zebra artifacts oscillator effects

[gentleclockdivider]

Some artifacts/zipper noise when playing low notes + ( certain )osc effect and max.resolution , in thid example ' the trajector 'effect .


-Choose osc , only fundamental , pure sine
-set osc effect to'trajector ' and modulate verry gently by an lfo or envelope , mod.value about 10 , or slowly change the trajector parameter manually ( -5 / +5 )
-Set osc to a low tuning
-Set osc resolution to max .
Result ..some nasty almost inaudible overtones are present , these overtones are not created by the trajector effect itself, but by the smoothing/interpolation of the resolution parameter , when resolution update rate is at max ,the side effects are most noticeable.
Even when resolution is set to a low value , the side effects are noticeable , there is just more smoothing going on.
Clearly audible when playing low notes .
Like I said , it does not happen with all osc.effects , the trajector effect is an example where it is most noticeable

Maybe a future version of zebra could benefit from a resolution parameter that is updated at audio rate ( yes I know more cpu ) , but minimazing the artifacts that occur with some osc. effects in the present version of zebra.

[trance_lucent]

Always thought this is a feature, not a bug. In real life, there is no truly harmonic sound. Harmonic sound means infinite time. And even if you just change the volume of steady sine, you get non-harmonic distortion, that is visible like bell-like noise around the sine.

Many OscFX (try scrambler or scatter!) in Zebra can greatly change the waveform even if being modulated slowly. Max resolution results in dynamic changing of waveform, that is rapidly morphed, all the harmonics change its levels very fast generating non-harmonic bell-like noise, so it can sound like noisy morphing with a lot of cool noisy distortion. "Eleven" the result and you get very interesting and stylish sound.

Edit: Just to clarify: by "bell-like noise" I mean band-passed noise that looks like bell on the spectral analyzer.

[gentleclockdivider]

urs can you please shed some light .
These are clearly artifactss of the smoothing parameter

[gentleclockdivider]

unbelievable , not even a response from the developer.
Sorry I have such a good hearing , but at least a confirmation from the dev's would be nice .

[EvilDragon]

The whole dev team is at Musikmesse, so you're gonna have to sit tight till they return. But trance_lucent is on a good trail - it's probably not a bug, but rather a byproduct.

[gentleclockdivider]

@ trance lucent ...the bell like noise is generated by the smoothing agloritm
It clearly is a side effect of the smoothing parameter creating ( almost inaudibble ) zipper noise ( bell like noise )
Look at the spectrum graph .
This happens with some osc.effects , the trajector being the most audible.

If this is an added feature , well that seems like a verry strange decision

[gentleclockdivider]

Urs?
Shed some light please

[Sendy]

Always thought this is a feature, not a bug. In real life, there is no truly harmonic sound. Harmonic sound means infinite time. And even if you just change the volume of steady sine, you get non-harmonic distortion, that is visible like bell-like noise around the sine.

Many OscFX (try scrambler or scatter!) in Zebra can greatly change the waveform even if being modulated slowly. Max resolution results in dynamic changing of waveform, that is rapidly morphed, all the harmonics change its levels very fast generating non-harmonic bell-like noise, so it can sound like noisy morphing with a lot of cool noisy distortion. "Eleven" the result and you get very interesting and stylish sound.

Edit: Just to clarify: by "bell-like noise" I mean band-passed noise that looks like bell on the spectral analyzer.

+1

Sine waves are the worst for realtime manipulation. Or best if you happen to like weird puffy artifacts

[Urs]

Hmm, by "smoothing parameter" I guess you mean "resolution". Latter implies that Zebra's oscillators are somewhat granular. And that's what they are. If I needed to come up with a buzzword for Zebra's oscillators I'd call it something like "Dynamic Granular Spectrum Synthesis".

Every so often (less with low resolution, more often with high resolution) Zebra calculates a waveform cycle and crossfades it with the previous one. Sometimes this transition can look like artifacts at, say, -60dB when in fact it's just a crossfade between two sine waves of different phase.

As was pointed out, some OscFX can cause dramatic transistions that appear inharmonic or outright chaotic, even though the single waveform looks absolutely perfect.