Cut 12 vs Cut 98 on HP filter -- question for the pros here

Nola

Well-known member
Hey guys. I'm having some trouble with my low end, and wondering the pros and cons of a 12cut vs a 98cut (I also have other options like 24, 48, etc). On the graph the 98 cut looks steeper. Are there downsides to using this? When I use the 12 cut I still hear some low end rumble in the guitars. With the 98 cut there is less of that low rumble. So the latter seems better, but are there cons?

Thanks!
 
Hi Nola. Steeper filters can produce artifacts and cause more abrupt dropouts in energy of stuff like a kick drum or a bass guitar. Sometimes we use special filters that compensate for this energy loss by boosting a portion of the frequency spectrum right infront of the cutoff. Here's a thread with an example where I was talking to Boz from Boz Digital Labs about this:

Bark of Dog plugin question - Recording, Mixing and Producing - indie recording depot

General rule of thumb would be to start with the shallowest filter you can deem effective. But filter as steep as is necessary to eliminate the low end rumble. But if you have to, go ahead and use the 98cut unless there's a compelling reason not to.
 
...12cut vs a 98cut...


When you say "12cut vs a 98cut" are you referring to the center frequencies of the low-end bands...or something else?

If so...what EQ are you using that has presets at 12, 24, 48 and 98 Hz....? Seems kinda odd.

Or are those the dB amounts of cut?
Seems like 48dB and 98dB would be pretty heavy...?
 
I believe it's the steepness of the curve of the filter, expressed in decibels per octave. For each octave of frequencies covered by the curve, the signal is attenuated by -12dB (for example).
 
Hi Nola. Steeper filters can produce artifacts and cause more abrupt dropouts in energy of stuff like a kick drum or a bass guitar. Sometimes we use special filters that compensate for this energy loss by boosting a portion of the frequency spectrum right infront of the cutoff. Here's a thread with an example where I was talking to Boz from Boz Digital Labs about this:

Bark of Dog plugin question - Recording, Mixing and Producing - indie recording depot

General rule of thumb would be to start with the shallowest filter you can deem effective. But filter as steep as is necessary to eliminate the low end rumble. But if you have to, go ahead and use the 98cut unless there's a compelling reason not to.

Thanks! I'll check out the thread. Are the artifacts right away noticeable if there's a problem? I didn't hear anything when I changed the cut. Would there be any reason why sometimes there are artifacts and other times not? I will read the thread and see if this is covered.

Miroslav, when I say cut 12 or 98 I'm just parroting what it says in the Cubase Frequency plugin (their default EQ). I assume they mean DB. If I choose cut 12, it is a gentle slop on the visual graph. If I do cut 98, it's a very steep slope where it seems to let little below it through.
 
Are the artifacts right away noticeable if there's a problem?
Yeah. Basically what you want to do is listen closely to the source as you apply the filter. And when you toggle it on and off or sweep you cutoffs you want to ask yourself "did I just do anything to hurt the source" and "did I just damage anything" and "do I hear it start to do anything that sounds weird or un-natural". In short, use your monitors and trust your ears. If you make a mistake no biggie...you can always go back and change it later.

I didn't hear anything when I changed the cut. Would there be any reason why sometimes there are artifacts and other times not? I will read the thread and see if this is covered.
Yes, this is partly because all filters are not designed equally. Part of what I asked Boz about in the other thread was why there was complex math in the coding algorithms for a filter on a plugin. I was surprised to hear that these are not simple devices.

It really has to do with the way filters interact with the source. Lets say we're hi-passing a kick drum to control the information below 60 hz and keep it from pegging your bus compressor disproportionately hard. You would need to pay much closer attention to the filter than if you're simply dumping an air conditioner hum out of a vocal mic in a bedroom. Because in a vocal, the isn't going to be anything usable under 100 hz anyway. Does that make sense? Like with a vocal you can almost always just chop hard with a steep slope and forget about it. Different story with a bass guitar or Kick drum or 808.
 
That does make sense, thanks. So is it fair to say the steeper the slope the more that resonant peak/artifact will be, and that's the downside of a bigger slope? So if I have two guitars and want a 98db slope at 200hz, I might actually get build up at 200hz? I have two guitars and wanted to cut all the rumble out since with two it was really building up. Conversely, the pro of using the steep slope is that you can get rid of more (possibly all) rumble in the low end?

If you do get a peak from using a steep slop, can you use another filter to bring that peak down?
 
It comes down to what you are trying to do with the HPF. Me, I pretty much restrict them entirely to eliminating anything in a track that has no reason to be there, so like to use a pretty steep curve on those. If I want to change the amount of energy down low, I might add a shelf (i.e., rather than tinker with the HPF's curve/slope). This is not an unbreakable rule, of course, but just my preference.
 
Miroslav, when I say cut 12 or 98 I'm just parroting what it says in the Cubase Frequency plugin (their default EQ).

OK...I get what you mean now. It was just the way it was worded, I wanted to be clear what you meant.


It comes down to what you are trying to do with the HPF. Me, I pretty much restrict them entirely to eliminating anything in a track that has no reason to be there, so like to use a pretty steep curve on those. If I want to change the amount of energy down low, I might add a shelf (i.e., rather than tinker with the HPF's curve/slope). This is not an unbreakable rule, of course, but just my preference.

I still like to fine tune any kind of broad stroke cuts...I mean, I don't apply anything "by default" until I can sit down and compare and see how it sits with/without.
Also, for vocals, I may apply a HPF on the preamp with vocals...depending on the mic and the vocal...so it's a lot of fine-tuning, especially when dealing with low baritone vocals. There may be harmonic content down there below....and the same thing with bass or kick.
I'm just not a "cut everything below and/or above by default" kind of guy. :D

I've seen people HPF shelf on just about every track...and then wonder what happened to the low end vibe.
I'll go with bell curve on the low-end often times, and tweak the slope of the curve and amount of cut...rather than just roll it all off with a shelf at a given point. Then sometimes a HPF shelf is the ticket....same on the other end, but I rarely do shelf cuts on the high end.
 
*Assuming* it's [dB/Octave] -- If you need anything more that 6 -- or 12 on an extreme scale, figure you've got problems that probably aren't easily solved. When you get into elliptic filters (24dB and more per octave), you sort of need a specific reason for that sort of stuff. Mid-side vinyl cutting for example. Still -- it's a rare occasion where you need something that steep even with vinyl.
 
*Assuming* it's [dB/Octave] -- If you need anything more that 6 -- or 12 on an extreme scale, figure you've got problems that probably aren't easily solved. When you get into elliptic filters (24dB and more per octave), you sort of need a specific reason for that sort of stuff. Mid-side vinyl cutting for example. Still -- it's a rare occasion where you need something that steep even with vinyl.

That was kinda my original curiosity and question when I saw the "98cut"...I mean, that is quite steep-n-fast.

I'll use very narrow bandwidth cuts to remove clicks/noises/artifacts from tracks...etc...but I'm not going much past 20dB, and even that is only on the worst offenders...but I don't use drastic stuff like that just for adjusting and EQ'ing for tone.

I guess in the end...if it sounds right to you...
 
I just wanted to get rid of all low end rumble on the guitars b/c there are double tracks, and it was adding up. I never saw the 98db cut so was curious what it was all about. At the 12db cut it still felt muddy around 200hz. Like the two guitars where building up there. I guess I can try a shallower HP and then notch the low end more. Mixing the low end is the hardest thing for me...it never sounds clear and defined to me no matter what I try. I recorded it under okay conditions so I don't think it's the recording. Might be the arrangement having too many guitar parts, that there isn't a lot of high end in the mix so maybe the low seems exaggerated, or maybe it's the drum sample and not really knowing how to EQ drums very well. I've tried a lot of things from books and YT and they just never sound right in my songs. When I compare my mixes to pro mixes the biggest problem is the lack of definition in the bass and mud, no matter how much I cut down there. Oh well, the original question was answered, so thanks!
 
Also remember that on a hi pass or low pass filter, the frequency you are dialing in is not the center frequency, it is the frequency below which the cutting takes place. So, if you need to cut 200 Hz, you will need to set it for above 200 Hz to cut it.

Gentler slopes tend to sound more natural, but in order to get a significant cut at a certain frequency, you will need to set it higher than that.

A steep slope will get rid of more unwanted stuff, but add more at the corner frequency. It's a compromise between shaving things down and chopping them off.
 
*Assuming* it's [dB/Octave] -- If you need anything more that 6 -- or 12 on an extreme scale, figure you've got problems that probably aren't easily solved. When you get into elliptic filters (24dB and more per octave), you sort of need a specific reason for that sort of stuff. Mid-side vinyl cutting for example. Still -- it's a rare occasion where you need something that steep even with vinyl.

"[FONT=&quot]while bands one and eight also offer cut filters, with slopes all the way up to a whopping 96dB per octave — that’s very handy if you want to filter off everything below/above a very specific frequency."

[/FONT]
 
Instead of a steep slope (with the attendant 'boost' at a frequency above the cutoff point), put several instances of EQ on the track, each with a shallow slope, but at the same frequency.
 
Back
Top