Final Song Volume/level, Where Should It Be --matching Reference Mixes

[AAHIHaveNoIdeaWhatImDoing]

Hey guys. I know this is not especially a mixing/mastering engineer kinda forum, but hoping someone has dealt with this before at the DIY/homebrew level.

Some my band is releasing our first EP. Recorded/mixed/"mastered" ourselves. I am seeing that with the songs I mixed myself, the final levels of each song are kinda all over the place. Here's the thing. . .I matched them to commercial releases volume-wise, BUT, those commercial releases themselves are not at the same level!

To go into more detail, I have two songs I'm concerned with here. One is a pop punk style song, and I used Green Day's "Welcome to Paradise" as my reference mix. Matched the output of that volume-wise pretty closely. Another song is supposed to sound basically like AC/DC wrote it, so of course I used an AC/DC mix ("You Shook Me All Night Long") as my reference. These are the actual CD/commercial tracks, one I bought via iTunes just to use, the other, I have Back in Black on CD.

So, You Shook Me is simply not as loud as Welcome to Paradise. I don't just mean as compressed/in your face/whatever, I mean that nowhere in the track does You Shook me approach Paradise's loudness. It's a pretty noticeable difference, as well, enough that one would be underwhelmed/overwhelmed while listening to, going from one song to the next and need to adjust their listening volume accordingly.

So the question is, being that we want all our tracks to be the same volume, generally speaking, and the listener NOT to have to adjust the volume to avoid having their ears murdered or finding one of the tracks too soft, which reference is "right?" How would I figure this all out?

 

[Ghostman]

Older production tracks, prior to mid 90's had a much lower recording volume, with much less compression. This maximized dynamic range, and gave the best bang for the buck. Now days, these mixes are compressed as much as possible, and the levels are pushed to the max. The producers want the sound to pop when someone hears it, and higher volumes does it.

You can't match two different track mixes from two different albums, from two vastly different eras. Stick with one, and go with it.

Changing volumes from one album to the next isn't so much of a problem, as changing them from track to track on the same album. I'd go the happy medium, and stick with that. Do not use a popular album produced within the last five years if you want to mix with the old and new.

Listen to Boston's first album on an original pressed CD, and then grab anything made in the last five years. The levels are insanely different. When mastering music to digital, the engineers have a fixed number of bits to work with, so they have found a trick of compression to get the most out of their mixes.

 

[AAHIHaveNoIdeaWhatImDoing]

Ok, thanks. Yes, these will all be on the same "album," so gotta be close. I feel like the louder reference is probably the safer bet since that's what a modern listener is probably used to, right? To complicate things, and I didn't mention this before, another of my songs was mixed by a friend, and was mastered louder even than the Green Day-sounding song! So I have that one to contend with too. I'd rather not re-"master" that one since I don't have access to the original session/tracks and would be working on an already mixed/"mastered" track, introducing possible degradation. Maybe raise the others to be as loud as that one? I like the happy medium idea in theory, but the Green Day-sounding song is at least kinda in the neighborhood of the one I didn't mix myself, so if I lower that and the AC DC-sounding one I put those two further apart in volume from the loudest one.

 

[Ghostman]

The trick is, getting a balance between solid level playback, and jacking up the levels too much that the high energy stuff starts clipping. So then, instead of backing things back down, you throw some compression on. This drops out the quiet stuff, and all of a sudden you have high level mush....

 

[AAHIHaveNoIdeaWhatImDoing]

I've been okay with the results of my limiting when I push it, just wondering WHERE to push it at this point.

 

[JimiRules]

I'm still in the learning phases of all this but what I've done on the CD we're working on is put a limiter on my master bus to where the track doesn't go over 0db at its loudest point. After doing this its still a little lower than radio volume, which is what I use as a guideline. Then I put the track in audacity and use the amplify feature and set it to 0db. After doing this each track is consistent and is pretty much radio volume.

 

[AAHIHaveNoIdeaWhatImDoing]

Fair enough. . .what is your reference for "radio volume", though? How are you matching your tracks to that level?

 

[JimiRules]

I load my tracks onto a flash drive and plug them into my car. Then I play them and switch back and forth between them and the radio and listen to see if there's any major differences in volume.

 

[Ghostman]

I use a known good track in my playlist. I've been using a few tracks from Extreme III-sides. I've found the levels are a good balance between Old School high dynamics, and the over compressed stuff from the last few years. Radio stations can be finicky at best, unless you find one that fits well.

 

[AAHIHaveNoIdeaWhatImDoing]

Ok thanks. cant really do a flash drive in the car but can load stuff into itunes/my phone and use the aux input. but then how loud should the phone output be???Its all very complicated.

 

[Dogs of Doom]

do not use "radio" as a reference... They have their own compressor/limiter system, which is not good for your commercial mastering project.

What Ghost somewhat touches on is that, in the days of analogue, the equipment had its inherent limitations, which included dynamic range, frequency & s/n ratio etc. In the day, they had what is the RIAA standard, which had a low standard of quality control. Basically, they figured if extraneous noises would be washed out by the limitations of the common playback media of the day.

But anyway, now to digital. W/, what in many cases is wrongly referred to as HD audio (market speek calls lo-fi digital HD), we really do have limitations unleashed. When you record soft things, they are more soft (w/o noise) than in the old days & when you have highs, you have to adhere to the headroom of the converter. Any clipping will be really bad. When you are driving an analogue medium (tube, tape, film, etc.) it can have surprising effects, as it reaches that clipping point, it compresses, breaks up & can create warmth & give you certain harmonic overtones, that are pleasing in effect.

So, w/ digital, you can not let that happen & if you reach that point, you have to control that level. If you simply lower the input volume, you also make the more quiet section more quiet. When you record in say 96khz/24bit, you could possibly have a dynamic range of 106 db. W/ tape, you would be more in the 86 db range. (pulling numbers off the top of my head). So, w/ the 20 db range of added dynamics, that means that when you get the digital signal to 0 vu (you more realistically want a -0.2 peak) that means that quiet parts are 20 db lower than if the project was recorded on tape.

This is where we are at today in modern production & mastering. People realize that, now, these quiet parts are too quiet & require some compression. It also goes hand in hand, that the analogue devices moreso have a limiting effect. The problem? People do not know how to use these effects, so they overcompensate & squash everything & it starts sounding like everything is loud (all the time), or... gasp! like it's being played on the radio! If it sounds like it's being played on the radio, before they put their processing on it, it will sound horrendous when played, well, on the radio.

It's an art. Just like everything else. There's a skill to it & finding out that fine balance between dynamics & listenability on a variety of systems, taking in consideration the whole radio dynamic, or lack thereof...

 

[AAHIHaveNoIdeaWhatImDoing]

I understand at least some of that, the issue I'm having is I have 3 different tracks with three different volumes, and they all need to be basically the same volume, but I don't know which is "right" or if none of them are, what's some objective standard to measure against.

 

[JimiRules]

When I say radio I'm talking XM radio which seems to be a little more stable than regular radio. I wouldn't trust regular radio seeing the signal can be inconsistent.

I also load regular songs on my flash drive and use the sound system in my car which is better than anything I have in my house.

 

[AAHIHaveNoIdeaWhatImDoing]

Thanks. I think I have it sort of figured out. . .found some posts somewhere where -10 DBFS RMS is recommended, and there's a relatively new metering system called "LUFS" (Loudness Units Full Scale) that is supposed to be more accurate in reflecting loudness than the older RMS metering. Itunes sets everything to -16 LUFS, regardless of how you bring it in (pushes it down if you give it to them too loud, brings it up if too low). Unfortunately my version of Logic doesn't have LUFS metering, and the OS I'm using keeps me from updating to the version that does have that metering, but fear of losing software compatibility with other programs makes me not want to upgrade the OS. At any rate, I checked the track our friend mastered, it's around -10 RMS, so I've made everything else kinda match that.

 

[JimiRules]

Glad you got it straightened out. Thing like that can be a pain. You take things like that for granted until you actually have a project to work on. Then it becomes clear that there is a lot of little things involved.

 

[AAHIHaveNoIdeaWhatImDoing]

Well, I am not SURE I'm right yet, but seems right at the moment. Maybe I can get a pro's opinion (I know a few professional sound designer/engineer types) before we release this stuff. One of the songs will probably have to be re-exported (the one a friend mixed) 'cause there are some dropouts/glitches in the copy I have, so hopefully I can talk to him then.

 

[Dogs of Doom]

there are plenty of free plug-ins available for LUFS metering...

Free LUFS Meters

 

[AAHIHaveNoIdeaWhatImDoing]

Oh thanks! Didnt know. Probably would have looked at some point, but who knows.

 

[AAHIHaveNoIdeaWhatImDoing]

Ok, the next question is, exactly HOW do you compare loudness between two songs? Take a lower/average volume part of the song, and match it with the lower/average volume part of another song? And do the same with the loudest parts? And which numbers do you go by with the LUFS meters? Seems like the momentary and short term meters provide the most useful information.

 

[Ghostman]

Ok, the next question is, exactly HOW do you compare loudness between two songs? Take a lower/average volume part of the song, and match it with the lower/average volume part of another song? And do the same with the loudest parts? And which numbers do you go by with the LUFS meters? Seems like the momentary and short term meters provide the most useful information.

That's a subjective question, and not applicable as a typical procedure for leveling a mix. A song's dynamics will constitute whether or not it's too loud for a Master or not. Simply averaging the tracks levels could go south, when one song is inherently quiet but has a high output crescendo in it, and it clips after thinking you set the average level properly. After mastering, a simple sit and listen will give the fastest and easiest way to determine if the Masters are out of wack.