Do I need compression now? (pre-mastering kind of question)

[travelingsales]

I am mixing my band's record right now (melodic rock, not too loud, not too quiet) and we are working with a "real" mastering engineer for the first time. One thing that has come up is that our mixes are not very loud. They sound pretty good, in my opinion anyway, but our bounces are currently sitting between -20 and -18 average RMS, with peak RMS about -12, and peak sample around -1 to -0.5. So we have a lot of dynamic range, probably more than is ideal. The ME has said our average RMS is on the low side and I should ideally be shooting for -16 average RMS, maybe up to -12.

I think the average volume is low because the songs are mixed with almost no compression. For example, I have one compressor on vox bus, 3.5:1 ratio, max of about 5-6dB gain reduction, maybe 1-2dB makeup gain, that's it. Then a similar setup on the drum bus. And that will be all the compressors for a whole song. I have volume automations, but nothing else compressing or limiting.

I mixed it this way mostly as an experiment, because I admittedly don't really know wtf I am doing with compressors and have squeezed the life out of mixes in the past when using them badly. So I decided to see how far I could get without using any compression, and the answer is apparently "pretty far", albeit resulting in some pretty "live" and raw sounding tracks.

Anyway I guess I am just looking for some advice. I like the sound of the recordings, but they could definitely use some "tightening" and I think compression is what we need. I just am not really sure where to start - I guess add compression to the bass? And maybe increase the compression a bit on the few things that already have it? Another confounding factor is that about half the tracks have just a stereo drum stem due to limited A->D during tracking, so I can't compress say, the kick independently from the OH's as they are all bounced to stereo already.

Thanks for any insight!

[MoreSpaceEcho]

if it sounds good to you i say just leave it. if you were saying "man my mixes are all over the place, especially the bass, some notes just jump out like crazy and other ones disappear, etc etc" i'd say "you need some compression". but it seems like in this case you'd be compressing mainly to get your RMS numbers up, which seems silly. turn off your monitor, forget what your ME said, just listen to everything and see if you really think it needs more compression.

personally as an ME i would loooooooove to get more records with RMS "on the low side". i can't honestly remember ever getting anything that was too dynamic to work with, but i certainly have gotten many things that were way too squashed.

can you have your ME do a test master of one song or half a song so you can see how it sounds post-mastering?

also, fwiw i pulled out faith no more's 'angel dust' recently, and was kind of amazed at how good it sounds, so i did some googling and found a couple posts from matt wallace where he said that he used almost no compression during tracking or mixing (just a little bit on bass, snare, and vox) and they did all the squash in mastering. considering that you say you don't really know wtf you're doing with compressors, maybe this sort of approach could work for you too.

[travelingsales]

thank you for your advice!

there are definitely some places where the bass (for example) is louder than other places, and notes that pop out a little more, etc. Personally I don't find this to be distracting as the performance is generally pretty solid. But as I mentioned the overall sound is definitely a little more raw than most good mixes I've heard.

I think what I may do is back up the project file as it exists now and then do another iteration where I allow myself some more freedom with compressors and see how that comes out (both subjectively and RMS-wise).

I think one thing that is getting me a bit is that I have not sent anything to mastering before and I feel kind of silly just ignoring the ME's advice. Advice is one of the things we are paying him for. But at the same time I don't want to compress things just to make them louder. If it tightens the overall sound and the RMS goes up a bit as a side effect that's fine, I am just not super confident I can achieve that.

[MoreSpaceEcho]

you know, if your RMS is hanging out around -20 but you're still peaking almost at 0, your "low rms" could just be that you have a few stray peaks that are sticking out way over everything else and killing your overall average level. watch the meters on the master fader and see what's up. if they don't peak over -6 the whole time except for one snare hit at the end, you can likely go in and automate that hit down a bunch and get some headroom/ a higher average rms.

[sir hills]

Hey all. New here...long time lurker...2nd time poster.

I just finished a mix with a similar approach to compression (minimal on vocal, kick, snare, bass & acoustic guitar...& just a touch on the mix buss) & automation proved to be my savior in mixing & prep for mastering.

if they don't peak over -6 the whole time except for one snare hit at the end, you can likely go in and automate that hit down a bunch and get some headroom/ a higher average rms.

Wise.

Also look at peaks that could be a build up of several instruments hitting at the same time or are creating a note/tone that peaks above. The band I was working with is VERY dynamic...from a single acoustic with flat wound strings being quietly finger-picked to rock band + horns, strings & a 50 person vocal choir playing/singing at full volume...all in the same song. During mixing I started experimenting with automating the master fader slightly so I didn't have to strain to hear the low volume stuff & so that the louder parts didn't get too out of hand. Usually never more than +1 or -1...maybe +2 for something almost inaudible. You have to be careful to be subtle & use transitions in the song to come up & down & still sound natural with a similar space. Also, if there are tracks with nothing happening (& without bleed that makes up part of the sound of what is playing) you'll want to cut or lower those parts so you don't get a rise in "dead air" along with the rise in what you want to hear more of.

That being said, it doesn't work in all situations but it did work for this one. The ME was pleased with what we brought in, mixes that are dynamic & not squashed but have a decent overall RMS...& the client is happy with the end result. Also, my client had a limited budget & it helped save a little time in mastering. Anyhow, I mentioned this since you are making new files specifically for experimenting. If it sounds good to you send examples of both to the ME & see which he/she prefers. Give it a shot.

[travelingsales]

update!

at least for the song I was initially talking about, I got the average RMS up to about -18 from -20 without adding any more compression. Just more volume automation and more attention to which tracks could be pushed without unbalancing the overall mix or clipping the master. So still a little below what the ME is saying is ideal, but louder than it was. Frankly I would rather be a little too quiet than a little too loud anyway.

thanks for the advice, now my new thing to obsess about is apparently harsh sounding cymbals :p

[lionaudio]

thanks for the advice, now my new thing to obsess about is apparently harsh sounding cymbals :p[/quote]

And the greatest antidote for that would, of course be, ribbon mics as overheads into a stereo tube pre!

[travelingsales]

Just to add a little more about my experience with this, having worked through another song now -

The hardest thing for me has been dealing with drum transients. We used Glyn Johns for the drum micing and it sounds good, nice natural picture of the drum set. Our drummer is good about mixing himself but there are random snare drum hits here and there that are 3-6dB greater than his average snare hit (which is itself louder than most of the cymbal crashes). The snare transients, both the "regular" snare crack and especially the loud ones, are the main thing forcing me to keep the overall level down. The other thing is that when he hits the low tom hard, we get gigantic spikes in the left OH mic. I don't know if this is a common problem with that micing technique, or if we had the left mic too close to the tom (probably the latter).

Anyway what I have been doing now is bouncing the mix to stereo and then looking at the resulting waveform. Usually the snare and low tom transients are pretty obvious. I have then been going back to the OH tracks and manually automating the volume down just on the transients. Sort of manual compression if you will.

I also experimented with using a limiter type thing on the OH's (really fast attack, fast release, high ratio, threshold set to catch snare hits but nothing else) but didn't like it for some reason I can't put my finger on. The volume automation method is tedious but seems to sound better so far.