Distorted corrected PC3 KB3 funky and Amazing G.

[Theo Verelst]

Hi all,

 

Working on my corrected (for sampling and A grade studio signal chain elements, mainly) PC3 sounds, which have gotten into the actually usable sound area, I was inspired by Chester Thompson's TOP teaching in a thread here (I'd seen the original one with maybe better quality video as well, years ago).

 

So after having worked on the sounds (this organ being not faithfully one-bag-of-tricks-only being transformed from a ROM sound) with the correction effects in their latest incarnation and order, I was Sunday afternooning this little piece if you like some entertainment:

 

pc3_funkorgan1.mp3 2min24 stereo .32mb/s float mp3 <6MB

 

Unfortunately, these corrected sounds aren't for download as it is, if you'd like a go with them (and most PC3 used would want to, I'm sure) maybe something can be arranged. They're pretty far away from ROM sounds and will sequence and record good, and probably I'll get all kinds of sensible variations at slider control soon.

 

Oh, the (free for non-commercial use) mp3 I played on the PC3-76's keyboard directly may sound a bit loud even if the 96 kHz --> single pole 20kHz low pass --> gate at -80dBr or so --> (probably unused) -0.1 dBr limiter --> ffmpeg downsampling to 44.1kS/s doesn't get near max mp3 headroom.

 

TV

[Markyboard]

That sounds good Theo. I've found that satisfaction with the end result is mostly about the in-depth process we go through in tailoring a sound to our exact individualistic preferences. And there's a whole lot that goes into that. Unfortunately most of that effort is not something that can be appreciated or shared with another user of that sound.

[Delaware Dave]

I'd like to get a copy of the leslie settings, it seems that the warble in the low rotor was really minimized, which seems to be everyone's complaint. Is it possible to make the leslie effect available?

[Theo Verelst]

I'd like to get a copy of the leslie settings,... Is it possible to make the leslie effect available?

 

As it is I've made nothing available yet, because the idea is that I've used built in deep sound DNA in a general way, so that all sounds are more sampling neutral and much better designed. So *if* I'd make some of this sound available it would be necessary to make the whole song+aux2 effect chain+ global effects (in the song) available. Or nothing would work right, like I'd normally describe these sounds.

 

There's a method prepared in the inners of the PC3 to allow this while correction to take place, and it will work on all the ROM sounds (albeit with automated-like editing of the ROM sounds at the moment), but it's an accurate correction, where sound elements are corrected by additional effects, and pretty exact settings of the global effects to iron out all kinds of standard mistakes.

 

T

[Delaware Dave]

When you create a sound in the PC3 the leslie is an effect object and it is stored separately in the PC3 that way, so it can be extracted separately. I was looking for the effects object that was included in the PC3 that should be stored in the PC3 on its own.

[Theo Verelst]

... mostly about the in-depth process we go through in tailoring a sound to our exact individualistic preferences...

 

My issues with most of the sounds are twofold: there's sampling related errors that are big distortions, and there's nothing about that kind of deviation from analogue or mechanical sound I myself like at all.

 

Then there's the idea that most instruments need to be put in a room, recorded right, will require effects, and in most cases will need some work to project right out of a set of speakers, which is a hell of a job, and doesn't work right ever it seems without applied science in the sampling and digital processing.

 

In the process of finding sound paths in this case in the PC3 where these problems receive solutions, I feel well and there's nothing much out there except the high grade music where that is the case as well..

 

T

[Al Coda]

My issues with most of the sounds are twofold: there's sampling related errors that are big distortions, and there's nothing about that kind of deviation from analogue or mechanical sound I myself like at all.

 

Maybe,- but KURZ KB3 sounds don´t use any samples at all !

 

A.C.

[jazzpiano88]

Hi all,

 

Working on my corrected (for sampling and A grade studio signal chain elements, mainly) PC3 sounds, which have gotten into the actually usable sound area, I was inspired by Chester Thompson's TOP teaching in a thread here (I'd seen the original one with maybe better quality video as well, years ago).

 

So after having worked on the sounds (this organ being not faithfully one-bag-of-tricks-only being transformed from a ROM sound) with the correction effects in their latest incarnation and order, I was Sunday afternooning this little piece if you like some entertainment:

 

pc3_funkorgan1.mp3 2min24 stereo .32mb/s float mp3 <6MB

 

Unfortunately, these corrected sounds aren't for download as it is, if you'd like a go with them (and most PC3 used would want to, I'm sure) maybe something can be arranged. They're pretty far away from ROM sounds and will sequence and record good, and probably I'll get all kinds of sensible variations at slider control soon.

 

Oh, the (free for non-commercial use) mp3 I played on the PC3-76's keyboard directly may sound a bit loud even if the 96 kHz --> single pole 20kHz low pass --> gate at -80dBr or so --> (probably unused) -0.1 dBr limiter --> ffmpeg downsampling to 44.1kS/s doesn't get near max mp3 headroom.

 

TV

The sounds are nice, but how are they different from say a Hammond clonewheel with Leslie?

 

I guess I'm not sure what your goal is, but I don't see how any organist would adopt your sounds as the subtleties are well within any organ to organ variability.

[Theo Verelst]

Make and record a sound like it...

[Delaware Dave]

My issues with most of the sounds are twofold: there's sampling related errors that are big distortions, and there's nothing about that kind of deviation from analogue or mechanical sound I myself like at all.

 

Maybe,- but KURZ KB3 sounds don´t use any samples at all !

 

A.C.

This is what I always thought but if you read the manual it appears that Kurzweil, like Hammond with the XK5, uses a combination of both physical modeling and samples. It's not clear to me as I read it as to how it works, they use the term "upper tonewheels"; I'm not sure what they mean by upper tonewheels, but these tonewheels are sampled based per the manual and adjustable. See the attachment. Can anyone explain what the upper tonewheels are or which specific tonewheels they are talking about?

[mate stubb]

I can't tell you how the Kurz works, but here's the deal with samples for organ:

 

A sampler/rompler general purpose playback engine will generate each pitch at each octave/fifth separately and you can end up with horrible harmonic beating due to the incompatibility of equal temperament vs the harmonic series.

 

What organ sims do is correctly model the 91 wheel tone generator by creating 91 continuously running waves which get reused everywhere that specific pitch is required in the different octaves. In addition, the tonewheel pitches on a real generator have some slight variances from a strictly equal temperament, caused by having to have fixed gear ratios and a fixed number of teeth on each wheel. A good sim models this and avoids the phasing/beating problem.

 

Now many sims read the waveform from a bank of numbers - effectively a wavetable. Whether the wavetable values contain a pure sine wave, or a sample representing a single cycle of an actual tonewheel matters not. For the second case, one could say it's sample based, but without the problems.

[Delaware Dave]

What organ sims do is correctly model the 91 wheel tone generator by creating 91 continuously running waves which get reused everywhere that specific pitch is required in the different octaves.

The Kurz KB3 model does do this and there is no harmonic beating. What I don't understand is why they also need samples to support the sound since they have the tonewheels continuously running waves. They don't do a good job in the manual of explaining how the two interact and what role the samples play.

[mate stubb]

Samples could be for artifacts - key on and off click maybe?

[Al Coda]

 

... Kurzweil, ... , uses a combination of both physical modeling and samples. It's not clear to me as I read it as to how it works, ...

 

Can be they use looped single cycle waveforms as a reference for physical modelling.

I remember NI B4 used similar tonewheel sets.

 

... they use the term "upper tonewheels"; I'm not sure what they mean by upper tonewheels, but these tonewheels are sampled based per the manual and adjustable. See the attachment. Can anyone explain what the upper tonewheels are or which specific tonewheels they are talking about?

 

Just speculation,- but maybe there´s not enough DSP power generating ALL the (24-91) tonewheels in realtime and make ´em user adjustable.

 

B.t.w.,- I´ve never read that page in the manual because dealing w/ KB3 parameters was self explaining and dead on simple for me.

So,- I´m now surprised too, especially because I don´t hear samples alias in top ranges of the keyboard in KB3 mode,- while w/ many "non KB3" sample based programs, you easily hear aliasing in top ranges of the keyboard.

 

A.C.

[Delaware Dave]

I don´t hear samples alias in top ranges of the keyboard in KB3 mode,- while w/ many "non KB3" sample based programs, you easily hear aliasing in top ranges of the keyboard.A.C.

 

I found the KB3 organ to be bright (like a late B3 model) which can get thin in the upper octave so I use the "Mellow" setting which tames the organ thinness to some degree. It appears to slightly dampen the last 4 drawbars' output.

 

I created a split manual and dual manual organ using KB3 and VAST organs with a double leslie.

 

https://forums.godlike.com.au/index.php?topic=3861.msg31515#msg31515

 

The lower split/lower manual uses VAST while the upper manual user KB3. There is no aliasing/phasing issues as the KB3 organ is phase-locked; the VAST organ for the 2nd manual or lower organ below the split is not phase locked, so I try to be careful not to expose the upper end of the bottom keyboard (where it is more prevalent) on the VAST organs. When in single manual split mode the VAST organ is much less exposed. You would have thought that the next generation organ architecture that should have been built into the Forte and PC4 would have accomodated the ability to have two KB3 organs running at the same time, eliminating this issue.

[Theo Verelst]

For the 60+ people with interest in the mp3, I forgot to mention there's a bit Lexicon in the sound, with a difficult electronic setup. It's not much short reverb, but it makes the recording noticeably different, becuase the Lexicon, when instructed right, understands about signals and sampling related distortion, and can often make a difference to make the sound better.

 

TV

[Al Coda]

... but these tonewheels are sampled based per the manual and adjustable. See the attachment. Can anyone explain what the upper tonewheels are or which specific tonewheels they are talking about?

 

I went thru several KB3 organ programs and found out they use the same 2 keymaps across ALL.

150 sinewave and 185 "altune",- a variant of sinewave w/ some different tuning,- perhaps.

When you´re in KB3 edit mode and set upper tonewheels volume to -96dB,- you´ll recognize it fades out completely @ (IIRC) D# (or E) in the highest octave of a PC361´s keyboard, representing the highest octave of an organ keyboard.

Will say,- D# -C (or E -C) are completely silent in the upper octave then

So, these "upper tonewheel" samples are for the last 9 (or 8) keys of a 61-keys keyboard action.

 

A.C.

[mate stubb]

Maybe the upper tonewheels are pitched differently to emulate foldback.

[Theo Verelst]

In the KB3, and it's strange combination of upper and lower manual mix rail combination into a single sound, there's a lot of hidden elements. The "samples" of the wavefroms for each coil being driven by a virtual tonewheel cannot be a whole number of sampling time steps long without taking special precautions. Say we sample at 48,000 HZ sampling frequency internally, and we want a certain virtual tone-wheel + coil to output a 1000 Hertz tone, we'd be able to sample a real tonewheel, and loop the "sample" of 48 discrete samples a 1000 times per second. Now, if we'd want the highest C at 2092.31 Hz, obviously there's no easy way to simply "loop" a short wave sample to do that, we'd have to interpolate between samples in a short "sample" or maybe take a long recording of that note until we can patch the beginning of the wave back to the end of the wave at the end of the table without too much of a "jump". So that whole idea is a bit more involved.

 

T

[Al Coda]

Maybe the upper tonewheels are pitched differently to emulate foldback.

 

 

Good point !

I´m not a Hammond tech and cannot explain how "foldback" really works.

But I know,- you can !

 

Leaves the question why they use 2 different sample sets/ keymaps then.

Are there 2 different types of "foldback" existing for Hammond organs,- p.ex. one for consoles and the other for some spinet models ?

 

A.C.

[Theo Verelst]

So, apart from the ability to pitch bend the organ, which clearly requires the samples to change frequency, there's the continuity of the organ, which requires all tones made by the wheels to run continuously, because the samples shouldn't be synchronized at key down. Then the waveform is imprinted in the tonewheel, which isn't a pure sine wave. After that, when the (at most) 9 draw bars per manual are connecting the key contacts of a pressed key with the mix rail, a transient wave occurs at all mix rails, and there are impedance effects when 2 or more tones connect up the coils near the wheels with the mix rails. That requires not just some sort of key click, but a electronic signal change related with the momentous phase and amplitude of the current in the wheel recording coils and the currents in the rest of the mix rails.

 

T.