Zero Latency Limiter plugin?

[pchase]

fircomp2 allows zero latency limiting

[Funky40]

TDR Limiter 6 GE
0.3 ms @ 48khz / in low latency mode
often at sale / $29 -$19,....sometimes $9.99 i suppose

[TB-ProAudio]

LA xLimit IV provides low latency modes (min 0.3ms@SR 48kHz).

[Medenka]

The most important consequence in practice is that if the signal exceeds the threshold for more than just a single isolated peak, then this sort of limiter can still sometimes pass some of the waveform as long as there is a higher peak first to trigger the GR.

The most important consequence is that you get distortion. It can't be a 'single isolated peak' because you would be slamming it, that's what limiters are for. That's why latency is necessary, the plugin needs time to calculate what's coming in and reduce it by how much - and keep it clean. This can't be done instantly.

DC8C has latency if you switch on the oversampling.

So latency = accuracy = audio quality.

[SredTotch]

I had good experience with Unlimited Sonic Anomaly

[SredTotch]

And FL limiter

[simon.a.billington]

AFAIK actual zero latency limiting does not exist - physically impossible. You can have low latency or clippers tho.

Well.. sorta.

One can think of a clipper with an instant attack and instant release. If you want zero latency, then your limiter has to have an instant attack (in the same sense as a clipper), but it can still have a longer release.

So in terms of the waveform, it essentially clips on one side, but it can still preserve the waveform on the other side after the strongest peak. The most important consequence in practice is that if the signal exceeds the threshold for more than just a single isolated peak, then this sort of limiter can still sometimes pass some of the waveform as long as there is a higher peak first to trigger the GR.

You can have a zero latency limiter, but it does cone with a few gotchas.

First thing is zero latency isn't actually true zero in digital audio land. There is always some kind of input buffer being applied. Thats because some operations wont work dealing with only one value at a time.

Gain does, you can in crease or decrease value without needing to know what the neighbouring values are. Frequencies, however is all about cycles per second so you need some value of time (multiple values) before you can work with it. Hence the need for a basic input buffer.

Sos we've already that very basic gain operations can be done instantly, the trouble is limiters aren't a simple increase or decrease in gain, that would only give us a basic volume control. So to make a limiter effectively you need to be able to compare a past value with a current value and add or subtract as a result. Fortunately, you can do that kind of thing because of the input buffer, so it doesn't have to add latency.

The trouble is it's limited in this simple state because what happens if there is a huge jump in values. It will either suddenly become overzealous and suck out those excess levels in a very instant and non-musical way, or it will do a bit more averaging and ramp into these quick changes. It's more musical, but it can't react instantly.

Therefore it is possible to make a limiter without adding extra latency, but it can't act instantly to change, so you wouldn't be able to make a brickwall or ISP limiter. You could make a soft clipper, or a basic har clip style limiter where it may not be quick enough to catch some of those quick, excessive peaks, but it will be good enough to the job you need it to.

That's okay for live use really, it's functional enough without adding extra latency, it can do the job you need it to do well enough. If you want more control, pore precision, more fidelity then that can come at a later stage in production where it really begins to matter and you trying to record or play live.

Logic's basic or Advanced Limiter set to zero latency will do this job fine, as will something like the Waves L1 which gets a lot use in both the studios and live. There will be other options out there too as you tell by the other responses.

[mixtur.se]

Most of PSP’s plugins are zero latency, ex https://www.pspaudioware.com/products/psp-twin-l

[simon.a.billington]

Most of PSP’s plugins are zero latency, ex https://www.pspaudioware.com/products/psp-twin-l

It's more achievable if your plugins aren't oversampling, at least it should give you an option to turn it off because that also adds latency.

[Echoes in the Attic]

Yohng W1 I think is pretty low. V2 been in beta a long time. Haven't used it in a while.

[electro]

Ice9?

[kPere]

The most important consequence in practice is that if the signal exceeds the threshold for more than just a single isolated peak, then this sort of limiter can still sometimes pass some of the waveform as long as there is a higher peak first to trigger the GR.

The most important consequence is that you get distortion. It can't be a 'single isolated peak' because you would be slamming it, that's what limiters are for. That's why latency is necessary, the plugin needs time to calculate what's coming in and reduce it by how much - and keep it clean. This can't be done instantly.

DC8C has latency if you switch on the oversampling.

So latency = accuracy = audio quality.

Fircomp2 does a great job at it though

[simon.a.billington]

[double post]

[@l3x!5]

X limit from SSL is zero latency and very good

Don't know about earlier versions, but SSL X-Limit v1.2.2 reports 46 samples of latency at 48 kHz on Cubase Pro 15.

My bad, for some reason that's what it shows in the plug-in manager, while it is indeed zero latency on track when TP and look-ahead are disabled.

[oobesan]

Maybe I don’t understand clearly enough, but turning off lookahead to reduce latency (all the way down to 0 ms) means your limiter will miss incoming peaks above the threshold. Unless you have a limiter that allows you to set the attack to instant. Which, at that point, you’ve turned your limiter into a hard clipper, and you get wave shaping distortion. Tell me how I’m wrong.