-----------------------------------------------------------------------------------------------
dpcon: My question is: exactly what string patches were used? It
only says 1st, 2nd Vlns., etc. I'm sure we would all love and
appreciate a detailed account of how the piece was realized as well
(including dynamics and so forth). How much time did it take?...
-----------------------------------------------------------------------------------------------

Dear dpcon
Here are the samplefiles I used for the E-Partita:

*_mV-sfz
*_mV-sffz
*_stac-1
*_stac-2
*_0_3s-1
*_mV_dim3_1_5s
*_pz_mV-1
1) *_tr-GS

*= VL-14,VA-10,VC-8, KB-6, Solo-VL
1) = not used for VC-8 and KB-6

Tonight I have no time to give more answers. Sorry! But tomorrow I'll be back with the "E-Partita-birth-story".
Tip: Please have also a look at herbs topic "First baroque music demo online, SAT Sep 27". There you will find some first answers to your questions.

Until tomorrow
Beat Kaufmann

My hat of blew me away too. The only thing that gives it away is that the 16ths are very even - the kind of thing one doesn't hear after working on a piece for long time. Have you tried using Groove Quantize? Tap in a couple of bars' worth of 16ths, create a groove from them, and quantize everything with a strength less than 100%.

But that's splitting hairs, because it's fantastic musically.

@Nick Batzdorf said:

....The only thing that gives it away is that the 16ths are very even...

Hi Nick

You are right and I agree.
For example I had to play the following notes:
b a cis a dis a e a dis a cis a b a a a
How to play them? I think that there are two possibilities:
1. stac-1 stac-2 stac-1 stac-2 .... every a will be played with stac-2 >>> EVEN
2. stac-1 stac-1 stac-2 stac-2 stac-1... you have a little variation every 2nd a

I took both of them. In case of the a-note you sometimes have a "leere Saite"-situation, specialy with repetitions. The direct translation is "empty string". I don't know the word in English. But it means that the violinplayer can play the a-tone without pressing a finger with the A-string. So with "leere Saite"-situations I followed the concept 1, because the "leere Saite" sounds rather even. For other repetition-situations I used scheme 2. I tryed to make the best... The best solution would be to have stac-1 stac-2 stac-3 stac-4... and harddisks... or
to buy a professional orchestra. [:O] So the Partita shows us also "some borders" between computer- and livemusic. [H]

Good idea to quantize with a groove... but don't tell Mr. Bach about it! [[;)]]


Beat Kaufmann

They're called open strings in English (e.g. "open A"), and I'm happy to know what "leere Saite" means!

Anyway, what I mean by "even 16ths" has to do with timing, not timbral variation - hence my suggestion that you try groove quantizing it to introduce some variation.

Again, this is very subtle, but I think it's that the attacks are exactly together between instruments - it's just a hair too precise (even for Bach). If you groove quantized just the violins, for example, I suspect it would eliminate the last clue that this isn't a group of live musicians. Or perhaps you could create different grooves for the different sections.

Bottom line, though, I think I'm being very cheeky to criticize your performance. I'm not at all sure I could come close to programming your level of detail!

@dpcon said:

I'm sure we would all love and appreciate a detailed account of how the piece was realized as well (including dynamics and so forth). How much time did it take?

Hi again
As I promised: Here is story about the creation of the VSL- E-Partita-demo:
I think everybody has its own technic to get music with VSL. Maybe it also depends on the equipement you have. So the following story about the birth of the Partita in E is my personal way to produce music with VSL.

1. With the Logic-sequencer and the Bach/Mangiagalli-partitur I recorded voice by voice (Viol1, Viol2, Viola…). I always do this work without VSL but with my good old Roland SC-8850-Soundmodul. By the way: For the First Violin I used the step-record-mode in this case because the only length of most of the notes is 16ths.

2. Then I prepared the midi-file: I fixed all volume-values to 88 (One count before the important breakpoint in VSL: mf > f). Further on I quantized the miditracks and managed the tempo variations.

3a) Then the work with HALion and VSL began: With the actual Partita I started with „Violins1“. I played through their part while loading all the necessary VSL-files on the way to get the expected sound. In this case the first Violins did not give a lot of „VSL“-work (lots of 1/16). I marked every 2nd 1/16 in one bar, selected the menu-point „mark the same positions in all bars“ and in a few seconds all the stac-upbows and the stac-downbows where linked. For the first note in each bar I usually used the sfz-file. In a second round I corrected "bad-sound-problems" (see also answer to Nick).

3b) Now to the dynamics: With the drawing-mode I added the dynamics as they are written in the partiture (cresc / dim / p / pp …). I don’t have to say, that from now on all the corrections and changes are adapted by ears. In the end I exported the „Violins1-Track“ as a wave-file. Important for a good final result: Every instrument should be prepared so that you would be able to hear it as a soloist as well.

4) Doing the same procedure with every instrument I had finaly 7 wavetracks.
violins1, violins2, violas, cellos, doublebasses, soloviolin1 and 2.
The miditrack from Violins1 I also used for playing the soloviolin1. This is possible if you take the same midichannel-sample-links. (for example Violins> stac-1 = CH3 , Soloviolin > stac-1 = CH3). The final mix with the wavefiles I always make with monitors…using Logics effects and mixer-automation… using samplitudes reverb…

Time-management:
Point 1 and 2: around 4hours, Point 3a)b): around 8hours, Point 4: 2hours
In total 14 – 16 hours.
…To get the E-Partita-Mangiagalli-score tooks months some years ago.

Best wishes
Beat Kaufmann

@Nick Batzdorf said:

Anyway, what I mean by "even 16ths" has to do with timing, not timbral variation...

Hi Nick

Thank you for the "open A"!


...Now I think I've got it > you mean "humanize"! ?
Good idea... with the next track.


Beat Kaufmann

Beat Kaufman,
Thanks for putting that description up so clearly - it is a good method, and has some similarities to what I've been trying recently, like using an older sampler loaded with basic sounds to get the notes and tempos down before going into the extreme detail of the VSL sounds. I wondered if anyone else was doing this as a normal working method.

Your method also shows the same thing Herb talked about using - sfz and accented or changing dynamic notes on normal lines, which might not be so obvious at first but is extremely important to get the natural variation of dynamics.

I don't agree with the criticism that there is any excessive regularity of 16th notes - it sounds just right the way it is.
Sincerely
William Kersten

William,
Thank you for responding to my explanations. As you know I'm Swiss, so my mainlanguage is "Schwiizerdüütsch". Therefore I'm also very happy about your sentens "Thank you for putting that description up so clearly".

@William said:

I don't agree with the criticism that there is any excessive regularity of 16th notes - it sounds just right the way it is.

Thanks again!! [:D]

In 1977 I bought the record UMBRELLA / Special Limitid Edition Direct-To-Disc Recordings / Toronto Chamber Orchestra, Conducted by Boyd Neel, UMB DD9.
There the musicians are playing this Mangiagalli-Bach-Partita.
Comparing both tracks:
1977 The Torontos Partita >> st-tt-tt-tac-cato - and often chaos - but it is music as well!
2003 Kaufmanns VSL-version >> quantized - always clear - but ???

So on our way to the reality Nicks idea quantize/humanize the midi-notes should be checked out ... [H]

Sincerely
Beat Kaufmann

Beat,

Thanks for such a generous and thorough breakdown of your Bach Partita realization (as well as the list of instruments.) I'm sure it will be studied by many since the results confirm the method so completely. I know I'm going to try that approach. I was amazed at the short amount of time it took considering the days and weeks some of us spend trying to get things right.

Thanks again for brilliant stuff that would make the Master proud.

Dave Connor

dpcon@pacificnet.net

@Another User said:

As you know I'm Swiss, so my mainlanguage is "Schwiizerdüütsch".

Gruetsi! (Spelling?)

@Nick Batzdorf said:

Gruetsi! (Spelling?)

written:
"Grüezi"

spoken:
G = G in garden
r = r in rolling
ü = [Y] (International Phoneticalphabet) you don't have this vocal in your language!
If you know, how to spell the french word "jus d'orange" you get it.
e = a like "a house"
z = [ts] (International Phoneticalphabet)
i = i in international
So it's easy: G - r - ü - e - z - i
This only word shows the difficults with "outlandish" languages.

Maybe Herb will produce some day a DVD with phonethics.
At that time I'll send you the midifile for playing Grüezi. [[;)]] [:D]

Have fun while trying "Grüezi" yourself.
Beat

I know it as "grüezigrüe-zigrüe-zigrüe...." from hiking near Pontresina as a child. You say it to every person coming the opposite direction, in other words.

@Nick Batzdorf said:

I know it as "grüezigrüe-zigrüe-zigrüe...." from hiking near Pontresina as a child. You say it to every person coming the opposite direction, in other words.

Grüezi means hello, hi. But also - if you want - good evening, good morning... it's the original swiss-greating-phrase for every situation.

"tschau" (goodbye)

Beat Kaufmann

Oh, I understood perfectly what it means! But that doesn't mean I knew how to spell it.

I noticed on the download page that you used a light reverb from Samplitude. Could you try using the real time Room Simulator, Halls, and the IR with the 2.9 tail? It would be interesting to hear how it sounds with a convolution reverb approach.

Peter Alexander
peter@truespec.com
310-559-3779