Leveling Amps; Fact or Fiction?

[Music Engineer]

Any chance you could "polish" or "verify" the Leveller beta you posted here and make a "safe" (release) version of the very same plugin? I think it "just works" as-is.

the project is so old that it is not compatible with my current development environment anymore, so i would have to start a completely new project to change anything. and my current environment also demands that i implement my own GUI. so the *very same* plugin has to stay as is. but i just checked the code and seems pretty safe to me.

[brok landers]

robin, thanks for the limiter!
i can hear bmanic running here with big steps!
i still love your phaser as one of the best ever made as a vst ...
greets to berlin!

[Music Engineer]

robin, thanks for the limiter!

or leveller?

i can hear bmanic running here with big steps!

ah, that's the dynamics guru here on KVR, right?

i still love your phaser as one of the best ever made as a vst ...

whoa, i almost feel a bit for all the appreciation around here

greets to berlin!

yup, greets to weil as well

[brok landers]

robin, thanks for the limiter!

or leveller?

stupid me ... of course, leveller ...

i can hear bmanic running here with big steps!

ah, that's the dynamics guru here on KVR, right?

yes, one of the few including me, hehe ...

i still love your phaser as one of the best ever made as a vst ...

whoa, i almost feel a bit for all the appreciation around here

well, honour to whoom ... etc ...

greets to berlin!

yup, greets to weil as well

no, i'm back to stuttgart since 1 1/2 years now (will do a new website soon, will change the adress then) ... the distance to the studio was simply too long ... every day 80km (40 there, 40 back) ... the gasoline prices here are nuts ...

btw, your compressor needs a on/off switchable 0db brickwall limiter (not a saturator) at the very end of the signal chain ... imo ...
that's the only reason i don't use it ... but i'd really like to, as it smacks the hell out of snares, when the attack is a bit open, love it ...
but i can't reach the rms level of the unprocessed signal again, as the transient already shoots to 0db, so i have to always use another limiter after the compressor ... which sucks ... well, it doesn't, but i'd prefer to have it on one gui, simply driving the output of the compressor into the limiter ...
is that possible?

[Music Engineer]

btw, your compressor needs a on/off switchable 0db brickwall limiter (not a saturator) at the very end of the signal chain ...

i didn't really look deeper into the limiter business yet, as i regard it actually just as compression with extreme settings. and then the question arises: what should those settings exactly be? i ideally would let the user decide...and then i would basically have two compressors (with their full parameter sets) in series in one plugin. also, strictly speaking, true brickwall limiting whithout instantaneous attack is possible only with noncausal algorithms which means lookahead and therefore latency.

but i can't reach the rms level of the unprocessed signal again, as the transient already shoots to 0db, so i have to always use another limiter after the compressor

mmm, isn't the whole point of compression to actually raise the rms level? i mean, a compressor attenuates transients such that the overall level can be raised higher without clipping, no? or do you use the compressor in a way so as to emphasize transients?

i'd prefer to have it on one gui, simply driving the output of the compressor into the limiter ...
is that possible?

of course would it be possible - but as said: either i would have two compressors in one (which could be a bit confusing to some users) or i would have to implement the second compressor with some hardcoded ("hidden") parameters. so i would be happy to hear your (or others) suggestions how exactly that could be implemented.

[brok landers]

mmm, isn't the whole point of compression to actually raise the rms level? i mean, a compressor attenuates transients such that the overall level can be raised higher without clipping, no? or do you use the compressor in a way so as to emphasize transients?

well, no and yes ... there's 2 ways of using a compressor:

1. to tame dynamics over the whole timedomain that the signal lasts (i.e. a complete leadvocal track), so that the overall dynamic range stays in a sensful, controllable frame ...
2. to use it to "define" and colour the signal, i.e. to make a snare more snappy and punchy ...

i love to add colour to my stuff, and a compressor is (added with other elements like i.e. an analog like eq) really cool for this ...
actually i love it, if a compressor can be "heard" ... if it breathes with the groove ...
now, back to the snare:

assuming the snare is leveled to exact 0db, but it misses that snap ...
what the usual way is, to compress it, leaving the attack a fair bit open so that the transient slips through ... that way the "body" level of the snare (which defines the "felt" level) is taken down, up to the point the release comes up ... this can sound wonderful ...
however, now the snare is mathematically as loud as the unprocessed signal, it still hits 0db, as the attack transient is unaffected by the compressor, but the rms level "the felt level" is much quieter ...
now, a common technique was, to use a limiter at the outputstage of the compressor (0db fixed threshold or adjustable, but the release was fixed but rather fast, just to tame the transients, making sure there's no overshoot), and driving that compressor output against the limiter, hence the name brickwall ... a lot of analog compressors had a limiter buildt in that way, to use it exactly like this ...
perhaps the most wideley known one is the behringer composer (though thatone is rather bad quality) ...
however, if done right, one ends up with just the perfect ratio between snap and rms level ...

so actually that way used, a compressor adds colour to the mix, it's done the way i described, commonly ... if one does "punchy" music, that is ... you won't find that way of working in classical or smooth bossanova jazz ...

[Music Engineer]

there's 2 ways of using a compressor:

1. to tame dynamics over the whole timedomain that the signal lasts (i.e. a complete leadvocal track), so that the overall dynamic range stays in a sensful, controllable frame ...
2. to use it to "define" and colour the signal, i.e. to make a snare more snappy and punchy ...

assuming the snare is leveled to exact 0db, but it misses that snap ...
what the usual way is, to compress it, leaving the attack a fair bit open so that the transient slips through ... that way the "body" level of the snare (which defines the "felt" level) is taken down

ah, i see. i kind of diffusely knew this but now its more clear, so thanks for claryfying. so basically, yes, you use the compressor to emphasize the transient in this case (how about a dedicated transient processor instead?).

it's done the way i described, commonly ... if one does "punchy" music, that is ... you won't find that way of working in classical or smooth bossanova jazz ...

yeah, i'm myself way more into 'punchy' music also (think techno'ish and/or rock/metal stuff)

O.K. but the question remains: if i include a brickwall, i had either to choose its parameters on the developer-side and hardcode them (such as behringer obviously did) or i would have a duplicate of the whole plugin inside the plugin. i've just gone through wikipedia's article on limiting:

http://en.wikipedia.org/wiki/Audio_level_compression

which seems rather good, though it also stays rather ambigous about the limiters parameters (high vs. infinite ratio, fast vs. zero attack-time). i - being a pedant - would probably insist on zero attack and infinite ratio. anything else cannot be justifyably called 'brickwall'. O.K. remains the release...

[brok landers]

ah, i see. i kind of diffusely knew this but now its more clear, so thanks for claryfying. so basically, yes, you use the compressor to emphasize the transient in this case (how about a dedicated transient processor instead?).

i knew you'd come up with that ...
well:

1. i often use transient designers
2. though loving these on specific tasks, it's not the same ...
on a transient designer the transient is raised/lowered, but the rest remains more or less untouched, even if you raise/lower the decay/release ...
or let me say it in a different way:
it simply sounds different ... with a transient designer, what is missing , is, that "squashing, pressured" sound that you could do with a compressor/limiter ...
the transient designer is used for only "redesigning" the transient in a most effective, but clean way, as the envelope creates a static shaping of the dynamic ... whereas i love to use compressors to add that breathing, living groove ... forgive me the mythdropping here ...
i just don't have other words for it ...

O.K. but the question remains: if i include a brickwall, i had either to choose its parameters on the developer-side and hardcode them (such as behringer obviously did)

yes, that's exactly how you should imo do it ... you could give the limiter a threshold, if you want, but the ratio must be infinite, the attack instantly, not allowing for overshoots ... the release ... there's 2 ways:

- either fixed (a lot of tweaking-by-ear is necessary, to find the right shape/time)
- or so called "program dependent", means, the release length is automatically choosen by the inputlevel of the limiter, the outputlevel of the compressor ...

in any case:
- the compressor output should be adjustable (ignore the fact that you might have a threshold on the limiter), as you'd often just want to drive the compressor-outlevel against the zeroed threshold ...
- there should be a final output level-attentuation slider at the very end of the signal chain after the limiterstage, so after you've compressed the signal and drove it against the 0db threshold, you often achieve a louder rms level than the original signal has ... so in order to _not_ beeing fooled by the loudness increase, the "wow" factor, it's always a good idea to temporarily attentuate the limiter output to the same "audible" level that the unprocessed signal has (meaning lowering the limiter output by ear, so that it fits to the same level of the unprocessed signal, ignoring the meters) ...that way you can actually hear, what the compression/limiting combi does to your signal _exept_ the loudness increase (by simply bypassing the plug), which often tends to mask reality ... that way you can avoid to actually overcompress, hence damaging the signal ...

[Music Engineer]

thanks brok for these suggestions. these will for sure find their way into my scheduled dynamics processor. i will have to look into algorithms for program dependent release - if you know any good resources for this, i'd appreciate your help.

[zeep]

Just heard the mp3 demo in this thread, leveller sounds so fat! I like it.

[garfield78]

Hi to all , anyone can reupload leveller vst plugin ? i don't find on the web .

A great thank you

[zeep]

Hi to all , anyone can reupload leveller vst plugin ? i don't find on the web .

A great thank you

If Braindoc doesn't mind.. Here it is.

https://we.tl/N2u9Rv1Syd

[garfield78]

Hi zeep , a great thank you !! work fine !