deleted by uploader

[MyCo]

AGC / Limiter / Expander / Compressor are internally all the same. In fact most of them aren't standalone effects because they can just swap a comparison and you have the other type of effect.
All of them measure amplitude/loudness and adjust to that. Limiter adjusts only when higher, expander adjust only when lower. AGC is a Limiter / Expander (same effect, just with different comparison) without comparison at all. However it needs a dead zone so they don't adjust to noise levels. That's the point where AGC can get extremely complicated, because you have to find out if the current input is noise or not. There is no such thing as "standard noise floor". In fact the plugin I showed above can determin what is noise at any level up to -20dB below signal.

And no AGC is nothing different, it's just the most stupid compressor with a noise detector. But there are a lot of chances to go really over the top. eg. one of this is: how to adjust quickly to spikes without any noticable distortion. You can just go linear/log over time, and in most cases this is abolutely enough... or find a complicated, time consuming,, experimentell (and I think it's even patented now) way...

[mrking]

AGC / Limiter / Expander / Compressor are internally all the same. In fact most of them aren't standalone effects because they can just swap a comparison and you have the other type of effect.
All of them measure amplitude/loudness and adjust to that. Limiter adjusts only when higher, expander adjust only when lower. AGC is a Limiter / Expander (same effect, just with different comparison) without comparison at all. However it needs a dead zone so they don't adjust to noise levels. That's the point where AGC can get extremely complicated, because you have to find out if the current input is noise or not. There is no such thing as "standard noise floor". In fact the plugin I showed above can determin what is noise at any level up to -20dB below signal.

And no AGC is nothing different, it's just the most stupid compressor with a noise detector. But there are a lot of chances to go really over the top. eg. one of this is: how to adjust quickly to spikes without any noticable distortion. You can just go linear/log over time, and in most cases this is abolutely enough... or find a complicated, time consuming,, experimentell (and I think it's even patented now) way...

Lol ok. USA have made actually noisefloor standards which world follow. Infact every dsp effects follow noisefloor standard and internally not everything is same, some limiters made up by macro power ratios, and agc design is not correct. Anyhow chill bro. Sorry if i hurt ur feelings bro.

[MyCo]

USA have made actually noisefloor standards which world follow.

Hahaha... nice try. Must be the reason why everyone in the industry is migrating to EBU-R128 which is an European standard When you read the tech specs of EBU-R128 you'll notice that they came to the conclusion there is no noisefloor standard (as they would have used it, if it existed). That's why they proposed a -10LU relative gate to cancel out the adjustments to noise. This means everything that drops 10LU (=10dB) below the ungated measured loudness is considered noise and therefor not taken into account.

[martinvicanek]

Mrking, you seem to have knowledge on the subject that we don't. It would be interesting to see it in practce. Would you like to share a demo schematic?

[mrking]

Mrking, you seem to have knowledge on the subject that we don't. It would be interesting to see it in practce. Would you like to share a demo schematic?

Well, no doubt you guys are great programmers, and LU is different when avoiding Noise issuses, observe & try to understand dear friend, NOISE FLOOR is "DYNAMIC RANGE". -20 dB means 20 Hz equivalent, so -40 dB is 40 Hz, which was suggested in USA, if you observe rather arguing my friend you will notice GROUND/LOOP/CHASIS/TRANSISTORS (lots of types) have their own noises, that is why -40 dB was used. Of course you will not believe. But no problem, I just work in this industry, I'm familiar with lots of standards & nobody in Asia, recognize european standard, 90% asian market use japanese standards or american standards, works very well. AGC is all about MMMMMMAAAAAGGGGGNNNNNNIIIIIIIITTTTTTUUUUUUUDDDDDDEEEEEEEE.

Of course I will see, if i get time to make one, I use FLOWSTONE for engineering based tools so I use it lesser and only for my needs for small collections or big collections of valuable tools for customers. I am not a good programmer in it, since I am java based programmer. Like I said I can try and see when I can make one. You guys can optimize it, I see lots of guys wrong coding done here(only few), yet there are my favs like acrobat & others.

[tulamide]

I try to ignore the arrogant tone of the post and concentrate on the content. Could anyone please explain this sentence to me?

NOISE FLOOR is "DYNAMIC RANGE". -20 dB means 20 Hz equivalent, so -40 dB is 40 Hz,

It confuses me. I'm no audio engineer or the like, so it's not easy to understand the meaning of this. For example, 'NOISE FLOOR is "DYNAMIC RANGE"'. I learned that dynamic range is the difference between the loudest possible undistorted signal and the RMS. Further, noise floor was known to me as a signal strength, which is the sum of all noises. So, how can a certain strength be a range at the same time?

Even more confusing is the second part. dB is a relative measure, -20 dB means roughly a quarter as loud as the original (or roughly 7 times less power). Hz expresses oscillations per second (and makes no assumptions on power or loudness). So, why is -20 dB equivalent to 20 Hz? And why is something that is even less loud (-40 dB) double as fast (40 Hz)? I'm very sorry, if this is something normal for the pros among us, but to me this sounds like saying "two meters are the same as two litre". I just don't get it.

[MyCo]

The whole post doesn't make any sense at all.
1. Noise floor is not dynamic range, noise floor is a signal level/strength. He might mean Signal-Noise-Ratio (SNR) that is in the broadest sense a dynamic range
2. AGC is not about Magnitude, AGC is about perceived loudness.
3. decibels have nothing to do with frequencies. dB is just a ratio
4. Noise floor has absolutely nothing to do with electronics. Just open your window and you'll notice a different noise floor than the one you have with closed window. It is just the loudness of the environment that you can not control.

The fact that he doesn't provide any links to "the noise floor standard" he is claiming to exist, tells you all.

And the "no one cares about european standards" statement is just ridiculous. There is not a single professional audio gear, that's not using AES/EBU <-- see E in EBU means European and as far is I know Japanese use that too and don't have any own standards.

But hey... who am I... he claims to be a professional in the industry, so he must be right.

[RJHollins]

Hi Tulamide,

I've been a Professional Audio Engineer for over 30 years. Including formal education/degrees.

I've read the post, but as yet, I'm am not following the statements. I think it may be something of a language/translation issue that is adding to this.

Upon first read, beside the BS arrogance which we could do without [thank-you], some of these statements sound mis-informed at best ... or plain wrong.

Again, I yield that this could be language misinterpretation. If he comes back with a clearer explanation, I'll look it over ... or it may hit the 'IGNORE' list if worthy.

[Perfect Human Interface]

I haven't been following the thread too closely, but I will chip in briefly. It's not incorrect to say noise floor is "dynamic range." The noise floor in an audio signal chain dictates the potential dynamic range. In other words, it's the dynamic range of a system vs. the dynamic range of an actual audio signal. Dynamic range of a system is the difference between the noise floor and the clipping/saturation limit.

That said I don't know how it fits into this context exactly.

Aside from that, I will say that I would think if you were going to be super picky about the perceived quality of an AGC program then utilizing some kind of lookahead would probably be important. That's the only way to account for fast changes in volume without things like brickwall limiting and associated distortions.

[RJHollins]

Personally, I would think an AGC is purely a GAIN control that looks to maintain either a average RMS or a PEAK ceiling. I do not believe it should affect the 'Crest factor', changing the dynamic range of an audio signal.

There are the limiter and compressors for that.

Point in case. There are many plugins that use -18db RMS as a nominal operation level. There have been several 'gain structure' control plugins that look to help balance these levels. Ian Shepard's, 'Perception' is one of them. There is a free JS plugin for REAPER that also provides this function.

In essence, these 'support' plugins help maintain UNITY gain across a plugin or a chain. With their built-in BYPASS function, allows a more accurate A/B comparison of how the plugin effect [or chain]. Helping to make a balanced sonic decision MINUS the level change influence.

It would be nice to have this in a FS module that also uses the EBU-128 standard.

Just my opinion. I already own the Perception plug AND use the Reaper plug as standard routine. Very, very useful.