Aktualisiert: 21. Nov 2020
If you are using multiband compressors or dynamic EQs, both of which are similar in the way they work, you have to make sure that you get your attack and release times right, dependent on the frequency range you're working on...probably more so than in a conventional single band compressor.
You might be using a dynamic EQ to tame low mids by reducing 200hz when it exceeds a certain threshold. You might also be using a dynamic EQ to control 4000hz when they get too harsh. I'm known to call it 4000hurts...you can assume why :) And for a third use case, let’s assume you are trying to balance out 12000hz when it gets too “sizzly”.
Generally speaking, relative to each other, you will want to use slower attack and release times for the lower frequency work and conversely, faster attack and release times in higher frequency ranges. That is due to musicality and "naturalness" and the fundamental nature of sounds in the different frequency ranges, how they are perceived as slower and faster. Imagine (or play, if you can) a bass and notice, even if it’s just in your head, how bass is slow and smooth. Even a hard hitting bass drum is slower and smoother, in comparison to a lead sound or even a hihat, which has a fast and loud or “hard” transient and is over with so much faster.
BTW: A bass drum is one of those instruments that can have a high pitch short sounding “tick” to it with a loud transient. But it will also have a much longer and more sustained “body”, which makes the punch and pressure. If you were to work on both parts dynamically, I recommend to use slower and faster reaction times for the according parts.
Let’s look at all three examples in a plugin called TDR Nova, a free and absolutely amazing parallel dynamic EQ. This is a mighty weapon in my mastering arsenal but it’s useful mixing as well, especially with recorded material.
If you look at the lower section, on the right hand, right after Threshold and Ration, the Attack and Release times of each band can be set to different amounts. For lower frequencies I will use higher amounts, which result in a slower response of the dynamic EQing (EQ cuts in this case). Attack of 10ms or even up to 20ms is where I have it very often, with a Release that’s always way over 100ms and can go up to 200ms. When I go much lower in frequency, into the sub bass region, I might even use 30ms & 200-250ms.
If I was using this on a bass or guitar (maybe when there are some resonant palm muted parts) or on vocals or really anything that has been recorded, the build up in that low and low mid area is not one with a very “fast” character. If you imagine a hihat or a crashing cymbal as “fast” and transient-heavy, this lower region is rather the opposite. A slower gain reduction in the lower frequencies sounds much less obvious and more “natural”...whatever meaning you want to insert into the word "natural" is up to you ;)
In recorded instruments, sometimes it's a ring that you want to reduce but if dynamic gain reduction is too rapid in the low end and on a punchy sound (like a bass drum or tom or a low snare for example), a too fast reduction can suffocate some of that initial punch that you want to let go through.
Next in line is the mid or upper mid range. Very often, and I never make the conscious decision, it just turns out that way, but the Attack and Release times will be around half of what I would use in lower frequency bands. 5-10ms & 80-150ms will often be just fine for me. The musical content in this region is not the very fastest, so the reaction to it doesn't have to be superfast, but it sure is much faster and less smooth than in the lows. This could be a vocal or guitar being too resonant and harsh. Another application could be in mastering, when there is too much - again - “harshness” overbearing to the ear. Especially if you’re processing the whole stereo sum dynamically, any fast and rapid changes become much more apparent...hence, different reaction times to different frequency regions. Thinking of the spectrum going from slow (low end) to fast (high end) makes sense in that regard.
This is a relatively fast region, figuratively as well as literally. I have mentioned a hihat a few times but I want to include any of those transient-rich sounds that come in fast, don’t last long and disappear quickly. Even non-musical sounds like clicks and crackles are transient-rich. I would often use 0.5-2ms & 15-30ms to tame those sounds. Take these amounts as a guideline and play around and try but to listen really carefully and focussed.
Getting sensitive to little details takes some time but it can be trained and it's crucial for becoming excellent at your art/craft.
So the principle is the same for multiband compressors because they do a similar thing. They reduce gain at a chosen frequency range, maybe just not as specific as a dynamic EQ. From my experience I would say a multiband compressor is often used for broader gain control. But depending on which frequency range(s) you want to set up, you will want to make sure to have your attack and release times set up well for the same reasons mentioned above.
If you open up any stock or third party dynamic EQ or multiband compressor, you will most likely see the default preset using slower to faster reaction times from low end to high end. Any of those plugin types that don't have these time adjustments as a feature, in my opinion they are missing the point.
TIP: The heavier a gain reduction you let your dynamic EQ or multiband compressor apply, the more careful you will want to be with being too fast because the change will become more and more audible and apparent. And again, especially when used in mastering, on the whole of a stereo sum, that can end up sounding disturbing.
Any questions ? Please hit me up at via email or leave a comment / use the contact form. If you think I haven't explained it properly, please let me know.
Until next time
Author: Robert Hundt // Date: April 16nd, 2020