TL;DW: I’ve revisited Not Just Another Dither, to add a 16/24 bit switch and DeRez.

NotJustAnotherDither

It’s late and I’ve been cranking away on a whole pile of plugins here, because it turned out I had to revise Beam and Dark too. (re-download them from where you got ’em if you’d like the updates)

This is because I started doing the more ‘normal’ dither plugins, and found that my DeRez code was making them go insanely loud at far reaches of the DeRez control: I put in a clamp so the maximum roar of bitcrush is 12 db down, even when it’s one bit or less. This might adjust the loudness of extreme bitcrush in your projects: any normal use, or use as a ‘simulated old school sampler’, will be unaffected. I’ll be having other dithers come out that need this volume clamp more: so far, Beam and Dark and NJAD gate off to silence so there’s a limit to how loud they can go.

I’d like people’s opinions on whether Dark is a suitable replacement for NJAD, in Monitoring. Though they don’t work exactly the same, and though there are things about NJAD that are pleasing, I still feel that especially at 24 bit but even at 16 bit, Dark is simply better. You can also compare Beam (I’ve updated that with a volume clamp too, re-download it) but I doubt that is quite as good. Each plugin has identical DeRez code, so you can set them identically to compare.

This work is supported by Patreon, and thanks to that I’m able to do many things. Like blow a month’s income on a microphone. Do you like the mic audio this week? If you do, I have some ideas on how I can make a simultaneous recording of that and my lavalier mic (very clean, though more noisy) and then do a plugin to apply the kinds of second harmonic distortion and transformer goodness my fancy pants mic boasts of. I’ll still prefer the real mic… buuuut would you like me to take a whack at the ol’ faux Elam plugin? Apologies for not trying out a real Elam: that would be several years of income, ain’t happening :D also, I keep learning more about what you’d need if you can make a break for the big time, and you know I share all my information. See ya Monday, those who like to compare notes.

TL;DW: Beam is a wordlength reducer that tries to heighten sonority.

Beam.zip(333k)

So here we are again: Beam is a radical, unusual plugin. I’m not certain it works, at least not as intended. I do know that it does what I built it to do: the usefulness of that is more in question.

So, that’s why I’m asking. Be critical, and tell me what you think. I’ve already got Dark and I know that does what I intended, and it’s arguably better than NJAD. I don’t know whether Beam is just a weird experiment… or better still.

What does it do?

Beam’s a wordlength reducer like Dark, but instead of seeking to always give you the least departure from whatever trajectory the audio’s on (like Dark: and that suppresses highs, including noise in the highs), Beam seeks to make all waveforms converge on the same angle (either ascending, or descending). Imagine a world of triangle-waves, the deepest frequencies taking the most energy, and progressively quieter until high frequencies are in balance with the lows. That’s Beam. Beam has a ‘focus’ control that sort of optimizes the frequency range it’s providing a window into (as in, lower settings SEEM to encourage focus on deeper sounds, higher settings SEEMS to highlight the treble, and the 0.5 is set to zero in on where our hearing’s most sensitive.) And it uses those frequencies, across a broad range of possible sounds, to emphasize the energy and depth of the audio to shocking effect.

Except, it’s not. It doesn’t know what a frequency is. It’s only ‘dithering’ in such a way as to try and accentuate certain waveform slope angles, and has no real power to do even that beyond statistical averaging of a bunch of samples.

On top of that, its noise floor is WEIRD. Dark gates into silence, sort of well-behavedly. Normal dithers become noise seamlessly. Beam goes nuts and screams, and the only thing I can say for it is, raw truncation is worse. It’s super weird… and yet, even while it’s doing that, the background audio retains a startling depth and personality. Or at least so it seems… even more than Dark, with this one you turn up DeRez and nothing happens to the sound at all. You don’t even hear the noise until it’s silly loud. (there will be a series of dither reissues with DeRez, both for auditioning and for lo-fi duties.) It might even be a ‘sonic maximizer’, putting tone qualities into the audio that weren’t there to begin with.

I admit I don’t know what to make of this at all. I know how I got it, and I know it’s doing what I made it to do. But nobody knows what you get if you sculpt audio, not by frequencies or loudness, but by reinforcing certain waveform slope angles, because to my knowledge nobody has ever wanted to do that or had a way to make it happen, even in subtle ways like this. And now that I’ve done it, I’m not certain it’s better than Dark. I understand what Dark’s doing, and why that’s useful. This is a wilder beast. Looking forward to getting people’s reactions… because one of these may replace NJAD in my flagship Airwindows plugin, Monitoring.

BTW, I’ve already took apart my Lavry I was griping about, and reseated the ground screw connections (which is what can happen after more than ten years of use: same thing happened with my DA10) and it turned out the funny noises were actually coming from the ‘Transformer’ setting on the AD10, so I’m going back to ‘non-distorted’ while I sort stuff out. It was fun while it lasted :)

This work is supported by Patreon. I hope you like it. The stuff about the CMOS chips and reselling it at cost, is a plan that I have for a future Patreon goal, should I get there.