job. It's clearly stereo and modern, yet does have some of the spatial concepts and feel of the original mono. 

Here’s a link to an NPR interview with Martin about Sgt Pepper in which he explains more about his intent.  He started with the final takes of the instrumental parts, which had been transferred from the first generation tapes to the 4 track tapes that were the sources for the original mixing and mastering.  The only way they could mix and master in those days was to tweak playback of the source tapes and record the adjusted signals on fresh tape - so what went onto a master could have been re-recorded multiple times with attendant degradation.

 

For the 50th anniversary remix, Martin started with digital rips of the original tapes.  He believed that the original mono mix sounded better because, even 8 years after the first stereo records became widely available, Sgt Pepper was “ designed for mono” as an album.  He believed that the original stereo mix was given short shrift because of the industry focus on mono, which seems very odd to me.  He seems to base this opinion on the fact that the mono mix took 3 weeks and the stereo mix was done in 3 days.  
 

I can’t help but wonder if they didn’t spend the first 3 weeks determining gain, pan, balance, EQ etc for every note of every track - there was no automation for this, so they had to do it all manually in real time while playing back the source tapes and re-recording the result.  Once they had the roadmap, which could easily take 2 to 3 weeks, they undoubtedly used most or all of it for both mono and stereo mixes.  Giles was 19 when Sgt P was made - I suspect he’s not quite right in some of his recollections.
 

I remember being struck by his comment that the original tapes he used had no hiss.  This is simply not possible - tape hiss is integral to the medium.   It’s the random noise generated by the magnetized particles that are not organized into a signal archive by the record head and are therefore randomly oriented.  Even brand new blank tape makes some hiss.

 

 

[firedog]

3 hours ago, bluesman said:

 

For the 50th anniversary remix, Martin started with digital rips of the original tapes.  He believed that the original mono mix sounded better because, even 8 years after the first stereo records became widely available, Sgt Pepper was “ designed for mono” as an album.  He believed that the original stereo mix was given short shrift because of the industry focus on mono, which seems very odd to me.  He seems to base this opinion on the fact that the mono mix took 3 weeks and the stereo mix was done in 3 days.  
 

I can’t help but wonder if they didn’t spend the first 3 weeks determining gain, pan, balance, EQ etc for every note of every track - there was no automation for this, so they had to do it all manually in real time while playing back the source tapes and re-recording the result.  Once they had the roadmap, which could easily take 2 to 3 weeks, they undoubtedly used most or all of it for both mono and stereo mixes.  Giles was 19 when Sgt P was made - I suspect he’s not quite right in some of his recollections.

Those aren't recollections, but Martin wasn't wrong. The Beatles, and their engineers, such as Geoff Emerick, said stereo was an afterthought for them till the White Album. The reason was the English market, that was still mono based. The Beatles based their marketing and recording around the UK market. May sound strange to Americans, but that was their orientation. In fact, Penny Lane and Strawberry fields were originally recorded for Pepper, but were released as a separate single because management felt the Beatles needed another single on the UK market (which was still singles based). The Beatles' approach in the UK was not to repeat songs released previously as singles on albums - they felt that was bad value for their British fans, so those two songs were left off of Pepper. In the UK, stereo was considered too high end and audiophile for them to worry about at the time. 

 

The Beatles were tailoring the sound for their fans, who they thought were still mostly mono listeners. Much of British Rock at the time was the same. That's why Giles said he was trying to get a "stereo version of the original mono" mix for the 50th anniversary remix.

The Beatles conceived of the sound of the album in mono, were actively involved in the details of the mix, and once the mono was done - they left. The Abbey Road crew did the stereo on their own, without input from the Beatles, and the Beatles didn't listen to it or approve it before it was finalized. This is all pretty well documented in interviews and books by the principals. In fact, Emerick says they played around with the stereo mix and did some stuff (extra use of automated double tracking, etc.) that the Beatles probably wouldn't have liked. There are a lot of discrepancies between the two mixes.  That's why a lot of Beatles afficianados (and people such as Geoff Emerick and Ken Scott) always considered the mono version the "authentic" one. 

Giles, BTW, wasn't even born when Sgt. Pepper was made. Not sure where you got the 19.

[bluesman]

3 minutes ago, firedog said:

Giles, BTW, wasn't even born when Sgt. Pepper was made. Not sure where you got the 19.

I remembered the right number but the wrong era - "as his dad was losing his hearing, the 19-year-old Giles Martin served as his father's ears".  Now I obviously have to re-read Summer of Love.

 

12 minutes ago, firedog said:

they played around with the stereo mix and did some stuff (extra use of automated double tracking, etc.) that the Beatles probably wouldn't have liked

I'm not so sure they wouldn't have liked it.  As I recall (and the age 19 thing proves that we may not be able to rely completely on my recollection...) ADT was invented at the Beatles' request the year before Sgt Pepper, and was used extensively on Revolver.

[firedog]

double post

[firedog]

20 minutes ago, bluesman said:

I'm not so sure they wouldn't have liked it.  As I recall (and the age 19 thing proves that we may not be able to rely completely on my recollection...) ADT was invented at the Beatles' request the year before Sgt Pepper, and was used extensively on Revolver.

I was paraphrasing what Geoff Emerick said in his book when writing about the Pepper mixes. He didn't seem to think much of the stereo mix he and the others made. 

 

BTW, most Beatles albums are a good example of recordings made with zero intention of "sounding like a real performance". Most of the albums are total studio creations, and the Beatles and George Martin were consciously and intentionally trying to create some sort of aural tapestry that couldn't exist outside the studio and didn't at all sound like a live band performance.