Paul Haneberg writes:
>>My patch of the week suggestion was to try to find a way for us all to learn from the masters of patching without asking them to give away their experience without receiving some compensation.<<
We're all here to answer reasonable questions for free. When somebody asks how I did something or other specifics I have no problem telling them if it's easy to explain. The problem comes when they want to see the patch. The only motm patch I have on the computer is my recorder patch. All my other motm patches are on sheets of paper that contain my entire synth on it, all marked up, with notes on the back, and too small and detailed to get a clear scan of at a reasonable size. I'd have to get into Corel draw and create nice clean readable ones from scratch, and that's not practical unless it's for a bigger project like a book or CD.
>>I'm personally interested in expanding the number of people who are doing real synthesis.<<
People start yahoo message boards for the most narrow niches, and yet when I suggested somebody set one up for synth programming, nothing happened. Too bad, because there are probably some people from many other groups that might be interested. The board would be for those who want to advance synthesis and do great things, not techno, tb303s, ghetto beats, or any of that stuff that we're neck deep in on other boards. A prerequisite to getting on the site would be to own most of Carlos' and Tomita's analog recordings (a list could be posted). Those are the two synth masters and things they've done would be refered to in conversation. Those interested in synthesis already have those. Those who have avoided those two or who still avoid those two show by their very actions a lack of interest in advanced synthesis and the board probably isn't for them. That's what I'd like to see.
>>real synthesis is in danger of going the way of the Krumhorn. (which I'm sure Elhardt could synthesize)<<
At about the 00:26 point in my MOTM_Renaissance3.mp3 piece most noticable in the left channel is a buzzy Krumhorn sound.
Mike Estee writes:
>>There's a rather large group of people currently engaged in trying to figure this synthesis stuff out all over again. There's a lot of twenty-somethings (such as myself) currently re-inventing the wheel. We don't really know what we're doing, we're not sure how to get there<<
Unfortunately, people aren't listening to masters from the past and seeing what a synthesizer can sound like and what's already been done. As a result, they're starting over from scratch as you say. Back when I was on the Access Virus list I'd hear people say they don't think a synth can do this or that, when in fact, it was doing those things 25 years ago. Ignorance of the past is a major problem.
Andrew Sanchez writes:
>>This is one aspect (module complement) that may hamper efforts of a patch of the week endeavor. Modules in peoples' systems probably vary greatly. Patches that used more than the run-of-the-mill modules might not be possible with some systems, though it could offer starting points for other sounds, as is mentioned above.<<
That is a problem. Some of my patches use outboard gear too. Some aren't even on the motm. But sometimes seeing what's there and how it's used might apply to something else.
Neil Bradley writes:
>>Makes me wonder how much popular music would be popular if people were actually educated musically.<<
Case in point. The latest song by Beck that made its rounds on the late night shows (usually means it's the best song on the CD, yikes) consisted of something like 4 measures of material stretched into an entire song. It went something like this: No introduction, repeat first 2 measures 4 times, repeat second 2 measures 4 times, repeat first 2 measures 4 times again, repeat second 2 measures 4 times again, repeat first 2 measures 4 times yet again but without vocals, repeat second 2 meastures 4 times yet again, maybe repeat first 2 measures 4 times still again, song over. Those who can't write, repeat over and over and over again. No memorable tune, no substance, no thought or effort went into it at all. Why put in any effort when millions will buy it anyway.
Aardvark writes:
>>Hey Ken, how about posting the patch for the vocal/chorus demos?????<<
I had somebody ask for those, but I had to point out to him that the male voices were done on a Doepfer synth and I didn't write down the patch, only the filter bank settings. And as per the earlier part of this post, the motm stuff is on paper, not in the computer. But I can paste below an old Analogue Heaven post where I verbally explained those patches. I forgot to mention below that the female also had some portamento on it.
---------- AH Post -----------
Les Mizzell writes:
>>As usual, the vocal here is very nice, IMHO. I'm beginning to think that
you have way too much free time on your hands!<<
Thanks for the compliment. Being a fast synthesist it only took a couple of
hours to come up with that sound despite it being quite complex. I was
influenced by the vocal of a real woman Tomita used in the beginning of
Venus from "The Planets". With no vibrato it becomes much more difficult to
sound real.
>>So, care to enlighten us all and share the patch?<<
I can explain the elements involved. It's an MOTM synth using a couple of
420 filters in series with high resonance creating a bandpass filter with a
couple of peaks, plus a 410 triple filter in parallel forming more
resonances (with some additional feedback to boost those resonances a bit),
plus a 440 lowpass to act as a treble control. I couldn't tell you what
frequencies are involved because it was done by ear, and the motm only gives
0-10 tick marks anyway. I'm using noise modulated PWM for noisy harmonic
modulation, slewed s/h for wavy pitch motion (hard to tell amongst all the
reverb and echo). That goes through a 31 band EQ with a complex setting
then into a Boss VT-1 voice transformer where I can easily shift the
formants that I synthesized, up or down with a single slider. I shifted
them up a bit until I got the particular quality of female voice I liked,
kind of moved it from an alto to a soprano type sound. That's the general
overall scheme.
Unknown Freak writes:
>>Awesome, Ken. Very nice rendition of Vaughn Williams too. I found this
chart on Mark Smart's site (Is this basically what you're following<<
Thanks. The Vaughn Williams came off a CD just incase you confused that
part as mine. As for the motm female voice, I'm not following anything,
just doing that by ear. For the Doepfer male voices I'm following something
roughly like his charts. I don't have any URL handy, but if you do an
internet search for vocal formants you can probably find a chart that I've
seen in several places that shows the 5 main vocal formants of different
vowel sounds and for male and female. Things don't need to be all too
accurate though, and in the analog world usually can't be anyway, especially
if using a fixed filter bank. I'm running a Doepfer fixed filter bank into
the synthesis vocoder module (another fixed filter bank) and boosting the
same frequencies for an "ah" type sound. This gives me steeper filter
slopes. But I'm also running a 3 band filter in parallel and EQ and a
Roland chorus unit. I still need to experiment with different
configurations to get a stronger and brighter sound while trying to avoid
that vocoder frequency limited bandpass filter sound.
Jason Proctor writes:
>>can you post the actual frequencies you used for the solo female
voice? - i'm using 730, 1090, 2440hz, with a bit of 3100hz but that
seems to make it too buzzy.
what resonance on the filters? - i'm using about 60%.
one oscillator or multiple? - i'm using one.
what duty cycle on the pulse wave? - narrower seems to sound better on the
ion.
the formant filters sound better but i want to do this from the ground up.<<
I'm just using one osc, but as per above I'm moduling the pulsewidth, but
I'm also mixing in some sawtooth from the same osc. Since I'm using
different methods for different voices I don't know the resonance settings.
For example there are no resonance settings on a fixed filter bank or the
motm triple filter. But you are experiencing the problems of analog
filters. They suck when it comes to this kind of stuff. They let too much
through and so your timbre will sound buzzy and synthetic. EQ helps to
further shape the sound. Running filters in series sometimes helps. If I
get a spectrum of a male voice saying "ah", there is a really steep almost
vertical dropoff at about 1KHz. No analog filter can do that. But using a
digital filter that can go to extreme settings I can set up formants that
are far more extreme than in the analog realm.
Following is an audio example that literally only took me a few seconds to
set up formants on. The first me recording my voice without formants by
holding the mic to my throat and humming. Then using a digital EQ that can
go far beyond +/- 12dB I added formants to my humming, after which it then
sounds like my mouth is open and I'm saying "ah".
http://home.att.net/~elhardt2/Hum_With_Artificial_Formants.mp3-Elhardt