Waveform Crossfading

EDIT2: there are other approaches to morph waveforms besides panning, it’s useful if you want to use instrument envelopes to morph tho.

  1. Meta Devices: see post#4 for linear morph (by gova); post#5 for XY vector morph.

  2. Phrases: the new Phrase system has the ability to locate individual samples; hence detailed volume control can be achieved.

  3. Panning + mono mixdown (see original post below)

— original post — (2-way wave morphing)

I read about how wavetable synths work, it occurred to me that a simple crossfade will do it cheaply. That means, if you don’t mind a mono output and only two-variations to morph about, one can use the panning modulation of sampler to play the “morph.”

How to do it:

1. Stick a Stereo Expander set to "mono"(minimize sliders)/"L+R" to the fx chain.

2. Import two samples; at the sample properties box, hard-panning them to 50L and 50R.

3. Use the Panning modulation to adjust/automate your "morph."

You can get creative for the two samples used, some suggestions I came-up with:

  • different single-cycles (old school wavetable morph)

  • different pitch (detuning, chords, addictive waveforms)

  • same sample; different interpolation (ex. mix-in just enough of dirt)

  • same sample; different modulation set (i.e. your dry/wet mix for filters)

  • same sample; one is processed (a dry/wet mix for off-line processors, like CDP)

You can load 12 or a much higher amount of samples with phrases per key

just crossfade those, no need to touch the pan.

Yeah was too exited and forgot about phrases, for fixed stuff it works miles better. Renoise is DEEP.

6147 Screen Shot 2015-10-18 at 7.06.54 AM.png

The phrase will only be necessary if you want to go over the 12 sample per note per instrument limitation.

It is not necessary otherwise

Vector Synth!!

Play the X-Y pad at fx section.

Edit: NM. Cool stuff!

Erm, but morphing and crossfading are different things. This is crossfading, there are vst effects that allowing to really morph sounds. Like prosoniq morph which uses kind of fft vocoder to spectral-crossfade sounds.

The korg wavestation vst can crossfade multiple waveforms in a sequence, they called it “adv. vector synthesis”. Ah, and synthmaster2 of course, too (not sure about sequence though)

Zebrallete is good free morphing vsti for example

Erm, but morphing and crossfading are different things. This is crossfading, there are vst effects that allowing to really morph sounds. Like prosoniq morph which uses kind of fft vocoder to spectral-crossfade sounds.

morph (verb) - to change gradually and completely from one thing into another thing

any kind of crossfading is fine

Just no. For comparison see image morph vs. image alpha crossfade. While morph, also the structure changes.

Or in audio, while crossfade you can identify two sounds more or less clearly, while morph not.

That is straight from a dictionary and crossfading was the way image morphing was achieved before computers came along, and you still crossfade anyway.

Ok call it morphing, if you like. Just wanted to mention that there is also “another kind of morphing”. Morphing also effects pitch for example.

Would be interesting would your dictionary says about “crossfade” :stuck_out_tongue:

There’s a tutorial video by DLR on making bass on FM8. He then uses the x-y to morph between 4 different presets he’s already made and also added some randomicity to the way they “crossfade” between themselves in time and timbre. So, that is some serious morphing.

Should I change the title to “Waveform Morphing” or something? If it’s confusing? How about “Waveform Crossfading?”

“Wavetable Morphing” was considered but two or three samples doesn’t really counts as “wavetable,” though it’s closer to what this thread’s about.

As far as i know, simple waveform cant be morphed.
Look like Serum or Waldorf Nave works, it is different concept.
So i would call this topic “waveform crossfading”

Morphing, crossfading, whatever you want to call it only cooks down to how smooth you can make a transition from one sound to another. Is there really a right way or wrong way?

From Xfer Serums main page:

Serum has a variety of methods and options for analyzing audio for breaking it apart into individual waveforms.(…) Morph between various wavetables using standard linear interpolation (crossfading) or via harmonic/spectral morphing.

From urs’s old response to a kvr thread regarding Zebra:

The synthesis basically swaps back and forth between wavetable, additive and whatever the osc fx do. The end result always is a single, bandlimited wavetable that will be crossfaded with its predecessor. This happens on the fly, depending on demand (modulations) and the Resolution parameter (up to 100 times each second or so).

Hence, to be precise, it’s a form of granular synthesis where each grain is obtained from a spectrum or a wavetable of different sources (spline curves, waves) which get furtherly mangled by various modifiers.

There’s much stuff to morph about these days, FFT, preset settings, wavetables, wave shaper curves, each product has it’s own way to do their own things… however Renoise is a sampler so we play with crossfades :stuck_out_tongue:

btw I think preset morphing is possible (to a degree) with some clever macro mapping and meta devices… that’s another story.

I would say that gain isn’t the only thing you can crossfade :wink:

preset morphing is possible with macros and an LFO for each linked parameter (for a fancier envelope), With tools, since 3.1 this mapping process can be greatly automated, maybe if someone could politely permanently borrow VV’s plugin parameter tool code , it would be almost as simple as moving a knob, could have smart randomization here too