KVR Audio

fluffy_little_something · Sun Oct 22, 2017 8:29 pm

Ah, ok

And is it worth sacrificing an entire LFO for?

Mutant · Post by **Mutant** » Sun Oct 22, 2017 8:34 pm

fluffy_little_something wrote:Ah, ok
And is it worth sacrificing an entire LFO for?

I don't know.

Load The Legend, listen to 1 oscillator on saw wave, decide for yourself.

aciddose · Post by **aciddose** » Mon Oct 23, 2017 2:07 am

Yes these are simple side-bands generated by modulation of the oscillator reset point. Frequency modulation is actually a more expensive way to do it, there is near zero frequency modulation in the real thing.

The difference though is zero as all the important harmonics are covered up by the noise-floor. If it were possible to measure the signal with a much lower noise-floor (-200 dB) we'd see the harmonics aren't so trivial and there are far more than just one of them.

This would be part of the PSU noise generator. If you look at the "analog modeling" threads I've posted to recently I talked about this and also about Gaussian noise which in fact sounds much different than uniform noise.

All analog signals have a baseline level of Gaussian noise that adds a special sort of "shimmer" to the sound. Uniform white noise doesn't provide the same effect because it is much more audible in comparison.

In fact I've already been looking at adding these things... although I really don't like the idea of turning Xhip into any sort of "analog model" it would be very interesting to have such signals available in a signal routing system (misnomer: "mod matrix") that could be applied as desired.

Is the effect audible? Yes it does in fact have a non-trivial effect upon the way self-oscillating filters "latch"/sync on to harmonics as they're swept. It creates a smoother effect and more interesting "wobble" (same effect you get with Xhip Q+saturation = high.) It is however a lot more minimal than clearly audible things like non-linear VCA CV response or asymmetric pulse waves (a hard-clipped ramp, actually.)

This exact "wobble" effect is already used in the "factory" bank presets such as "Verb" in "Lead" #35. You might notice as you increase the saturation the noise becomes far more important: actually this is an example where Gaussian noise would be far better because it would create a more interesting effect while being far less audible even at "low" levels like -36 dB vs. the filter's oscillation.

yul · Post by **yul** » Thu Dec 21, 2017 2:54 am

Hello, can someone please help me with the x-mod section it doesn't seem to do much in my case?

Perhaps I need to read to manual although it should be rather straightfwd?

Loving Xhip btw with a bit of effort it can get close if not beat quite a lot of big names (if you know what you are doing..)

Cheers!

Mutant · Post by **Mutant** » Thu Dec 21, 2017 3:44 am

yul wrote:Hello, can someone please help me with the x-mod section it doesn't seem to do much in my case?

Perhaps I need to read to manual although it should be rather straightfwd?

Loving Xhip btw with a bit of effort it can get close if not beat quite a lot of big names (if you know what you are doing..)

Cheers!

Use very low notes and experiment with the Xmod amount knobs in oscillator section and osc2 pitch.
You will quickly discover what does what.

aciddose · Post by **aciddose** » Thu Dec 21, 2017 4:14 am

See here: http://xhip.net/synth/releases/xhip_8_u ... df#page=12 this page describes X-mod Mode, and the next page (see the bottom page 13) tells you that the oscillators each have a X-mod depth parameter.

Also a quick description of the controls:

1) The X(cross)-mod mode control in the oscillators section:

This selects the source. You need to imagine this like taking the output of one or both oscillators, passing it through some filters and plugging it into the FM input in the oscillators. It would help if you have used modular synthesizers before as you'd be able to imagine how different parts are connected.

The mode is just choosing if the source is Osc.A, B, or both A + B at once. It also selects whether the signal is passed through a high-pass filter (differentiator) and whether it is input to the pitch input (like keys/notes) or linear input (Hz) on the oscillators.

A common example is that with vibrato the depth is in pitch, so if you use 10 cents and play 100 Hz, the modulation will be 10 Hz wide. If you play 1000 Hz it will be 100 Hz wide, so by using the pitch input the depth "tracks" with the note played and automatically increases or decreases. For example if you detune two oscillators by 10 cents at 100 Hz (95 Hz and 105 Hz) the phasing will happen at 10 Hz. At 1000 Hz though it will happen at 100 Hz.

With a linear detune you detune by Hz. So say you detune by 5 Hz, when you play a 100 Hz note the phasing will happen at 5 Hz. At 1000 Hz the phasing will still happen at exactly the same speed: 5 Hz. So you can play all over the keyboard and get the same speed for the phasing: this is a very cool effect not available yet in Xhip because only detuning in pitch is made available in the current version.

2) The X-mod depth in both oscillators:

This controls how much modulation is applied. For example if you select Osc.A as the source and adjust Osc.A X-mod depth, you're applying FM feedback to Osc.A. If you change the depth for Osc.B, A is then modulating B.

If you leave the depth on zero for both oscillators it won't matter which mode you select. Nothing will happen exactly like changing oscillator waveforms while the mixer levels are set to "off".

yul · Post by **yul** » Thu Dec 21, 2017 4:21 am

Holy smokes thanks so much I am SORRY I just COMPLETELY overlooked the depth knob in osc A.

I am very tired I need a vacation

Thanks also for the fantastic FM on XHIP explanations for ACIDDOSE this is truly appreciated!

aciddose · Post by **aciddose** » Thu Dec 21, 2017 5:11 am

Well, the GUI layout isn't 100% clear because it doesn't distinguish "xmod (mode)" from "xmod (depth)".

Hopefully if I ever get around to it I might be able to improve the documentation. I think a youtube video running through 90% of the features would probably help a ton if I could manage to set up a mic + studio to produce anything of decent quality, but I don't know if I'll ever do that.

yul · Post by **yul** » Thu Dec 21, 2017 2:17 pm

Hi thanks again

can you also please confirm that the mixer or otherwise the filter drives/saturates even without the use of saturation knob? I was trying to match the tone of some vintage sound and seems that the right setting would be at around 50% in the mixer.

I don't think you would need to go the extent of the YT videos. Some devs do have their manual in HTML straight on the website which I like although I realize that yours is also updated (yay!) in PDF which is also nice!

Cheers!

Mutant · Post by **Mutant** » Thu Dec 21, 2017 6:59 pm

yul wrote:I was trying to match the tone of some vintage sound

Which sound was it ?

YT link ?

aciddose · Post by **aciddose** » Thu Dec 21, 2017 8:44 pm

Yes the filter is sensitive to input level. The "0 db" level is slightly more saturated (by about +3 dB) than typical filters because I hate the very loud self-oscillation some synthesizers produce.

You'll probably want mixer levels from anywhere between -3 dB and +12 dB to roughly duplicate most sounds, but you won't be able to reproduce heavily saturated moog sounds or similar.

Example of heavily saturated "moog" bass sound:
http://xhip.net/temp/hangock.mp3

That one you could probably get with +12 dB in addition to the waveshaper. It's hard to guess exactly how the original was created, although they likely used an arp2600 it's semi-modular and also no reason to believe they didn't use a pedal of some sort on top.

My X1 analog synthesizer always sounds like that when driven, so it isn't just the "moog filter", but the combination of saturation at multiple points. Xhip only saturates feedback (hard clip) and the low-pass integrator (2nd order hyperbola.)

fluffy_little_something · Thu Dec 21, 2017 9:09 pm

Pity the other instruments are so loud, hard to hear the funny bass as such...

yul · Post by **yul** » Thu Dec 21, 2017 9:13 pm

Mutant wrote:
yul wrote:I was trying to match the tone of some vintage sound
Which sound was it ?
YT link ?

It's your common analog 2 osc poly on the raw osc side.

It's not scientific for sure but by ear it's close.

It's when applying filter rez that the difference shows up if you sweep it.
If the oscs are set too high the sweep changes in character a lot.

aciddose · Post by **aciddose** » Thu Dec 21, 2017 9:39 pm

fluffy_little_something wrote:Pity the other instruments are so loud, hard to hear the funny bass as such...

The radio version doesn't have the long solo intro to the bass, but I wasn't only demonstrating the bass. The lead sound is a high mixture of a couple oscillators and self-oscillation driven hard which creates that "ring-mod" effect, but I don't think it uses any ring-mod (although it may, the 2600 had one.)

Those ultra-heavy sounds with loads of distortion won't be possible in any software, period. That's due to the aliasing that occurs with sampled signals and is unavoidable. You can get slowly better by doubling the sample rate, but once you're running at 64x over-sampling it gets a little bit hard on the CPU

fluffy_little_something · Thu Dec 21, 2017 10:49 pm

Fusion was one cool genre back then

Of course it is not possible to make the exact same bass patch, but I guess with some effort one could get close.
Was just playing around a bit with the saturation and distortion effects in my synth

https://app.box.com/s/3oom514jlc05s3wdks1t4w944o4xq7rx

Thinking about it, the new Repro 1 must be ideal for that type of sound thanks to its Jaws effect.

Xhip vs Analog Hardware