Volume issues

[sephfire]

I've had this problem since I first started remixing. I thought it was a problem with Reason, but now it's doing it in Ableton as well.

Whenever I render out a wav of a mix, the volume is usually extremely low and I can't find a way to fix it. If I push the master level any higher, it'll clip or just get butchered by my master limiter. But when I take the track into iTunes, my mix is half the volume of every other track in my library. Even when I turn the volume up 100% for that particular track, sometimes it's STILL too quiet.

SGX gave me some good ideas to help with this, but I still can't find any complete solutions to this problem. Please help.

[zircon]

Are we talking perceived volume, or ACTUAL volume level (like.. -16db vs 0db)?

[sephfire]

Well, the actual volume (db) is pushed as high as possible in Reason/Ableton without clipping. But when I render it out, the perceived volume is much lower than almost every other music file in my library. I did have some success with a track today that came out louder than usual, but it's compressed pretty thoroughly.

I just can't figure out how to boost the perceived volume without touching the actual volume. I know I can just crank up my own speakers, but there has a to be a real way to fix this in the file itself.

[zircon]

Compression can do it. But I think the problem is that some frequencies are making it harder to max out the perceived loudness. You tend to use ridiculous amounts of rumbling bass/drums (from what I've heard) and so that's taking up all your headroom. Some selective EQ will really fix that, or different timbral choices.

[sephfire]

Compression can do it. But I think the problem is that some frequencies are making it harder to max out the perceived loudness. You tend to use ridiculous amounts of rumbling bass/drums (from what I've heard) and so that's taking up all your headroom. Some selective EQ will really fix that, or different timbral choices.

Ain't that the truth. I'm getting better about that. Thanks for the tip!

[Compyfox]

The problem is:

Mastered tracks nowadays are pushed to it's limits (Loudness of about -6dB RMS). This is usually the stuff you refer to, right? Well I have the same problems in Cubase, but because engineering is one of my final steps, I know how to work around it.

The problem lies within the loudness levels. Chances are that the bass uses up a lot of it, and on the meter it looks totally overpowered, but while comparing to "chart productions" (I don't say professional on pupose, cause there's nothing professional on it anymore), it totally looses. This is a typical Loudness issue.

Another thing is Reason and Ableton. If you render something, so that it barely hits 0dB (peak), then you can't reach an RMS level louder than -12dB RMS (refer to the K-System by Bob Katz for example). This however can be "readjusted" if you master your track and push the loudness to your desired level (maximizer, compressor, caraful engineering). Then you can get the loudness you like (even -6dB RMS, even if it's squashed into a squarewave now).

IF you want to push the track right from the start, that's your thing. But if you want to engineer it later, you will get problems as your track is too squashed and already limited to a certain level. But if you don't want your track to clip, you have to go for the actual "mastering" step, or let it handle by somebody else.

Simple issue, large influences.

[Splunkle]

Compy, you mentioned Bob Katz and his K-system. I looked over at his site the other day, and read some of the stuff on the K system - and was wondering how I could master my stuff so to one of the K standards - I like the idea of never having to touch the volume controls on my speaker.

The thing is, Katz keeps talking about a VU meter, and how it wasn't the same as a peak meter, because the VU meter's needle is delayed a few hundred ms so it gives an average loudness reading rather than a peak loudness reading. This makes sense. But I don't have a VU meter - only the loudness meters in FL.

Now I can render my track and get the RMS for any part of it - but is that the same as the VU meter reading? Is there a VST you know of that give me a VU meter, or an equivalent?

And to make my track fit the K-20 standard, do I just need to make sure it stays around -20 dB RMS, getting a little louder (4dB I think he said) for the loud parts?

I know, I know, lots of questions. But important ones, I think.

@ Sephfire: What other music are you listening to? As Compy said, mainstream music is getting louder and louder. To illustrate this, get an old, but still availible CD (Say, Led Zepplin - Remasters) and a brand spanking new CD, (say, The Strokes - First Impressions of Earth) preferably pop or pop-rock for both. Compare percieved loudness.

So if you want to check if you track is really quiet, compare it to some older stuff - preferably either straight off the CD, or ripped yourself by a program you know won't screw with the loudness. Or you could compare it with some classical stuff - that's always very quiet, becuase classical fans demand that the dynamics aren't screwed with.

[Compyfox]

I guess it's really time I start writing PDFs with certain topics, but not before end July, i fear. Just a quick stab,cause I'm pretty sure this stuff vanished in the depth of the "board cropping", too.

Now I can render my track and get the RMS for any part of it - but is that the same as the VU meter reading? Is there a VST you know of that give me a VU meter, or an equivalent?

Katz also mentiones that the VU meter is just a "guide". The ears are still your most important weapon. The problem with the RMS meter of FL is, you can't set it up to fit it to the standards of the K-System. So you need something that works.

In this case, I would have recommended you "Inspector XL" from Elemental Audio, but it was recently "swallowed up" and their plugins are sold for 3times as much money. Their Metering Tool however is still not available, but it had the K-System built in. But fear not, there is an alternative, and it doesn't cost you a thing: PSP Audio Vintage Meter.

This is a Dx Plugin, but still works like a charm. Load the plugin after your limiter, then click on the name on the bottom of this plugin, so you can see the backside. Setup the VU Integr. Time to 600ms, the Meter Delay can be left "as is" (also the right side). The 0VU reference Level sets your K-System. -20dBFS for K-20, -14dBFS for K-14 and -12dB FS for K-12. How you setup your speakers, is described on Katz page.

If you go to back to the mainpage now, you can click on the meter field to change the scale from +3dB to +6dB (beter for acourate checking). The dipswitch sticks to VU of course.

And to make my track fit the K-20 standard, do I just need to make sure it stays around -20 dB RMS, getting a little louder (4dB I think he said) for the loud parts?

That's the idea, yes. If the needle hovers around the "0-point" of the PSP VU, then you're where you want the track to be. The additional +4dB are for mezzoforte parts. If you get into the reds (the last 2dB from the additional 6dB), your track is not in the standard anymore. However "how long" you stay in the amber zone, or what is "mezzoforte" and what not, is a matter of preferences. Everybody does it differently.

Remember however... The K-System is dynamic, but people will bitch about it that your tracks are "too quiet". If you have no problem with that, but fight for your tracks "clarity", then go with it. If you want to check some stuff in the K-System, the "Chrono Symphonic" project is in K-12 (full amber zone used up for mezzoforte tracks). Or just check Reuben Kee's and pixietricks "The Place We Knew" (also from the Project), that's also seperately available on OCR directly, which is the same as the "project" release but with a different name.

I know, I know, lots of questions. But important ones, I think.

This is what we engineers are for. To teach others. Unfortunately a lot of so called "sounddesigners" and "engineers" seem to forget thatand rather stick to their secrets to remain "special".

So if you want to check if you track is really quiet, compare it to some older stuff - preferably either straight off the CD, or ripped yourself by a program you know won't screw with the loudness. Or you could compare it with some classical stuff - that's always very quiet, becuase classical fans demand that the dynamics aren't screwed with.

Then you didn't listen to the "Final Fantasy 7 - Advent Children" soundtrack. That thing is pushed to -8dB RMS and doesn't fluctuate that much. Same with Soundtracks by Hans Zimmer and Harry Gregson Williams. Sometimes they are in the K-System, sometimes not.

And yeah, I wouldn't go for remasterings. Some are pushed to their limits on purpose - even though they're remasterings.

[sephfire]

I love you guys.

[Splunkle]

MASSIVE POST OF WIN AND LOVE

Thank you very much, Compy. I've grabbed that PSP vintage meter now, and it seems to be exactly what I need. For those of you you too lazy to google it, it can be found here. Also, if you are a mac user, don't panic there is a version for you guys too. 3 versions, in fact. It also does VST, so maybe they have updated it since Compy's version.

Actually, I would recomend everyone gets this plugin - it is quite remarkable the different readings I am getting. -6 dB on the Fruity dB meter is coming in at 0 dB when I have it set to K-14, which would be -14 dB. Quite a difference.

Then you didn't listen to the "Final Fantasy 7 - Advent Children" soundtrack. That thing is pushed to -8dB RMS and doesn't fluctuate that much. Same with Soundtracks by Hans Zimmer and Harry Gregson Williams. Sometimes they are in the K-System, sometimes not.

And yeah, I wouldn't go for remasterings. Some are pushed to their limits on purpose - even though they're remasterings.

Aye, perhaps it is an exception, but Led Zepplin's "Remasters" was done back in 1990, so it is a reasonable volume. Especially in comparisson to The Stroke's latest single. My ear is not the best, or even anywhere near the best, and even I can hear the painful distortion going on there. But I shouldn't rant in Seph's Thread.

[Compyfox]

Also, if you are a mac user, don't panic there is a version for you guys too. 3 versions, in fact. It also does VST, so maybe they have updated it since Compy's version.

I guess I got mine 1-2 years go. Since then, nothing much changed and I switched to iXL anyway. Like I said... a pity that it's gone.

Actually, I would recomend everyone gets this plugin - it is quite remarkable the different readings I am getting. -6 dB on the Fruity dB meter is coming in at 0 dB when I have it set to K-14, which would be -14 dB. Quite a difference.

That comes from the integration time, and the incapability of setting up the "0-point" of your metering tool.

Maybe I should mention here, as soon as the red LEDs in this plugin (under the needle) go off, it means that your track clips. What to do next should be clear, I hope.

[Mark7]

Read this:

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

Anyway, Reason has a new compressor from version 3.

First just increase the master volume as high as possible without clipping. Then put the MClass Maximizer between the mixer and hardware interface. And then you can increase the volume (and decrease the quality) with the input gain button on the Maximizer. I'd say going much higher than 4.5dB will sound like shit.

Don't use softclipping, it sucks.

[Hy Bound]

This is SUCH a noob question but wtf does RMS mean?

[zircon]

RMS = Root Mean Square. In the context of audio, it's a more accurate way of measuring the volume of a track than just using the peak volume. The concept behind it is that because you have a constantly fluctuating waveform, just trying to measure the peaks isn't going to give you a good idea of what the overall, "effective" volume is like. So that's where RMS comes in. I believe at it's simplest, for a sine wave, RMS is the peak vol times .707 .

(Correct me if I'm wrong, guys)

[klm09]

RMS = Root Mean Square. In the context of audio, it's a more accurate way of measuring the volume of a track than just using the peak volume. The concept behind it is that because you have a constantly fluctuating waveform, just trying to measure the peaks isn't going to give you a good idea of what the overall, "effective" volume is like. So that's where RMS comes in. I believe at it's simplest, for a sine wave, RMS is the peak vol times .707 .

(Correct me if I'm wrong, guys)

In terms of data / math / digital audio, what it means is taking the square root of a given amount of samples (don't know how many) counting the average (ie. mean) of those, and then multiplying the number by itself (squaring it). This way large, quick fluctuations in the audio are averaged out, sort of like a slew generator in modular analogue synths.. or like a low pass filter, where it's taking out the the high frequencies (quick fluctuations).

[Salluz]

Make the computer volume fairly loud. Make the master volume pretty loud. Next, turn all of your instruments down to about a half or a fourth as loud as they get. Then, you can properly hear everything without getting too crunk/soft.

Well, this is what I do with FL5 and 6. Also, a computer's nature could bring fault as well.

[Cerrax]

Maybe this is terribly obvious and I'm a n00b but if your song has a great dynamic range (very soft to very loud) you can fool the listener into believing that the track is louder than it actually is. During a very soft spot in the song bring the master volume way up so that it is very clear and (possibly loud) then as it nears a part where the volume jumps up (like a cymbal crash into a huge allegro section of a song) gradually decrease the master volume. The listener's ear will unknowingly adjust to the lower sound and percieve it as the same volume as before. Then when the crash comes in, the sudden spike in volume seems to be very loud, but it's not really that loud at all. I do it all the time to push the perceived loudness of my songs further.

[Compyfox]

That's not dynamic at all, it's the other way. Then why not clip everything to it's maximum anyway? You won't hear any difference at all and it's the same effect you were just talking about.

Make the computer volume fairly loud. Make the master volume pretty loud. then, turn all of your instruments down to about a half or a fourth as loud as they get. Then, you can properly hear everything without egtting too crunk/soft.

Well, this is what I do with FL5 and 6. Aslo, a computer's nature could bring fault as well.

Ever heard of "adjusing your setup properly"? Setting everything up correctly right from the start is a good thing. Borrow an SPL (sound pressure level) and use the K-System from Bob Katz as guide, then set it up "correctly" according to your room.

Pushing everything right from the start doesn't make it better. Your "technique" even hints to pushing the loudness even more, cause the producer is like "Okay, my volume slider is at 0dB, my sequencer volume slider is at 0dB but the amps still sound too quiet, let's push it more, where is my multiband compressor/limiter?!".

THIS... is a total NONO in terms of engineering!!!

Mix everything together so that nothing clips at all (which means, you barely even hit -0,5dB in the master bus). After you're finished, do whatever the hell you want with it (aka overcompress it). But if you compress and boost right from the start, chances are that your music won't sound any better after the mastering stage. And not even an audio engineer can fix that.

Learning the wrong way is easy. Learning the right way is like a crusade that's happening in the shadows, but nobody wants in reality. We "careful engineers" fight our war alone.

[Eon_Blue]

Learning the wrong way is easy. Learning the right way is like a crusade that's happening in the shadows, but nobody wants in reality. We "careful engineers" fight our war alone.

I am choking back a mouthful of vomit as I read this.

While your advice is often useful, you are so full of yourself it's sickening.

Flame on.

[SnappleMan]

I like the duck in our sig, eon.

[Eon_Blue]

I like the duck in our sig, eon.

Thanks, man.

...I like the implied homoeroticism in yours.

[Salluz]

Learning the wrong way is easy. Learning the right way is like a crusade that's happening in the shadows, but nobody wants in reality. We "careful engineers" fight our war alone.

Mm-kay. Oh, and I edited my crappy, hasty post for more clarity.

[Cerrax]

Compyfox: Clipping everything at its maximum is not the same as what I said. That would just make the mix sound jerky and unbalanced (volume-wise). My suggestion works (for an example of this listen to a digital recording of "Money For Nothing" by Dire Straights and then a vinyl recording of it. The vinyl recording does what you say and it sounds horrible, while the digital recording slowly reduces in volume as the different layers come in. The result is a song that sounds louder than it actually is with out being absolutely obvious that the master volume was changed.

In addition, I just remembered that you may need to do a low and high end cutoff of frequencies. Sound below 20Hz and above 20kHz is inaudible to human ears, so there's no point in keepoing those frequencies in the song. Many "professional" recordings do just this to give a little bit of head room for more volume.

As others have said, compression and peak limiting can also help, especially if applied to individual instruments (say for example a very punchy bass drum that just needs a little compression, instead of compressing the whole song).

Plus most "professional" are pushed to the 0 dB mark (which IMO as well as others is not a good idea).