The problem with high pass filters

I actually hear this when mixing but never saw what it did to the sound wave.

cant you just back it down 3 db to compensate?

Quote from christianbad

the thing is: if for example you you make a high pass at 80hz, you are also boosting around 80hz +3db

Well, it depends on the slope of the filter. Might be more, or it might be less.

I wouldn't say this is a "problem" of high pass filtering, just the way they work. It can be advantageous or disadvantageous. This is why we mix with our ears, not our eyes

Quote from Michael_dk

Well, it depends on the slope of the filter. Might be more, or it might be less.

I wouldn't say this is a "problem" of high pass filtering, just the way they work. It can be advantageous or disadvantageous. This is why we mix with our ears, not our eyes

If it sounds good it is good 👍

Filters have been used in recording almost since it’s inception. Just because someone shows you how they actually work, should not mean you should suddenly be scared of using them. They are a very necesary part of production and sound sculpting.

And you are certainly not “degrading” a guitar sound by applying hi or low pass filters. Guitar amps are dirty and degraded by their very nature. All you are doing is helping the FOH engineer to avoid clash in the bottom and “distortion bees” in the highs. The advantages FAR outweigh any “mathematical danger”

Btw, any EQ is also filtering

Quote from Kim_Olesen

Filters have been used in recording almost since it’s inception. Just because someone shows you how they actually work, should not mean you should suddenly be scared of using them. They are a very necesary part of production and sound sculpting.

And you are certainly not “degrading” a guitar sound by applying hi or low pass filters. Guitar amps are dirty and degraded by their very nature. All you are doing is helping the FOH engineer to avoid clash in the bottom and “distortion bees” in the highs. The advantages FAR outweigh any “mathematical danger”

Btw, any EQ is also filtering

Spot on. Use your ears and "watch" out for phase issues when using filters parallel to other paths. What is shown on the scope is phase shift of lower frequencies.

What I thought the presenter was really saying is that we should be wary of just automatically using a HPF at a set frequency on every track (because youtube “experts” say so). This would have the danger of creating a significant buildup of unwanted frequencies at the crossover point. Whereas using your ears and applying HPF where required at the appropriate frequency for that specific signal would avoid this as you would have little boosts in many diff places.

Quote from Wheresthedug

What I thought the presenter was really saying is that we should be wary of just automatically using a HPF at a set frequency on every track (because youtube “experts” say so). This would have the danger of creating a significant buildup of unwanted frequencies at the crossover point. Whereas using your ears and applying HPF where required at the appropriate frequency for that specific signal would avoid this as you would have little boosts in many diff places.

That's also what I got from it

Honestly, I'm not sure of the relevance for using a HP on the Kemper.

It's interesting that a filter way below the fundamental of a sound should significantly change the waveform; that would be an unintended effect. Will probably depend on the specific implementation of the highpass filter algorithm; there may be others that will shape the sound differently. But it seems that a lot of highpass filter implementations do create harmonic distortion, even when set to low frequencies, which also explains the increase in level. As far as I understand, this may have to do with the phase response of a highpass filter - it changes the phase of a signal differently, depending on frequency.

Overall, the increased level can be easily countered by just reducing the level after the EQ, so no real damage done. And if you don't want a highpass filter to change your signal, then don't use it. If OTOH you need to remove some lower frequencies in your signal, a HPF is your friend, and you don't care about any side effects.

What the video does show is that using a HPF when you don't need it isn't good for your signal - fair enough...

The video is interesting at some level, but it doesn't tell the full story regarding linear system theory and filtering. There is no ideal filter in the real world, whether it is implemented in hardware or software. There is also potential to be misled if one does not understand the difference between the frequency and time domain. There are many concepts at play here and this video omits almost all of them. If you really want understand this then you need to understand linear system theory, Fourier transforms, the relationship between the frequency and time domain, magnitude, phase, group delay, sinusoidal superposition, Gibbs phenomenon, etc... Point being this is not a simple subject and this video does not begin to explain what is really happening. In fact, it creates more questions than answers and does not equip the user with any information to address the 'problem'. Probably best to file it in the circular file.

Oh and the these types of effects are not limited to high pass filters. Similar issues exist with all filters.

Or… move the little knobby thing until it sounds good.