Kilohearts Plug-ins (My Misunderstanding)

[WackyZoundz]

Denormals haven't been a thing for a long time now. Or did they make a comeback? And if so why not using a much lower noise floor at -600 dBFS or lower (which I did back then)? -150 dBFS would be a bad choice considering how much gain staging can happen in a complex project, it can reach audible ranges quite fast.

[dionenoid]

In Cubase you can choose to show plugins that support 64-bit float processing. There ain't much. So i wonder why the OP only picks Khs from a huge list of plugin devs that don't support it.

Just to name a few which don't support it : Fabfilter, Waves, Softube, UAD, Melda, Tone Projects, Arturia, NI, Izotope, IK, PA, Kilohearts, Sonible, Toneboosters, and so on. Some of the best sounding, and most used, plugins in the industry, which all don't support 64-bit float processing.

In fact, there are only a few plugin devs that support 64-bit. I'd say roughly about 10-15% of everything that anyone has installed. Amongst them are DMG, Sonnox, TDR, Voxengo, Acustica Audio, Steinberg, and a few others. And then not even all plugins that they offer.

So, iow : Don't worry about it.

[limitlesssss]

In Cubase you can choose to show plugins that support 64-bit float processing. There ain't much. So i wonder why the OP only picks Khs from a huge list of plugin devs that don't support it.

Because I love Kilohearts. I was planning to make their 33 snapins my default plug-ins.

Just to name a few which don't support it : Fabfilter, Waves, Softube, UAD, Melda, Tone Projects, Arturia, NI, Izotope, IK, PA, Kilohearts, Sonible, Toneboosters, and so on. Some of the best sounding, and most used, plugins in the industry, which all don't support 64-bit float processing.

FabFilter: Love FabFilter, but I'm not planning to make them my default plug-ins.

Waves: Will never use them. If they were the last plug-in manufacturer in the world, I'd rather use my DAW's stock plug-ins then.

Softube: Will never do iLok.

UAD: Not interested.

Melda: Never used. Don't like the GUI either.

Tone Projects: Lovely, but too complicated. You need a PhD to use Unisum.

Arturia: Not interested in most of their plug-ins. ColdFire is great though.

NI: Love Reaktor and Monark.

iZotope: Not interested, except for Trash 2.

IK: Not interested.

PA: Not interested.

Kilohearts: LOVE KILOHEARTS.

Sonible: Not interested.

Toneboosters: I like TB Compressor for its upward compression capability.

and so on: Mostly not interested.

In fact, there are only a few plugin devs that support 64-bit. I'd say roughly about 10-15% of everything that anyone has installed. Amongst them are DMG, Sonnox, TDR, Voxengo, Acustica Audio, Steinberg, and a few others. And then not even all plugins that they offer.

So, iow : Don't worry about it.

DMG: Great company. I don't like their GUI though.

Sonnox: Great company, but I won't do iLok.

TDR: Are you sure they're 64-Bit FP?

Voxengo: I don't like the GUI.

Acustica Audio: Not really interested.

I hope this answered your question why I'm singling out Kilohearts. Because I love Kilohearts.

[foosnark]

It introduces noise at -150dBFS

If I worried about noise at -150dBFS I would never be able to use any of my hardware.

[limitlesssss]

Just did the same null test with TDR Kotelnikov, TDR Nova and TDR VOS SlickEQ. Absolute silence, negative infinity. Their output gain can therefore be used as a gain plug-in without engaging the compression and eq features.

So it *is* possible. Back to Kilohearts.

[Tj Shredder]

ah well, good you have this obsession, but i dont think adressing this in forum will benefit that much since really your "problem" with synth is your "problem" which is so miniscule i doubt anyone cares

Don't forget that it's not just a hobby for everyone, some make music and record/mix/master for a living. For this group it's a catastrophe to deal with software which performs that poorly. The requirements are much higher if you work with audio in a professional way.

Obviously you are not a professional. Just the idea to boost something by 150 dB shows that not even the basic knowledge for audio engineering (or math) is in place.
Searching for inaudible flaws is a hobby obviously…
The cost of improving something inaudible is too high. As a professional I‘d rather get more tracks in my project… And I prefer also lower cost for the developers. Your ears should judge not your intellect…
Please do not contact Kilohearts support, they can only laugh at you, which would still waste their time…
Don‘t get me wrong, there are areas where 64-bit precision is absolutely mandatory. The time domain and the accessibility of RAM. But all developers who are creating 64-bit plugins cover that correctly…

[kraster]

Just did the same null test with TDR Kotelnikov, TDR Nova and TDR VOS SlickEQ. Absolute silence, negative infinity. Their output gain can therefore be used as a gain plug-in without engaging the compression and eq features.

So it *is* possible. Back to Kilohearts.

I tried the test in Pro Tools.

Sine wave recorded to an audio track. No processing.

On a second track sine wave processed with Kilohearts gain and inverted.

Summed the Kilohearts inverted and original to a new track and normalised the resulting audio file.

Nothing.

Tried it with multiple instances of Khz gain and still nothing.

[SamDi]

This request and discussion is pointless. To reduce my effort it‘s easier to cite a guy from fabfilter, answering a similar request better and more concise, than I ever could.

Link: https://prod.fabfilter.com/forum/topic/ ... processing

We have done extensive experiments with 64-bit processing and we concluded that it is just not necessary to use it as an audio format. You do need it internally in a plug-in for some types of processing. You don't need it when scaling the gain, or adding channels, which is exactly what the mixer in a host application does.

Keep in mind that the floating-point format is very different from a fixed-point format. It's true that you get no more than 23-24 bits of precision, which gives you a dynamic range of around 140 dB. (We never claim to have 32 bits of precision by the way.) That means that when your audio is at the 0 dB level, the tiniest detail that can be represented at the same time is at -140 dB, way beyond the threshold of human hearing which is around -90 dB at its best. But it gets even better, because when the overall volume gets lower, the precision scales with it so the precision actually increases.

With a fixed-point 16-bit format like a CD, you can force quantization noise during low-volume sections of a song (mostly with classical music which has a higher dynamic range), if you increase the volume enough so that loud sections will be far too loud. With floating-point, that doesn't happen because during soft sections, the precision increases and you don't get quantization noise.

Why does Cubase offer 64-bit processing? Because they can, and because people are asking for it. It is overkill, and we don't support it because we would need to have an entire separate chain of audio components for 64-bit audio in our plug-ins which would significantly increase our development, testing and support load. And we are convinced that you wouldn't hear the difference.

Cheers,

Frederik (FabFilter) — Dec 14, 2020

[WackyZoundz]

Obviously you are not a professional. Just the idea to boost something by 150 dB

...is entirely yours because I never wrote that. If you were a professional you would never come up with such a ridiculous idea. And you would know very well how quickly noise can add up when chaining several noisy plugins together, making a clean 24 bit integer signal impossible which is necessary for high-quality music formats, commercial sample libraries and films/video games with high-dynamic range audio. Obviously never worked in any such fields which means you never heard of 24-bit dithering either.

Searching for inaudible flaws is a hobby obviously…

It's not my problem if your hearing is so damaged that you can't differentiate between a signal with 16 and 24 bit integer. It's not my problem that you lack fundamental knowledge about audio engineering. And your stupid comments aren't my problem either because I put you on the block list where all the other wannabes and trolls are. Goodbye.

[mixtur.se]

64 bit double float makes no audible difference, but requires more ram and cpu, for nothing. So instead, all the vendors, please remove 64 bit double float immediately and return to 32 bit.

You measured differences can't be caused by 32 vs. 64bit float. Maybe there is some weird resampling involved? Actually I have no idea what you are doing there... Seems to be negligible for real life, too.

If kilohearts stays 32 bit, very good! Keep fighting the marketing dummies.

64-bit typically requires less or the same ammonit of CPU, but twice the ram to load compiled applications (not data obviously). That’s what the AMD64 instructions are for.

[limitlesssss]

Obviously you are not a professional. Just the idea to boost something by 150 dB

...is entirely yours because I never wrote that. If you were a professional you would never come up with such a ridiculous idea. And you would know very well how quickly noise can add up when chaining several noisy plugins together, making a clean 24 bit integer signal impossible which is necessary for high-quality music formats, commercial sample libraries and films/video games with high-dynamic range audio. Obviously never worked in any such fields which means you never heard of 24-bit dithering either.

Yeah, he does not understand that the reason I boosted the null signal +150dB is to demonstrate how the noise sounds and also at what noise floor it occurs.

[limitlesssss]

Just did the same null test with TDR Kotelnikov, TDR Nova and TDR VOS SlickEQ. Absolute silence, negative infinity. Their output gain can therefore be used as a gain plug-in without engaging the compression and eq features.

So it *is* possible. Back to Kilohearts.

I tried the test in Pro Tools.

Sine wave recorded to an audio track. No processing.

On a second track sine wave processed with Kilohearts gain and inverted.

Summed the Kilohearts inverted and original to a new track and normalised the resulting audio file.

Nothing.

Tried it with multiple instances of Khz gain and still nothing.

I've never used Pro Tools so I'll take your word for it? All I know is I also installed other DAWs to make sure it was not a REAPER thing. I did it in Studio One, and Ardour with the same result which meant REAPER was not the cause. I wanted to test it in Cubase/Nuendo but the installer gave me so much headache that I gave up.

[limitlesssss]

This request and discussion is pointless. To reduce my effort it‘s easier to cite a guy from fabfilter, answering a similar request better and more concise, than I ever could.

Link: https://prod.fabfilter.com/forum/topic/ ... processing

We have done extensive experiments with 64-bit processing and we concluded that it is just not necessary to use it as an audio format. You do need it internally in a plug-in for some types of processing. You don't need it when scaling the gain, or adding channels, which is exactly what the mixer in a host application does.

Keep in mind that the floating-point format is very different from a fixed-point format. It's true that you get no more than 23-24 bits of precision, which gives you a dynamic range of around 140 dB. (We never claim to have 32 bits of precision by the way.) That means that when your audio is at the 0 dB level, the tiniest detail that can be represented at the same time is at -140 dB, way beyond the threshold of human hearing which is around -90 dB at its best. But it gets even better, because when the overall volume gets lower, the precision scales with it so the precision actually increases.

With a fixed-point 16-bit format like a CD, you can force quantization noise during low-volume sections of a song (mostly with classical music which has a higher dynamic range), if you increase the volume enough so that loud sections will be far too loud. With floating-point, that doesn't happen because during soft sections, the precision increases and you don't get quantization noise.

Why does Cubase offer 64-bit processing? Because they can, and because people are asking for it. It is overkill, and we don't support it because we would need to have an entire separate chain of audio components for 64-bit audio in our plug-ins which would significantly increase our development, testing and support load. And we are convinced that you wouldn't hear the difference.

Cheers,

Frederik (FabFilter) — Dec 14, 2020

I've read all their answers as to why they don't go full 64-Bit FP in multiple threads on their website. I wish they eventually change their mind. But that's another story from another developer.

[WackyZoundz]

I'm shocked how few users here understand how summing works. It doesn't take many instances of those plugins to end up with a noise floor that ruins even a 16 bit integer signal.

[limitlesssss]

I'm shocked how few users here understand how summing works. It doesn't take many instances of those plugins to end up with a noise floor that ruins even a 16 bit integer signal.

Yes, and as I said before, I was ready to just accept this situation and use them as my default plug-ins. I said to myself, as long as the accumulated noise of 64 Kilohearts plug-ins won't exceed 24-Bit noise floor (-144.5dB), I'll just won't care and use them.

Why 64 plug-ins? Let's say the average project has 16 tracks (which is very conservative by the way), and the average track has 4 plug-ins on it: therefore 64 plug-ins.

So that was my criterion. But as I showed, even *A SINGLE* Kilohearts Gain plug-in (doing nothing), raises the noise floor to -150dBFS.

It only takes 4 Kilohearts Gain plug-ins (cutting and boosting around 3dB each) for the noise floor to reach -124dBFS which is around 20.7 bits. And after 64 Kilohearts Gain plug-ins (cutting and boosting around 3dB each), the noise floor has already reached -101dBFS which is around 16.8 bits.

So as you said, no it doesn't take much. And keep in mind this was their simplest plug-in just cutting and boosting gain.