Continue from Part 12

Older HIP vs newer and (supposedly) more uniform

Bradley Lehman wrote (March 30, 2004):
< Not especially referring to the SMP I have heard many modern recordings which I compared with older recordings - from the early HIP days - and concluded that the older ones are a lot better. >
I suspect that at least some of the difference in the past 20 or 30 years comes from the facility of digital editing (as opposed to tape-splicing). A completely flawless surface is now easier to obtain, and that turns into a goal, for some people.

Sure, there were (and are) wizards of conventional tape-editing able to clean things up with the blade and splicing tape, but it's even easier with digital where one can adjust individual notes in time, pitch, dynamics, and more. The regularity of some performers' meter makes it even easier to cross-fade from one take into another. One wonders about the interrelationship of cause and effect, here in all this.

And along with that, the use of electronic tuning devices even when the goal is purportedly an old-style keyboard temperament: the scientific measurements ensure that the results are reproducible across hours, days, weeks, months, years.... So, those temperaments themselves mutate from being a by-ear style into being a fixed set of pitch relationships, all mathematically correct.

When everything's a constant, and reproducible, it's all interchangeable. Welcome to the aesthetic changes of the Industrial Revolution. Unpredictable irregularity isn't valued so much anymore. It's reinterpreted as sloppiness, instead of character and soulfulness.

Pianists could already do this very close editing of individual notes in piano-roll days, and the companies (with or without a pianist) could assemble performances that were never played by a human being in real time, but worked out to take advantage of the machine's precision. Examples: Stravinsky's rolls of "Rite of Spring", or all of Nancarrow's pieces for player piano. Real-time performances could be cleaned up too, limited only by a person's diligence and patience and goals of regularity.

The march of technology since then has just made it easier for everybody else also. With digital there's no danger involved in even trying an edit (as there was with tape, physically cutting the master tape): stick it in there, mess with it, take it all back out if it doesn't work, no harm done. Mix and match. This goes way beyond the laundering of unwanted noises.

In the production of a recording, it's a very important decision: when is the editing finished, time to stop tinkering with it, where any further work would make the overall result worse rather than better? The extra time and money thrown into the editing rounds doesn't necessarily improve the finished product very much.

My own admittedly reactionary philosophy is: performers should only record pieces they really can play/sing well enough without any editing tricks. Give it a couple of goes, whole or in big sections, put it together, and don't mess with it much at the closest levels. I wouldn't forbid editing altogether, but I would (and do in my own recordings) keep it very minimal, to preserve the sense of real performance. Irregularity is part of the music, not something I'd like to whitewash out. This is also part of the reason I so much enjoy listening to recordings from the 1920s-40s: they aren't edited, they're real performances...at least in 3- or 4-minute chunks.

Gabriel Jackson wrote (March 30, 2004):
Bradley Lehman wrote: < My own admittedly reactionary philosophy is: performers should only record pieces they really can play/sing well enough without any editing tricks. >
This is a bit simplistic, surely? However great a performance is, experienced live and in real time, it will also contain a myriad of 'imperfections' which, on repeated hearing and in the close scrutiny that recordings receive, are not acceptable to many/most listeners. The odd moment of shaky intonation, a slightly scrappy entry, a badly-balanced chord etc. etc matter not one iota (and are usually barely noticed) in a concert but to many, they do matter on recordings. And is there anything wrong with that?

Donald Satz wrote (March 30, 2004):
[To Bradley Lehman] ] In these days of 'smooth and similar', a pianist like Burkhard Schliessmann is a treasure. He is an individual in the best sense of the word, and his Bayer recordings of Liszt, Chopin, and Schumann compare well with the greats of the past. As an example, his Schumann reminds me some of Walter Giesekiing's Schumann - impetuous as can be and totally immersed in the Florestan/Eusebius interaction.

I was conversing with him a little by e-mail, and he told me that he will soon record either the WTC or Goldberg Variations (I don't remember which). Anyways, I'm really looking forward to this recording that I think will be released with a year's time.

Donald Satz wrote (March 30, 2004):
[To Gabriel Jackson] I think the world has become obsessed with perfection, the type that only applies to surface values.

Gabriel Jackson wrote (March 30, 2004):
[To Donald Satz] Maybe so, but a recording is not the same as a concert - they are two quite different things and should be treated as such, surely?

Donald Satz wrote (March 30, 2004):
[To Gabriel Jackson] I'd like to agree with Gabriel, but I can't. I don't have different expectations from recordings and concerts. I 'want it all', whatever the venue. I don't mind errors in concert or on record; other matters are much more important to me.

Gabriel Jackson wrote (March 30, 2004):
[To Donald Satz] I don't mind 'errors' either, up to a point, but the use of editing to create the best possible result for home listening seems both legitimate and, to me, desirable. (And this means selecting, say, take a) here because that phrase was better shaped and b) here because the line has a better momentum (etc. etc.) as much as eliminating an untidy entry or an out-of-tune note). I can't think of many completely unedited live recordings that afford much satisfaction on repeated listening.

Uri Golomb wrote (March 30, 2004):
I agree with Johan's descritipn of recent changes in HIP: there is a greater tendency towards legato, and fewer harsh edges (though this is a generalisation, and not without notable exceptions). In some cases (for instance, Herreweghe), legato largely predominates; in others (for instance, Harnoncourt), there is a "simply" greater range of articulation -- the harsh edges are still there, alongside (and often simultaneously with) smoother articulation. Another feature is a greater dynamic range. However, for me, these develompents (and other, related ones) are not necessarily negative.

On the issue of recording (spliced vs. continuous etc.) my position is, quite simply, that I only care what the results sound like; I might be curious to know the process by which they came about, but such knowledge doesn't usually affect my appreciation of the performance, one way or another. I want the performances to sound continuous and spontaneous, but it's possible for an "edited" version to create such an impression. On the other hand, even live performances can sometiems sound as if they were artifically stitched together. I also count myself among those listeners who do have different expectations from recordings and from concerts. I don't mind the odd "fluff" on a recording, either; but I can definitely tolerate and ignore lot more mistakes in a live concert than in a recording which I'm likely to hear again on several occasions.

I also remember hearing a lecture by Susan Tomes, one of the foremost chamber-music pianists today (she used to be a member of the piano quartet Domus; now she is a member of the Florestan Piano Trio), in which she described her recording process, and presented examples of two types of recordings -- formal, edited studio recordings, and unedited radio recordings of live concerts, made before the ensemble went to the studio to record the same works. (In reading this , bear in mind that I heard this lecture about two years ago -- it's possible that my recollection is not entirely precise. but I think I remember the general spirit of things). In several cases, the studio performance sounded more alive, dynamic and musically conviincing then the live recording -- which, while definitely not stitched from myriad small takes, did feature some degree of editing. The primary reason, as she explained it, is that before entering the studio, they didn't really have a chance to hear what their performance sounded like: the image a performer gets in the process of performance is not the same that the audience gets. (A conductor can step back and listen to the orchestra from the other side of the hall in rehearsal; soloists and chamber musicians don't have that luxury). Listening to playbacks made them realise that certain strategies did not have quite the same effect that they thought they had; and they were able to re-think their interpretations accordingly. (I should add that Tomes's ensembles do sound terrific in live concerts, too; but it's possible that their studio experience partly assisted them in the process).

Well, that's getting Off-Topic. But the general point remains: for me, all that ultimately matters is the result. If the performance sounds stitched together, then you won't make me think any better of it by proving that it was all one long take; and if it sounds concinvinclyg continuous and spontaneous, I won't think any worse of it if I find it's been edited.

What I do mind is stories like Malcolm Bilson's, who says that producers deliberately removed his attempts at rubato (where he deliberately disengage let his right hand be rhytmically freer than his left hand) from one of his recordings, throwing away his more lively takes and retaining more "precise" yet dull ones. But all that proves is that performers should, whenever possible, attend the editing sessions and make their own choice of takes -- rather than leaving that responsibility solely to the production team.

Gabriel Jackson wrote (March 30, 2004):
Uri Golomb wrote: < What I do mind is stories like Malcolm Bilson's, who says that producers deliberately removed his attempts at rubato (where he deliberately disengage let his right hand be rhytmically freer than his left hand) from one of his recordings, throwing away his more lively takes and retaining more "precise" yet dull ones. But all that proves is that performers should, whenever possible, attend the editing sessions and make their own choice of takes -- rather than leaving that responsibility solely to the production team. >
I know a pianist to whom this was done by a producer, and she (quite rightly) insisted the way she shaped the phrase originally be reinstated. If he didn't have final approval of the finished edit, Malcolm Bilson has only himself to blame - he should have insisted on it from the outset.

John Pike wrote (March 30, 2004):
Uri Golomb wrote: < The primary reason, as she explained it, is that before entering the studio, they didn't really have a chance to hear what their performance sounded like: the image a performer gets in the process of performance is not the same that the audience gets. >
I certainly agree with this. I have heard occasional recordings of my own playing and am always surprised by errors heard on playback that I was unaware of during performance, especially intonation.

Bradley Lehman wrote (March 30, 2004):
<< My own admittedly reactionary philosophy is: performers should only record pieces they really can play/sing well enough without any editing tricks. >>
< This is a bit simplistic, surely? However great a performance is, experienced live and in real time, it will also contain a myriad of 'imperfections' which, on repeated hearing and in the close scrutiny that recordings receive, are not acceptable to many/most listeners. The odd moment of shaky intonation, a slightly scrappy entry, a badly-balanced chord etc. etc matter not one iota (and are usually barely noticed) in a concert but to many, they do matter on recordings. And is there anything wrong with that? >
I didn't say that the resulting recording should be unedited. I said that I believe performers should record pieces they CAN perform, and not rely on studio tricks to put together things they're actually unable to do. (This is of course in music that was written before the invention of recording: music where the only way to do it would have been to work up a straight-through rendition in real time.)

In the recording session it's good to play the whole piece at least once or twice, to get the sweep of it all, before getting down to sections or smaller fixes. It's a different sort of attention both for the musicians and the producer, presenting the whole thing vs a careful/cautious working through minutiae. The former may have more "imperfections", sure, but it also may have more shape; and (at least in my experience, both doing my own and producing other people's) the big takes usually have more freshness and natural flow. Getting down to do half a dozen takes of the same few phrases, it often starts to sound overly cautious and tense.

The whole take also shows the producer/editor how to synchronize the connection of any smaller sections that are put together artificially. He can run them in parallel to get the unwritten timing modifications the same way the musicians do them in real life: the breathing, the settling at cadences, the transitions at tempo changes. Or, if it's good enough, the whole take can be the master one, and bring the smaller fixes into it only as needed. That way the sweep is still there while the product sounds clean enough. It's a lot easier to get a performance that sounds real by this method, than to stitch together section by section.

It's also good, after all the piece is safely "in the can" according to the producer's notes, to give it one more whole go if there's time left, and if the performers still feel up to it, just for enjoyment...and let the tapes run. That one sometimes turns out to be "the one" going back out to the big picture, after the safety net of "we're already done" is in place, and after all the too-close attention to detail.

And as I said yesterday, it's a crucial decision to know when the editing is done, stop messing with it. That's an artistic choice that (naturally) the performers should have final say on: does this sound like the way we really play this piece with the luck of a good day, or is it sounding artificial in any way?

Douglas Cowling wrote (March 30, 2004):
Bradley Lehman wrote: < I didn't say that the resulting recording should be unedited. >
The next question is the editing of timings, especially in oratorios. Robert King's recording of Handel's "Joshua" was just a string of recorded tracks with no dramatic links. Of course, King's limp, uninvolved performances may have had something to do with the problem.

Donald Satz wrote (March 30, 2004):
[To Douglas Cowling] I find Robert King a wonderful conductor of Handel's music. His Hyperion disc of Heroic Arias featuring James Bowman is my desert island Handel recording.

Gabriel Jackson wrote (March 31, 2004):
Bradley Lehman wrote: < I didn't say that the resulting recording should be unedited. I said that I believe performers should record pieces they CAN perform, and not rely on studio tricks to put together things they're actually unable to do. >
I know you didn't, but how often does the latter happen, really?!

Gabriel Jackson wrote (March 31, 2004):
Douglas Cowling wrote: < Robert King's recording of Handel's "Joshua" was just a string of recorded tracks with no dramatic links. >
What does this mean?

Douglas Cowling wrote (March 31, 2004):
[To Gabriel Jackson] All of the movements had clearly been recorded separately and in the editing little dramatic sense was created by the silence between them. Recits that should have jumped into arias or followed choruses just meandered along with no dramatic urgency. But then I find King's performances full of ill-considered tempos, flacid energy and remarkably little musicological editing: the recording oJoshua just plays the note as printed. Extrordinary.

Donald Satz wrote (March 31, 2004):
[To Douglas Cowling] Another great King recording is Alexander Balus - happens to be an oratorio.

Charles Francis wrote (March 31, 2004):
[To Douglas Cowling] I recall reading that the Harnoncourt / Leonhard cantatas were also performed in this manner. Arias recorded on one occasion, choruses on another; then everything strung together in the editing room. Other HIP conductors such as Leusink followed this example.

Uri Golomb wrote (March 31, 2004):
Charles Francis wrote: < I recall reading that the Harnoncourt / Leonhard cantatas were also performed in this manner. Arias recorded on one occasion, choruses on another; then everything strung together in the editing room. Other HIP conductors such as Leusink followed this example. >
Let's not pretend, however, that this phenomenon is in any way unique to HIP. Herbert von Karajan, Otto Klemperer, Helmuth Rilling, Karl Richter and many others have adopted the same method. Just look at the recording data for Ricther's 1970s recordings of Bach cantatas -- each cantata was recorded over a span of 3-4 years. No details are given in booklet as to what was recorded when, but it's fairly easy to guess what happened: all the soprano arias for all the cantatas in the relevant box (Chrismtas cantatas, Easter cantatas, etc.) were recorded in one session when Edith Mathis was available, all the bass arias when Dietrich Fischer-Dieskau was available, etc. (with special sessions arranged for duets and trios). And then the choruses were recorded in separate choral sessions. And let's not forget Karajan's 1952 B minor Mass (BWV 232) -- recorded with two different orchestra, in two different cities! I would also be very surprised indeed if the phenomenon is unique to Bach recordings.

It is true that, in the 1970s, Harnoncourt's and Leonhardt's ensembles were not able to mount continuous concert performances of cantatas -- the "old" instruments were still new to many of the players, and the technical difficulties were great. However, this has no bearing whatsoever on the situation today, when HIP ensembles can -- and do -- give highly successful concert performances of entire passions and oratorios, and can certainly handle a cantata! If they still record arias and choruses separately (and -- again -- so do most modern-instrument performers of Bach's choral music), it is for economic reasons (record companies prefer not to pay the soloists to sit through a choral session, and vice versa), and the artists themselves would probably prefer to do it differently.

Uri Golomb wrote (March 31, 2004):
PS

Philippe Herreweghe made a point of recording his latest SMP continuously, despite misgivings from his record company. That recording also came after a series of concerts. It is probably one of the very few studio recordings of the Passion -- on modern or period instruments -- to be recorded in this manner.

Bradley Lehman wrote (March 31, 2004):
<< I recall reading that the Harnoncourt / Leonhard cantatas were also performed in this manner. Arias recorded on one occasion, choruses on another; then everything strung together in the editing room. Other HIP conductors such as Leusink followed this example. >>
The most convenient way to mislead is to tell part of the truth and then stop talking. The correspondent who wrote those sentences uses this technique frequently to take his potshots at "HIP" musicians. Here it is again.

=====

< Let's not pretend, however, that this phenomenon is in any way unique to HIP.Herbert von Karajan, Otto Klemperer, Helmuth Rilling, Karl Richter and many others have adopted the same method. Just look at the recording data for Ricther's 1970s recordings of Bach cantatas -- each cantata was recorded over a span of 3-4 years. No details are given in booklet as to what was recorded when, but it's fairly easy to guess what happened: all the soprano arias for all the cantatas in the relevant box (Chrismtas cantatas, Easter cantatas, etc.) were recorded in one session when Edith Mathis was available, all the bass arias when Dietrich Fischer-Dieskau was available, etc. (with special sessions arranged for duets and trios). And then the choruses were recorded in separate choral sessions. >
From an LP of Richter's cantatas for Pentecost ("Pfingst-Kantaten"):
Cantata BWV 34: March 3 & April 25 1974, January 3 & 26 1975.
Cantata BWV 68: May 19 & 21 1974, January 4 & 24 & 26 1975.
Cantata BWV 175: February 27 & 28 1974, January 23 & 26 1975.

=====

In Otto Klemperer's bio, about the SMP: "As the sessions depended on the availability of a number of eminent soloists, the work had to be recorded in dribs and drabs, a process that Klemperer abhorred. There were also interpretative differences. Peter Pears, who was singing the Evangelist, found Klemperer's tempi unduly slow, and was upset by his insistence on conducting the recitatives. (Once Klemperer's back was turned, these were re-recorded without his knowledge or permission.) Klemperer for his part disliked the pathos that both Pears and Dietrich Fischer-Dieskau brought to their singing; in his view the Evangelist's role was to relate a story."

The footnote after the parenthesis goes to author Peter Heyworth's interview with George Malcolm, 1990: Malcolm being the continuo harpsichordist in that recording.

Donald Satz wrote (March 31, 2004):
[Ti Bradley Lehman & Uri Golomb] Thanks to Brad and Uri for clarifying and enlarging on the statement made by Charles.

Charles Francis wrote (March 31, 2004):
Charles Francis wrote: << I recall reading that the Harnoncourt / Leonhard cantatas were also performed in this manner. Arias recorded on one occasion, choruses on another; then everything strung together in the editing room. Other HIP conductors such as Leusink followed this example. >>
Uri Golomb wrote: < Let's not pretend, however, that this phenomenon is in any way unique to HIP. Herbert von Karajan, Otto Klemperer, Helmuth Rilling, Karl Richter and many others have adopted the same method. >
Since Herbert von Karajan, Otto Klemperer, Helmuth Rilling and Karl Richter have never claimed to apply Historically Informed Performance practice, this point is somewhat mute. But, having clarified there is no hypocrisy at work in these cases, can you back up the specific claim about Rilling?

Much, of course, has to do with recording best practice. At the time of Harnoncourt / Leonhardt it was usual to deploy multiple microphones for capturing choir, soloists, instrumentalists and ambience. All such sonic inputs were captured on multi-track tape to be mastered and balanced by the studio engineers: perhaps, compression here and there to control unwanted dynamics, equalisation and enhancers to bring out the colour of certain instruments, not to mention adding ambience from the building for authenticity. Indeed Andrew Parrott admits his early One Voice Per Part recordings were recorded with multiple microphones - his "Heart's Solace" CD recorded in 1998 is an exception: http://www.amazon.com/exec/obidos/ASIN/B0000062CX

Uri Golomb wrote (March 31, 2004):
Charles Francis wrote: < Since Herbert von Karajan, Otto Klemperer, Helmuth Rilling and Karl Richter have never claimed to apply Historically Informed Performance practice, this point is somewhat mute. >
I assume you mean "moot"; and, in any case, I don't think that's entirely relevant. Up until now, the topic of our discussion focused on artistic issues (the damage to continuity), not historical. And they all -- HIP and non-HIP alike -- do it for the same reasons: the logistics and finances of recording sessions.

< can you back up the specific claim about Rilling? >
It's easy enough to prove this: just look at the indications in some of his cantata recordings. In a few cases (admittedly rare), the recording sessions were more than ten years apart (Cantata BWV 70 was recorded in 1970 and 1982). Mofrequently, they are months apart: Cantata BWV 46 was recorded in September & December 1977 and January 1978; Cantatas BWV 27 and BWV 47 were recorded in February & October 1982; Cantata BWV 28 was recorded in November 1981 and February 1982. Perhaps, say, Cantata BWV 29 was recorded in sequence (the indication in the liner notes is "February 1982") -- though I wouldn't rule out the possibility that it was done over several days.

Besides, Rilling told me so himself when I interviewed him three years ago: I asked him about his three recordings of the B minor Mass (BWV 232), and he readily conceded that he recorded choruses and arias separately, and that this was his practice in several large-scale works, not just in the Mass. At least with his latest recording, however, he tried to keep these separate sessions close together (within a week or less of each other).

In general, we're not talking here about a closely-guarded secret. My data on Richter and Rilling alike was published within their CD booklets, so the record companies certainly made no effort to conceal it; and performers are usually quite frank about it in press interviews. Recording arias and choruses separately is standard practice -- which is why Herreweghe made a special point of mentioning that his second SMP was done in sequence.

This doesn't necessarily mean that we should condone it; but it doesn't make sense to single out just one group of performers for censure in this matter. With regards to this, and to the issue of miking and mixing which Charles mentions later, we should also remember that perforemrs often don't have total control over what the production crew is doing; sometimes, perhaps, the choice is to acquiesce to the producer's demands -- or not have the recording done at all. Koopman is married to his producer, Tini Mathot; Heltmuh Rilling has enjoyed a long partnership with one record company who allow him considerable control over proceedings (this doesn't apply, however, to his CBS recordings); and, in a different repertoire, Peter Phillips, director of the Tallis Scholars, owns his record company, Gimell. But not all musicians are so lucky...

Donald Satz wrote (March 31, 2004):
[To Charles Francis] No, the point is not moot. For some inexplicable reason, Charles now gives us the premise that that it's fine to have multiple non-HIP recording sessions, but multiple HIP sessions are inconsistent with the hip approach. I submit that the number of sessions is irrelevant to this matter.

Gabriel Jackson wrote (March 31, 2004):
Charles Francis wrote: < Since Herbert von Karajan, Otto Klemperer, Helmuth Rilling and Karl Richter have never claimed to apply Historically Informed Performance practice, this point is somewhat mute. >
What on earth difference does it make whether someone is an HIP conductor or not?! Either you approve of the practice, or you don't. I personally couldn't give a damn how/when a piece was recorded as long as the results are persuasive and stimulating.

Gabriel Jackson wrote (March 31, 2004):
Uri Golomb wrote: < And let's not forget Karajan's 1952 B minor Mass -- recorded with two different orchestra, in two different cities! >
Quite! And Fritz Wunderlich died during the making of Karajan's recording of The creation (or was it The Seasons?) and they just finished the recording with a different tenor.

Gabriel Jackson wrote (March 31, 2004):
Charles Francis wrote: < Much, of course, has to do with recording best practice. At the time of Harnoncourt / Leonhardt it was usual to deploy multiple microphones for capturing choir, soloists, instrumentalists and ambience. All such sonic inputs were captured on multi-track tape to be mastered and balanced by the studio engineers: perhaps, compression here and there to control unwanted dynamics, equalisation and enhancers to bring out the colour of certain instruments, not to mention adding ambience from the building for authenticity. Indeed Andrew Parrott admits his early One Voice Per Part recordings were recorded with multiple microphones - his "Heart's Solace" CD recorded in 1998 is an exception: >
What's the problem here? It's the result that counts - not the method.

(Although presumably whatever Harnoncourt did must, by definition, be bad........?)

Gabriel Jackson wrote (March 31, 2004):
Uri Golomb wrote: < Peter Phillips, director of the Tallis Scholars, owns his record company, Gimell. But not all musicians are so lucky... >
And if those who take a more purist (even moralistic) view of what is appropriate practice in recording and editing sessions knew how much manipulation is involved in creating a Tallis Scholars recording they would be appalled! (I'm not, by the way......)

Uri Golomb wrote (March 31, 2004):
Gabriel Jackson wrote: < < What on earth difference does it make whether someone is an HIP conductor or not?! Either you approve of the practice, or you don't. I personally couldn't give a damn how/when a piece was recorded as long as the results are persuasive and stimulating. >
I agree on both counts. The practice of recording items separately might make it more difficult to achieve a satisifying continuity, but that's a difficulty which performers can -- and often do -- overcome (whereas, in other cases, even a life performance can sound "stitched").

Gabriel Jackson wrote (March 31, 2004):
Donald Satz wrote: < I submit that the number of sessions is irrelevant to this matter. >
Quite!

Charles Francis wrote (March 31, 2004):
Donald Satz wrote: < I submit that the number of sessions is irrelevant to this matter. >
A perfectly consistent position, if there is no objection to radio stations playing isolated movements from Bach's works. Such reproduction does, after all, most accurately reflect the performance.

Gabriel Jackson wrote (March 31, 2004):
[To Charles Francis] What a bizarre thing to say! Why?

Douglas Cowling wrote (April 1, 2004):
Charles Francis wrote: < Since Herbert von Karajan, Otto Klemperer, Helmuth Rilling and Karl Richter have never claimed to apply Historically Informed Performance practice, this point is somewhat mute. >
Richter's recordings in the 60's were certainly perceived as a revolutionary new way of performing Bach. I remember the critics fulminating about his fast tempos, scaled down choir and orchestra (!), and lack of Romantic "interpretation".

Charles Francis wrote (April 1, 2004):
Gabriel Jackson wrote: < What a bizarre thing to say! Why? >
Well if Mr. Harnoncourt, for example, conducted a single Bach aria at a session with Max van Egmond, then the most accurate reproduction of this performance would be to recreate the event. Bizarre perhaps, but self-evident to the audiophile.

Gabriel Jackson wrote (April 1, 2004):
[To Charles Francis] How very silly! You are against all editing in recordings, presumably? Otherwise, by this logic, you should listen only to unedited stretches of music in isolation (be they of several minutes' duration, a few bars, or even one note). Do you turn your CD player off at the end of a take, wherever it may occur? Because that would be "most accurate reproduction of the performance" wouldn't it? Recording is an artificial medium. Your idea of what is self-evident is, in this case, curious.

Donald Satz wrote (April 1, 2004):
[To Charles Francis] Let me be clear. I don't care if the 2nd Movement of a Bach Concerto is recorded 10 seconds after the 1st Movement, 10 minutes, 10 hours, 10 days, etc. What I care about are the results.

Charles seems to feel that multiple sessions result in a lack of continuity that I assume he thinks he hears when listening to a recording. Given that most recordings are not made at one session, there must be quite a lack of continuity that Charles must deal with.

Charles Francis wrote (April 1, 2004):
Gabriel Jackson wrote: < How very silly! You are against all editing in recordings, presumably? Otherwise, by this logic, you should listen only to unedited stretches of music in isolation (be of several minutes' duration, a few bars, or even one note). Do you turn your CD player off at the end of a take, wherever it may occur? Because that would be "most accurate reproduction of the performance" wouldn't it? Recording is an artificial medium. Your idea of what is self-evident is, in this case, curious. >
Isn't the audiophile goal to overcome artificiality, by closely approximating the performance parameters (volume, acoustics, context etc.), so that one is there, so to speak? And why do you think Philippe Herreweghe felt the need to perform his 2nd recording of the Matthew Passion in one take? Why, in your opinion, did Paul McCreesh create his Epiphany Mass, restoring Bach's works to their performance context? Futile efforts in your opinion? Do you prefer the cut and paste of Harnoncourt/Leonhardt?

Gabriel Jackson wrote (April 1, 2004):
Charles Francis wrote: < Isn't the audiophile goal to overcome artificiality, by closely approximating the performance parameters (volume, acoustics, context etc.), so that one is there, so to speak? >
If you believe that listening to a recording is ANY way comparable to hearing the same piece live, I cannot help but wonder....

"And why do you think Philippe Herreweghe felt the need to perform his 2nd recording of the Matthew Passion in one take?"
He didn't record in one take, for goodness' sake. The recording took place over 5 days!

"Why, in your opinion, did Paul McCreesh create his Epiphany Mass, restoring Bach's works to their performance context? Futile efforts in your opinion?"
Not at all, but that has nothing whatsoever to do with the merits of edited and unedited recordings. And McCreesh's recordings are heavily edited, like just about everyone else's.

"Do you prefer the cut and paste of Harnoncourt/Leonhardt?"
Are you so blinded by your antipathy to Harnoncourt that you actually believe it is only his recordings that are edited? The onlycompletely edited recordings out there are a very few live recordings - even most 'live' recordings use material from more than one performance.

Robert Sherman wrote (April 1, 2004):
[To Donald Satz] I agree with Don, of course, but let me throw another element into the mix.

On Marriner's first Messiah recording with Ely Ameling, her "Behold and See" is magnificent, but her "I know that my redeemer liveth" is beneath poor. (These are my judgments and you may not agree with them, but accept them as a hypotheses for purposes of this discussion.)

Suppose, now, I splice in a beautiful Redeemer (Harper/Jackson or Blegen/Westengerg). Which has more continuity: The recording with the same soprano at both points giving one great and one awful performance, or the recording with great soprano singing, in the same style, at both points?

Donald Satz wrote (April 1, 2004):
[To Robert Sherman] The continuity would likely be greater with just the one singer, but I'd rather hear the great soprano singing.

Johan van Veen wrote (April 1, 2004):
Gabriel Jackson wrote: < The onlycompletely edited recordings out there are a very few live recordings - even most 'live' recordings use material from more than one performance. >
Or they are 'repaired' in the studio.

Uri Golomb wrote (April 1, 2004):
Charles Francis wrote: < Why, in your opinion, did Paul McCreesh create his Epiphany Mass, restoring Bach's works to their performance context? >
As Gabriel pointed out, that recording is probably as much "cut-and-paste" as everything else. Here is the recording data: "Brand-Erbisdorf (Saxony), Parish Church, 4 & 5/ 1997; Freiberg (Saxony), Cathedral, 11/1997". Hardly a sequential, continuous recording! And, again, the information comes from the recording's own booklet -- no attempt at disguising anything.

McCrees's purpose was for the listener to experience the original performance context; it applies to teh end-product, not to his means of producing that end-product.

< Futile efforts in your opinion? >
NO -- it's a fascinating recording. But I find myself doing a lot of cut-and-paste listening to it -- I usually skip the non-Bach bits. I did hear it in sequence, once or twice; it gives you an insight into the original context, and helps you understand. But ultimately, I prefer the non-historical way. I much more enjoy listening in sequence to some of McCreesh's other reconstructions. But perhaps the best medium for such a
reconstruction is a video recording.

< Do you prefer the cut and paste of Harnoncourt/Leonhardt? >
I still don't understand why you insist on writing as if this is something only Harnoncourt and Leonhardt did. It is the standard practice for anything except truly live recordings (and even most live recordings feature some editing: they record several concerts, paste together the best bits from each, and sometimes do patching sessions as well).

If you want to dismiss H/L for that reason, then, by the same token, you must throw away Richter, Rilling, Klemperer, Karajan, Koopman, Suzuki, McCreesh and just about everyone else. (Not to mention Glenn Gould, who turned cut-and-paste into an ideology...)

Gabriel Jackson wrote (April 1, 2004):
[To Johan van Veen] Of course I meant to say "completely unedited"!

Charles Francis wrote (April 1, 2004):
<< "And why do you think Philippe Herreweghe felt the need to perform his 2nd recording of the Matthew Passion (BWV 244) in one take?" >>
Gabriel Jackson wrote: < He didn't record in one take, for goodness' sake. The recording took place over 5 days! >
First, Harnoncourt, then Leusink and now Herreweghe! Even the Epiphany Mass of Paul McCreesh appears to be a cut-and-paste job. Will there be no end to the unhistorical performance practices of HIP-conductors?

Shocked
Charles

Johan van Veen wrote (April 1, 2004):
[To Charles Francis] I wonder how many people on this list continue to take this clown seriously.

Gabriel Jackson wrote (April 1, 2004):
[To Charles Francis] Oh, I see - it's not unhistorical, as you put it, to make a recording, but only as long as it's unedited.

Drew Pierce wrote (April 2, 2004):
[To Charles Francis] Indeed, what is the world coming to? Terrorism and unhistoricism abound!!!

Didn't Bach himself record his works in one take? Truly, for HIP conductors to truly be hip, they should take their cue from the Maestro, no?

Epicentre of HIPdom

Charles Francis wrote (June 1, 2004):
Johan van Veen wrote: < A mark of desperation: if you are out of arguments you can always fall back on labeling. What HIP has to do with Calvinism is beyond me. Most HIP musicians have no roots in Calvinism whatsoever. And nobody forces 'the poor guy' to play in a HIP ensemble. I am sure there are other things to do. The possibility that he chooses to play in a HIP ensemble seems to be unthinkable to some individuals on this list. If someone looks for an example of a 'closed mind': don't look further. >
I am naturally open to correction, but my perception (and it is only a perception) is that Holland stands at the very centre of the revival of Historically Informed Performance practice. I think of the recitals I have attended by Leonhardt and Koopman, and the numerous keyboard players who mention in their CV a period of study under Leonhardt. I think of Rifkin's criticism of the HIP movement for building on the principles of historical keyboard practice rather than correct choral practice (a reference I take it to the Dutch fathers of HIP). I grant you that Koopman is Catholic, but when it comes to cantatas don't the majority of Dutch performances happen in the context of Calvinism rather than Catholicism? What is the official state religion in Holland?

And on the freedom of choice issue for the trumpet player, just how many professional orchestras in Holland perform Bach cantatas on modern instruments?

Riccardo Nughes wrote (June 1, 2004):
< I grant you that Koopman is Catholic, >
Are you sure?

Uri Golomb wrote (June 1, 2004):
Charles Francis wrote: < I grant you that Koopman is Catholic, but when it comes to cantatas don't the majority of Dutch performances happen in the context of Calvinism rather than Catholicism? What is the official state religion in Holland? >
I'm curious: does Holland have an official state religion? And does that mean anything significant? (England's official state religion is Anglican Christianity -- does that mean that the majority of the population are committed Anglican believers?)

Beyond this: in a radio interview in 2000, Ton Koompan indeed complained about the existence of a Calvinist approach to Bach -- but what he had in mind was not quite what Charels had in mind... He directed his complaint against those musicians who refused to acknowledge the existence of secular influences in Bach's sacred music, who denied, for example, the existence of dance-like elements in the sacred cantatas. The "Calvinist" in this context is more likely to be someone like Karl Richter, rather than Gustav Leonhardt -- not to mention Koopman himself... In a later interview, he repeated the same message, though expressed reservations about his own use of the term "calvinist" (see: http://www.goldbergweb.com/en/magazine/interviews/2003/09/16243_2.php),

Johan van Veen wrote (June 1, 2004):
Charles Francis wrote: < I am naturally open to correction, but my perception (and it is only a perception) is that Holland stands at the very centre of the revival of Historically Informed Performance practice. >
That is open for debate. One could also argue that Britain is the centre of HIP, or Switserland (remember that someone like Leonhardt came into touch with the historical performance practice through Eduard Müller and August
Wenzinger at the Schola Cantorum Basiliensis).

True, in the 1970's Dutch ensembles were very prominent, but among the members of those ensembles were many players from Belgium, which is a almost completely Catholic country. I have only to mention the Kuijken brothers, Paul Dombrecht, René Jacobs etc etc.

< I grant you that Koopman is Catholic, but when it comes to cantatas don't the majority of Dutch performances happen in the context of Calvinism rather than Catholicism? >
Most performances of Bach's sacred works in the Netherlands take place in form of concerts - completely 'secular' affairs. Only at rare occasions cantatas are performed as part of a service.

< What is the official state religion in Holland? >
There is no official religion. The Netherlands never had a 'state church'. And the largest church, the United Protestant Church of the Netherlands, can hardly be considered 'calvinist'.

Apart from that, I don't think many professional musicians in the Netherlands can be considered 'calvinist' either, even not Christian or 'religious'. And I can't see any religious influence in Dutch music making.

And I still haven't read a clear explanation why 'calvinist' should have anything to do with musicians' preferences for the historical performance practice.

< And on the freedom of choice issue for the trumpet player, just how many professional orchestras in Holland perform Bach cantatas on modern instruments? >
I can't give any statistics, but there are a number of orchestras in the Netherlands, which specialise in taking part in performances of the many choirs our country has. How many of them perform Bach cantatas is something I can't tell.



Continue on Part 14

HIP (Historically Informed Performance): Part 1 | Part 2 | Part 3 | Part 4 | Part 5 | Part 6 | Part 7 | Part 8 | Part 9 | Part 10 | Part 11 | Part 12 | Part 13 | Part 14 | Part 15 | Part 16 | Part 17