Monday, March 31, 2014

How to Mix and Master with EQ – CDSoundMaster.com – 12 EQ Issues Part 16

Audio Recording Issues – Nice sound when tracking, but way too much "-----" when blending parts together, Part Two


So, what happens when we get too much overlap of a given range of Frequencies? Obviously, it sounds wrong. The overall volume of a mix is compromised because it has to make room for a lot of a certain Frequency range, thus the other Frequencies are too low in comparison. Some people attempt to rectify a good mix that is too heavy in a given range by squeezing it flat with compression. This results in one of several bad endings. The mix may distort in reaction to an abundance of spectral tones, or we may get pumping and breathing from Frequencies that were fine otherwise. The point being, that if there is too much build-up in a certain range of Frequencies, they have to be dealt with before a good mix can happen. It is better to identify any conflicting tracks before mixing down, or the result will have to be dealt with at mastering, at which point other parts of the mix may be compromised that didn't have to be.

Why does this build-up tend to happen? As I mentioned in the previous post, sometimes it is the mere fact that multiple parts of the instrumentation or music performance involve instruments or vocalists that are in the same range as each other. This can also happen by using the same microphones and preamps over and over again. It is likely that some of your favorite “go-to” tools are not only the super-flat, super-precise ones, but often are chosen because of their personality. “I love the ----- for bass and the ----- for vocals, etc.” These choices can create intentional boosts and cuts in our mixes by preference, but come mix time, the same hills and valleys are already there. If we don't know when this is a good thing, we might end up doing some truly awful things to the mix.

Sometimes people have no reason to run into the issue until mix time, because of their routine mix practice. Some people like to use high or low pass filters on every track. They say that it cleans things up and always leads to a better mix. I understand the logic and agree that this process is logical and people that are successful with it have their reasons. It is a well planned process for selecting certain Frequencies to cut out of each portion of a mix, so that there is plenty of space for each instrument in the blending together of the mix. However, this can also be the first stage that a person realizes the over-use of other Frequencies, or there may have been better mixing options before removing certain sections of extreme spectral ranges, and now the focus brings rough central frequencies to the surface. I personally avoid using high/low pass on all tracks as a practice for one main reason: there is a lot of intricate timing information captured across all Frequencies. I tend to only change or eliminate things if I know I want the color, musical, or surgical result, and only if it is not sacrificing details that give the brain lots of feedback about timing, placement, distance, etc. But, regardless of what got you there, we are talking about the situation where nothing was a mistake, but there is simply too much x, y, or z happening. So, the toughest question: “what do you get rid of?” You like the balance and you like the individual sound. You don't really want to get rid of any of it, but the mix is simply too heavy on certain Frequencies. I recommend going to the center of the issue, which is to identify the bump.

“Huh? Identify the bump?” It is very likely that the overlap that is happening has a complex texture that is not all built up in a perfect slope across all of the exact same Frequencies. There are bound to be smaller patterns within the overall offending range. For instance, where an earlier example there was too much mid range coming from a lead vocalist, guitar, and tom tom section that were all awesome but sharing lots of mids, I recommend listening to only these elements and follow the rhythmic choices for the specific song. How often is each element running in unison? Which of these are more prominent? Is it possible to turn an instrument track down to lessen the load, or does it need to be up in the mix? Can you carve only one small part of each of them out to make room for the other? Although I will go into that in more detail with a future post, what I recommend is to think of this like the range of Frequencies that you are dealing with is a mountain. We are looking for the molehills, or to put it more accurately, we want to dig little chunks out of the least important sections within that mountain range.

Let's pretend the mountain starts at 500Hz, peaks at 2kHz, and recedes back down at 5kHz. This is not a consistent bell shape, but has the most build-up towards the middle Frequencies. The highest peak represents the Frequencies where the most overlap occurs between the multiple tracks. The greatest issue occurs from notes that sustain the longest, but if any notes stick out that are harsh or cause peaks that force the rest of the mix to a lower average than needed. The goal is to isolate how much to reduce, only from the Frequencies that overlap the most. You may want to only reduce a little of the guitar in one part of the range, a little from the vocal, and a little from the tom toms. This will lead to an overall reduction of the problem Frequencies without having to reduce them all the same amount, or to reduce any other parts of the mix by the same amount. You can also do a small amount of multi-band compression to slightly reduce the peaks in these areas, especially if any of the instruments are short notes instead of sustain. This allows short peaks to come into balance with less noticeable effect on the mix. The entire mix can increase by the resulting change. I have spent many years developing my own process of a more complex complementary Frequency process as well, but for the sake of staying on topic, let's stick with the effective process of reducing the bumps of Frequencies from combined sources by identifying them to their sources, and reducing them at the source, in order to make the mix smoother.

 

No comments:

Post a Comment