Mastering tips?

I´m trying to do sort of a master of a song. Just for learning purposes. I haven´t done to much of mastering.

 

The song is about 8 minutes long, sort of post-progressive stuff. The first couple of minutes has drums and some guitars, there´s a lot of space and air in the mix. Then the bass kicks in, and then some heavy distorted guitars and in midway through the vocals enters.

 

I´ve chained 4 compressors the Waves C1, the Renaissance compressor and two more C1´s, with between 1-1,5 dB attenuation each. Slow attack, faster release. After that i have some subtle EQing and the adaptive limiter at the end with i think 1.2 on the scale. Perhaps 8 dB on the gain knob.

 

It sounds pretty good, but one thing "goes wrong". When the heavy distorted guitars and bass kicks in, it´s naturally supposed to lift the track and make it sound louder, which it does in the mix. Of course the compressors will make the levels of the tracks lower, so when more instruments enters the audio the opposite happens and the track sounds lower. This is all logical to me. But the question is how to avoid this; keep the loud spacious drums in the intro, keep the dynamics from the mix and still compete in the loudness war, at least a bit. hehe. Should i exaggerate the dynamics more in the mix? Or is there another route here?

 

Once again i turn to the wisdom of this forum.

I generally get the loud part of the song sounding the way I want it. I also do a lot in the mixing process to get dynamics right and to tame peaks so that in mastering I'm not limiting/compressing things that much. Also, Bass triggers compression faster than the higher frequencies so you might want to use a multiband compressor with less compression applied to the bass region.

When you say you've chained four compressors together do you mean that the first compressor compresses the song then the second compresses the song which is already compressed, etc. etc?

You'd be far better off with a dedicated three or four band compressor where you can do as Daveyboy suggests and target your compression to individual frequency bands.

yup, that´s what i meant. Each compressor doing just a little bit. I guess I´ve read about this technique in far too many places. Maybe not as common technique as many sources would have me believe?

 

I´ll try utilizing multibands instead and work on the mix so that it doesn´t get screwed up when i do the limiting. Thanks guys. More input is valued!

I don't really see the benifit of chaining if any. All you're doing is compressing an already compressed track, even though each compressor is around 1.2 the accumulative figure is around 5, Thinking about this I would guess the figure is much higher because again of the accumulation, not only is this too high but the sound will suffer.

You should notice a big difference if you use a multi band, you are then working on separate frequencies instead of the overall volume of the track.

If you can automate it for different sections all the better but I have found I rarely need to do this.

I see your point. Makes sense. I will take your advices into account and experiment a bit more. Trying to get it right, hehe. thanks again!

You can be sneaky and start the quiet part fairly high in level, and then gradually lower it as you get to the loud part. When the loud stuff kicks in, you'll have more impact.

I can try that too. I´ll see what gives the best result...

I am not very experienced but are multiple compressors that odd of a thing? I would think in some cases it would be necessary to use more than one compressor on a track. For example if the song was very peaky one compressor with a fast attack that only shaves a few db off the more extreme transients and then a second compressor set up with a slower attack to add cohesion. Am I off base?

 

I guess what i am trying to say is that 2 comps with the same settings in a row seems to me like it would be un-needed but 2 comps set up to accomplish different tasks seems like it would be ok to me.

Using several compressors in mastering is quite common.

 

You need to think about why you're applying each stage of compression and what that particular type of compressor and the settings used is bringing to the table. Maybe you're doing upward or parallel compression to bring out details or fatten up the track, or maybe you're trying to create punch with a slower compressor, or simply trying to reduce dynamics - or create a tighter feel.

 

So each stage isn't simply about reducing dynamics and spreading out the gain reduction but rather trying to achieve a variety of things, depending on what the track needs.

 

Personally, I have a wide variety of compression options available. I use a combination of digital and analog. I like to use the Flux Solera for low level/low ratio compression, and the Flux Syrah for parallel compression and parallel de-expansion. I use either the Flux Alchemist or Waves Linear Phase MB for multiband dynamic processing. I use the SSL XLogic G Series hardware compressor for mix glue or punch, and the Crane Song STC-8/M for general mastering compressor duties. Sometimes I even use the Gyraf Gyratec tube equalizer for compression by running it flat and slightly overloading the inputs, in which case I use it early in the mastering chain since it also affects the frequency and phase quite a bit.

 

Using several compressors is also very common during mixing on individual tracks.

 

It's not uncommon to use 1 compressor/limiter during tracking (such as a hardware 1176), then 2 during mixing (one with a very low ratio and low threshold, and one with a higher threshold and higher ratio), and then top it off with a bit of brickwall limiting.

 

That channel could even be routed to a bus for sum compression and another limiter, the latter mostly for headroom- not sound.

 

If you do it right and the source sound benefits from compression (many pre-processed samples don't) then it can work very well. I often do just that when I'm mixing.

Thanks a lot for your reply Lagerfeldt. Very insightful. You seem to know quite a bit. Why don´t you come over to my place and we´ll talk more and have a beer? I´m buying.

One of my cousins live in Norway but you can pop by my studio in Copenhagen if you're in the neighborhood.

Damn, i was in copenhagen a month ago.

I´ll keep that in mind anyway. What´s the name of the studio?

Popmusic.dk

Awesome!

Interesting threat....

 

Is it feasible to use automation in mastering (thresholds, EQs...) to follow the dynamics and frequency spectrum of the mix? Or is it better trying to find one "ultimate" setting throughout the song?

 

TX

Is it feasible to use automation in mastering (thresholds, EQs...) to follow the dynamics and frequency spectrum of the mix? Or is it better trying to find one "ultimate" setting throughout the song?

Automation can sometimes be necessary in mastering but it's rarely necessary to do advanced automation.

 

If your solution is something that apparently requires constant and therefore advanced automation then there's likely another and better approach. This could be using some type of adaptive processing, e.g. multiband compression/expansion or dynamic EQ.

 

Simple automation of EQ, thresholds or gain/volume is fairly common in mastering, e.g. changes between intro/verse/chorus or perhaps a particular spot that needs attention.

 

In the latter case you sometimes do a separate bounced edit which is spliced into the the final master or mix at the offending point in time.

Thanks for sharing your expertise with us Mr Lagerfeldt. Very informative and straight to the point, like all of your post. I am learning so much here...

 

Tx

You're welcome!