Very good understanding of the subject. I'm sorry for poor choice of words in the video - this could be explained a bit more precisely. Though I'd still lean towards saying that if the current correlation between the two sounds is so bad that some of the frequencies cancel out (let's take the low frequencies), removing the low-end from either of sounds could be taken as improving that phase correlation (since the frequencies don't cancel out anymore). Well, one could say that makes sense, someone else could call that "mental gymnastics". It's true the phase shift is non-linear - this is something I show in my other video about OTT compression that I refer to at some point of this tutorial. It's difficult to say whether flipping the phase reduces the correlation or not - it depends on both sounds being used when layering. Also, it's up to one's preferences to say whether flipping the phase of either of the sample changes the correlation in a way the resulting sound is better or not. Using an all-pass filter is an interesting take I've never seen in use yet, though it's still not a perfect solution. We can take for granted that whenever we take 2 kick drum samples, they are going to have a different pitch envelope. Each kick sound can "start" from a different pitch and they can end up with the low-end occupying a bit different frequency range after a different period of time. At some point during playback, the correlation must change for the better or worse. If we used an all-pass filter, at one point the correlation would be ok, but after let's say 50 miliseconds, the correlation would be off again. It's true that the correlation between the kick or bass and other instruments is also something that exists (even though we only talk about that in the context of kick and bass sounds), but the thing is that even if that "correlation" would be wrong in the mid or high frequencies (think of a reese bass sound + a snare drum/hi-hat on top of it), its effects won't be as harmful as a bad correlation between the kick and bass (which ends up with for example a lacking low-end across the whole track). Sidechain compression and gluing things with compression/saturation is of course one of greater ways to mix, but how is one supposed to use that when layering kick drums for a better transient?

@@MarcinRajski- 1. naa, it's really not. if you are removing frequencies from one and layering with another, you are reducing the interference (that's arguable as well)by reducing the phase correlation between the two. for 2 signals to correlate they need to share similar frequencies to begin with for them to correlate in phase at those frequencies. 2.first video of yours I've seen in my feed, so wasn't aware of that. 3.well, that was kind of the point of my closing statement. there is no perfect solution and it really doesn't matter as much as people claim. phase is only the derivative of pitch. they do not follow each other linearly, so your pitch envelope analogy doesn't make as much sense as you think it makes. i can design two kicks with the same pitch envelope trajectory to their fundamental and have them have completely different phase relationships either fixed or overtime if you use an FM/PM synth or do micro-adjustments of the cycles with the pitch envelope or edits in post. they might sound the same in isolation depending on your monitoring setup, but they would correlate differently with everything else in the mix that they interact in time and in unison with. I am also not sure what you mean with your argument against allpass filters. that's exactly what is going to happen if you simply flip the polarity and use minimum phase filters, especially as steep as you are recommending here(with 4 pole and above you are decorrelating the frequencies a considerable amount to the point the correlation between the two amounts to a comb filtered response at the stopband and some ways into the passband. it's the complete antonym of correlated or "in phase). you aren't getting a better correlation by splitting the sound with minimum phase filters and polarity flipping. at least with allpass filters the frequencies will actually be in correlation though if you set them up properly albeit at the expense of group delay. your solution always results in less correlation between the two sounds at those frequencies. if one of the sounds only contains the transient, then what exactly is it correlating with in the other layer? absolutely nothing. therefore if you wish to go down this rabbit hole, its correlation with other elements that coincide with it are more important than the kick thud/doof/ tail itself or bass sound(i am not arguing for this). this is not a bad thing is what i am trying to put across though, but it's also not accurate to call it "more in phase"or whatever you wish to call it. 4. who's "we"? you mean other KZbinrs? I already alluded as to the reason why that is and it's largely because they do not know what they are talking about. I can assure you especially in Techno nobody is really focussing on this stuff. most of the marriage of sounds comes from dynamics processing and distortion/saturation treatments. i've did a few final stem and multitrack mixdowns for well known Techno labels in the past and absolutely none of them did anything that i would consider phase aware. very bread and butter mixing The problem with your understanding of phase is that because phase is nonlinear in complex sounds, any kind of minimum phase filter you add adds more and more local delay at those frequencies. the steeper the filters you use, the more delayed those frequencies become decreasing their correlation with the sounds that play in unison or within the same pocket of time. because the delay is local, you cannot simply just shift it back with a sample delay or whatever, therefore trying to correct phase this way is a destructive process that actually accomplishes the complete opposite of what you set out to do. listen, if people truly gave one about phase, and could actually hear its detriments, then linear phase and FIR filters would be far more common in peoples repertoire when they are producing. but alas, they are not. the only way you can improve the phase relationship between two sounds is to improve their radius. you cannot do it by introducing yet more group delay outside of an allpass filter/ minimum phase phase rotator (1st to 2nd order allpass essentially with pos/neg phase) which lets you set a phase rotation amount independent of the phase shift introduced by the filter you introduced, but at the expense of yet more group delay. unless you want to deconstruct a sound into its individual partials and rotate them individually, there is no way to truly put something "in phase" especially in a complex arrangement, so its better to focus on reducing interference and increasing correlation other ways e.g stereo, aggressive compression to force two sounds to share one envelope (the most common one i know of by Techno producers)parallel saturation to marry their crest factors, reverb, but mostly importantly, being more mindful of the pitches you use when shaping the entire track is another as well, and no. i do not mean have the kick and bass/sub share the same fundamental frequency(this is one example of where absolute correlation would screw you). that sounds horrid on big systems.

Hi! In this video, the reverb is added to hi-hats only.

Thank you for watching =)

Thank you for the suggestions! It's difficult to get artists on my channel, but I'm going to write a few more people soon 8)

Thank you for the recommendation =)

Yes, there aren't many videos like that

Thank you for the appreciation ;)

@@MarcinRajski- You already added distortion with amp and pedal after the adding a low pass filter, could you please explain why you added more distortion with the decapitator after compression? If the goal was to have more distortion in the first place, why not further adjust the amp and pedal? Why bother with the amp and pedal if the decapitator is a higher quality plug in? Thanks, I'm new to this.

:D :D :D

what's wrong with it :-D

@@duppedoore Do I really have to tell you why?

Thanks bro 8)

Dzięki! Nie wiem, skąd GWELD wziął wokal, ale ja polecam mocno brać wokale ze splice. Wpisz tam "male/female vocal", zaznacz sobie tagi w rodzaju "rap", "hip-hop" i zaznacz, że interesują Cię sample w okolicach 140/70bpm i tyle :)

Thank you :D :D :D

Thanks bro

Thank you!