I'm using Hornet's TheNormalizer for this. I like the workflow. I set it to -18dbFS for emulation type plugs. Hit the Analyse button, play the audio and let it do it's adjustment. Sometimes you'll get a crazy adjustment, like INF. If that happens, just stop the audio, hit Analyse again, and repeat and it usually sorts it out.
It does have a "rider" like function on it where it will continually adjust, but I'm not using that.
I'm using Nebula and Acustica Aquas so I'm thinking the -18dbFS is important with those due to the hardware sampling technology they use.
I've got to say, it helped out with clarity, punch, and the transients.
-18dbFS just seems so quiet. Guess I'll get used to it. I'm mixing a lot less "hot" now which is making some of the mix decisions easier.

 

i thought the overloud stuff would want -18 but no , didnt find anything in manual..but there are some emus which demand -18 in manual..i have read in forum..
could somebody pls just state a few plugs witch demand -18DBFS for analog emu sweetspot ..
please..

also in nomad factory anal studio rack manual i didnt found it

 

Last edited: Aug 30, 2019

It's a bad case of Gain Staging Snakeoil!

 

very handy thanks man

 

hah yeah. its weird -18 is now the magic number. it is definitely from old VU meter calibration but -18 is where the red started. needle to be pointing straight up or less was conventionally considered to be a 'good' level which would be more like -20 or 21, tops. pushing levels like -18VU on to some tape machines can really start to bloom and compress the signal, which can sound great but it is hardly audiophile stuff. on some tape -18VU and up can also flatten the sound and lose the etherial, floating quality of tape that some people like. even expensive +24dbV desks and gear may start to futz at -18VU. i know distortion ('colourful *glueing* staturation' in marketing terms) is why people like these things, i am only talking from a purely technical standpoint. maybe they commissioned a marketing group to peer test various levels and -18 won out.

VUs were also an RMS style reading and were non-linear with respect to frequency, people who only know very fast peak meters could be easily confused.

i noticed waves in its 'plugins for dummies' series has a simple led type thing that glows green for good and red for STOPPPP. this is perhaps a better strategy than giving a user something they have no tangible reference to ;)

 

if you search around the studio there is probably a dusty old potentiometer somewhere that has'nt been touched in years. you will need to crank that clockwise. it will communicate a purely analogue, "make loud", signal to the monitoring system.

 

I know the Waves SSL Series is calibrated around -18 dB VU, but I never care about that, all I have a look at is red lights. And even if it kisses the red it doesn't sound digitally distorted. And then there's input gain control to turn the signal down. Needless to say that the signal you bring into a plugin should not be clipping beforehand. Because a clipped signal stays clipped even if you turn it down.

To summerize gain staging:

Rule #1: Don't drive clipped signals into ANY kind of plugin (unless it's a desired effect)
Rule #2: Don't trust Audio YouTubers who are just YouTubers, trust your ears

 

I'm so glad I early on learned about gain-staging, AD/DA converters, 32bit float internal processing and all that. It ought to be the first thing any beginner learns about digital recording.

 

For some ikm emulations, such a level is recommended. not because the plugin is calibrated that way, but because of internal clipping.

 

hardware and tape was the only option at the time. perhaps i did not make it clear that i was talking about the past.

 

pretty sure the waves ssl stuff is linear to digital zero. non-linear response with respect to signal level in not modeled. you could boost as much as you want on the way in as long as you compensate adequately. as long as you use 32 or 64 bit internal processing you will have to push some extreme levels into it to make it futz out. this is easily tested by placing a boosting plugin and some meters (if you want measurements) before and after the plugin -- do not engage any compression for obvious reasons. the 'analogue' modeling is probably just an eq curve and that generic waves random noise but this is mostly just speculation on my part. you can just joyfully pin those meters without consequence.

the introduction of artificial harmonics by clipped signals can be a real problem on digital systems. if you run at 32 or 64 float internally it is theoretically impossible (within reason) to digitally clip while processing unless you are using plugins or processing that are modeled with a non-linear response with respect to signal level. the problem lies in the input and output stages (usually files, printing and rendering) and if you are intentionally using hard, soft clipping, distortion type processing for control, affectation or effect.

however, as i stated before i value content way (way way way) above technical form. go ahead and clip, or whatever, if it is in service to communication of a unique perspective with respect to an individual observation of process of living and all that entails.

 

Quote from the Waves SSL Channel Manual: "The plug-in is aligned so that -18 dBFS = 0 dbu".

 

what exactly has that quote got do with what i wrote? or at least what you think it has to do with what i wrote. i do not see any correlation.