BuzzCut (The Erosion Clipper)

[LusiD_Music_UK]

Being the most average among all the average Joes - while making a living with these kind of tools - the only thing I can say is that the results I get from using this clipper sound more open and "clean" than any of my - much more expensive - clippers.

I´m not using it for extreme gain reduction and mostly with 16x os, though...

Love that! A lot of people saying they can get an extra dB or so before it starts choking. Which was the idea (being a degenerate bass music producer occasionally as high as 0LUFS at times). Happy it's delivering for you!

[xx JPRacer xx]

After reading this, for me it's like:
- "Hey, try this deodorant, your armpits will smell good."
- "No, it won't. Only washing them will clean them."
- "Ok. I'm not saying they'll be clean, but it will mask the bad smell".
- "No. Only cleaning them will."
- "But masking the bad smell with a better smell is better than nothing and you can still wash them if you want."
- "No."

* English is not my native language.

[martinvankoek]

Since when is calling out a developer on their deceptive marketing and flawed DSP claims (they still refuse to drop) considered bullying? I don't think you were bullied, you just got caught.

[pekbro]

I think it’s amazing how many dsp experts we have around here who are still only speculating without seeing the actual code.

[El°HYM]

Lovely 1st post and you waited a year to drop it.

[martinvankoek]

I think it’s amazing how many dsp experts we have around here who are still only speculating without seeing the actual code.

Sorry if I've missed something but I haven't seen a single speculative post in this entire thread from anyone regarding the DSP going on in this plugin. The developer already openly talked about their technique so there is no need for any speculation, since there is no secret algorithm being discussed.

From what I've read, the critique was about the mathematical reality of what their technique isn't capable of doing which is in contrast to what was claimed. There was also the very obviously intentional and misleading "apples to oranges" graph on their website, which doesn't require speculation either. The funniest bit for me was that even after the developer finally conceded to the "misleading graph" point, they still had to blame the other guy/girl of bullying.

[Andreya_Autumn]

Come on, the dev backpedaled on that one thing, fairly. That doesn't change the fact the critique was based on a combination of two things:
A somewhat half-baked understanding of the nyquist theorem.
A pedantic interpretation of the copy text.

"Bullying" is a bit strong maybe, but it is pretty hostile in my opinion.

Again, the devs claims *are* true. Replacing (yes, replacing. And yes, even without oversampling) the reflected harmonics with noise is in fact what happens. It's what the nyquist theorem predicts should happen, and it does! Try the plugin yourself to find out.

Is resulting sound is pretty darn aggressively noisy? Yup! Is it "better quality" than just oversampling the clipper. Absolutely not. But if you try the thing for yourself it'll immediately become obvious to you that a clean high-quality sound isn't the goal. I personally won't buy it though I did try it and think it's good. I won't have much use for it for the music I make. But the reception the dev got is undeserved. It's a good product and does do what it says on the tin (unless you are committed to a pedantic read of the "eliminate aliasing wordings).

[pekbro]

That guys been nothing but nice to people, giving away way more than original small number of free lics.

As for here I only saw some douche attack him call him a liar while dismissing his responses.

[Zoopy]

Plugin sounds good btw

[martinvankoek]

Come on, the dev backpedaled on that one thing, fairly.

Oh sure. I already acknowledged their concession on that point.

"Bullying" is a bit strong maybe, but it is pretty hostile in my opinion.

Having read the whole thing again now, Zeisner's tone could have been softer indeed.

...A pedantic interpretation of the copy text...
...and does do what it says on the tin (unless you are committed to a pedantic read of the "eliminate aliasing wordings).

See, I don't think my takeaway requires a "pedantic" reading. The marketing language on the website really naturally reads as "solving" the aliasing problem. For example:

The Problem with Standard Clippers
Standard hard clipper VSTs have a fatal flaw: Math. When you drive your signal into a hard ceiling, traditional plugins introduce digital aliasing, crush your high-end, and turn your mix into a harsh metallic mess.

The Solution: Erosion
BuzzCut is the world’s first "Erosion Clipper". It modulates the clipping threshold with shaped noise. This replaces inharmonic foldback aliasing with superior "analog noise" (much like how dithering masks truncation).
Finally, you have the freedom to choose: "digital aliasing" or "analog air".

I think the most straightforward reading here is: "Standard clippers have a fatal flaw(aliasing). We have a solution: Buzzcut (which implies the aliasing problem is solved)

Add to that the unfair apples-to-oranges comparison graph (later acknowledged), the whole presentation becomes pretty shady.

Again, the devs claims are true. Replacing (yes, replacing. And yes, even without oversampling) the reflected harmonics with noise is in fact what happens. It's what the nyquist theorem predicts should happen, and it does!

I don't really follow the appeal to "Nyquist predicts that replacement" here. Assuming I haven't misunderstood you, I think Nyquist predicts the opposite, that is: if you generate content above Nyquist, foldback is unavoidable at that sample rate. Here, the threshold is modulated and so the nonlinearity becomes time-varying/stochastic. As a result of that we get more broadband stochastic components (along with deterministic ones). So even in the "Nyquist-free world" (continuous time), the stochastic threshold method would still be generating all those components. With Nyquist present (discrete time), any part of both the deterministic distortion and the stoachastic broadband stuff that lands above Nyquist will fold back. The foldback isn't "replaced", everything is still aliasing exactly like before. Any harmonics that land above Nyquist will also "reflect" exactly as they do with "traditional clippers" too and won't be "replaced". This method only changes what the nonlinearity generates which is: the signal being fed into the aliasing mechanism. It changes absolutely nothing about the aliasing mechanism.

[billinder33]

I demoed this. For now, I'll give it an A for effort/ingenuity and a D for execution.

For reference, I'm the target user for this device - loud Bass Music/Trap/D&B/etc. I run a typical clip-to-zero loudness mix style, with TBPAclip as my mix and summing buss clipper and Krafter as my 2-buss/mastering clipper. I've got my production techniques refined to the point where I'm generally hitting -5.5 to -3.5 LUFs on drops before any master processing, so "more loudness" isn't really a problem I'm trying to solve. But I'm always open to new ideas, so wanted to give this one a try.

On the positive side, this BuzzClip seemed to have a nice rounding of transients that I've never heard in other clippers. I think this would be the main selling point and why some people claim it sounds more "analog" or "3D" than other clippers. Not sure if that's the 'erosion' or the interaction of all the processing stages, but that softening sound is very compelling.

On the negative side, lots of issues here. First, I don't find the controls all that intuitive and a decent manual would help. I rarely feel like I need a manual, but this one is an exception. I'd like to see detailed descriptions about how each control is implemented, because the tool tips aren't enough. It's not entirely clear what the internal signal flow is. Or what the interaction between the soft clipper, erosion, and hard clipper is. Or what all this means as it is represented by the visualizer. The visualizer is confusing - it looks like the erosion is happening in db range between the soft clip curve and the hard clip cutoff? But that logically doesn't make any sense in my primitive brain.

Second, even with True Peak engaged this plugin bleeds peak overs. I'm not even sure the TP actually works at all. Not good.

Third, it crashed Bitwig 4 times before I finally threw in the towel. Twice when changing the oversampling, once when messing with the noise function, and once seemingly for no reason at all. Granted, Bitwig 6 is a real picky ass and crashes a lot, but with BuzzCut the crashing is next level.

Finally, while the rounded transient sound BuzzCut provides is nice, it's also a lot of work and tinkering to get all the levels right. You can do something similar in Darkplace Sloth (admittedly not at all the same type of process) with basically one click-drag. IMO BuzzCut is a lot of work for a relatively minimal payoff. I'd probably be into this plugin if it was 1 or 2 quick moves to paydirt, but there are just too many things to keep an eye on when working with it - like it needs a 'safe' mode.

The improvements I would make (manual notwithstanding) is fixing the true peak, and coming up with a better set of visuals. Maybe separate visuals and metering for the soft clipping, erosion, and hard clipping processes. I find it really hard to understand what's going on with everything just jammed together and no explanation. Where is the signal in relation to the soft clipper? Am I pushing into the erosion via output from the soft clipper, or adding it into the signal? Or both? What are the levels moving as the signal moves from one stage to another?

There's a really good idea and some very nice DSP hidden inside BuzzCut, but this one feels like a prototype at this point. I wish the developer luck because it's really hard to bring a genuinely new idea into this crowded space, but he did it. I'll definitely be keeping an eye on the evolution of this one.

[LusiD_Music_UK]

Since when is calling out a developer on their deceptive marketing and flawed DSP claims (they still refuse to drop) considered bullying? I don't think you were bullied, you just got caught.

What claims am I making that I "still refuse to drop"?

I never claimed oversampling wasn't used or isn't required.
I showed "worst case scenario" vs "best case scenario" to market the plugin because it looked better. That's as far as my thought process went here.

You're interpreting the decision to show a better comparison on my website (because marketing) in the least charitable way possible.

[LusiD_Music_UK]

Plugin sounds good btw

Thank you! <3

[LusiD_Music_UK]

I demoed this. For now, I'll give it an A for effort/ingenuity and a D for execution.

For reference, I'm the target user for this device - loud Bass Music/Trap/D&B/etc. I run a typical clip-to-zero loudness mix style, with TBPAclip as my mix and summing buss clipper and Krafter as my 2-buss/mastering clipper. I've got my production techniques refined to the point where I'm generally hitting -5.5 to -3.5 LUFs on drops before any master processing, so "more loudness" isn't really a problem I'm trying to solve. But I'm always open to new ideas, so wanted to give this one a try.

On the positive side, this BuzzClip seemed to have a nice rounding of transients that I've never heard in other clippers. I think this would be the main selling point and why some people claim it sounds more "analog" or "3D" than other clippers. Not sure if that's the 'erosion' or the interaction of all the processing stages, but that softening sound is very compelling.

On the negative side, lots of issues here. First, I don't find the controls all that intuitive and a decent manual would help. I rarely feel like I need a manual, but this one is an exception. I'd like to see detailed descriptions about how each control is implemented, because the tool tips aren't enough. It's not entirely clear what the internal signal flow is. Or what the interaction between the soft clipper, erosion, and hard clipper is. Or what all this means as it is represented by the visualizer. The visualizer is confusing - it looks like the erosion is happening in db range between the soft clip curve and the hard clip cutoff? But that logically doesn't make any sense in my primitive brain.

Second, even with True Peak engaged this plugin bleeds peak overs. I'm not even sure the TP actually works at all. Not good.

Third, it crashed Bitwig 4 times before I finally threw in the towel. Twice when changing the oversampling, once when messing with the noise function, and once seemingly for no reason at all. Granted, Bitwig 6 is a real picky ass and crashes a lot, but with BuzzCut the crashing is next level.

Finally, while the rounded transient sound BuzzCut provides is nice, it's also a lot of work and tinkering to get all the levels right. You can do something similar in Darkplace Sloth (admittedly not at all the same type of process) with basically one click-drag. IMO BuzzCut is a lot of work for a relatively minimal payoff. I'd probably be into this plugin if it was 1 or 2 quick moves to paydirt, but there are just too many things to keep an eye on when working with it - like it needs a 'safe' mode.

The improvements I would make (manual notwithstanding) is fixing the true peak, and coming up with a better set of visuals. Maybe separate visuals and metering for the soft clipping, erosion, and hard clipping processes. I find it really hard to understand what's going on with everything just jammed together and no explanation. Where is the signal in relation to the soft clipper? Am I pushing into the erosion via output from the soft clipper, or adding it into the signal? Or both? What are the levels moving as the signal moves from one stage to another?

There's a really good idea and some very nice DSP hidden inside BuzzCut, but this one feels like a prototype at this point. I wish the developer luck because it's really hard to bring a genuinely new idea into this crowded space, but he did it. I'll definitely be keeping an eye on the evolution of this one.

Just thought i'd copy my response from Gearspace here too:

Hi mate! Thanks for this amazing (hardball) review. I'll send a couple of 50% codes your way!

Really nice to hear you like the sound of the clipping and how the multiple stages interact! That's the "main thing" (sounding good) so I'll take it! lol

First thing that sticks out! "True Ceiling" is not "True Peak Limiting" and I can see how that's easy to misunderstand. I may need to adjust the tooltip to clarify that. (and a manual will be here in a few days) It's simply a ceiling that leaves the minimum necessary 0.3db of headroom for the antialiasing filter so that oversampling isn't undermined that way "what you see is what you get".

And thanks for the Bitwig heads up. I'll need to look into that! Embarrassing.

For me that quick payoff is select the $120 preset and then push the input gain to taste (linked) if that's any help?

So the signal flow goes left to right.
Soft, Buzz, Hard, Clamp

Thank you! I appreciate the balanced feedback! It will go a long way to making 1.1 better <3 Manual in a few days

[LusiD_Music_UK]

Come on, the dev backpedaled on that one thing, fairly.

Oh sure. I already acknowledged their concession on that point.

"Bullying" is a bit strong maybe, but it is pretty hostile in my opinion.

Having read the whole thing again now, Zeisner's tone could have been softer indeed.

...A pedantic interpretation of the copy text...
...and does do what it says on the tin (unless you are committed to a pedantic read of the "eliminate aliasing wordings).

See, I don't think my takeaway requires a "pedantic" reading. The marketing language on the website really naturally reads as "solving" the aliasing problem. For example:

The Problem with Standard Clippers
Standard hard clipper VSTs have a fatal flaw: Math. When you drive your signal into a hard ceiling, traditional plugins introduce digital aliasing, crush your high-end, and turn your mix into a harsh metallic mess.

The Solution: Erosion
BuzzCut is the world’s first "Erosion Clipper". It modulates the clipping threshold with shaped noise. This replaces inharmonic foldback aliasing with superior "analog noise" (much like how dithering masks truncation).
Finally, you have the freedom to choose: "digital aliasing" or "analog air".

I think the most straightforward reading here is: "Standard clippers have a fatal flaw(aliasing). We have a solution: Buzzcut (which implies the aliasing problem is solved)

Add to that the unfair apples-to-oranges comparison graph (later acknowledged), the whole presentation becomes pretty shady.

Again, the devs claims are true. Replacing (yes, replacing. And yes, even without oversampling) the reflected harmonics with noise is in fact what happens. It's what the nyquist theorem predicts should happen, and it does!

I don't really follow the appeal to "Nyquist predicts that replacement" here. Assuming I haven't misunderstood you, I think Nyquist predicts the opposite, that is: if you generate content above Nyquist, foldback is unavoidable at that sample rate. Here, the threshold is modulated and so the nonlinearity becomes time-varying/stochastic. As a result of that we get more broadband stochastic components (along with deterministic ones). So even in the "Nyquist-free world" (continuous time), the stochastic threshold method would still be generating all those components. With Nyquist present (discrete time), any part of both the deterministic distortion and the stoachastic broadband stuff that lands above Nyquist will fold back. The foldback isn't "replaced", everything is still aliasing exactly like before. Any harmonics that land above Nyquist will also "reflect" exactly as they do with "traditional clippers" too and won't be "replaced". This method only changes what the nonlinearity generates which is: the signal being fed into the aliasing mechanism. It changes absolutely nothing about the aliasing mechanism.

I'd like to know if you'd take issue with the following since the concept is the same:

Dithering replaces truncation distortion with superior "analog noise".
Finally, you have the freedom to choose: "digital aliasing" or "analog air".

That seems 100% accurate and fair to me.
(I acknowledge "analog air" could sterilized and described as "noise" but obviously it's designed to be marketing copy, not a DSP thesis)