Please help with some mastering common practices.

[bluedaffy]

This is my first go at mastering a home project just to see what its all about and hopefully learn some things about it and understand the art a bit more than I had before. A few questions I would love to get some input on. Any and all responses are welcomed!

The songs that I'm working on are primarily rock songs. The way you approach mastering a song (or album) depends on the song (or album). This is something that's been ingrained into my skull from reading around on the mastering forum threads, and this is probably the most important thing I've read, so I will keep it in mind during my process. Having said that, it'd be nice to get some input on how some of you find yourselves commonly approaching different problems in the mix to a arrive at your mastered tracks. I see it this way, if you can help me by offering some of the common techniques you like to use, I can implement them, maybe they will work, maybe they wont, but they will most certainly help me to find out what about it didn't work, or what about it did work. Kinda like a science experiment thought process, no data is bad data, by getting some starting points, I will be able to tailor it to the albums needs.

Here are few questions that would be a big help to me to hear some opinions/advice on:

1. Say the song is very dynamic where there are very loud sections followed by very quiet sections. Is it ideal, or a common practice to have a compressor on to help control the loud sections, but then do nothing at all during the quiet sections? During mixing I would have no reservations about something like this, but because mastering is more about cohesion and making everything feel as one, I'm kind of inclined to believe that all the parts of should be subjected to the same processing? Am I off base on this?

2. If the song has a 'peaky portion' (one song of ours has a bridge where every instrument slams down on every quarter note for 2 measures) how might you go about dealing with it? The problem being that the peaks during that portion of the song significantly restrict the possibility of the rest of the song from being brought up in volume.

3. If a song is being treated twice by compression, one serving as a leveling compressor, and one serving as more of a conventional style feedback compressor that works on the dynamics but leaves the transients alone more or less, which is commonly placed first in the signal chain?

4. In your opinion, generally how much gain reduction on a limiter does it take to make you start to feel like 'that's too much'? And here's a big one I've been struggling with! -> What's the deal with the 'soft clip' function on a limiter?!?! I've read that it makes the limiting sound more natural, like analog clipping. If so, why is there even an option to use it?!? If it sounds more natural what would be the benefit of not using it?!??

5. My limiter has INPUT, CEILING, THRESHOLD, and RELEASE knobs. How do the first three interact exactly? I've noticed from messing around with the knobs that if I set the ceiling to say -.02, I can turn the input up as much as I want and it will never clip the master output (although it sounds nasty-squished). So where does the threshold come into play? If I turn the threshold down, it also makes the song sound louder, much like if I'm turning the input up, so I'm assuming that what is happening is that the limiter is 'limiting' where the threshold is. So what is happening between the threshold and the celling? What is the purpose of having a threshold AND a ceiling?

Thanks so much for reading this over and hopefully giving me your $.02, I apologize for the abundance of questions, if you can help me out with responses on even one or 2 of them, it would be so helpful.

Thanks again -bluedaffy

 

[Massive Master]

1) Do what sounds right.

2) Do what sounds right.

3) Do what sounds right. That said, typically, I go for transients first. THAT said, I never know how many stages of gain change or dynamics processing I'm going to use until I hear it (then I do what sounds right).

4) How much it can take is completely dependent on the material and the limiter. Some recordings sound like crap (dynamically speaking) before a limiter is even in the chain. Some give up with a dB or two, some sound reasonably decent after 10dB. Totally dependent on what's hitting it and how it deals with it.

5) I'd start by reading the literature from that particular limiter.

I almost wish there was a sub-forum that was "This really isn't about mastering, I'm just obsessed with loudness" or something. "Loud" is easy -- Loud is an afterthought. Not that it can't be relatively challenging occasionally with some mixes -- But the mix determines its volume potential -- Sure, it's usually up to the mastering engineer to find that potential -- but there isn't a set protocol. You know the tools, you listen, you pick the tools that are appropriate for that particular mix (keeping all the other mixes in mind), you set it how the mix needs it. Just like every other phase of audio production.

If this is your own project (I'm not even going to go into how I feel about mastering one's own mixes), get the mixes sounding like you want them and then slap a limiter on and drive it till you're happy with it. If it "deserves" the loudness, it'll let you know. If it doesn't (and many, many, many mixes do not), it'll let you know that too. Slap a limiter across the mix buss and smack it 6 or 8dB in when you're mixing and see how it reacts. It might react well, it might not. Hopefully you'll hear why it reacts well or not and you can go in and try to make compromises -- For something that few care about (except bands and labels).

[/rant] Sorry. One of those days...

 

[bartekkck]

2. you can manually turn down those peaks or use compressor
sometimes is better to use 2 compressors than one working hard for example you want to turn down peaks by 6db and if you'll do it with one compressor you can hear it working and that's not what you want in mastering, it's better to use 2 compressors that do 3db of gain reduction
you can also compress side signal if that "peaking" instrument are not panned in the middle

4. it depends but you don't want constant gain reduction. it should be sporadic and not too long

 

[k2shei]

5. Just a guess but your limiter might act as a loudness maximizer - basically a limiter that instead of cutting off peaks rounds them off. I wouldn't know unless i played with it but I would guess that the input controls how much input comes in.

If the threshold is below the ceiling it rounds off any of the sounds that reach the threshold. Finally the ceiling is where it acts as the actual limiter which will stop any sound reaching a specific decibel.

Test my theory out and tell me.

 

[bluedaffy]

k2shei- I looked into it further and finally found some documentation on that limiter and I think you're right. It said when the signal passes the threshold it is hit with a 20:1 ratio, so I believe that is what your guess was. It sounds like the threshold acts as a very steep compressor and the ceiling is the true upper limit like you stated. Thanks.