[Release] DephazEAudi0 — real-time geometric audio processor

[Cuauhtli]

Holy moly. Someone needs to lay off of the lysergic acid....

[whassup]

del

[tonycore]

Any chance of a Mac compatible build?

[dephazeaudio]

Yes — a macOS build is already complete and currently in final testing.

It will be included in the Gumroad package very soon alongside the Windows version.

An Android build is also in progress as an experimental platform release.

Thanks for the interest — platform support is expanding shortly.

[bmanic]

One thing that is apparent, you need a final output volume control in the plugin. Otherwise it'll be super annoying to use.

[dephazeaudio]

Good point — from a workflow perspective I completely understand why an output trim is convenient.

The processor itself isn’t designed around level shaping; it operates structurally on the signal, so the internal behavior is intentionally independent from a traditional “makeup gain” stage. In normal use the host trim or track gain handles final level, which keeps the processor mathematically consistent.

That said, usability always matters. If enough users feel an integrated output trim would improve workflow without compromising the processing model, it’s something I’m open to adding as an optional stage.

Thanks for pointing it out — practical workflow feedback is always valuable.

[Tj Shredder]

Good questions — let me translate the abstract description into something practical.

The plugin isn’t trying to “fix” a traditional audio problem…

Its still abstract, but the web site says its a “Solution to Astrodynamical Anomalies”
That explains it more precisely:
Its either a psychological test if its possible to fool the KVR community to believe its valuable if you just burry the description into technical sounding meaningless words, or its an experiment of an extraterrestrial intelligence to understand what the hell is happening on planet earth…

[jamcat]

If you can’t describe what it actually does, it’s because it doesn’t actually do anything.

[dephazeaudio]

To add some context — everyone who has actually spent time using the plugin so far has reported finding it useful in practice. That doesn’t mean everyone will feel the same, of course, but real-world feedback has been consistently positive.

Practical experience matters more than abstract discussion. If there’s curiosity or skepticism, I’d encourage talking to users who’ve worked with it or trying it firsthand and forming an opinion based on what you actually hear.

And if someone prefers a more technical angle, there’s nothing stopping you from analyzing the processed signal — spectrum measurements, waveform comparisons, or A/B renders are all valid ways to explore what the processor is doing.

At the end of the day, conclusions based on listening and measurable observation tend to be more meaningful than theory alone.

[Surgo]

This is what would have happened if TimeCube guy had access to Claude Code.

[Debutante]

Good questions — let me translate the abstract description into something practical.

The plugin isn’t trying to “fix” a traditional audio problem like compression or EQ. The idea is to apply a nonlinear geometric transform to each sample, which subtly alters phase and amplitude relationships. The audible result is mostly about perceived spatial coherence and texture rather than loudness shaping.

There’s no lookahead, buffering, or spectral processing — it’s strictly sample-by-sample math. So the character comes from how the waveform is reshaped in real time, not from analysis or prediction.

GAIN is not a simple input gain stage. It controls the strength of that transform. At lower values the effect is subtle; at higher values the waveform deviation becomes more audible.

Stereo spread is implemented via mid/side scaling after the transform. It’s independent of the gain control — they’re not chained in a typical “gain → stereo” way.

If both parameters are effectively neutral, the output approaches the input, but because this isn’t a linear pass-through stage, it won’t null perfectly like a bypass.
Hope that clarifies what the plugin is actually doing under the hood.

NVM

[bermudagold]

To add some context — everyone who has actually spent time using the plugin so far has reported finding it useful in practice. That doesn’t mean everyone will feel the same, of course, but real-world feedback has been consistently positive.

Practical experience matters more than abstract discussion. If there’s curiosity or skepticism, I’d encourage talking to users who’ve worked with it or trying it firsthand and forming an opinion based on what you actually hear.

And if someone prefers a more technical angle, there’s nothing stopping you from analyzing the processed signal — spectrum measurements, waveform comparisons, or A/B renders are all valid ways to explore what the processor is doing.

At the end of the day, conclusions based on listening and measurable observation tend to be more meaningful than theory alone.

I apologize you had to experience kvr default state...juvenile trolling...rational behavior and cordial objective discourse can be hard to come by...there is some modicum of value if you put on ur rubber boots and wade through the sewage...that said

"DephazEAudi0 processes audio as geometry, in real time. DephazEAudi0 guides the signal back toward the geometric coherence it had at the moment of generation. Not by reconstruction, but by removing the structural drift introduced by digital processing."

So is this meant to undo the subtle waveshaping distortion induced by dsp processes that came before it to the original waveform shape?...if its placed as last plugin, how does it know where t=0 in signal chain or "moment of generation"?

[dephazeaudio]

Good question — the wording can make it sound like the processor operates in time, but mathematically it doesn’t.

There is no timeline reference, no stored history, and no attempt to reconstruct a past waveform. The processor applies a deterministic sample-wise mapping:

  y = f(x)

Each output sample depends only on the current input sample and fixed parameters. There is no memory, no look-ahead, and no notion of a signal origin or t=0 inside the algorithm.

So when I describe it as guiding the signal toward “geometric coherence,” that’s a metaphor for how the instantaneous numeric mapping reshapes amplitude relationships — not a reversal of earlier DSP or restoration of an original waveform.

Placing the plugin last in a chain simply means it processes the signal state that exists at that moment. It doesn’t know or care what produced that signal — it applies the same mathematical transform to whatever numeric input it receives.

In short: it’s a local geometric transform, not a temporal correction system.

[Vortifex]

So, in a sentence, using plain non-technical language, what effect does it have on the sound?

[dephazeaudio]

It subtly reshapes the signal so the sound feels clearer, more coherent, and spatially more stable — without obvious coloration or distortion.