Hmmmmmm...multiband compression. Seems this is all the rage these days! Or course, I have not seen a mastering studio (professional) that uses them yet. But, that doesn't mean they couldn't have a use either!
A multiband compressor may or may not do a thing to help with this problem. If it is just a kick drum that is messing things up, well, it might help. BUT, the problem sounds like the drum loops is just too loud eh?
Let's say that it is just the kick drum on that loop that is the problem. Let's say that we apply a multiband compressor to deal with that. We would probably focus on the 60-150Hz range of the audio to deal with that kick drum. It might be a little tighter of a range, and that range might shift up or down a bit, but indeed, it will be somewhere around there. Cool.
Let's say that kick drum has attack at around 3KHz. Hmmmmmm...I don't see the multiband taking care of THAT. Yes, we got the boom out of the kick, but the attack will be unmolested. The attack will be just as present as it always was. In addition, while the compressor is doing it's gain reduction in that frequency range, ANYTHING else in that frequency range is going to be gain reduced too. Oops. There goes quite possibly the meat of the bass guitar. There goes the resonance of the guitars.
Well darn it, let's complicate this whole thing! JUST the kick being too loud is too simple. More than likely, the snare is popping to much too! So then we concentrate yet another band of compression on?............Well, do we need to get rid of the boom of the snare? The crack of the snare? The sizzle of the snare? Which? All of them? Hmmmmmm...another two bands of compression. 4 bands of compression now. Oops....But wait! What about all that OTHER audio going on? ALL that other audio sharing the frequency ranges you use the multiband on will of course be effected. Is that acceptable? Probably not.
A mix is a mix friends! If the relative level of something in a mix is way too loud, unless that part it restricted to a VERY defined range frequency wise, ANYTHING you do to lower it will effect other parts that use that frequency range while that processing is doing it's thing.
Now, let's fall back a bit. I KNOW that a few here are already formulating their "that sonusman shit doesn't know what the hell he is talking about at all!" reply.
Feel free to post that reply. No problem here. BUT, I think an investigation into the actual PROBLEM is in order here before multiband compression is the issued cure! MB compression is NOT the cure all in the mastering phase, and is used far LESS than most would protend. I have messed with it a LOT in mastering, and found it mostly uneffective. While it sometimes takes care of a frequency range's overly hyped dynamics, usually used for low frequency control, I found that it makes the higher frequencies sound harsher.
Why?
Interesting question.
I had an interesting problem a year ago with a project I mixed/mastered for a client. You see, the producer wanted to mix with a LOT of low end. He was quite unfamiliar with the room and the monitors we mixed on, and kept insisting on more low end. Fine and dandy Mr.Producerguy! You can have your low end! But with that flabby low end killing everything, we of course started lossing "definition" in the mix, which was a byproduct I warned about over and over again. So, along with way too much low end, all other instruments need a more "cutting" sound to compete. We needed definition in the mix so all other instruments had a lot of "beef" carved out. Well, aside from the fact that the overall bass level was too much for "real life", the mix of course sounded "phat" and "big" in the control room. Of course in "real life" playback systems, it was very muddy.
So, in mastering, a LOT of low frequencies needed to be cut from the mix. Multiband compression wasn't the ticket here because frankly, it wasn't an issue of a few certain frequencies jumping out here and there. THAT would be a good use of multiband compression, to grab a very narrowly defined frequency range that SOMETIMES, jumps out over the rest of the mix.
In the case I am referring to, eq needed to be cut.
What was wrong with this scenario is that when those frequencies were cut, well, it is gone for EVERYTHING. You see, he wanted to bass guitar and kick drum to be really bassy. No problem with that, except that they were unappropriately too loud. Thus, with the bass being bassy all the time, the only solution was to use cut eq on it.
That of course pulled low end away from other instruments that sounded just fine in the mix the way they were. But now, those instruments sound much harsher because their guts were carved out from the cut eq.
Now, while I had the bass guitar and kick mostly even with other instruments in the mix, the problem was that all the other stuff sounded more or less harsh and hard.
A bad mix is a bad mix. Applying ANY processing to a overall mix will created problems with balance in the rest of the mix.
If the problem is that a part of the music is just plain too loud, and that part fills in a lot of space in the time line (continuous notes, etc...) ANYTHING you do to deal with that is going to effect everything else.
Here is another example that may be a little more pertinent to this case listed above.
I had the priviledge to work on a bands project doing some digital editing to the tracks last summer. They mixed with somebody else at a pretty nice studio in town. They mastered with a VERY big name mastering engineer.
The mixing engineer did something sort of interesting. EVERY kick drum hit came in about 6-8dB louder than anything else in the mix. I could see some logic in that IF the attack would have been a lot lower than it was. It was of course the "boom" of the kick that accounted for the huge increase in volume. But to accompany the rather large "boom" the attack on the kick was in addition very "present". Almost as much too loud as the kick was. This is where the problem started.
When the mastering engineer got to work on it, he of course had to do something about that kick drum being 6-8 dB louder than anything else. I know it was mixed this way by design. Fair enough. They WANTED to kick drum to trigger some compression and create a "pumping effect" in the mix. I have heard this approach used many times and if you are into that type of thing, it works very well. The problem was the attack of the kick was about 3-4dB too loud compared to the rest of the instruments in the mix that share the same frequency range as the attack. Oops....
So, the mastering engineer succesfully compressed/limited the kick drum, and got the desired "pumping" effect from that compression. He was then of course able to bring up the whole overall volume of the audio. Balance was more or less create now, EXCEPT that the attack of the kick drum was VERY annoying now because it was STILL 3-4dB louder than anything else sharing that frequency range. Oops..... Of course quickening the attack of the compression could have helped, but that tends to take away from the "pumping effect". So, the attack of the kick drum came on through the compression unmolested. Bummer.
I could go on and on and on about how dynamic processing has it's drawbacks in audio. They really DO when used to the extreme. Even in moderation, they can have problems too if not CAREFULLY applied. Even when carefully applied, they can create other problems.
Mastering is not easy task. Making decisions about what problems you SHOULD take care of is no easy task, and is VERY subjective. How you solve that problem is every bit as subjective too.
No easy answers. A lot of "depends". A lot of "lesser of two evils" IF pristine audio is sought. Of course, if just solving one part of an overall complex problem is sought, there is usually easier answers.
A drum loops could potentially fill in a lot of the time line in music. This is a problem, not unlike if a vocal part was just plain too loud. Trying to fix a drum loop that fills a lot of space in the overall music and is too loud is going to have drastic effects on the overall audio because potentially, the drums fill up a lot of sonic space too. It can potentially be too busy of a part to effectively "turn it down" after the song has been mixed. With the range of sonic space it probably takes up, ANY solution to "turn it down" would be mostly unsuccessful because what ever you do to it to "turn it down" will effect everything else that shares the frequency range. If those loops are specific to a narrow frequency range, such as possibly a 808 kick drum, which has very few overtones and little attack, you could zero in on the fundamental frequency of that kick and use multiband compression successfully on it. But most kick drums have a more complex wave form than that. Once you compond this with other drums in that loops that are probably too loud, then you have more frequencies to deal with. Soon, it just really comes down to the fact that it is just plain too loud! It potentially covers far too much of the frequency range to deal with it.
Sounds like the song needs to be remixed. If that isn't possible, well, you need to make some sacrifices in the overall song or just leave it alone and call it good. There is NO way you can make that drum loop "mix" better with the rest of the parts unless you remix it to be that way.
I feel like the grim reeper right now. LOL
Ed