What am I doing wrong???

[tonesponge]

Everytime I get my mix done and bounce it down and transfer it to my jump drive and listen to it on my laptop, the loudness is so low I can't even hear it. Normalizing it to zero dB doesn't help much because I have some ridiculously huge transient peaks on the beat.

If I put a compressor and limiter on the stereo bus almost to the point of audible artifacts, and bounce down through that, I can get it quite a bit louder, but then I still have to turn the volumes on my laptop up all the way to hear it, and other people in the room can't hear it well enough to even understand the lyrics.

Recently, I tried normalizing the mix to zero dB instead of compressing, and then applying multiple stages of the backwards compression technique to address those transient peaks (compressing, bouncing down, importing track, reversing, compressing, bouncing down, importing track, reversing again). That really got it loud without obvious artifacts, but it really seemed to mess up the balance of the mix; for example, the formerly crisp drums sound muddled and boomy.

I'm considering trying just bouncing down the stems, and then using backwards compression to bring those individually up to whatever point they each can tolerate well, then mix them, bounce down, and normalize. But that will take a lot of time to do, so I would appreciate some guidance here before I go potentially waste a bunch of time.

Thanks in advance.

 

[Massive Master]

As much as the whole "mastering" = "getting it loud" thing just continues to bug the hell out of me (as it's really the least significant part of the entire process), it doesn't just 'happen' by any means.

The ridiculous levels that people are looking for --- I'm sorry --- That artists and labels are looking for (the "people" never asked for this and if they knew what was really happening, there would probably be some sort of uprising) --- Those levels don't come easy (especially digitally).

I suppose the obvious thing would be to find out the source of those transients and start there (in the mix).

 

[SouthSIDE Glen]

If I put a compressor and limiter on the stereo bus almost to the point of audible artifacts, and bounce down through that, I can get it quite a bit louder, but then I still have to turn the volumes on my laptop up all the way to hear it, and other people in the room can't hear it well enough to even understand the lyrics.

Recently, I tried normalizing the mix to zero dB instead of compressing, and then applying multiple stages of the backwards compression technique to address those transient peaks (compressing, bouncing down, importing track, reversing, compressing, bouncing down, importing track, reversing again). That really got it loud without obvious artifacts, but it really seemed to mess up the balance of the mix; for example, the formerly crisp drums sound muddled and boomy.

A few points that stick out right off the bat to me:

First, the assumption that a mix should be perfectly audible on laptop speakers is somewhat misguided. A lot depends upon the laptop of course, but I know that on my laptop (a pretty nice Toshiba Satellite), most perfectly good commercial songs from my playlist - not homemade stuff, even, but Grammy-class commercial stuff - play back at low volume with barely audible vocals. It's the stuff that sounds like absolute crap everywhere else that is nicely audible on laptop speakers.

Second, I'm with John. Take care of some of the out-of-control transients in the mix - and handle at least some of them manually - instead of waiting to clobber them with compression in mastering. This will increase the effectiveness of what compression you do apply in mastering with less loss of tonal control when you do so.

Third, while the "backwards compression" technique is an interesting one, and can have it's special purpose uses, there should be no reason why you should have to go to the extremes of multiple passes of backwards compression just to get some volume out of your master.

It's not surprising in the least that the "balance in the mix" gets screwed up; that is exactly what such extreme smashing *does*. That *is* high-compression artifating; you say you have no artifacts but that in fact is the number one artifact of extreme compression. You need to address the lower-level issues that are coming to the fore under compression by first cleaning up your individual tracks, and second, using some corrective EQ if/as needed before each compression stage in mastering in order to get rid of those problem low-level frequencies and distortions introduced by each pass of compression.

Fourth, there is absolutely no need or advantage to using peak normalizing as part of your process. When you're throwing that much compression into the mix (with associated makeup gain), any potential effects of peak normalization - which are almost nil to begin with - are completely washed out anyway. It's like spitting in the ocean as the tide is rising; there's no point to it.

The results: Tame your most unruly peaks in mixing, using no compression to do so. Skip the normalization. Clean the low-level garbage in your tracks before you mixdown, and EQ before each stage of compression. Keep each stage of compression gentile, and stop compressing when you start sacrificing too much sound quality for loudness. It's juts not that important. And finally, stop mixing and mastering on laptop speakers or using them as a judge for whether your mixes work or not Great mixes DO sound like crap on those.

G.

 

[RAMI]

I'm not saying the only way to do it is to slap a limiter on it. But, if you do slap a limiter and you can't get it to a decent volume without messing up the mix, then I don't think mastering is your problem. The problem is your mix. It's not ready for mastering.

 

[tonesponge]

I know that a laptop sounds like a strange metric, but I'm talkign about a pretty good gaming laptop, and pro soundtracks always sound just fine on it. I guess I should say that modern pro soundtracks always sound good on it. I'll take a listen to some older works in a bit to see what's going on there.

In any event, my mixes, and a lot of other amateur mixes I hear on it, are too low in volume to perform and also lose so much bass that the bass track is practically inaudible. So I don't think it's completely unreasonable for me to include the laptop in my collection of platforms to determine whether my mix is up to modern pro standards.

I've started working on the stems, and it's become clear that the drum loops I'm using don't respond well to much compression. I don't see any way to deal with the snare spikes in that premixed drum track without doing some compression or limiting, and that brings up those frequencies that make the track really boomy.

For now, I've done some motown style compression on that track, to load it with frequencies that can handle compression (while making it louder). I'm basically using a parallel track that I put a brickwall limiter on and smashed way too much and then EQed out all the bad stuff. Then I mixed that back in with the original track. I guess we'll see how it turns out.

Thanks for the input. Got some more?

 

[SouthSIDE Glen]

So I don't think it's completely unreasonable for me to include the laptop in my collection of platforms to determine whether my mix is up to modern pro standards.

It doesn't manner what artificial "standards" you wish to use to judge your mixes - as if what other songs sound like has even the slightest, tiniest bit to do with what your own mix sounds like - using a laptop to make or to even judge your mixes is a just plain awful idea.

That aside, since I doubt I will ever change your mind on that ladder, don't attack your most wild peaks with compression. I'm talking the highest 10% or so of them that seem to have gone wild over the averake of most of your peak levels.

When they are peaks that are caused by a combination of peak in mixdown, go back inside the mix before mixdown and manually - bot with compression - knock down the component peaks in the individual tracks that are causing those wild compound peaks in the mixdown.

Also, similarly manually knock down the individual rogue peaks in the individual tracks in the same way.

Sweep all your tracks with a paraEQ and remove all the honker frequencies in the tracks before you mix down, especially in those frequency areas that the mixdown compression seems to badly emphasize. And for god's sake, don't do this step on a laptop or on any computer speakers.

When you do compress the master buss to bring down the rest of the peaks, do it in small bites, using corrective EQ in between to keep the sound in line.

And stop comparing your mix to commercial mixes. If you want your song to exist in a playlist with the big boys then you need to work the whole process from the beginning like the big boys, or just give it to them to do for you. It ain't gonna happen by just throwing some mastering tricks at your stuff and checking it on laptop speakers.

Or better yet, figure out for yourself what sounds good instead of judging your own original work by whether it sounds like someone else or not.

G.

 

[SouthSIDE Glen]

Accidental double-post. Mods please delete.

G.

 

[tonesponge]

don't attack your most wild peaks with compression. I'm talking the highest 10% or so of them that seem to have gone wild over the averake of most of your peak levels.

When they are peaks that are caused by a combination of peak in mixdown, go back inside the mix before mixdown and manually - bot with compression - knock down the component peaks in the individual tracks that are causing those wild compound peaks in the mixdown.

Also, similarly manually knock down the individual rogue peaks in the individual tracks in the same way.

Thanks for this tip. I thought I was supposed to use the pencil tool and that has not been working for me at all.

Sweep all your tracks with a paraEQ and remove all the honker frequencies in the tracks before you mix down, especially in those frequency areas that the mixdown compression seems to badly emphasize.

I wasn't sure how to find these frequencies, but I guess using the sweep in combination with a compressor that is set way too hard, and then backing off on the compression once the frequencies are found, is doing the trick. I wonder, is it generally better to sweep with an eq placed before or after the compressor?

And for god's sake, don't do this step on a laptop or on any computer speakers.

Don't worry, my Daw is running on a desktop with real monitors (if M-Audio SP5Bs can be called "real monitors"). I don't really have a good room though, so I like to try and check my results by porting to many platforms to see what the mix sounds like.

When you do compress the master buss to bring down the rest of the peaks, do it in small bites, using corrective EQ in between to keep the sound in line.

Oh yeah! That's the kind of guidance I can really use. Thanks very much for this one.

I'm a bit concerend about transients forming through the attack window on the compressor. It seems that part of the waveform doesn't get compressed at all, and it just gets spikier and spikier with each stage of compression. How do you stop that from happening? Do you use a brick wall limiter or do backwards compression in between compression stages? Or do you bounce it down and then automate the fader on the result and compress again?

 

[RAMI]

I wasn't sure how to find these frequencies, but I guess using the sweep in combination with a compressor that is set way too hard, and then backing off on the compression once the frequencies are found, is doing the trick. I wonder, is it generally better to sweep with an eq placed before or after the compressor?

Forget about the compressor. For a parametric sweep, just narrow the Q and boost the gain and sweep through the frequencies using your EARS to LISTEN for the frequencies that "honk" or "sing" too much.

 

[SouthSIDE Glen]

I'm a bit concerend about transients forming through the attack window on the compressor. It seems that part of the waveform doesn't get compressed at all, and it just gets spikier and spikier with each stage of compression. How do you stop that from happening? Do you use a brick wall limiter or do backwards compression in between compression stages? Or do you bounce it down and then automate the fader on the result and compress again?

I don't use anything. I don't use smashed up crap as my reference. When it's a choice between sounding good or sounding loud, I'll take good 10 times out of 10.

But if you think that asking the listener to actually use their volume control so you don't have to destroy your mix just to satisfy their lazyness is asking too much, then stop compressing the whole signal and concentrate on what needs to be done. If your transients get out of control, get rid of them. Compressing the non-transients won't do you any good if your transients are your problem.

Compression and limiting is neither your problem or your answer. The problem and the answer is in your mix. Get rid of those transients and clean up the rest of your signal before you even consider compressing and limiting. Because if you don't do that, all the compression and limiting is going to do is just what you're getting; a mix that turns to mud.

G.

 

[RAMI]

To reiterate what Glen's saying, I'll repeat what I said earlier...maybe it will be ignored again.

The problem is your mix. It's not ready for mastering.

 

[tonesponge]

To reiterate what Glen's saying, I'll repeat what I said earlier...maybe it will be ignored again.

Not ignored Rami! I went back to my stems and figured otu that the premixed drum track is causing the problem. So I guess I basically have to master that stem more or less separately. If what I'm doing now doesn't work, I'll try something else.

But part of my problem is this.

Get rid of those transients and clean up the rest of your signal before ...

OK, so now I know to automate faders to address transients, but what does it mean to clean up the signal? Are we talking about reducing peaks, cutting overlap frequencies between parts, increasing SNR somehow? A lot of the time, I'm just not sure what stuff means.

 

[tonesponge]

I finally finished remixing my tune. I basically did parallel compression on each track with a lot of EQ, and did some tedious track automation. I got a mix I liked, and then tried to maximize the loudness with compression and it fell apart again.

So then I went all hillbilly on it and put that remix in a session and used a trim plugin to ramp up the gain on the master WAY into the red until I could HEAR distortion and then backed it off until it sounded tolerable. The result is the loudest mix I've ever produced, and it seems a lot better to me than my previous attempt.

If anyone wants to listen to the before and after before getting all mouthy about what a retarded hillbilly I am, please see this thread: https://homerecording.com/bbs/general-discussions/mp3-mixing-clinic/moar-fuk-pleez-316274/

 

[buzzard bass]

To reiterate what Glen's saying, I'll repeat what I said earlier...maybe it will be ignored again.

Not by everyone me ole China.

 

[Lt. Bob]

wow ........ off topic but ya'll have made me feel a lot better about not using compression!