High pass on mastering does not equal to more headroom

[chagzuki]

When it comes to mixing, yes, pretty much high pass everything besides kick and bass. The art is figuring out how much to remove, and it can be a little tricky around the 200-ish Hz range making decisions. Depends on what the style demands in terms of priority.

I could never figure out whether it was good practice in mastering since I don't really have the monitoring tools to be tinkering around the threshold of inaudibility.

[imrae]

I don't think any respected mastering engineers would recommend a super-steep highpass like the one in the thumbnail.

Did I say that?

You said "no one can be right or wrong", which is... a bit unhelpful in the circumstances? Yes there are valid uses for steep filters to intentionally disrupt phase and create peaks. But if you use such filters with the intent of creating headroom and then act surprised that they interact with bass frequencies to create peaks... you were wrong.

[mitchiemasha]

It also messes with your tight side chain ducking one may have set up. Any low cut after the sidechain will move the low frequency later, meaning the tail moves more into the transient.

Discussing this on another forum... I made this rough video.





It's worth adding, this phase shift of the lowcut could also make things line up better, moving the bass ever so slightly back, filling the "gap" more perfectly, this would be very freq specific of the cut, not simply adding it, and would take serious critical listening if you wanted to go for that level of actually caring!!!

[legendCNCD]

High/Lowpass or/and High/LowSHELF what is needed. Its just that simple. Filters when needed, shelves when needed. Nothing when not needed

[ramseysounds]

I don't think any respected mastering engineers would recommend a super-steep highpass like the one in the thumbnail.

Did I say that?

You said "no one can be right or wrong", which is... a bit unhelpful in the circumstances? Yes there are valid uses for steep filters to intentionally disrupt phase and create peaks. But if you use such filters with the intent of creating headroom and then act surprised that they interact with bass frequencies to create peaks... you were wrong.

So 2 times you make something up that I said and 2 times you call me out for it?

f**king hell; thats gotta be a record huh

Please show me where I mentioned anything about steep filters. I’ll wait…

[imrae]

OP showed steep filters. It's the thread topic.

For the second post, you are reading the word "you" too literally.

But if one uses such filters with the intent of creating headroom and then acts surprised that they interact with bass frequencies to create peaks... then one would be wrong.

Is that less offensive?

[vurt]

i might high pass this thread, at about 40k. should get rid of all the noise.

[ramseysounds]

Good shout. Au revoir

[camsr]

Some (or most) FIR highpass filters do not completely remove DC offset, but take it down so much that it is to be considered negligible.

[The Chase]

I agree with the general consensus that "blindly highpassing everything is harmful but circumstantially it can help a lot, depending on what you're doing."

In DnB production it's practically gospel because sampling vinyl has a lot of low frequency content, as do heavily distorted complex signals where intermodulation creates low frequencies. So does getting weird with unconventional synth mangling. So most of your channels will warrant highpassing.

I'd also like to echo the warning of phase in this thread. The oft-repeated conventional wisdom to highpass right under your content might be twisting the phase of your fundamentals by a quarter period or more, which is a disaster for punch and clarity.

Like anything else it comes down to the circumstances and the details of the execution of the technique.

[mystran]

I'd also like to echo the warning of phase in this thread. The oft-repeated conventional wisdom to highpass right under your content might be twisting the phase of your fundamentals by a quarter period or more, which is a disaster for punch and clarity.

It's also good to keep in mind that significant non-linearity of phase typically extends quite far into the passband where the magnitude response is already more or less perfectly flat .. and the steeper you make the filter, the worse the phase is typically going to look.

[legendCNCD]

I'd also like to echo the warning of phase in this thread. The oft-repeated conventional wisdom to highpass right under your content might be twisting the phase of your fundamentals by a quarter period or more, which is a disaster for punch and clarity.

Yeah, and more often than not, just lowshelf at 0.71 is enough to get the needed frequencies off the way. Sometimes only highpass - some synth patches emit under 20hz signals "nice"

[ON BEATS]

No filtering on the 2-bus here either. When I put one of my fav producer's tracks through a spectrum analyzer I noticed he has a lot of frequencies in the below 30Hz area. I stopped giving a shit since then.

[Ah_Dziz]

I bet I can drop your mix to any amount of headroom using a high pass filter... Lol!!

But yeah. You have to have something to remove to have any major effect on the level of a track with a filter. Then you have phase to consider. When it comes to processing an entire track there are very few "always do x" type rules. If you are attempting to produce a rough master or a self release though it is generally true that you may want to eliminate signals that are there and have no audible change when removed. This would normally be before any major dynamic processing and literally only something that you can measure but not hear when attenuated or fully removed.

JJ

[camsr]

A good way to make a mix more "low-ending" is to start with sounds that have low end. If highpasses are used everywhere in the production chain, all the low end eventually disappears. Steep highpass filters only make this exclusion of low end happen faster, to the point those low frequencies will never be recoverable.