The Cleanest, Most Accurate EQ Possible (and you’ve never tried it)

Click the link below to start making mixes that accurately translate what you hear in your head.
www.sageaudio.com

[@kostyakonstantinoff]

After using this secret method, my mixes became much cleaner. I just sent one of them to one of the major labels and after 20 years of silence they finally offered me a contract, a castle by the sea and a Lamborghini. This is how one video can change your life. Amazing.

[@Studeep]

I just hope you were paid handsomely for this exceptional marketing work

[@MrAydinminer]

@@Studeep 😂😂😂😂😂

[@sadhill38again]

amazing i hope you have a great and long carrer

[@shaferbeats4478]

😂😂😂😂

[@Bboy_cult]

LOL!!!!

[@theCloneman5]

As to your guess, what RX is doing is Fourier editing/filtering. There is no value for certain frequencies in an audio file by default, just amplitude per sample, but this can be Fourier transformed to represent amplitude per frequency, which is what you see in RX. Now editing this and transforming it back to audio yields the results you hear. This also isn't perfect though, as the audio cannot be transformed at infinitely high resolution. The signal is split into "bins", certain frequency bands of which there can be fewer or more of, resulting in a higher or lower resolution transform, especially in the low end. Higher bin counts also mean much more processing power/time is required. This is, why you cannot reasonably do this type of filtering in an insert/online, but rather as an offline process that has to be rendered.

[@LYSHEmusic]

Actually, Reaper has the ability to do non-destructive spectral edits to the audio items.

[@steamer2k319]

Regular digital EQs should also be using Fourier transforms... If RX is doing something different, maybe it's the size of the default look-ahead /look-behind windows used by each? I don't think RX worries about operating in real-time at all.

[@nj1255]

This. I don't know what resolution the FFT in RX is, but I would guess that it's at or above 2048 bins. That's a moderate resolution for a 44.1kHz source, but it's still far from perfect. If it's 2048 bins, that would mean that each bin spans ~21.5Hz.

[@Marfix_ds]

Sorry but i don’t really understand, what is the difference between Fourier editing/filtering and regular EQ editing/filtering?

[@NoahHornberger]

@@Marfix_ds regular filtering sends the whole wave through the same filter and removes what does not pass it. FFT splits a wave into its mathematical components to build hundreds of little waves that combine to form the original signal. Then the processing is done on a smaller wave, thus not effecting the whole system as much. Still, far from perfect, but a change to say 80hz only effects the wave that is passing through that region, not all the other ones in tandem.

[@CypiXmusic]

The more technical videos of yours are the most interesting and insightful!

[@StevieBoyesmusic]

The reason for the leftover content when you eq the modulated 'sine wave'... Is a real sine wave is infinitely long with no level changes. Amplitude modulating the signal, by changing the gain, adds sideband frequencies, which extend higher the faster the transition occurs.

[@baptistechatel2569]

At last, somebody here understands. This is not the right test signal for the job.

[@Keroser1983]

@@baptistechatel2569 How about the examples at the end? For the real songs and music. On the instrumentals I could not pick up so much difference but on the vocals it was so clear.

[@baptistechatel2569]

@@Keroser1983 i can definitely hear differences. This is not unexpected, as the two filtering methods wildly differ. Classic EQ have a slope (in dB/octave), which means that you cannot have, say, a 6dB difference from 1000Hz to 1010 hertz. The FFT EQ applies gains to small frequency ranges (called "bins") ; from one bin to the next the gain difference can be huge. I'd love to compare the spectrum content of each clip. Having the dry clip would be interesting.

[@happyshadow]

@@baptistechatel2569 the beauty in changing the amplitude of a portion of the spectral range in offline FFT editors is you can tailor the window size and type for the source material. Definitely not a convenient method but can come in handy in some situations

[@cachelesssociety5187]

Thank God you exist bro! I knew this video was misleading and was about to go on a wild goose chase trying to verbalize what you said in one sentence! DDMF talks about this issue with their EQ's & I think DMB as well. "In its minimum phase setting, GrandEQ manages to do just that to an extent larger than most other EQs, and notably both in the AMPLITUDE and in the phase domain, while introducing only a very small amount of latency. The result is a very open and clear sound with, at the same time, excellent CPU efficiency. Its linear phase setting stays completely in the time domain, using an elaborate time-reversal algorithm and 2-fold oversampling to ensure excellent amplitude and (flat) phase response with the lowest possible amount of digital distortion." LP10 is also a great DDMF EQ with controllable "flexible phase".

[@teddycook1299]

RX filtering is definitely still introducing pre-ringing, and a lot of ringing in general. It really is equivalent to a very high quality linear phase EQ. The visual feedback from the spectral display can help point out issues quickly, but the only away to alleviate the ringing is the same as in a brick wall Fabfilter HPF: make the slope gentler. To do this in RX, there is a "Feather selection" tool. You can try something like 0.5-octave feathering. Not sure what the equivalent db/oct is, but I'm sure you could easily get a similar result with a high quality linear-phase EQ. For example, I tried cutting ground loop noise from a high-gain guitar part, using RX's "select harmonics tool", and after filtering, whenever signal is present there is audible ringing at the "notched" frequencies. Tried again with a feathered selection and the ringing was less audible but the part was missing too much. And too bad spectral de-noise sounds like a crappy mp3, even at conservative settings. I found it was better to leave the noise in the mix for that particular guitar part, it was only for one section anyway, and it sounded great in the mix. For critical low end filtering of the sub-bass found in bass and kicks, I've been turning to TDR Infrasonic. It's not any better than other digital HPF methods, but its focus on options specific to that region helps me to get the results I'm going for quickly.

[@Reneromero08]

You're overthinking it.

[@Freedom4Ever420]

If you turn up the gain to like 120 db , you can see the ringing and other artifacts caused by doing this.

[@manofgoat]

In dubstep we adapt to these extra harmonics the saturation and distortion helps us build the sound but in most genres they try to keep it perfect and clean, great video!

[@GianlucaSibaldi]

RX like all spectral tools uses FFT (fast Fourier transform) and inverse FFT to go back and forth from wave data to spectrum. I would not do it too many times because that process has a limited “frequency resolution”, which can degrade the signal if applied repeatedly. BTW I am already annoyed by one round of FFT and inverse FFT processing… Actually it’s the same kind of processing which mp3 encoding-decoding is based on.

[@Bboy_cult]

i can hear it

[@donotoliver]

Bruh my comment got deleted Anyway in short, fl's Edison has the same thing, it uses linear phase filters, only introduces some pre-ringing (less with smooth mode). No fft bins, no lossy artefacts. Very easy to integrate into workflow, ive been using it for 5years now to low cut raw vocals

[@Keroser1983]

I would be interested to know that more. Would you recommend any tutorial? Or would you care to email me for tips & tricks on Edison? I am sooo behind with Edison to be honest 😒

[@hipethenomad]

The Normal EQ sounds more musical, but RX just sounds so crispy clean!

[@joesalyers]

Yes, using dual modulating all-pass filters is how digital EQs make HP/LP filters work in the digital domain. I could write a book on how this was a good way to solve the problem in 1993 when Waves came up with it in Q10 but in the modern world 30 years later we still haven't figured out how to solve this issue. Every DSP cookbook explains this in detail. But you took the longest most arbitrary and frankly more confusing for way to explain it to musicians and engineers who aren't DSP coders. RX has spectral artifacting that is in my opinion worse since 99% of those plugins are based on the Opus codec standard from Xiph, since it is open source and free to use and no extra cost to the developer its the standard. Spectral processing has its downsides too which can be heard as lossy artifacts similar to using a codec filtering for AAC/Mp3 or Vorbis. HP/LP Filtering continues to be a debate in the audio coding space since there is simply no way to truly mimic it in the digital domain with a 1 to 1 like other tools such as compression or even regular EQ curves like shelves and bells which digital does perfectly well.

[@happyshadow]

@@joesalyers The guy who developed Spectralayers says you can mitigate pre/post ringing/all artefacts caused by drastic amplitude changes if you understand how to use the software properly.

[@OTBB047]

When do you use speclayer, ? In normal processing or mastering?

[@joesalyers]

@@happyshadow I'm not talking about pre-ringing whatsoever. I'm talking about how that high/low pass filters work in the digital domain. HP/LP filters are actually made in the digital realm by using dual offset modulating allpass filters. 2 all pass filters 180 degrees off set based on frequency divided by time. The other issue was just Spectral processing uses an open source codec that is Lossy (opus) so it can work in real time. If Spectral plugins used lossless codecs under the hood it would take an hour to render a 3 minute stereo audio file, so until we get a more efficient lossless codec spectral plugins are always going to have that weird spectral sound from the Opus codec running under the hood. Don't get me wrong Opus is the best codec for lossy audio since a 64kbps Opus file will sound as good as a 320kbps Mp3 or AAC file but its still lossy and how much of that creeps into the output of the spectral plugins is anyone's guess. So I have no problem with digital EQs but I think we need a bit of innovation since most of the code we use is from general ideas pioneered nearly 30 years ago in the early 1990s.

[@happyshadow]

@@joesalyers who told you Opus was used? That's for audio transmission over the internet.. Neither RX or Spectralayers encode audio into Opus otherwise it wouldn't null/form the original file after the back and forth conversion. Also Spectral editing is an offline process that converts the file into FFT so that the full spectrum doesn't get altered (which I've null tested and holds true). The whole point of the video is that filters that change the phase aren't used. If the FFT window is set correctly and the image is focused in the correct way phase shifts, pre/post ringing can be avoided. As it's an offline process it's much more tedious but for people looking for an artefact free method is desirable.

[@joesalyers]

@@happyshadow To answer your first question Who told me Opus was used its not just for transmission over the internet stop repeating the first thing that pops up on a google search? The Open Source GPL licenses that all spectral plugins and applications are built upon which are public domain. Opus isn't the output but the real-time sidechain of the actual processing. Xiph hold the ONLY open patent on spectral audio analysis in the public domain so they all use it. You are being combative just to argue on a subject that isn't even my original posts concern. I was highlighting the fact that there is no such thing as High and Low pass filters in the digital domain, it is a clever trick of using all-pass filters to achieve the same goal. You are hung up on spectral processing using the Opus codec (which can process lossless but not in real time) This is why most of the spectral apps use the lossy codec for realtime processing or its offline like the apps you mention. But still no matter if its Spectralayers, Clarity or RX or anything else. HIGH AND LOW PASS FILTERS UNDER THE HOOD ARE ALL-PASS FILTERS NOT ACTUAL HIGH AND LOW PASS FILTERS. I'm done this is clearly you arguing about something I don't care to repeat ten thousand times. I don't care about spectral apps or plugins. The original video didn't have the knowledge to notice the filters in plugin-doctor were being modulated because they are ALL PASS FILTERS! Spectral plugins are cool, use them if it works for you thats great. But the simple fact that he missed what makes HP/LP filter modulate is astonishing. But I don't care about spectral plugins I know how they work, I've worked as an audio engineer and a DSP coder for over 20 years. I was trying to explain why he was having the modulation from "digital HP/LP filters" but he didn't even know the reason why and it is clear what is going on under the hood. Apulsoft built one of the very first spectral plugins for audio and then improved it to their spectral exciter called apUnmask. Apulsoft explains spectral processing somewhat in the manual and more in depth in their white papers on the subject if you want to know more. Unlike Izotope, Waves & Steinberg who make you write them a letter to receive a physical copy of their open source licenses (which is a violation of GPL3 but no one is going to sue them) Apulsoft do not hide it and display it proudly. Some companies like to hide the fact they use open source tools but others do not.

[@mfcfbro]

This makes a way bigger difference than I expected.

[@lukedelalio3155]

Interesting video and a good, clear explanation. I hate to say it, but I found the RX examples less appealing than the EQ'd examples. The bulk of my experience with EQ is analog, and the RX didn't have a pleasant sound to me. I'm also not sure if clean and accurate was something I ever set as an objective when making records. Still, very interesting and I appreciate you taking the time to do this. Luke Korneff Audio

[@alsoulmusic]

Wow! Great Video. Never thought to do this. I have RX but I also have spectraLayers that I can use with ARA in Cubase.

[@alexdiakov_ambient]

Same is useful for resonances cleaning - easy to see the bright area, easy to choose it and put the gain down. RX is a best tool for such actions. Quite often it is faster and easier than using Soothe or EQ dip.

[@anatomicallymodernhuman5175]

You’re wrong. I have tried it. It can be the best way to eliminate a steadystate tone. It is not artifact-free, though. And EQ sounds more musical for 99% of all applications.

[@kenvives]

I really like this idea for my initial repair stage eq. Not only does it seem cleaner, I think I’ll be able to avoid having to automate eq changes from song section to section. Thanks!

[@sageaudio]

You're very welcome!

[@CreativeMindsAudio]

I mean rx is an incredible tool! Been using it forever now. Not necessarily for EQ but if I want to remove some unwanted sound or noise, it’s definitely the easiest way to EQ it out. I’ve met several top level engineers who spend hours cleaning up files on RX like this. Also for pro tools users they got it built into the daw now with ARA support.

[@TheRussianGenius]

This is crazy. I read years ago on forums about linear phase EQ, so I knew the benefits of it, but wild to really hear them two side by side, and plus the RX is a game changer. First couple of mins of your video is by far the best mins I've spent learning this year on EQ. The visualization is great!

[@paulluckey6997]

Woah. Thank you. i have Spectra Layers in Cubase and just never played with it. Now I have a direction, :)

[@m4gn3t.0]

awesome... ive never seen those problems but have always heard them :p some blokes tried to tell me there's no peaks at the cutoff... thanks for keeping on top of these details :)

[@GustavoM.D.]

This Izotope RX Spectral Delete trick can be done with the same quality and accuracy in Audacity's "Filter Curve" Effect. I already do this mastering trick in Audacity since 2020.

[@iam-music]

Great vid, thanks. You knew the counterpart of RX is built into spectral layers for free? As an ARA? Spectral Layers 1, cut down but does all that out of the box. Really appreciate the time you took to do this.

[@americanantagon1st]

RX applies a normal linear-phase filter to separate the selection. The only difference with FabFilter's linear-phase EQ is the frequency response shape. RX is more surgical in frequency, while Fabfilter uses more "analog" shapes, with a certain number of dB/oct. Still, RX has similar "artifacts" to any other linear-phase EQ (pre- and post-ringing). Like me, you appear to favor RX's lack of frequency leakage, which is probably due to surgical EQ shapes. Very similar EQ shapes are also available in Kirchhoff Eq, which measured the have the least noise @ 96dB/oct. of all the popular EQs that I tested. However, I did modify the settings, by using Linear phase and 117bits internal processing option. I hope to hear back from you on this!

[@cuttybang415]

Wow… I’ve used RX for this type of thing, but never very much and certainly not very often. Might have to revisit

[@mbayasi]

great video. i have been doing something similar for over a decade now. for vocals, i just go to the frequency spectrum, highlight and delete everything below 60. I don't know how different it is than using izotope, but it works

[@Keroser1983]

How do you do that? I mean which tool are you using?

[@mbayasi]

@@Keroser1983 I use adobe audition for mixing. they have a frequency analysis spectrum. there, i do it. I also do a lot of de essing, removing unwanted clicks or voices there with a lot of trial and error:)

[@mbayasi]

@@Keroser1983 frequency spectrum analysis on adobe audition

[@TotalOrcAnnihilation]

Stunning difference, thank you

[@KingTreyOhOne]

Can you cover Phase issues and what they are and what to listen for? I want to understand this better, Please and Thank You!!

[@sageaudio]

We already have some videos on this topic! If you just search "Phase" in our video search bar on our channel you should be able to find them

[@philippgrunert8776]

Now I wonder if the spectral editing in reaper basically does the same thing... And if I could translate Eq moves to the spectral filters

[@fizzorff903]

This is absolutely fantastic. Such a huge difference

[@olivermaier-landshut3047]

Beside that using RX or Spetrasonics for eqing is a great idea, I never saw great mixing engineers do that. So overall there is more to this art the just the technical aspect.

[@paul35ism]

Bro this is so geeky and in depth. I love it. But for all practical mixing the “cleanest” method won’t be used 99.99% of the time, just slap linear phase and finish the mix lol

[@iam-music]

I think it would be worthwhile ie as a batch process on definites like vox and guitar etc as a totally clean check before starting the mix

[@negvey]

Omg I've been doing this for few years and I noticed how much cleaner it came out... I thought I might have been wrong but watching this makes so much sense now.... I don't using izotope but Edison in FL studio, but I'm sure it does the same, even before the video was going to reveal this method I was thinking hmm is he going to do what I've been doing...

[@AlexPicciafuochiRainbox]

A game changer!!!

[@TracingFlares]

This would be worth checkin if you are already good at mixing and mastering. For those who have problems with mixing.. this is not going to fix any. In other words there are literally millions of wonderful mixes out there "suffering" from artefacts and phase issues both harware and in the box... learn to mix, to reference and you may not care at all about this. But hey...if you are at a good level of mixing and mastering this can be very useful and plugin manufacturers should incorporate this approach in their EQs, comps and more. This could be a method that completely takes place...as a better method. So Steve... you may hire someone to program this for you... if you are quick enough.. this could be a bestseller😊😅

[@alephestudios]

it's too CPU heavy, and yes, mainstream music "suffers" from artifacts, that doesn't mean they're good. Actually a lot of loud music sounds like digital sh1t in some speakers, I guess due to this artifacts becoming more apparent

[@richardsisk1770]

That’s so interesting. I was wondering why after EQ and compressing voiceover using various EQ and compressors that I was hearing distortion. Also wondering about using a low cut filter in a microphone does that introduce distortion I wonder if EQ on a traditional mixing board also introduces distortion or not do you have any thoughts on that?

[@sqcaraudio]

MEqualizerLP has linearphase fft mode, its non realtime when you adjust a band but then its realtime. It has normal linearphase and minimum phase.

[@unforg1venofficialmusic758]

This is very interesting but not at all practical.

[@ErixSamson]

Wow, that's incredible. I suppsoe we can do the same with Steinberg SpectraLayers...

[@NoahHornberger]

maybe it's my ears in the morning or my headphones but the 'better' method sounded more harsh to me

[@Nosa_ade]

It not the better is the most accurate

[@fl7210]

This is very well made

[@sageaudio]

Appreciate your view!

[@CoopMusic247]

I find it amazing how much better informed we are on a wider scale about technical stuff in the audio world compared to 50 years ago, even if we have new (and old) problems to deal with. There is a 14 year old watching this video somewhere. I'm gonna just keep using a shelf instead of a low cut to minimize issues like this which unless I'm mistaken helps. (Anyone feel free to correct me on this point) Thanks tho Sage for giving something else to consider as a reason why my snare sucks.

[@dante8677]

using a shelf vs using a low cut is a never ending debate, but honestly it's never gonna impact your career

[@CoopMusic247]

@@dante8677 It doesn't mess with my career, but it does make life easier to not have the build up from a low cut affecting my compressor afterward. I never know what was happening, but then I figured it out and it's nice to know.

[@fluorescent2403]

this is literally just the infamous byproduct of doing real time fft processing which introduces problems like spectral leakage and low spectral resolution due to having to work with small time windows rx isnt doing anything particularly different here, its just that it has the privilege of having access to the entire signal which means it doesnt run into these issues as opposed to eqs doing live-processing

[@youngpopsz4622]

You’re the goat 🐐

[@seangill2413]

excellent video

[@Trappin21]

Great Video!

[@leonardofavaretto2289]

Great vídeo!

[@flamesintheattic]

I noticed years ago that plugin eq causes massive super low freq distortion. It's really annoying to avoid it and to get rid of it!

[@Migesteban]

I love this Chanel !!

[@sageaudio]

Thanks for watching!

[@_TeaDj_]

❤

[@Migesteban]

@@sageaudio Thanks for this videos !

[@GaussianWonder]

I don't think there's such a thing as removing 70Hz information from a file. All audio files are kept in some form or another as PCM data. A wave. So what RX may be doing is spectral decomposition (into frequencies and phase for each frequency) edit, then recomposition. This process is also very time consuming. That's why it's not available as a plugin, you can't "live" preview this without massive delay. But this allows for very precise and surgical edits without affecting phase at all.

[@ALOISC1]

Maybe the Kirchhoff-EQ gives cleaner results?

[@moskva-kassiopeya]

Damn that’s a pain in the a** method

[@davidpereira4455]

Just intonation solves it 😅

[@oscar1353-e7c]

When you make a video like this, diving deep into the inner workings of an equalizer, why then speculate about how RX works? Why not ask an expert in DSP to help explain it properly? The algorithms behind both methods are not exactly secret, so it must be possible to compare the two methods properly. This video now raises more questions than it answers.

[@sageaudio]

Because I make 2 of these videos a week 🤣

[@iam-music]

Because enough of the rabbit holes...lets just make music lol The results speak for themselves and there are no other options that I know of...I just upgraded spectral layers for a bunch of other drum bleed reasons and its wild, just another use for it as a house keeping tool...well for those that actually record music these days

[@SnoopDoggBlast]

If you really wanna brickwall like cut the frequencies, okay. But EQ changes the sound and brings character into the sound, right? So this method is only for a specific use, I'd say.

[@RyTheUnDefined]

I use SO many EQ filters. This video gives me way too much anxiety 😰😭

[@riq4724]

what causes these anomalies with pro Q ?

[@cachelesssociety5187]

I'll take a stab at helping relate it to an example product: Read about and demo Grand EQ by DDMF (DMG Equilibrium could be in the ballpark but it was too complex and with higher latency so I stashed it away in the cellar). Grand EQ minimal phase sounds accurate to me in the freq. & amplitude domain without adding unworkable latency. "The goal in digital equalizer design is always to reduce the amount of digital artefacts which are introduced due to the necessarily finite sampling frequency. As with everything in life, the last few percents of optimization are the hardest, so a lot of effort has gone into creating an EQ which does the least possible damage to your audio material." "In its minimum phase setting, GrandEQ manages to do just that to an extent larger than most other EQs, and notably BOTH in the amplitude and in the phase domain, while introducing only a very small amount of latency.

[@privacee1845]

Spectral filtering is the cleanest 😂

[@НикитаЧеркунов-п6я]

RX. If you make spectrum brighter, you will see at is linear phase. It has pre ringing and post ringing. But you can edit, cut out any sh"t in any frequency range, any time range. In recorded audio you can copy selected noise and paste where you want, instead of selecting and filtering. Invisible method of editing

[@Bboy_cult]

I absolutely prefer the eq. Listening on JBL 708s.....

[@lun4rkid]

Actually u can see it in the q3 curve

[@shuya4104]

isnt that first part cause at the amplitude change there is distortion introduced that causes all kinds of aliasing harmonics across the entire spectrum thus they are affected by any filter whereever? amplitude cannot change without introducing additional frequencies. that signal definetly does NOT just contain the frequency set within the plugin tester.

[@herbie5263]

This is like watching a science channel... just use whatever sounds good :) It's an art

[@mando3022]

Thanks for the video. What about the MAAT EQs? Did you test them yet? They’re supposed to be very accurate as well. I’m using the SantaCruz EQ and am quite happy with the results.

[@Autogolem]

Very interesting!

[@drrodopszin]

I was thinking about RX throughout the video and was pleasantly surprised! The question is, things like ReaFIR are using the same Fourier transformation method?

[@naimelhabbas9100]

jacob collier used it on moon river

[@joelibretti]

Pro Tools?? I Thank You.

[@rafalvarezsevilla]

Kirchhoff EQ is much cleaner than Pro Q, and a much better Workflow than Rx :)

[@nothing7ish]

What about ZL EQ?

[@CRayBeats]

Why not just use volume automation to turn the low sine down along side a filter cutoff for anything that isn't bass?

[@alephestudios]

missing the point

[@CRayBeats]

@@alephestudios no i get it trust me.

[@alephestudios]

@@CRayBeats no you don't, this isn't a bass technique, it's about manipulating any part of the spectrum. How would you do what you say if you want to modify the high mids?

[@beatbanger27]

Good vid!!! My question is what is the best mode in pro q2 for instruments and vocals? I was told the linear phase is better for bus/group processing while natural or minimum phase is gods for individual tracks. Is this a good assessment? What about zero latency, is that only good for live set up?

[@havila1906]

My man, humanity has been mixing great music for decades without ever worrying about these technical aspects you keep mentioning. Just do whatever sounds good and don't let the ocean of information get too much into your head, it really doesn't matter and no consumer can hear the difference.

[@AndJesusSayWTF]

Use it in zero latency, maybe minimal phase. Linear phase eq really affect the dynamics. For mastering it sounds "better" but use it really gently

[@alephestudios]

@@havila1906 yes they can, but some artifacts or distortion are easy too ignore if everything else is in place

[@ParanormalArson]

In the vast majority of cases, even zero latency mode is totally fine. If you're using aggressive high or low pass filters, you may have issues with phase shift, which natural phase or linear phase can help mitigate.

[@happyshadow]

@@ParanormalArsonexactly this. Also if you're doing any mid side EQ it's best to stick it into linear phase mode as during the decoding some time domain issues arise.

[@Fwuzeem]

I hate to say it but I've had bad vibes about Pro Q3 for years now.

[@dystopia-0616]

spline eq?

[@lusid_music_uk]

Does this mean I can finally high pass my masters (say at < 20hz) without introducing phase distortion or pre-ringing? This is HUGE! I do wonder what happens at the cutoff freq though? Surely just having the frequency content fall off a cliff at 20hz will have some weird effect no?

[@sageaudio]

Some artifacts will still be present, but it looks like it'll be a lot less substantial than if you used a typical linear phase EQ!

[@lusid_music_uk]

@@sageaudio think I’ll need to do some null testing to see what exactly those artefacts are from this method. Linear phase without pre-ringing and no real downside seems too good to be true

[@teddycook1299]

@@sageaudio RX filtering is definitely still introducing pre-ringing, and a lot of ringing in general. It really is equivalent to a very high quality linear phase EQ. The visual feedback from the spectral display can help point out issues quickly, but the only away to alleviate the ringing is the same as in a brick wall Fabfilter HPF: make the slope gentler. To do this in RX, there is a "Feather selection" tool. You can do something like 0.5 octave. Not sure what the equivalent db/oct is but I'm sure you could easily get a similar result with a high quality linear-phase EQ. For example, I tried cutting ground loop noise from a high-gain guitar part, using RX's "select harmonics tool", and whenever signal is present there is audible ringing at the "notched" frequencies. Tried again with a feathered selection and ringing was less audible but the part was missing too much. And too bad spectral de-noise sounds like a crappy mp3, even at conservative settings. For critical low end filtering of the sub-bass found in bass and kicks I've been turning to TDR Infrasonic. It's not any better than other digital HPF methods, but the focus on options specific to that region helps me get the results I'm going for quickly.

[@TheTonyTitan]

I'll admit RX sounds super clean, but it also looks tedious

[@sageaudio]

Definitely a little tedious! I noticed that the more I worked with the platform, the more I got used to making changes quickly. Still not as convenient as EQ, but worth it in certain situations in my opinion. Thanks for watching!

[@Shijota]

how are you guys get like little to absolutely NO whitenoise in your videos o: Is there a VST you can reccomend? Or is it my Wave XLR interface? im Using it with a Fethead and my Shure sm7b.. But everyone has a cleaner sound then i have..

[@KaiMusicProductions]

Expanders, noise gates, turn off all fans. High chance they use audacity noise profile supression or ai. im pretty okay with my noise but i have very bad reverb in my recordings i use waves dereverb on my vocals. Maybe confirm you got a real sm7b too. Lots of fakes.

[@Shijota]

@@KaiMusicProductionsis there an expander vst i could use on my Wave xlr?

[@viktorlarzev3528]

Did someone tried to do the same with Console Emulation low cut?

[@Daniel1987H]

"on a binary level" It's digital audio... How could it be done differently? Also you just said that the pro q would have audible artifacts but didn't gave an actual example. This video seems very in-depth but it doesn't explain anything.

[@Walid.OnTheTrack6725]

oooow myyyy god

[@TheKatt08]

Clickbait title.

[@aninatoneklikej]

I guess u have already tried it

[@drofdissonance]

you don't understand filtering or DSP sorry, you've modulated the input sine, which will generate content all the way up to nyquist, so surprise surprise its gone by the time you hit the max bandwith. the spectral filters of izotope cause many artifacts that you haven't even tried to detect. Take a free MIT course in digital signal processing, and then come back with some informed opinions

[@sageaudio]

I appreciate the input! You're right; I could definitely learn more about the math behind what's happening here. Still, this video has enough new info that it's helpful to those more focused on mixing and mastering - but perhaps not helpful to someone such as yourself, with a complex understanding of DSP

[@Keroser1983]

@@sageaudio Top pro and kind comment!

[@Samurai-zc4bm]

WHAT IS THAT SONG IVE BEEN WONDERING FOR MONTHS SOMEONE HELP

[@slingshotjamaica]

Same here

[@sageaudio]

Thank you all for watching! Unfortunately, this is a small musical snippet I made for showing ideas in these videos, so it's not a full song or released anywhere.

[@Samurai-zc4bm]

@@sageaudio can it be 😭🙏🏾

[@slingshotjamaica]

@@sageaudio it’s a great idea to release that’s why more than one of us have caught onto it .

[@PassingTheDog]

Wow

[@QueMusiQ]

9:23 wait, in your daw, you can “unrender”? Rendering is by definition destructive, as in irreversible. But there is an intermediate way. I only use studio one, so that’s what I use as reference: you can “freeze” a track, meaning create a reversible waveform with GC and then ghost the effects to preserve the insert settings during a mix. It does not render or overwrite. Is your daw creating a SEPARATE bounced render that turns off RX? So that if you change your mind it puts back the original file and turns RX back ON? Because if it’s unrendering sh💩, there are gonna be questions about what sort of magic hoodoo have you conjured, and can it be used to unpregnant annoying ex girlfriends, or unbeer goggle last night. 🤣🤣🤣 After seeing what algorithmic ai can do in the music industry, I wouldn’t count out the possibility to “unrender” stuff. We can do stems with greater and greater granularity, even seeing it go from RX 8ish “Drums/Bass/Music/Vox” poorly to some nailing it even removing digital reverb tails seamlessly when removing a vox track, to -I wanna say something like 🤔 divx? SOME company - that can pull out specific tonal INSTRUMENTS and specific drum SOUNDS. 🤷🏽‍♂️ So it is POSSIBLE in the year since I last turned on my music equipment after losing EVERYTHING IN A SSD CRASH (😡) - it’s possible that you can unrender.

[@kjgkgkgkgkgkgkgk]

What's saving the previous audio is the audio editor not the daw. by going into rx you can undo the changes and resave to have It update in logic.

[@mttlsa686]

The minimum phase example doesn't sound like a brick wall at all. Are you sure it's correct and it's not 48dboct? I can hear too many upper frequencies that are leaking to be a brick wall filter... Anyway, you might be right on Izotope rx (maybe not about the fact that it "physically" delete information but about its better, audible, results), because the example sounds like the vocal has been re-recorded with a better microphone and a better placement instead of being just equalized, and the difference is so clearly noticeable...It's not even subtle! Really interesting!

[@LYSHEmusic]

He showed us the sound of the difference, the delta caused by applying the filter. The filter also changed the phase, so the difference is happening in a wider spectrum range than just the filtered-out frequencies.

[@mttlsa686]

@@LYSHEmusic I know it's the delta but it's a huge amount of leaking frequencies to be a brickwall, even if minimum phase...I'm going to test it by myself to see if I'm wrong anyway. 🤝

[@sageaudio]

I just checked the session and yep! Brickwall minimum phase filter set to 68.8Hz. But, might be worth checking on your end just I missed something. But yeah I agree it's a lot more than anticipated - the "bleed" seems to be due to the aggressive phase rotation and reshaping of waveforms introduced by the filter. Thanks for watching!

[@peterorany]

Fucking amazin method . I waiste time cos who the fuck is buying isotope Rx for 900pounds

[@Hogboy345]

Interesting but how is this useful information? I doubt a single song in the history of music has been mixed using RX to EQ, hasn't seemed to bother anybody yet.

[@sageaudio]

Sometimes looking into something is fun :] I agree, I don't think it'll impact if listeners enjoy a track or not (some of my favorite songs have terrible production), but this stuff is interesting nonetheless imo.

[@ParanormalArson]

The first time anyone uses a technique, that technique was never used by a single song in the history of music. It doesn't make it less useful or valuable.

[@Keroser1983]

This is just a tool bro. Relax. Just keep it in your arsenal and use it when needed.

[@gnomerod]

I cannot hear any difference between EQ and RX to be honest. Is it because the method is secret?

[@bestdisco1979]

So what your saying is fab filter pro Q 3 is shit...yes ?

[@clarenegracey]

Hey buddy, all music created by humans is flawed by human error. Software errors have little impact compared to this.

[@ramalshebl60]

if music can't be perfect, i'll just never make music again...

[@ParanormalArson]

Music can't be perfect and you should make more.

[@lilwombat]

that is a horrible idea nothing in existence can be perfect, always do your best

[@sparella]

:P

[@Durkhead]

Its the imperfections that make it perfect