Sorry, are my questions stupid or am I invisible?

I found this https://www.recordingrevolution.com/do-you-know-how-to-read-your-meters/

At the end it says: "Remember, we’re not talking about your tracks peaking at -18dBfs. In fact they should peak higher. We’re talking about the average volume (RMS) here."

Is this correct? Is this how you guys are going about it?

Also I've re-watched the video here https://audiosex.pro/threads/in-the-box-tutorial.23103/ and that guy is aiming for -10 peak in PT, so why use the the VU meter at all (besides using it as a trim plug as he does) if you're just going to aim for -10 peaks?

Sorry to be a pain here but I'm just not getting it.

 

There's nothing wrong with aiming for -10dBFS peak value for your drum and bass buses. I said already that it can work, too. But still it is useful to have a VU meter for measuring the audio level that goes into the analogue e-mu plugins, because they're usually calibrated for -18dBFS/0dBVU *peak* levels. 0 on the VU meter is at -18dB full scale and it is a RMS value averaged every 300ms.

Nobody is ignoring you, freefeet12, but it seems like you are ignoring what the people said here or just don't understand it? Sorry. Maybe just read more on the subject. Web is full of knowledge, you just have to look for it.

 

Thank you SineWave. Believe it or not I've been obsessed with this over the last few days, reading, watching videos, so many videos, for hours. I haven't yet found a source that details how this is done in practice since I'm not working with a sine wave.

Waves CLA-2A: "The hardware was modeled at reference levels of -18 dBFS = +4 dBu, meaning that a signal of -18 dBFS from the DAW to the hardware unit will display a meter reading of 0 VU (+4 dBu)."

Does this mean literately -18 dBFS peaking in the DAW's peak meter equals 0 VU on the hardware? If so, I can not achieve this without a sine wave on VUMT. Or does this mean that an average of -18 dBFS on the peak meter with peaks lower or higher equals 0VU on average on the hardware?

In practice, should I be avoiding peaks higher than -18dB full scale on the DAW's peak meter when going into a analogue e-mu plugin calibrated for -18dBFS even if I do not hit 0 VU on the VU meter calibrated at -18? Or just ignore the peaks on my DAW's peak meter and aim for an average of 0 VU on the VU meter regardless.

This is what I'm not understanding. How it's actually implemented in practice.

 

In practice, on old analog gear peak metering wasn't that important or even feasible from a practicality standpoint. I am taking a guess that the concept that is alluding you is that of headroom and how it relates to that calibration point of -18dBFS=0 VU, since for the most part this only partially affects creative audio process (you are mixing and processing with your ears and making judgements for the most part.)

+4dBu (0VU) is more of less the linearity threshold of the circuitry... Claiming your circuitry has more headroom than the competition is basically saying that even if you exceed 0VU your processor will behave in a more linear fashion without (or with much less) distortion than your competitors. In broadcast, post and film levels and calibrations, distortion as well as loudness are much more controlled and having a consistent point of calibration from say for example from a live broadcast van in a stadium --> local affiliate --> national news desk matters a bit more than when one is chasing a particular sound.

Driving a processor is the process of exceeding it's headroom in order to get saturation and distortion, so older analog units distorted more audibly and with less drive over 0VU than those that came after them with increased headroom. Some of units weren't even linear bellow 0VU and since the ballistics of VU / PPM / BBC meters can let fast transients peaks through quite easily, sometimes the sound coming out was a recovery from a nonlinear event that the meter didn't even register...

This is also why you sometimes see old analogue consoles being driven into the red... sometimes they were chasing distortion, but in some examples the console just had the headroom to push the signal harder while remaining linear enough to not turn the sound coming out into crap. Even in some of the best examples you probably were still only looking at peaks in the -9 to -6 dBFS with an occasional clipping event near the limit of the systems maximum voltage if a transient was fast enough and powerful enough to get through before the physics of capacitors and circuitry washed it out into distortion.

 

Finally a thread speaking my language. I've been using -18dbfs as a goal for the past year or so, and it always works out in my favor at the main mix bus stage. Question I've had for a long time, though: plugin companies recommend -18 as the sweet spot. Some, like UA's Studer and a couple others, recommend -12. Fine. Still, the problem I face is that all presets, which I use as starting points, and even default settings are not even close to utilizing such a low signal. On UA plugins, I end up having to crank the headroom screw on some plugins to even get the signal to read. This seems wrong to me because they're the ones recommending -18...yet then I'm cranking the input gain of that plugin or lowering the threshold of a compressor to a ridiculous extreme. Anyone have similar experiences?

 

Yes, instrument presets are usually cranked up to hell and obviously made for and by the "loudness warriors". I noticed that u-he presets are generally lower in volume, depending on who made the presets, but you usually have to turn them down a bit anyway.

I usually just turn the volume of the whole track down about -10dB down until I find the sound I like and then I turn it down in the instrument according to -18dBVU or according to mix/ear so I can put the FX on it.

Mixing so the loudest track peaks to -18dBFS [peak metering] might be a bit too low, but it's not too far from the optimal volume and a few dB up-down don't matter anyway. It could be too low for plugins, though, since they are optimised for average level [VU-300ms window], not peak level.

 

Having a VU meter on the master bus is really handy because you can check any track's level easily by soloing it, at any time and it saves a bit of CPU not having many instances of it. Although, one could argue that having it on the buses is useful, too.

-12 VU-RMS is perfectly fine level, not to say - perfect, for the whole mix. That's why UA recommend that level for the Studer plugin.

 

Hi, thanks for the additional information, everyone,& mozee. I do appreciate you taking the time.

After some more reading from different sources it turns out that VU meters aren't the greatest at reacting to sudden peaks or transient rich material by itself, and or in isolation, and knowing this helps as this is what I've been trying to use it for thus far on my way into a analog emulation plugin. It seems the more appropriate use of a VU meter is to measure averages of combined and or sustained sounds/signals/instuments/ect. If this is true, my exercise of processing a single transient rich sound using the -18 = 0VU method is an exercise in futility.

Even if the above is true, I've been essentially asking the same question in many different ways without getting a direct answer to any version of the question. Instead I'm getting something more analogous to bread crumbs. Bits of the information I'm asking for spread out here and there, none of it inspiring confidence or relieving any doubt. All of my questions boil down to how I should be reading the peak meters in my DAW in conjunction with a VU meter (Klanghelm VUMT) calibrated to -18=0 VU, in the digital realm as to achieve the best possible digital signal going into a digital analog emulation plugin calibrated to -18 dBFS , never leaving the digital realm, and more specifically what to make out of the peaks showing on the peak meter VS trying to get that needle "dancing" near or around 0VU. <--- See that, there I go again, rephrasing the same question. Should peaks matter, should they not matter, is 0 VU more important going into the plugin than the peaks, if I'm peaking at -6 but reading 0 VU going into the plugin, am I good, if the peaks are about -6 but barely moving the needle on the VU meter, than what? On and on, the same question in many ways.

Since I've gotten no direct answer to any of these questions I can only conclude that:

A: My question, no matter how phrase it, does not make sense.
B: There is no direct and or simple answer to the question.
C: No one actually knows the answer.
D. One or all of you have directly answered my question and I'm on an upside down world and do not speak the language spoken here.


In any case. I give up. The VU meter has now become a source of stress and distraction. I'm just going in circles, asking the same question in various forms, (not only here) reading the same information over and over, watching someone in some video regurgitate that same information, ad nauseam, and I'm no closer to grasping how to put this to use in a real world situation without a strong sense of doubt, completely in the box.

If I'm lucky, I'll eventually see someone actually use it in a session and then I'll have my aha moment, if one is to come at all. For now I'll keep going at it blindly, as I have, so long as the individual signals and their sum aren't on fire. That much has always been clear to me.

Again, thanks!

 

Peaks do not matter with these old analog emulations.

You'll get the clearest signal from analog emulation plugins bellow +4dBU, no need to keep the needle pinned at 0VU the processor should work reliably in a somewhat linear fashion from it's noise floor to 0VU.

If you want saturation and even more color (distortion) you get can it by overloading the circuity and pushing your signal over 0VU. The amount of saturation and how quickly it becomes audible will depend on the circuit being modeled as circuits with more headroom will handle more signal before distorting. By the time you get -9dBFS or +9dBU you should be getting audible distortion on even the best circuits though.

Peaks didn't really matter in audio until digital, the meters, ceilings and ballistics were set conservatively enough that you really had to go out of your way to overload a signal. How fast the a meter responds, how fast it falls back and what constitutes an overload in the analogue metering isn't as precise as in digital but the safety margin was large enough that even with slow ballistics the probability an overload when staying within parameters was low enough to be of no concern outside of extra-ordinary circumstances.

TLDR

For analogue gear calibrated at +4dBU
peaks don't matter
anything under 0VU = clean,
dirty happens close to or somewhere after 0VU
You don't have to drive your old analog emus at any specific level, unless you want them to distort.

 

Regarding peak meters, I'll try to be more to the point this time.

If you're using a VU meter for assessing the loudness levels of tracks, specifically the loudest tracks, usually drums and bass and they are not exceeding 0dB on the VU meter, peak meters don't matter.

By using VU meter you will never see clipping on the master. You can forget about the peak meters and when you finish mixing you will be surprised at how good the whole level of the track is and ready for mastering.

 

@mozee @SineWave

That was it, it makes sense to me now. I can't thank you enough. Here, have an internet hug.

Phew, what a load off. I'm off to put this to use and see what happens!

I was at the end of my rope with this. Really, thanks a lot!

 

Glad to be of some use.

Cheers.