Mixing Loud with CTZ - Ep. 23 - Considerations for mastering

[@809rdl]

Yeeeeeees we are back! The most beautiful voice i have hear today!

[@Nova_Afterglow]

I was chatting up the Guitar Center lady and she let me know she was going to school for audio engineering. When I was describing how I was using clippers to control dynamic range she looked at me like I was crazy. I could see this type of engineering was completely foreign to her. Saturation and clipping will do more than a single mastering limiter could ever hope to do. Makes me wonder what they are teaching nowadays. I'm glad Baphometrix is my instructor.

[@josepapiii]

And now my day revolves around watching this in full

[@jshstuff]

Same here

[@proaudionajia]

Great to have you back. Missed your deep insight on production and mixing.

[@Jedimasta21]

This whole tutorial series is the GOAT.

[@ArtardiDebernuro]

We need you back man all love 4 u.

[@vertecsrecordings]

Hey Baphy, haven´t seen it yet, but I KNOW, this will be my highlight this weekend, when I will sit down, relax an learn from Ep. 23. Thanks to have You back!

[@tombrazy3496]

I love this series Correct. Critical. Comprehensive. and best of all... CUTTING EDGE!

[@tyeblee23]

Thank you for all of this awesome information Baph!

[@Six2Nine]

You are my mentor. thx for this knowledge.

[@Noisownd]

Thanks for this video. Excited to check out the rest of the series, this is the first I"ve found. Amazing knowledge, learned a ton about mastering. I've been struggling to figure out how to do this for some time, keep it up!!!

[@a.t.onthebeat2016]

I've learned so much from this CTZ series bro. Thanks for helping me improve my beats baphy. Much appreciated

[@calvinbotha4859]

Baphometrix. Can't thank you enough! CTZ has literally change my music for the better in the best way possible!!! The penny dropped when you showed the way I used to mix (slapping a limiter on and not taking summing into account). Again, Seriously can't thank you enough!!!

[@GetSongsDone]

Same.

[@wintourmusic]

Nice to see you back, I look forward to watching this video later on! It’s holiday in the uk due to the queens 70th year. Nice timing haha

[@matthewchavezm.b.s5503]

These are all very good tips and well explained to get the best sounding track. I would definitely say that there is no substitute for the real car test. If your like me, then you listen to music all the time in your vehicle and your ears are tuned to hearing well produced music by the pros in this environment. So playing your track through those speakers in whatever shape your vehicle cab is, will give you the best understanding on what your track is doing.

[@MokerB]

The Slate VSX does that for you, it really gives you the substitute for your car. Just like you're vehicle your ears need to be tuned to it, but that's with every environment.

[@matthewchavezm.b.s5503]

@@MokerB how can a plugin emulate your car speakers plus the shape of the inside of your vehicle through your studio speakers in a room shaped much different than your vehicle?

[@finnhuman1012]

The king is back

[@OdoSendaidokai]

Thank you, especially for the True Peak and the 16/24 Bit topics 🌻

[@MissingLynxMusic]

I still get flack for my video against True Peak limiting, so thank you(!) for going deeper and better into this topic. We've had true peaks as long as we've had digital audio, and now it's suddenly a problem?! TP limiting really hurts the openness of a mix, and it rounds off your transients and robs them of some of the snap. Re: multiple masters, Seth Drake discovered that Kaskade (or their engineer) has been making separate masters for normalized streaming that are NOT just turned down, but are optimized for the bigger dynamic range. The "real" master you can buy is way louder with reduced dynamics. I don't think it's necessary and Seth only sometimes does it, but it is known to happen. I have actually been making a streaming master, not by limiting down the true peak, but literally just by turning down the whole track 1-2 db, so I'm more like -9 or -8 which is still pretty loud unnormalized. My thinking: they were going to turn me down anyway, now they'll turn me down less and I'm at -1 TP. I have a full volume master for myself (and in theory for beatport/bandcamp where I haven't actually released anything yet). But I'll probably stop this practice and just use streamliner to check for distortion and trust my ears.

[@mattydatboi]

I couldn't agree more with a lot of statements in this video regarding final output levels and the like. Bear in mind to anyone reading this, I am a nobody and have nothing to stand behind my name other than what I am about to talk about. Let's just say I know and have seen some stuff so you can either take what I say as valid or not. Up to you. Anyway, I am not going to name names, but suffice it to say that I attended what is supposed to be a pretty good music school here in the great state of California. Don't get me wrong, they had some really great instructors and I am thankful very much so for the information I obtained. That being said, there was some info that was given to me that really hits on a lot of what you are talking about with a dated viewpoint on things and a practical usage that may have been the norm when platforms like Spotify, Deezer, Tidal, iTunes, Soundcloud, were just becoming mainstream and published certain criteria to be met for their platforms. However, as you said, times have changed and I have scoured the internet looking for tunes not even with loudness in mind but just based on music I like for the potential application of maybe playing them in a club (if I can ever even get to that level but that is a different topic). Just out of sheer curiosity given the tools I now know about regarding mastering, loudness, and the like, I ran some stuff through Ableton (especially for making edits) putting YouLean Loudness Meter on the channel (something I highly recommend besides just dpMeter5 for monitoring loudness and for anyone reading this is a free utlity). I CONSISTENTLY saw songs in EDM hitting AT LEAST -9 LUFS with most averaging -7 to -5. THIS WAS EVEN IN THE STREAMING PLATFORMS when I later got a hold of Minimeters (a real cool visualizer that a lot of people use and is only like $10 that gives you a spectrogram, a wave display, width monitor, LUFS Meter, oscilloscope, and spectrum analyzer in one strip)! Mind you volume doesn't effect this as you can play back at any level on your speaker and the LUFS do not move with the exception of if you turn down the volume slider in the platform say for example Apple Music; but, again, for most cases the volume/production of a track is for the volume at max and usually people change the volume within their PC or speakers and not the app/utility. Artist like Svdden Death or Yoinks have songs that hit pretty damn close to 0 LUFS WITH NO AUDIBLE BAD CLIPPING/DISTORTION (check out Kino Der Toten by Yoinks and you can see what I am talking about)! So, whatever people (or a single person) told me about levels needing to be true peak at -14, yeah I'm not buying it as the average loudness was from all major platforms (Spotify, Apple Music, Beatport, Tidal, Soundcloud). We also, at the school, used to have case studies where we had to match the loudness level of an example given to us that EVEN THOSE WERE AT LEAST -8 LUFS and a lot of us struggles because we were told to use true peak. Granted, I did it using tricks, different plugins I had, and parallel processing/bussing to get there, but it was a pain in the ass. We were even told in one class we might not even be able to match the level because one of the songs was processed through a console. Well no wonder we couldn't match the levels as those DACs and consoles have some serious oopmh and add more harmonics on the upper end of things vs. audible distortion. To this day, I wonder (for one guy in particular) if it was a game and he knew the secret sauce but was trying to make a point/a moral high ground statement. I have noticed a lot of that floating around. It seems there are too many people from prior generations really complaining about the loudness levels and instead of telling/showing people how to get there, they gatekeep based on a moral stance that it shouldn't be that way. I'm not saying they're wrong, but it's all a pile of shit if the market doesn't work that way. How are you going to help someone get in the door by giving them bad advice that won't even be able to step foot in the ring with the big players just because of a moral stance? In a lot of ways, I agree: songs should be finalized with more dynamics and less squashing for more clarity. It DOES sound better. But, again, THE MARKET DOESN'T SELL THAT! At least, not right now it doesn't. Maybe things will change in the future and the industry will demand lower levels and set new precedents. But, until we are there, people SHOULD BE INSTRUCTED HOW TO PLAY AND NOT HOW THEY SHOULD PLAY and we can worry about it later. I know this is long and you may not read this, but this is ONE OF THE BIGGEST PROBLEMS I have with a lot of the music industry and people trying to feed their ego instead of the art. I know it will never go away because some people get into this for money and accolades which is a separate problem. To me, though, the loudness wars never died they just transformed and got normalized by streaming platforms to play out at the same loudness levels.

[@mattydatboi]

Also don't get me wrong the guy I am referencing is very knowledgeable/a nice guy overall and easy to work with. I just think where he is coming from and what he might be saying is flawed.

[@lautarocastro]

Wow what a nice track! Any chance to hear the full version one time? Thank you 🙌

[@Baphometrix]

kzbin.info/www/bejne/iJTUZImPbdlrqsU

[@Nova_Afterglow]

I think i'm just gonna comment as I go along. I love this stuff. Anyways. I would posit that a bricked waveform isn't necessarily a bad thing. I know you didn't say it was, Baphy. At 2:50 you talk about how the -7.5 waveform still is pretty dynamic. I have Virtual Riot's REDLINE in Ableton and some sections are BRICKED. A person would be hard pressed to say it sounded bad though.

[@augustauseil]

Im enjoying watching with Nova commentary turned on! :)

[@DjFreefall]

Welcome back Baphy 🚀

[@Nova_Afterglow]

I have actually moved away from pretty processing as of late. I try and do everything in the mix, so my Master track has only an Izotope Imager (to make sure one last time everything under 200hz is mono) and my Master limiter. I used to saturate by bands using Izotope's Exciter, but because I saturate so much already following CTZ, I just haven't felt the need to come back to it. I would be open to trying it again tho. I would be open to any mix candy if it did something to improve the track.

[@Studio22mix]

Never watched a video like this one ☝️ Got me subscribed definitely going to watch more of your content really good stuff ❤ A lot to learn again 😊

[@jshstuff]

Hey Baphy, do you know of anyone who strongly disagrees with some or all of the concepts you teach in the CTZ method? Like in any subject (politics, science, religion...) I like hearing any counterpoints direct from the other side to keep myself sane, as well as strengthen my convictions. Any channels / reading material you recommend? Even if you don't agree with them. Thanks!

[@Baphometrix]

Honestly, no, I don't, other than ANY producer/engineer who feels strongly that "more dynamic" is always better than "loud".

[@panosfillipou14]

Nice ! Thx bro , good job , very useful !

[@sopastar]

Thanks for doing this! Just a heads-up: The written guide asks the reader to check out your DubSkald stuff for examples of music mixed with the CTZ strategy, but the linked website seems to be taken over by someone spreading malware

[@Baphometrix]

Oh noes! Thanks for letting me know.

[@Baphometrix]

I appreciate you pointing that out to me. Link changed to my Spotify page instead. Thanks again!

[@djkymaera]

You really break this stuff down well, appreciate it :)

[@MatamiMusician]

Hello, Baphometrix! I very pleased to see you back!!! Thank you for a tons of knowledge! Please, help me to understand one moment in CTZ strategy 🙏🏻🙏🏻🙏🏻 . When i gainstaging my kick and sub, the gap between my Kick and sub is - 6 Db . My kick clip to zero & my sub clip to zero. Then, I put my sub buss to - 6 db, right? if i want to push my instruments levels up , my kick is going to clipper and becomes louder, and peaks at the same zero level. Please, help me to understand what i need to do with my sub? Do i need to push up the level of my sub with the clipper, or on the sub buss? Thank you !!!

[@Baphometrix]

Your sub does NOT need to go anywhere near the ceiling at 0 dBFS. You should never need to actually clip your sub. (Because it never needs nor wants to get very close to 0 dBFS.) So leave the fader for your sub track at Unity, and just use whatever synth or sampler etc to make your sub peak somewhere around -6 dB relative to where your kick is peaking (or being chopped off at 0). Your kick should be the **anchor** for your entire mix (if it's dance/electronic music, anyway). So you should never be changing the level of your kick. Your kick is the ANCHOR; it stays where it is. You adjust everything else **relative** to the kick. That said, yes, if for some reason you decide your kick wasn't loud enough when you started out, then sure, you might need to push it further up into the clipper at 0 dBFS on that kick track. And yes, this might indeed squash out that particular kick too much. And you'll need to either find a different kick that can be made loud enough without sounding bad, or you'll need to decide to NOT make the song louder than the kick--the ANCHOR--can handle. Hope that helps.

[@MatamiMusician]

@@Baphometrix Thank you for the detailed answer! 🙏🏻 Now I understand! And just one more question… Do you use plugins like R-Bass on your subs ( in region from 30hz to 80hz) to make it warmer, or it doesn’t need ?

[@Nova_Afterglow]

I believe letting your ears rest is one of the most important things in production/engineering. I have simply lost track of how many times I thought something was fucking awesome, only to come back an hour later and realize its trash. a mix of rotating between a project, reference and no music at all keeps me in check and prevents me from letting too much wack shit pass. However, I will say this: Even up to a month ago I was so concerned with proving myself that I would rush tracks just to post to the Discord and prove I was serious. This led to several rushed tracks. Even my only public track has a part that I would change, even though the track as a whole is legit. It's very hard early on to take all the time you need, but I can see now producing/engineering fast for fast's sake is bad way to go about it. Who cares how many songs you can bang out in whatever amount of time? Are those songs built and engineered at your highest level?

[@Baphometrix]

THIS

[@chrishaake8126]

What a relief 😌

[@owenbulliss2378]

Hey great video as always. I'm interested in the mention of spectral clippers reducing low end. I'll make sure to compensate for that when I use elevate and such. Do you know if New Fangled Audio - Saturate also reduces some low-end? I figure that is also some sort of spectral fft magic? I'm considering using saturate for some bus summing but I wouldn't want to mess with the low end by using like 2, one on the premaster and one before the master limiter or something. Btw now sure if you've seen but saturate has an update with asymetrical distortion and antialiasing with a hard clipper at the end just like kclip has. Interested to hear your thoughts and thank you for taking the time to make these videos.

[@Baphometrix]

Spectral compressors and limiters tend to hit the very lowest FFTs (the sub region) harder, so that they don't have to hit the more audible ranges as hard (therefore "cleaner and less distortion"). So if you make bass-heavy music, it can be problematic. In Elevate, you just use their EQ section to boost the sub FFTs back where they originally were. (Which of course makes the rest of the limiting/compression LESS "clean".) In general, if I like the sound of a compressor/limiter that does NOT use spectral FFTs, I'll usually just keep my life simple and use something else. But that's because I make bass-heavy music. For a lot of music types, the spectral products can sound really good.

[@owenbulliss2378]

@@Baphometrix thank you! I think that's what I've been hearing in practice. I make more bass heavy stuff so I'm fine with a db or so of summing on my pre master bus, but I've been mixing a friend's music and the clipping was a bit too intense. I tried a limiter and it is smoother but the vocals get a pumping sound. Saturate should be perfect because the track needs less lows anyways so great! Thanks again and I'm looking forward to future videos!

[@helina-keijo9561]

Pro l2 is great but i nowadays use voxengo elephant for mastering limiter 😎 it is absolutely stunning.

[@RasShango]

Been patiently waiting 😎🙌🏾

[@nikolaudio]

Thank you queen🙏🏻🙏🏻🙏🏻 :)))

[@talksykart5468]

Just curious. Have you had a chance to try the Devious Machines UrsaDSP Boost Limiter? I've been using that one instead of ProL after trying SmartLimit. You can turn True Peak and oversampling off. You can bring dynamics back with the dry/wet. It's fantastic. Give it a go if you have a chance to. Or I'd be curious to know what you think of it if you've tried it.

[@Baphometrix]

I love nearly everything Devious Machines puts out. But not UrsaDSP Boost. I found it too much of a "black box" and didn't like the workflow.

[@talksykart5468]

@@Baphometrix I just turn it on. I don't know what happens under the hood for sure. But it's not doing much so I'm not too worried about it. Everything at zero and using CTZ. Every mix I've put it through so far sounds better with it on than off. It's almost like a Gullfoss which I think is what's going on under the hood. Magic Beef is what I like to call it. Lol.

[@Baphometrix]

@@talksykart5468 Good to know. I might check it out more closely at some point in the future, to see how it's come along since its first release.

[@talksykart5468]

@@Baphometrix I've built a template for Ableton using CTZ. You should see it. It's beeeuuutiful!🥲 I rerouted the whole thing. No more busses. Created my own Master channel that routes into the OG one so I got around the listen channel issue that way. Busses routed directly into the groups so you can control level. It's a work of art. Thanks dude. If no one else appreciates you I do. I've been doing this 20+ years and I could never really get that last 2-3db out of my mixes. Not anymore. You have glued all of the pieces together more thoroughly than my Master's from Berklee. Lol. 🙏

[@Baphometrix]

@@talksykart5468 I love hearing feedback from folks for whom the CTZ approach finally 'clicks' into place like this. Stoked for you! Go make music now!

[@Ein-Stuck]

Great stuff!! What episode has the VCA trick?

[@vminus7505]

Ep. 8 - Hitting ANY loudness target with "the VCA trick"

[@809rdl]

Hi Baphy; how should I approach the gain staging in recording in a CTZ project? I also have another question. i am film composer, and I am preparing my self to use the CTZ for my next score; the main benefits will be to stay predictable and well organized since early production, between tens of music ‘’cues’’. The genre is inspired by SPAGUETTI WESTERN orchestral music. My doubts: 1- How to approach the gain staging in recording instruments in these sessions? 2 (related to CTZ) I have to emulate the EQ spectrum, the color, timber of a 60s recording. Should I listen to isolated frecuencies of these old recordings, and may use EQ in aLL MY TRACKS and or buses, or put an EQ in the mix bus?...Sometimes I don’t know whats the right question to do, so I prefer to explain you my context. 1 -RECORDING INSTRUMENT WITH CTZ. 2 Got the same dynamic range as my reference, now how to steal the same EQ VINTAGE color?

[@Baphometrix]

I mean, this question has many aspects, which I've made 23 episodes so far to answer.

[@809rdl]

@@Baphometrix ok; you are right. I am watching all the series; i am in Episode 7 now. Ill continue! Ok

[@martinlarrivee5081]

Do you do mixing and mastering? I find everything you do so precise and thought out. The results are amazing.

[@paulbocostudio]

Have you had a chance to test Black Salt Audio's Clipper? I'm really liking it, but I would like to know your thoughts on it with the CTZ strategy in mind.

[@Baphometrix]

I doubt I'll be buying it. But from looking at the manual it's clearly a hard clipper (that's good), and you can set the clipping threshold at 0 dBFS (that's good), and you you can drive input up past 0 and clip off whatever passes 0 (that's good). What it doesn't do is automatically compensate the gain so that you can **hear** the clipping at equal loudness like you can in KClip. So it's going to be a little harder to judge where the "transparent" results stop and the "colored" sound change is happening.

[@paulbocostudio]

@@Baphometrix Thanks for the insight. I knew you would know what's going on under the hood.

[@wintourmusic]

A really good video full of knowledge. I find I rarely care about true peak levels too, the distortion from the DAC gives transients a nice modern snappy quality. If you have true peaks in sub information or sustained overs it does sound terrible. Momentary events of a few ms above 0 do not matter and I'd argue sound better in some cases. Aggressive DNB kicks and snares definitely benefit. Where it will matter is in aggressive conversion for social media - which will most likely played back on phone speakers which do insane quality destroying processes anyway so its not big of an issue really haha. I'm super interested in your opinion of RESO - its pretty cheap and I have been considering adding it to my arsenal.

[@Baphometrix]

I'll definitely be talking about RESO soon. Bottom line for now is that Soothe and iZotope Spectral Shaper are just "spray and pray" approaches to toning down bad-sounding resonance buildup. You have a rake. You can make the tines of the rake wide and flat, or narrow and skinny, and you can fan the rake wider or narrower (fewer or more tines). You can make some of the tines dig deeper than others. But it's still just a "blunt instrument". RESO first of all sets its starting "points" very smartly, and doesn't over do it. And then you can INDIVIDUALLY remove or adjust how deep any of those points dig. And then you can effortlessly sweep around and find other specific points you want to set for yourself, because you can hear an ugly whistle there. And then you can easily and effortlessly fine-tune everything and make sure you're not digging too deep or not deep enough in any one area. It's brilliant.

[@wintourmusic]

@@Baphometrix Thanks for getting back to me with that. It certainly seems a more transparent approach to frequency taming. Sounds easier than just using an EQ as well, perhaps also the filters are more tuned for this purpose instead of a generic filter bell notch. I maintain that Soothe2 is still brilliant but when used more sparingly. Larger amounts of processing cause the sounds to become less natural sounding. MSpectralDynamics is a beast with a lot of control (probably the most configurable) - but it's CPU demand is too high to put on multiple channels throughout a project.

[@daybit1014]

Hey Baphy! Thanks for all these videos. I’ve learned so much! Question - you say for CTZ users their choice of mastering limiter is irrelevant since loudness is established within the mix. For mastering engineers getting stereo mixes done without CTZ method, what limiter would you suggest? i.e., a mastering engineer gets a mix where the loudest rhythmic element isn’t slightly dancing on the mastering compressor/limiter. Whats the best course of action? Thanks :)

[@Baphometrix]

Talking about mastering limiters is like talking about wine tasting. Every limiter sounds different when you start working it hard. Most pro mastering engineers will have many (if not all) of the better limiter plugins out there, and they'll quickly grab the one they know will sound best on a certain type of song. Or they'll just try a few and pick the one they like best by ear. They're all good dogs, Brandt. 13/10 would limit again. That said, if you have a tight budget and can afford only one, IMO the safest all-around choice for nearly any genre is Fabfilter Pro-L2. If you DON'T work in super loud, bass-heavy dance genres, then I'd recommend smart:limit by Sonible.

[@knoerdklokibril]

Hey Bapho, Nice video as always. I was wondering if you have opinions about multiband compression on the master?

[@Baphometrix]

Multiband compression has its place. But it's easy to overdo it or over use it, and hurt your sound. especially on the full song signal happening at the mixbus. MB compression (even extreme versions of it like in an OTT process) can sound GREAT as part of your sound design. But on a full mixbus? It's a good "band-aid" tool for a pro mastering engineer who doesn't have access to the full project and all its tracks and busses, but IMO it's not the right tool for most home producers who are engineering their own work. Why not go fix problems where they **originate** (on individual tracks and lower-level upstream busses), instead of screwing with the phase relationships of major bands of the spectrum up on your 2 bus? That said, one of my favorite mastering limiters (Limitless) works in part by employing 5 different limiters in a multiband manner (for part of the signal being limited, in their first stage limiter). But again, this is more of a "problem solving" tool for pro mastering engineers and not something you should need to directly mess with if you're a home producer using Limitless. As I said, it has its place and its uses, but you need to be careful and know what you're doing and why you're using MB compression when you're in the later mixdown/mastering stage of things. For early sound design, though? Anything goes, as long as it makes interesting sounds.

[@demonix7515]

Hey Baphy, If you were to make a track at lets say 0LUFS during the drops would everything you teach in this CTZ series remain the same? Would essentially the only difference be the clippers are being slammed harder? Also at a target of 0LUFS would it perhaps opt you towards more soft clipping than hard? I ask this because I've seen some recent releases from artists I like having there drops hit 0LUFS and I enjoy the way it sounds and its quite fitting for the genre.

[@darotm7628]

Can you tell me one of the songs that hits -0 LUFS

[@demonix7515]

@@darotm7628 Svdden Death - Kereberot Borgore - Si Here are some pretty extreme examples. The Borgore song peaks at +2LUFS

[@yuulian4776]

When I upload songs on instagram (they are mp4 from wave files) they get downgraded a little bit and also it sounds quieter.... Do instagram lowers the volume? I once uploaded a beat and it sounds loud and nice, and all others beats even if they are at same LUFS they are quieter

[@Mookae5]

Why the hate on Opus? From what I've read on the audiophile side, everyone seems to like it for performing slightly better at the usual bitrates and not falling apart at ridiculously low bitrates.

[@Baphometrix]

Opus is fine. 64-bit lossy anything is not so fine.

[@jaumecabalgante7178]

Why cutting -0.3dB in the Limiter output? Why not scalating these dB cut in the last two mixing buses and then push the limiter in the Master Bus? Why not using the CTZ technique as a sound check for a better consistent and dinamyc mix? Thanks for sharing so much knowledge. These are the most 23 revealing modern production videos ever posted on youtube. 🙌🏻📖

[@Baphometrix]

I'm not sure what some of your questions are actually asking, but as for the -0.3 ceiling in the Mastering Limiter, that's just a final absolute ceiling. Some people like using -0.1 for that. Some like using -0.3 (which was the Redbook standard for CDs) Some people like -0.5. For me I guess it's just habit to keep the sample peaks at -0.3.

[@jaumecabalgante7178]

@@Baphometrix Thanks for your answer, now I get it. Forget about the other questions, it’s just that those -0,3db broke my schemes forged in the 1st video. Thanks again for all what you do.

[@Baphometrix]

@@jaumecabalgante7178 I see. Okay, so the very last mastering limiter you use in the signal path (on the Master bus), you can set to any final value you want. The reason I all the clippers and limiters I've shown you so far clip right at 0 dBFS is because A) that's the easiest way to control everything and to use the VCA trick, and B) there's no point leaving any headroom on your tracks and busses. You can run the signal right up to 0 dBFS until it's time to export to a master WAV or AIFF. At that point, you probably want to set your limiter ceiling to either -0.1 or -0.3.

[@kylehegger5549]

​@@Baphometrix Why would you want to set a ceiling of -0.1 or -0.3 especially when trying to achieve loudness? Are those ceiling standards from people who are trying to prevent there true peak from going over 0? To simplify what I am trying to ask, lets say you have a true peak of +2db, wouldn't having a ceiling of -0.3 just lower your true peak to +1.7? What good does the limiter ceiling do when your true peak is going over regardless?

[@Baphometrix]

@@kylehegger5549 Don't overthink it. -0.3 sample peak is a "fudge factor" rule of thumb for minimizing the occurrence of ISP overshoots for the higher-quality lossy codecs. (320 Kbps to 256 Kbps). Note that you can't do squat about the huge ISP overshoots you'll see on low-quality codecs (and the low quality aspects themselves far overshadow any longer runs of ISP events that might actually be noticeable). But it was observed and passed around as a rule of thumb that leaving -0.3 dB of headroom (sample peak) generally resulted in very few ISP overshoots on higher quality MP3s and AACs once upon a time. So the number just kinda "stuck" as tribal knowledge. If a -0.3 ceiling vs a -0.1 ceiling vs a 0 dBFS ceiling bugs you, by all means go with the -0.1 or 0 as you see fit. At the end of the day, a difference of -0.2 or -0.3 dB is **neglible** on the perceived loudness/dynamics of the signal. At the end of the day, who cares if a lossy codec--even a higher-quality one--has some ISP overshoots. If they are spread out in time and not happening in dense clusters, you can't hear them. So it's no big deal. On the other hand, if you trust the tribal knowledge rule of thumb from the early days of MP3s.... then does it really matter if your final LUFS is -7.5 versus -7.2? I hope I've explained well enough that LUFS is only a **rough measure** of perceived loudness. There's a lot more to "loudness" than just a measured LUFS or RMS value.

[@aretwodeetoo1181]

Any update on sonible smart limit?

[@citadelo5ricks]

I noticed that Sonible added "True Peak off/on" switch, what was Bapho's other issue with smart:limit?

[@kylehegger5549]

Does anyone know how I can setup the vca one knob controlling all the clipper gain in ableton live 11? Thanks

[@Baphometrix]

It should work the same as in Bitwig. Have you seen Episode 8 in this playlist yet?

[@kylehegger5549]

@@Baphometrix I’ve seen every episode but I’ll have to reference back to that episode, thanks!

[@Six2Nine]

finally ur back!! how r u?

[@Baphometrix]

I'm fine. Just a super busy stretch at work. 🙂

[@Richard-Raw]

Thank you very much, I learn a lot from your channel ? 😄

[@Nova_Afterglow]

RIP True Peak.

[@jonathanoates1972]

hey baph, i deleted my last comment because i found the answer. It's blue cats patchwork! Got the demo and it works great as studiorack. I canhave one instance of this as the ctz rack and then insert on any channel i like. :-)

[@vincecrow4512]

Should somebody always export with dithering on, or only for the file that’s getting upload to streaming? Is it necessary on the file you export for DJing?

[@Baphometrix]

Any time you print music out of your DAW (which is 32-bit floating point) into a fixed bit-depth file, you should dither. This means you should always dither when exporting your "masters". Now, when you're exporting stems (for example to hand off to a collaborator), you almost always want to export those in 32-bit float format to prevent any clipping from getting baked into the files you're passing around. Same thing when you export your two-bus mix to give to a 3rd party mastering engineer; you should always be giving them a 32-bit float master. In those cases, you don't want dithering (and most DAWs will prevent you from even using dithering). For internal project bounces and resampling, you should likewise always bounce and resample in 32-bit float with no dithering. Otherwise you risk baking in some clipping. Even if potential clipping isn't an issue (because you work at very low gainstaging levels), you still don't want to be adding dithering noise. That said some DAWs force you to use 24-bit bounces and don't give you the option to bounce to 32-bit float. In that case you should indeed dither your 24-bit bounces.

[@vincecrow4512]

@@Baphometrix ok awesome! That was really helpful. Thank you!

[@hdea4072]

Great stuff as always! Thank you! Is there a bitwig device or a free vst that has that "side check" functionality with a click of a button? I use MetricAB and Proq3 to do the wall/high check and I suppose I could do the side check with proQ but it would be easier if theres just a click of a button. Hopefully someone knows of one....Cheers

[@Baphometrix]

If you have Metric AB, there's a "Sides" button right above the big blue/orange AB button. But yes, you can also do a side check with two native Bitwig devices. Grab a "Mid-Side Split" device. Then in the "Mid" channel, put a "Tool" device. Then turn down the Volume knob in the Tool device to -inf. Now you hear only the sides. Save the MS split device like this as a preset to your User Library.

[@hdea4072]

@@Baphometrix awesome thanks!

[@hdea4072]

@@Baphometrix Actually, I just saw that ISOL8 is free haha! I'm a bit braindead at the moment so this may be a dumb question: is selecting M in ISOL8 for mid or mono? Are those the same thing?

[@Baphometrix]

@@hdea4072 They are the same thing. Yeah, I know it's confusing. Some plugins will call it "Mono"; others will call it "Mid".

[@talksykart5468]

Midside by GoodHertz is awesome. 👍

[@jonathanoates1972]

Hi Baph, what do you recommend as a light non cpu intensive saturator to have on every track? I have several, Black Box, Decapitator, spectre, Saturn and a few others. I love Black box, but when i have one on every channel my asio guard hits the red. I't not my cpu as that runs at less then 50% and ram is the same. Would spectre be lighter than black box? Also i've experimented and i notice that kclip across 50 tracks and 20 odd buses takes a lot of asio. I bypasses all of them and it dropped a load. Not sure what the answer is here because its not my processing power.

[@Baphometrix]

First things first, if you're up to 50+ tracks and 20+ busses, you should be flattening most of those TRACKS to audio anyway (as you finalize the sound design for each track. That many plugins of any sort will require a lot of resources. Second, I've tested 50 instances of KClip with an ASIO buffer size of 256 and these were all consuming only 2.6 percent of my total audio CPU in Bitwig. (On an i9 system with 64 GB of RAM and a MOTU M4 audio interface.) You can see the tests here: kzbin.info/www/bejne/op28nICdr8afmsU So I'm not sure what you're running into, but it shouldn't really be KClip itself causing any significant load. As for "lightweight" saturators, your guess is as good as mine. My favorite-sounding wideband saturator is Pulsar Modular P42, but I wouldn't exactly call it lightweight. Spectre isn't _exactly_ a "saturator" so much as an alternative to an EQ that works IMO better for really loud genres when your goal is to get a little more perceived loudness out of a given sound by **boosting** one area of its spectrum. Spectre will give more of a noticeable "boost" in loudness with less phase change than an EQ boost would do.

[@jonathanoates1972]

@@Baphometrix Thanks Baphy, what i should have made clear is that i created a CTZ template of about 60 tracks compiling of audio, instrument tracks and groups. The template loads up with all the ctz plugins already inserted, dp meter, kclip, black cat gainj,black box, fab filter pro q3, and spectre. However the loop i have going isn't using all those tracks. In fact it's only using 18 of the 50 and they are just audio. All instruments have been taken out. And now asio is only using about 75% so loads of headroom left. I also found some tweeks in cubase that has helped. If you suggest flattening the source tracks to audio once designed, i presume we flatten them with the ctz in place on that track as well. So when it's flattened do i then put the ctz plugins back on that flattened track and gain stage it again, even tough it's already at ) 0.0db or just stick on a kclip? I'm using focasrite 8i6 3rd gen. And i'm wondering if it could be the interface and drivers because the computer isn't even using 50% of it's power. The asio guard is still high and i don;t think it should be with what's on the project in terms of plugins. I have by passed all the ones not currently on use. It's a bit pricy to go and buy another interface just to try it and see. Head scratcher!

[@Baphometrix]

@@jonathanoates1972 I don't know enough about Cubase to guess at settings there. You haven't mentioned what buffer size your focusrite is running at. As for when/how to flatten tracks.... Normally you do all your sound design (including getting it into the right general loudness range to fit your mix. And then you flatten everything (even the CTZ pair at the end) to audio. Does Cubase allow you to do the equivalent of an Ableton "Freeze"? Or a Bitwig "Bounce and Deactivate"? In both Ableton and Bitwig you can effectively flatten everything on a track to a single audio clip. But you can also get everything back and start over with some tweaks, and then another flattening/freezing. So if Cubase allows the same freeze > undo > tweak > freeze again, type of workflow, you should be able to make any downstream adjustments you _might_ need. Sound design > flatten. Sound design > flatten. It's just the way of life in big complex projects.

[@jonathanoates1972]

@@Baphometrix I've got my buffer size set to its max. 1024. I can find all i need about cubase with their tech help or forums. I don't think its the daw or my comp, which is a Ryzen 7 1800 X 8 core, 32gb ddr4 corsair 2666mhz, 1tb ssd Mp 50 M.2, 1tb slave drive. When i've done a perfornce check of the cpu and memory at the point of when the asio maxing out, the cpu amd memory arnt even trying, they're running at less than 50% of their power. Now unless i'm missing somthimg here i can't work it out. Fortunately this isnt happening all the time. However i guess that once i've sound designed a sound source i can render it out with current settings, and use that printed audio and release the original with all its inserts. I guess i can then insert dp meter , blue cat and kclip as before but they'll just sit there idle because the file was rendered out at its gain staged signal from the original?

[@Baphometrix]

@@jonathanoates1972 Well. Here's how you can rule out KClip as the culprit. Do a test like I showed in the video I linked above. Make an empty song project. Drag in one 4- or 8-bar sample and set up the DAW to loop that sample. Put KClip on that track and make sure it's clipping the sample by a few dB. Then dupe that track out 49 times. Then see how much CPU is being used up with all 50 tracks playing and being clipped. As you can see, in Bitwig on Windows 10 with an ASIO sound card with a 256 buffer size, I'm using less than 3 percent of my DAW CPU. If you get different results, it's either something about the way Cubase manages plugin resources, or it's something about your ASIO driver, or it's some kind of system latency in your computer and other running programs and drivers and services.

[@OctavianMusic]

thank you again :)

[@bbl7813]

Have you compared TBProAudio DSEQ to those other resonance suppressors?

[@Baphometrix]

Similar questions in my video about De-harshing (posted right after this video). Short answer is no, I like TBProAudio's stuff, but I already have tools that overlap with DSEQ.

[@bbl7813]

​@@Baphometrix would you mind giving the trial a spin to get your insights on how it compares to the other tools you mentioned?

[@Baphometrix]

@@bbl7813 Maybe at some point. No promises, though.

[@sokoleski]

@@Baphometrix Yes... i would love to hear your 'take' on DSEQ as well ! In context to RESO, and SOOTHE and Ozone Spectral Shaper. Similar tools for a similar job. Thank's for all you insight and love the CTZ which i'm slowly adapting. 💚

[@titusrivers5059]

Did I miss the "how to place sounds in space (depth, panorama)" video?

[@Baphometrix]

Nope. That will be the next episode.

[@titusrivers5059]

@@Baphometrix 💯💯💯

[@sussmusic8173]

what do you do for phone test

[@Baphometrix]

Slate VSX (the program I show in the video) has Apple air pod simulations (wired and wireless). As for the phone itself, everyone has one, right? And if you're a producer you probably also have Dropbox, because it's the easiest way to share stems with other producers for collabs, right? So you can drop a master in a Dropbox folder on your computer and be playing it from your phone a few minutes later.

[@Nova_Afterglow]

I've trained my ears to hear so much, but I cannot for the life of me tell you what dithering sounds like. I wouldn't even know what to listen for.

[@Baphometrix]

In most cases you won't hear dithering in the original master. That's the idea. You **might** hear truncation distortion affecting a super quiet, super long fade in or fade out. But that's why we use dithering--it replaces the truncation distortion with something less noticeable. You **might** REALLY hear truncation distortion if you resample bits of a mastered song that somebody did NOT use dithering on, and you process that resample in aggressive ways that amplify the noise floor, such as running it through OTT.

[@kiko8u]

WOOHOO!!!!

[@jonathanoates1972]

So i was watching a video from @Bthelick who was showing his method of ' no mastering needed' i dropped a message in saying how i follow the CTZ method. You may find the reply interesting and i would value your feedback on this. He goes on to say: It's simply the fact that he uses clipping on the channels and buses before reaching master. That's not how digital audio works, it doesn't 'save' any headroom in the long run and only creates more problems at the master buss stage and more aliased clipping down stream. If it's working for you then great don't worry, I'm only saying the science doesn't add up because his fundamental theory is flawed. But if you're confident in your ear training and you like the sound then go for it. I'm not about to enter a me vs x technical method battle with people sending me tracks haha. I've been releasing music for 10 years like this professionally to hundreds of millions of streams across accounts and it's literally paying the bills , I'm not seeking change because there's no problem right now. This video is not for comparing methods, it's to help those that are struggling . (And to clear up some misinformation) If you still have problems , i.e you don't trust your ear training yet and your releases aren't getting any plays then obviously start seeking changes sure you can definitely try my method it's a LOT less steps but only if there's a problem. Be careful off this academic back and forth of technical matters 'on paper', I see it everyday on gear space / Reddit and it's 99% hobbyists that have got obsessed with numbers instead of music , and usually don't actually release any music for a living. He also says: Baphomatrix's clip to zero involves pre clipping at channel and bus levels which only makes things worse imo (also in terms of digital sound science doesn't actually work as adding 2 clipped signals together = +6 which just creates more clipping)

[@SchneiderRicky]

doope!!

[@vincecrow4512]

I get that you can’t hear much difference when you toggle the codec on and off when you’re not using true peak, but why not use it just to make your mix distort even slightly less? You didn’t talk about the negatives of actually using true peak. And also, if you turn the output of your limiter down to -1dbtp, it doesn’t squash your mix more, it’s not actually doing more limiting and compressing/clipping to your mix, it’s just taking the whole song and turning its output down 1db. And that doesn’t matter when it’s getting uploaded to a streaming platform because the platform does loudness normalization. Any amount of “loudness”, which is actually just output level, that you’re losing from turning your track down 1db is then compensated for by the streaming platforms normalization. I’ve always just made 2 bounces for DJs. They’re both the same master, the same level of clipping, LUFS, RMS, etc., one just has the output turned down 1db, and the other doesn’t. One for streaming and one for DJing.

[@Baphometrix]

We'll have to agree to disagree on what happens when you tell your limiter to use "True Peak Limiting". It does in fact squash your mix harder. That setting is not a simple "ceiling" setting. It affects the degree to which the attack stage of the limiter squashes peaks. Think about it for a moment. If you set your limiter to raise the signal loudness to -7 LUFS with True Peak Limiting on or off, the resulting loudness is -7 either way. But the True Peak setting is keeping the LUFS at -7 while also pushing the peaks down much farther, compared to the non True Peak setting keeping the LUFS at -7 while letting True Peak jump as high as 1 or 2 dB over 0 dBFS. Clearly, that means the True Peak setting is squashing the signal more. And you can _hear_ this extra amount of squash. ^^^ And there's the explanation about what's "bad" about True Peak limiting.

[@vincecrow4512]

@@Baphometrix I see what you’re saying here, but didn’t you demonstrate in the video that the true peaks that are going over zero are very random, sporadic, and short? So wouldn’t it only be squashing those tiny little bits of random sporadic peaks harder? Also, you’ve demonstrated in the CTZ method that most of the time the master limiter is doing little to no gain reduction. There’s only gain reduction happening at all when you do little amounts of pretty processing. So say you’re not doing any pretty processing and therefore your master limiter is doing zero gain reduction. If you set your limiters attack and release really fast to make it behave closer to that of a limiter and then turned true peak on it would be doing something close to transparently clipping (limiting but sounding closer to clipping) those true peaks without making your mix sound more squashed. Am I right or wrong in assuming that?

[@Baphometrix]

@@vincecrow4512 What you're forgetting is how a limiter works. A limiter does not instantaneously squash a peak and then stop. The _attack_ is nearly instantaneous, but then there is a (typically quite long) _release envelope_ that is carved into the original signal after the peak that triggered the limiter finally falls below the threshold again. It's the act of carving all these release envelopes into the signal that soften up the mix. Also, the total amount that the limiter is pushing down those peaks is much further (with true peak limiting on), so the initial squash itself is _farther_ and therefore _more distorted-sounding_ than if you simply let the true peak clip the DAC down to 0 dBFS. Which is yet another nuance! Many modern DACs can handle several dB of overshoot at the DAC conversion stage _without audible clipping or distortion_, especially if the true peak events are very short in duration.

[@KOjoeBeats]

@Paphometrix: what about uploading lossless audio somewhere else, and then giving the link in the KZbin video description ? Then we would just need some sort of starting CLAP in the begining, so those of us that want have good audio can synch it and happily watch your work in a satisfying way ? 160kbps is so freaking sad, especially when we see someone that put so much effort into uploading the great things you teach us KZbin REALLY need to make a move. I analyzed some 40fps videos with a pixelezied JPG as video coded with 160kbps audio ? Wtf if i may ask ? Tha'ts such a shame for audiophiles. What they are waiting for ? It's often quite tough to notice sublte differences in those eductational videos about audio engenering. I often i feel like "well, i guess A is better than B, but well, LFU match properly, and then let me A/B blindly without those retarded codecs-. It's sooo easy to be tricked and there is no reason to continue that way. We really need me make some noize about this problem.

[@erkamau9629]

Too much long sorry

[@djvoid1]

Playback speed = 1.5

[@Noisownd]

There is gold in here. Learned so much

[@jonathanoates1972]

Isol8 is now free :-)

[@rxbflac_]

The king is back