A very newbie question:
In a track (a kick drum) I put consecutively two plugins:
a Compressor used as a limiter at -6 dB
a VU Meter
Like this:
Why, if in the compressor I set the limit at -6dB, on the meter do I see the pointer going into the red zone (+1, +2 dB)?
Shouldn’t it stop at -6dB?
Thank you,
a.
No, dbFS and VU aren’t the same
The compressor uses dBFS – digital peak level in decibel relative to Full Scale.
You want to compare this to a digital peak meter.
VU meters are different. They do not measure signal level (voltage), but as the name “VU” implies signal power [1].
A short overview is presented at The Ardour Manual - Metering in Ardour
The following image from https://x42-plugins.com/x42/static/meters.pdf may help:
–
[1] The reading shall be 0 VU for an AC voltage equal to 1.228 Volts RMS across a 600 Ohm resistance. Strictly speaking the VU meter shows signal energy, since the meter averages power over time.
Thanks, all clear.
So I keep an eye on the meter that is present in the compressor itself.
Bye,
a.
Yep. Except I would not trust the calf GUI. calf plugins are known to show different things in the GUI compared to what actually happens.
Robin may not want to self-promote, but in case you were looking for another compressor to use in place of the CALF one, he has one available at his site:
If you need finer-grained control, the LSP Plugin suite also offers a compressor:
Thanks @GuntherT – personally I do use the compressor I wrote, LSP is also a good choices. And well, if you like calf’s sound then by all means use it, just trust your ears more than their GUI.
Keep in mind that a compressor is not a limiter, even if you set the ratio to infinity and the curve looks like nothing passes the -6dB.
According to your screenshot you have an attack time of 20ms. Simply speaking this means: every time the kick drum makes “boom” with a peak level higher than -6dB (otherwise the compressor wouldn’t do anything, and I’m ignoring the knee here for the sake of simplicity), the compressor will take 20ms until the gain reduction is ramped up to trim down the output to -6dB. This means, the initial transient “b” of the “boom” will more or less go through with the original level.
If you want to keep sure, that not even a very short transient is slipping over the -6dB, you have to use a “real” limiter plugin, e.g. from x42, LSP, or Calf…
Actually a limiter IS a compressor, just a compressor with specific settings. Generally anything over 10:1 ratio is considered a limiter.
The rest of your post is correct however.
Seablade
That’s not true. Every compressor has the reaction time. It doesn’t start to compress immediately, otherwise we will hear distorted signal. If it does not start to compress immediately, then we can not guarantee that short peaks will be properly gain-reduced to desired level.
That is true of limiters as well. The difference is that often (Not always) digital limiters have a lookahead built in which obviously increases latency as well. But that is not always the case either, for instance the final limiter in Mixbus can have it’s lookahead bypassed, and yes if you do this it does sound distorted when driven, just like any compressor with to short of an attack.
Analog limiters, where all this originated, did not have this capability in the same sense.
Seablade
EDIT: So I was reminded technically now you are bypassing the limiter itself in Mixbus, not the lookahead, so I struck that out above. The point still stands though, a limiter is a compressor, in the digital world it has had a lookahead as part of standard practice for it now to improve performance in terms of attack, but is is still just a compressor. In the analog world no such practice exists for the record, at least not commonly that I know of.
The point still stands though, a limiter is a compressor, in the digital world it has had a lookahead as part of standard practice for it now to improve performance in terms of attack, but is is still just a compressor.
Not exactly. Compressor is the only one way of how the limiter can be implemented. For example, in LSP we have two-stage limiting: the first stage (ALR) acts as a compressor and compensates extra gain at the input. But the second stage behaves differently. It searches for peaks over the defined threshold and applies corresponding gain compensation ‘patches’ to the data collected in the lookahead buffer.
So, actually, no, limiter is not a compressor.
Got it…
Ok compressor…
How is this not a compressor? A compressor is a ‘dynamic range compressor’. It will apply gain compensation to anything over a set threshold to reduce it’s gain according to set parameters, in your case parameters set to not allow a peak over that threshold.
So really your first statement could be rewritten to, you have two stage compression with a lookahead buffer to accomplish the limiting.
Seablade
Your statement is clear. You operate on the ‘Compressor’ term as a definition of some device that ‘just compresses’ the dynamics range. Very common definition. Not completely agree but OK. There is no need to argue about some terms, it is much more productive to reach the agreement about these terms.
For me, the ‘Compressor’ is radically another device compared to the Limiter or distortion/saturator because of implementation. That’s why @atux could not reach the desired effect by applying Calf Compressor.
And yep this is exactly how another conversation on this topic I was having in chat ended as well, to paraphrase where that ended:
‘A limiter w/o lookahead is a compressor, and a compressor with lookahead can be a limiter’
Obviously applies to the digital world more than the analog world in that, but both statements are true, end result though is that a limiter really is a compressor, but in digital it often has a lookahead buffer associated with it. In the analog world this doesn’t really exist (Yes theoretically could be done, but I can’t think of one off hand) but you still had limiters, that were just compressors with specific settings.
And yes implementation can vary, and that is how you get differences in different plugins and equipment, and I am not saying that there can’t be more to it, but in the end it is a compression algorithm in one form or another that is driving all of this.
Seablade
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.