dude!! This actually happened to me, a plugin glitched out and spiked super loud. Forgot the actual number but luckily i always have a limiter on or my speakers probably wouldve be cooked
Maybe you keep your master channel free from any Compression or Limiting/Wave shaping plugins. And instead gainstage everything and apply limitiers just to BUS channels, which are rooted to a premaster channel. If it still clips and unintentionally overshoots then, either your plugins cause Oversaturation or you gotta check all of your plugins.
@@AVDREMelda Production's parametric EQ had a bug where it would cause absolutely insane ring spikes when using multiple bands at the smallest Q setting as a dynamic resonance suppressor; I recorded while inducing the bug and I got a spike of over 600db. Uninstalled it, and ensured that a limiter was on the master at all times in the future
If you're doing parallel comp, saturation and maximising then you can put a limiter on after all that stuff and it'll barely have to do anything but still come in useful every now and then.
The reason you aren't seeing the Inflator "square off" the waveform is because the input gain is 0 dB. If you started to increase that, assuming you have "clip 0 dB" on, you'd get noticeable softclipping. If you don't have clip 0 dB on, you get a kind of sinusoidal wave folding.
Excluding limiting means you have less control of where your desired cutoff point ends up, this might lead to overshoots when platforms make transcodes of your material. That's why you don't leave out limiting, but rather use it wisely.
Use other tools to achieve the bulk of your loudness / bigness / evenness, but still use a limiter at the end to catch overs and gain a touch more level (-3db reduction max)
You ain’t getting there with parallel compression only. I record hot out of my api all tracks clean but hot and when I get to my master my hardware is getting to -10 lufs clean (Manley/maselec). Plugins then get you louder but induce distortion. Parallel compression won’t take you from -10 to -8 integrated and that is why this video is a bit misleading. I hate that we have to get that loud these days and all our music sounds distorted as a result.
The tricky bit here is that since stuff like Inflator, Saturn, and Vwarmer are effectively wave shapers with transfer curves designed to add lots of harmonics, you do have to be extremely careful not to oversaturate the mix and change the sound of the original recording. And since inflator (and possibly PSP, maybe they’ve updated since I last used it) don’t oversample, there’s also a risk of adding aliasing noise if you’re working at lower sample rates. Parallel (and upwards) compression are usually more transparent but if hit too hard can rebalance the mix more than you (or your client) may want. Now, sure, limiting and clipping and even compression are also wave shaping that add harmonics too, but since there are use cases for those specifically requiring minimal coloration, they tend to be designed with transfer functions that keep it to a minimum. My favorite solution is to use a few different stages with minimal impact so as to apply the least unnecessary coloration. A limiter is usually involved at the end but only doing a few db of work.
Thanks for watching! True - these waveshapers will introduce more THD than limiting. Lookahead generally decreases THD significantly by circumventing attack times and "cutting into" the waveform during processing. PSP has the "fat" option, which is its weird name for oversampling - not sure why it's called this haha. Clippers, at least from what I've experienced and measured, add significant THD through the creation of a square wave (odd-order harmonics), but that's not necessarily bad, similar to how THD from waveshapers isn't necessarily bad. But yeah I see what you mean! At the very least I was hoping this video would keep people from relying too heavily on limiting and instead to focus on controlling dynamics in other ways that affect transient detail to a lesser extent. Your solution sounds like exactly that 😊
Oxford Inflator (and his it`s mock-up Hornet Thrust) on a master really alters the sound. it brings out mids, but kills the bass and highs. Oxford Inflator really kills the KICKS. Just tried it yesterday. Decided to never use it on master again and went on wth Kraftur.
it's a simple waveshaper with nothing special going on. you can get the exact same results from other free waveshapers. and i mean *exact,* passing a null test. kraftur is doing a lot more. much more tweakable. & a more unique sound i think.
seeing more and more popular music hitting as high as -3.5 integrated LUFS, the most popular and competitive loudness is still "as loud as physically possible"
@@puella_meiberu yet they sound clean asf. I mean I pushed my master to -5 the other day and I can hear distortion. Might be the clipper before the limiter, but if I don't clip you will clearly hear the limiter thus ruining the track.
@@deviousmind1092 masters that loud are made in the mix, and the mix is made in the arrangement, the only way to go that loud is to build it for that loudness from the moment the first instrument is picked up or the first midi note is put down, this is part of why the structure of pop music has changed so much over the past couple decades relatedly, particularly genres with full-spectrum non-transient elements sensitive to audible distortion and unable to meaningfully duck out of the way of anything else, such as most shoegaze, just cannot be pushed both loud and clean, -5 is about the wall for those kinds of mixes
Philips and Sony give us 16 bits of dynamic range, only for us to master tracks with dynamic ranges which would sound okay in 8 bits with a bit of noise shaping. *shakes head* - Thank you, car audio & iphone "speaker" listeners.
In my opinion, limiting should not be doing any of the heavy lifting in a master. Compression, saturation, clipping, maximizing should all come first. (Not necessarily in that order) By the end of the signal chain the limiter should only do a very very small amount of limiting, like 1 to 2 db if you really need it. And you can do two limiters if you really need to get your track that much louder. I stack the ozone vintage limiter with the ozone maximizer. The vintage limiter usually does 1 db of limiting and the maximizer does maybe .4 or .5 db of limiting.
I don't agree, you don't have to choose between one option or the other, you should use all of them if necessary. Parallel compression, saturation + limiting will make the limiter practically not need to work.
8:168:238:318:39 Not much transient difference, which was the main thrust of the purpose, but clearly different options arriving at end result options.
I like how you are able to describe other ways of mastering besides just brickwall limiting. Parallel compression sounds like a great idea, and maximization is definitely a great innovation!
I like to shave off a decibel or so with a transparent limiter nonetheless (Voxengo Elephant). After bouncing to tape and then using the plugin, I honestly could not tell any negative impact on the sound whatsoever.
These days, I'm studying and practicing a lot of mastering with upward compressors. I'm now using limiters only for outlier peaks. This approach has worked best for me in genres of music that require a higher level of transparency and dynamics in the mastering: classical music, jazz, and progressive/symphonic rock.
@@ILOVEDRAGO Its an answer, of course. You should not have oversampling on by default, IMHO the negatives I mentioned outweigh the benefits. And yeah, pre-ringing isn’t typically an issue with lookahead in compressors like it is in linear-phase EQ.
I admit, I haven’t watched the whole video yet, and maybe you’ll clarify this, but bringing up everything besides the peaks IS functionally equivalent to turning down the peaks. You may feel like you’re leaving the transients alone, but by turning up everything around them, you are decreasing their “impact,” potentially in a very similar way to what a limiter would do. Looking at the plug-in doctor transfer curve you show a lot, it feels like bringing up the quieter parts of the signal while leaving the loudest parts alone is a different operation from turning down the loudest parts, but, except for some makeup gain, they are similar if not equivalent operations. There’s no such thing as a free lunch. You have to balance loudness and transient integrity. You’re right that some limiters do this better than others, due to how their algorithms choose at what points to start and stop attenuating, and with what “shape.” (I think you know this, though you didn’t specify at least so far; look-ahead limiters, just like regular limiters, don’t JUST turn down the individual samples that are above their thresholds. This operation is functionally saturation/clipping. What makes something a limiter and not a clipper is that it turns down the samples AROUND the over-threshold samples as well, so as to preserve the transient better.) It’s an interesting watch so far, I’m just not convinced that it’s anything new.
Okay, yeah, the Oxford Inflator and the PSP plugin are also both limiters, just limiters with soft knees. If the video were “stop brickwall limiting,” I’d agree, I don’t really do that anymore either, but what you call “maximization” is just limiting with a soft knee
What you're describing is reducing the dynamic range. Maximization or bringing up quieter details reduces the dynamic range - limiters do this as well. You can decrease the dynamic range without attenuating peaks - that's the concept I'm trying to demonstrate here. The regular limiters you're describing here would have to operate as a clipper. A limiter, by its nature, utilizes lookahead to attenuate - otherwise, it would clip. There is no place in a waveform that a limiter can choose to start attenuating. To be a brick-wall limiter, it must attenuate the full "over," which necessitates the use of lookahead. Saturation and clipping are the same thing - waveshaping to peaks. Clipping is usually associated with hard clipping or a complete squaring of the waveform. Saturation is usually associated with a gradual squaring of the waveform (unless asymmetrical waveshaping is used like with tube saturation), but they're 2 sides of the same coin, so to speak. I bring this up because I don't understand what you mean by turning down peaks and saturation being functionally the same thing. The purpose of a limiter is to not saturate or clip
@ thanks for the reply. I’m actually glad to be discussing with you, you very well may know things I don’t about this, but here is my understanding: 1. Increasing the volume of everything but the peaks is functionally equivalent to decreasing the volume of the peaks, save for some makeup gain. Whether we call this limiting or not is a matter of convention, but a unique process is not occurring. Now, it could very well be that your method allows for bringing up quieter parts of the signal in a particular way that is less intrusive to the transients in your signal, but, in theory, you could be bringing up the quiet parts of the signal that results in nearly identical results to decreasing the peaks of the signal. If so, you would see the same decrease in transient “impact.” Transients don’t exist in a vacuum - if the samples surrounding a transient are quieter, that transient will have more “impact” than if the surrounding samples are approaching the same amplitude as the transient itself. Again, maybe you can do this in a cleverer way or have more control with certain methods than with others, but principally, depending on how you do it and how a given limiter is configured, the operations could be functionally equivalent. 2. Let’s consider clipping vs. limiting vs. lookahead limiting. A simple clipper is just a few lines of code: is the sample in question, regardless of its surroundings, above the threshold? If so, change the value of the sample to be identical to the threshold. The problem with this is that it audibly distorts the signal. Logic has (if I’m not mistaken) a limiter that can be configured to a 0 ms lookahead time. This limiter has a release time but no attack time. What (I imagine) is happening here, is that the signal is run through a similar circuit to the simple clipper, BUT, when the input signal dips back down below the threshold, the processing doesn’t just leave the signal alone again. It remembers how much it turned down previous samples and then “ramps down” (with some shape that depends on the release time and algorithm used) the gain reduction over some time frame, typically milliseconds. This preserves the transients a bit better, because the samples AFTER the over-threshold samples are also attenuated, preserving a bit of the transient amplitude relative to its surroundings (at least on one side). Finally, the look-ahead limiter. This requires the most clever processing. The limiter “looks ahead” into the future, detects some over-samples, but then considers the samples before AND after them as well. Depending on the algorithm used, it both “ramps up” AND “ramps down” the gain reduction. The loudest sample would then be at the threshold, and surrounding samples, whether they were over threshold or not at the input, are turned down a bit as well to preserve their amplitudes relative to the loudest sample of the transient. A limiter with true 0 attack time and 0 release time is a clipper. A non-brickwall compressor with true 0 attack time and 0 release time is a gentler saturator. The attack and release (or, in the case of the lookahead limiter, pre- and post- attenuation) are what makes compressors and limiters not saturators and clippers. If it weren’t for that ramping up and down of the gain reduction, it would just be clipping/saturating, and may sound like audible distortion.
I really liked this one! I'll definitely use the parallel compression method and the Oxford plugin too, sometimes limiters can change your sound not like you wanted, and these techniques are great options
Just 'wow'. And in addition to 'wow', also, thanks for sharing this rebellious info. Might've never the gall to consider limiterlessness. To '25 and new territory! And inflators are free to all in many forms, for the yet-uninformed, hopeful inflators out there. Keep on inflatin'!
Really interesting! I'm going to try to do this and compare how it sounds against the typical Pro L 2 on master, but I'm excited for the Dynamics processing technique since it'll keep the transients intact haha Great video and solid explanation, thank you as always! 🙌🏼
What about vintage style limiters - Abbey Road TGI or Fairchild, stuff with a tone and feel? The kind without look ahead? I’ll never make a commercial recording - but I want my songs to sound like tunes from 1955-1975. Any tips if I’m trying to emulate the sounds from that era?
I reckon it'll be just status quo for you assuming you're emulating the mixing and mastering chain of that time period (vintage analog mixing + mastering for vinyl)
I just tried this on a song I've been messing with, and i do feel like the overall mix felt much more alive. Going to be considering this every time I master!
I use clipping + limiting on my masters. Fat, punchy, and full. Limiters help pull the sound into a prettier image when utilized properly - similar to framing a picture.
What chain would you use for this? I'm getting ready to master my first track and limiting was the centerpiece of the whole thing I thought. Please lmk how I should approach a limitless master because honestly I really hate the squashed transient sound but want to bring up my low and increase clarity of the overall track.
Currently using an Ableton stock limiter on master just to tame what otherwise could cause distortion, mainly on vocals that I push to the limits. Only .1-.2 db. You think i should just soft clip on the channels and glue or parallel compress on master? Idk i will test it out.
Any tips for my mixes that are a little quiet compared to other songs on Spotify? LoudnessPenalty is saying it's at -4.5 reduction on Spotify so I thought I had that part right.
Awesome I was just trying to think of how to accomplish this. For the opposite outcome tho- i want to bring my mix (-3db peak and -12 Lufs) to streaming levels for the digital publishing. I want to make this mix of (-3db peak, -12 Lufs) to (-1db peak, and -14 Lufs). Then for the masters that I want to sell and distribute, I don’t mind the louder sounds.
I did notice that my pre-mix conscious songs ended up sounding a bit clearer albeit around -9 to -12 LUFS compared to clipped and limited songs around -6 to -7 LUFS. Honestly, the limiting might be more harm than good unless it's a distortion heavy genre, or a Rick Rubin production.
So many comments to refute this but Andrew Schultz has been using similar technique for years. He chooses not to parallel compressed drums during mastering. Worth a try anyway
The biggest takeaway that I got from Bob Katz' "Mastering Audio" book was that the best master is one where the ME has to do very little because the mix is so good. That sorta opened my bedroom musician eyes to the importance of "fixing it in the mix" before it even gets to the master bus, let alone the ME. So if I think I need to strap some low end transient beeferizer on the master bus to fix a kick that doesn't kick... then maybe I should go figure out the kick drum first. So in your example, maybe the vocals are overcompressed... OR the vocals are just right, but the rest of the track is limp.
This might be common sense to someone with your experience But how would fab c, or oxford inflator keep you from digital clipping I agree that it sounds much better But As far as i know, these plugins dont prevent you from going above 0 So while i preserve my transients better with these plugins, how do i maximize loudness while keeping my master fader from hitting red
@@SToXC_. I Just tested it and you are right. Thank you for the response my friend Ima test my luck again lmao, but what about inflator Any way to prevent peaks Tbh i might just use pro c as limit and push into it with an inflator, but id like to hear your opinion as well Thanks again
@@EversonBernardes You know what you're right. Idk why, i havent used it in a long time and i only used it on my drum bus so i had no experience. Thank you both Ill leave my comment up in case someone inexperienced like myself is wondering.
@@EversonBernardes ahh and to add as im testing it rn, when activating band split it doesnt prevent it but im sure there's a reason for that as well. Regardless thank you
What would you say about limiters on the mix bus during the mixing stage? I assume it’s a similar deal but are there any benefits in the mixing stage absent in the mastering stage?
it's not inherently wrong but mastering engineers generally don't like this, & if you're using a limiter on your mix bus, you are likely using it to "fix" things that are better off tweaked in the mix. if you take off the limiter and your mix falls apart, your mix needs work
@ Of course to your last point lol. But appreciate the feedback. Yea, I usually use the limiter for one particular section that seems a little out of control. But I was thinking that same thought process this morning (about just needing to tweak the mix in those areas anyway)
💥 The thing is, Skrillex`s masters look like a -1LUFS SAUSAGE, but the kicks sound clean and crisp. If THIS is the way how to achieve it, i`m in. But is this the way?💥
Bro Sage this is Gamechanger (what) the sound is much more pleasant . I Bought Limiter Elephant last time because like your video about him, but this is next level on the finally effect track !
You should definitely give that a try! Parallel compression has long been an industry standard for drum processing, but you can definitely try out some of the other techniques/tools we mentioned on your drum busses too!
Interesting concept and proof that perhaps a little bit of every thing (saturation, compression) can go a long way to achieving a better result than whacking a limiter straight on. Just one issue. At 0:59 I think you're doing the excellent Newfangled Audio Saturate plugin a disservice here by having it set to Clipper Shape 'Hard 100%' which is always going to produce that square wave. Setting it to 'Soft 0%' produces a lovely sine wave and in the other view a very gentle knee even when driven fairly hard. It's a great saturator!
Last time I've used a limiter on a master was probably 2007. A limiter on a master channel simply does not achieve anything that is not better achieved without.
But all this could have been done years ago, why is it suddenly the thing to do? Has something in the industry changed? (genuine questions not rhetorical) Thanks
I spent some time experimenting and realized I was overlooking this. Limiters are quicker/easier to use, but this method definitely has a big advantage with the retention of transient detail.
I tried it and it sounds good. But it doesn't fit my track (Orchestral liquid drum and bass). I think it depends on the situation. ->My track: Warm and soft with my Softclipper->limiter master. ->My track: Hard and I'll destroy your speakers with the compressor master.
-14 is loud? I don't think so. I would love to see these examples while reaching -9 or -8, as most clients expect, or even -6 for some styles, without limiting. All in all, the video is still about loudness war, you just didn't approach the problem raised by it.
Part of the sound of someone like Skrillex has been heavily reliant on proL2. I wonder what guys like that would say about this video. Interesting idea though.
Ive been following this channel for a while now and id say i agree for about 80% of the time, considering im a dnb producer. But in this case i have to disagree. In the age where every label tells me tp get a song to at least -5 lufs integrated it doesnt work without a limiter. I'd love to send you a premaster of mine and you try and master it however you see fit and i only use an instance of standardclip and a pro L.
There's also the imprinting to the brains of listeners you would need to factor in. If they got used to the squashed masters they'd expect it. It's not good, and maybe we could get used to cleaner masters.
Thanks for watching! In the traditional sense a limiter is anything with a 8:1 ratio or higher. But brickwall limiters behave much differently than compression. Whereas compressors have an attack time, which represents how long the compressor takes before the maximum amount of attenuation occurs, brickwall limiters have no attack time. They utilize lookahead to measure the peaks ahead of time, then attenuates accordingly (in a way that introduces significantly less distortion and waveshaping than compression). This results in significantly different treatment of transients.
I was always told to limit my master. When I was working with the Pro-L², I hated looking at the red crushing my transients. No matter how many times I saw people crushing them, I couldn't do that to mine. I always used the limiter in a subtle way, which kinda defeats the purpose of limiting. I think I'm just going to stick with compression & my Softube Bus Processor for the final stages.
i think one of the problems is that you were judging the way a meter looks. a limiter is meant to do what you're afraid of it doing. i think the problem is this anxiety of dynamics reduction. you know your compressor and softube stuff is reducing dynamics too. it's just not showing you that in a scary red color. don't be afraid of dynamics reduction! so many masters out there sound weak because someone was afraid of making the needle move too much, as if it's bad to do that. if it sounds bad then of course tweak it.
I really appreciate your content and knowledge I gained through years of watching this channel but I feel I need to disagree on this matter. Modern music standards requires you to use not only one, but sometimes two or even more different limiters in the chain. Music nowadays is 6 LUFS and streaming services requires 2 db of TP headroom with that loudness. Of course we still can master at by the book level of 14 LUFS but that will sound weak and practically no one talks about that. Every newbie mastering tutorial is about 14 LUFS and that’s so wrong. And even this level of loudness requires 1 db of TP headroom.
Yea sorry man, every actual mastering engineer out there uses limiters for a reason. If Ted Jensen is limiting masters, it's fine to limit masters. Get better balances, clip before limiting, manage frequency balance better.
Thanks for watching! I first learned this from John Mayfield about 10 years ago (he's mastered some of the most successful country albums over the past few decades). I sat in on a few of his sessions and asked "why aren't you using a limiter?" and he just laughed at me haha.
@@sageaudio Yea I'd say that's genre dependent, and engineer dependent. You're not going to hit -5 LUFS on a metal track without using clipping and limiting.
2nd kooky video I've seen from this channel, after the "EQ the fundamental" thing. Avoiding from now on. PS: Being aware of a sound's fundmental does help, but the thrust of the video was to move away from thinking of the EQ spectrum as a spectrum at all, and more toward looking at the fundamental. It was just too much "mix following an idea" rather than mix following your ears.
Thanks for watching! Sounds like the video made you consider a different approach for equalization, so it's genuinely cool to see it help another engineer. Kooky maybe, but calling a frequency range "warm" is also kind of kooky lol.
I always keep a limiter on the master purely because the sounds I make can sometimes spike up to 300db and I don't want my ears to explode
dude!! This actually happened to me, a plugin glitched out and spiked super loud. Forgot the actual number but luckily i always have a limiter on or my speakers probably wouldve be cooked
Maybe you keep your master channel free from any Compression or Limiting/Wave shaping plugins. And instead gainstage everything and apply limitiers just to BUS channels, which are rooted to a premaster channel. If it still clips and unintentionally overshoots then, either your plugins cause Oversaturation or you gotta check all of your plugins.
on Reaper I always have the problem of reaper auto muting, cause I get super loud spikes
@@AVDRE this but the plugins i use can intentionally do that instead of it being a bug lol
@@AVDREMelda Production's parametric EQ had a bug where it would cause absolutely insane ring spikes when using multiple bands at the smallest Q setting as a dynamic resonance suppressor; I recorded while inducing the bug and I got a spike of over 600db. Uninstalled it, and ensured that a limiter was on the master at all times in the future
I feel like all edm producers can ignore this video
EDM, Pop, Rock, Rap, otherwise yeah great advice ………
@@rkdmiiiyeah I was literally gonna say the same thing lol
@@nathanizzowhy ? I don't get the point
i feel like everyone can ignore this video cause 2 adv in 10 min video with very disputable info is greedy af imho
Why only edm ?
If you're doing parallel comp, saturation and maximising then you can put a limiter on after all that stuff and it'll barely have to do anything but still come in useful every now and then.
PSP is a gem when it comes to mastering
I'm more of a Gameboy guy myself
Interesting concept. I’d really appreciate a video of you running through a limiterless mastering chain 🙏
yes lets see it in action and comparison and please use a dance track :D
WE NEED IT! I mean - the preacher should listen to own prayers, right? =)
That's a cool idea! Thanks for the suggestion, and it's cool to see there's interest in this concept 😊 I'll do this for the next video I put out
at the end of the day to achieve a limiter-less master you would need to have a squeaky clean mixdown. like surgical room clean mix down.
The reason you aren't seeing the Inflator "square off" the waveform is because the input gain is 0 dB. If you started to increase that, assuming you have "clip 0 dB" on, you'd get noticeable softclipping. If you don't have clip 0 dB on, you get a kind of sinusoidal wave folding.
Excluding limiting means you have less control of where your desired cutoff point ends up, this might lead to overshoots when platforms make transcodes of your material. That's why you don't leave out limiting, but rather use it wisely.
Limiting sounds better for my ears
A good secondary test would be to sit on a subwoofer and see how it feels on your butt 😆
Use other tools to achieve the bulk of your loudness / bigness / evenness, but still use a limiter at the end to catch overs and gain a touch more level (-3db reduction max)
I would be interested to hear the results if instead of -14 LUFS, you tried this at around -8 LUFS
You ain’t getting there with parallel compression only. I record hot out of my api all tracks clean but hot and when I get to my master my hardware is getting to -10 lufs clean (Manley/maselec). Plugins then get you louder but induce distortion. Parallel compression won’t take you from -10 to -8 integrated and that is why this video is a bit misleading. I hate that we have to get that loud these days and all our music sounds distorted as a result.
Ain't happening. I was thinking this is great if you don't want a hot mix...which everyone does sooo....
@@rube6729+ @mollyoko. I agree with you both, It would be a cacophony. I raised the point to show how limiting is necessary on today's recordings.
@@mollyoko Agreed, on all points.
@@rube6729 And hence why 95% of music today sounds flat, two dimensional, lifeless and generic.
The tricky bit here is that since stuff like Inflator, Saturn, and Vwarmer are effectively wave shapers with transfer curves designed to add lots of harmonics, you do have to be extremely careful not to oversaturate the mix and change the sound of the original recording. And since inflator (and possibly PSP, maybe they’ve updated since I last used it) don’t oversample, there’s also a risk of adding aliasing noise if you’re working at lower sample rates.
Parallel (and upwards) compression are usually more transparent but if hit too hard can rebalance the mix more than you (or your client) may want.
Now, sure, limiting and clipping and even compression are also wave shaping that add harmonics too, but since there are use cases for those specifically requiring minimal coloration, they tend to be designed with transfer functions that keep it to a minimum.
My favorite solution is to use a few different stages with minimal impact so as to apply the least unnecessary coloration. A limiter is usually involved at the end but only doing a few db of work.
Thanks for watching! True - these waveshapers will introduce more THD than limiting. Lookahead generally decreases THD significantly by circumventing attack times and "cutting into" the waveform during processing. PSP has the "fat" option, which is its weird name for oversampling - not sure why it's called this haha.
Clippers, at least from what I've experienced and measured, add significant THD through the creation of a square wave (odd-order harmonics), but that's not necessarily bad, similar to how THD from waveshapers isn't necessarily bad. But yeah I see what you mean! At the very least I was hoping this video would keep people from relying too heavily on limiting and instead to focus on controlling dynamics in other ways that affect transient detail to a lesser extent. Your solution sounds like exactly that 😊
Oxford Inflator (and his it`s mock-up Hornet Thrust) on a master really alters the sound. it brings out mids, but kills the bass and highs. Oxford Inflator really kills the KICKS.
Just tried it yesterday. Decided to never use it on master again and went on wth Kraftur.
it's a simple waveshaper with nothing special going on. you can get the exact same results from other free waveshapers. and i mean *exact,* passing a null test. kraftur is doing a lot more. much more tweakable. & a more unique sound i think.
@Fire-Toolz yeah bruh. i just reportig what i heard yesterday why mastering a client's song. Threw waveshaper/inflatir out of mastering chain.
When the comments are way more helpful than the video. Thanks!!!! Kraftur is actually what I needed this whole time
@@dougleydorite enjoy the aliasing!
@ any alternatives you can recommend?
-14lufs is not "competitive" volume lol. My mixes without any compression or clipping are at -12 at mimimum.
I master to -7/-6 LUFS and still they come out less loud than the "competition". Their masters are rocking a solid -4.5 LUFS 😂
seeing more and more popular music hitting as high as -3.5 integrated LUFS, the most popular and competitive loudness is still "as loud as physically possible"
@@puella_meiberu yet they sound clean asf. I mean I pushed my master to -5 the other day and I can hear distortion. Might be the clipper before the limiter, but if I don't clip you will clearly hear the limiter thus ruining the track.
@@deviousmind1092 masters that loud are made in the mix, and the mix is made in the arrangement, the only way to go that loud is to build it for that loudness from the moment the first instrument is picked up or the first midi note is put down, this is part of why the structure of pop music has changed so much over the past couple decades
relatedly, particularly genres with full-spectrum non-transient elements sensitive to audible distortion and unable to meaningfully duck out of the way of anything else, such as most shoegaze, just cannot be pushed both loud and clean, -5 is about the wall for those kinds of mixes
@@deviousmind1092 let me guess, electronic dance music that all sounds the same and has no dynamics?
Philips and Sony give us 16 bits of dynamic range, only for us to master tracks with dynamic ranges which would sound okay in 8 bits with a bit of noise shaping. *shakes head* - Thank you, car audio & iphone "speaker" listeners.
In my opinion, limiting should not be doing any of the heavy lifting in a master. Compression, saturation, clipping, maximizing should all come first. (Not necessarily in that order) By the end of the signal chain the limiter should only do a very very small amount of limiting, like 1 to 2 db if you really need it. And you can do two limiters if you really need to get your track that much louder. I stack the ozone vintage limiter with the ozone maximizer. The vintage limiter usually does 1 db of limiting and the maximizer does maybe .4 or .5 db of limiting.
I absolutely agree!
Exactly!
I don't agree, you don't have to choose between one option or the other, you should use all of them if necessary. Parallel compression, saturation + limiting will make the limiter practically not need to work.
I agree! I am doing exactly that prior to the limiter and my limiter barely limits if at all.
You don't have to agree. You do you, he does he. 🤣
@@nunojorgenjstudio7954 that's hilarious
parallel compression on the master ?
8:16 8:23 8:31 8:39 Not much transient difference, which was the main thrust of the purpose, but clearly different options arriving at end result options.
I like how you are able to describe other ways of mastering besides just brickwall limiting. Parallel compression sounds like a great idea, and maximization is definitely a great innovation!
Thanks for watching! I'm glad you enjoyed the video, and I'm happy to have helped you discover some new techniques!
I like to shave off a decibel or so with a transparent limiter nonetheless (Voxengo Elephant). After bouncing to tape and then using the plugin, I honestly could not tell any negative impact on the sound whatsoever.
Now you upload this. I have been resisting using one for years. Just starting mastering with one last week. Damn you You Tube.
I think the best way to use a limiter is for the sound and movement it makes, not for loudness.
Hip Hop production: What is the best way to execute mastering? Limit or no limit
just went back to an old master where I used the j37 as a limiter
J37 us nice sometimes it's a little too much
I work with a limiter on the master as a safely layer!
These days, I'm studying and practicing a lot of mastering with upward compressors. I'm now using limiters only for outlier peaks. This approach has worked best for me in genres of music that require a higher level of transparency and dynamics in the mastering: classical music, jazz, and progressive/symphonic rock.
Upward compression is the real deal,especially when paired with clipping for punchy transients.
4:52 - dear Sage, why do you have Lookahead and Oversampling at "OFF" on the Pro C2? DOes it have any purpose or did you forget to switch it on? 🤔
Less CPU usage, no pre ringing?
@djcata7474 Is it a question as an aswer? i don`t know, that`s why i asked in the first place
@djcata7474 Look ahead causes pre-ringing? (akin to linear phase processing)
@@ILOVEDRAGO Its an answer, of course. You should not have oversampling on by default, IMHO the negatives I mentioned outweigh the benefits.
And yeah, pre-ringing isn’t typically an issue with lookahead in compressors like it is in linear-phase EQ.
@@happylittlesynth no, not tipically.
awesome, hence Melda has its MutiBand Limeter and Maximizer.
I admit, I haven’t watched the whole video yet, and maybe you’ll clarify this, but bringing up everything besides the peaks IS functionally equivalent to turning down the peaks.
You may feel like you’re leaving the transients alone, but by turning up everything around them, you are decreasing their “impact,” potentially in a very similar way to what a limiter would do.
Looking at the plug-in doctor transfer curve you show a lot, it feels like bringing up the quieter parts of the signal while leaving the loudest parts alone is a different operation from turning down the loudest parts, but, except for some makeup gain, they are similar if not equivalent operations.
There’s no such thing as a free lunch. You have to balance loudness and transient integrity. You’re right that some limiters do this better than others, due to how their algorithms choose at what points to start and stop attenuating, and with what “shape.” (I think you know this, though you didn’t specify at least so far; look-ahead limiters, just like regular limiters, don’t JUST turn down the individual samples that are above their thresholds. This operation is functionally saturation/clipping. What makes something a limiter and not a clipper is that it turns down the samples AROUND the over-threshold samples as well, so as to preserve the transient better.)
It’s an interesting watch so far, I’m just not convinced that it’s anything new.
Okay, yeah, the Oxford Inflator and the PSP plugin are also both limiters, just limiters with soft knees. If the video were “stop brickwall limiting,” I’d agree, I don’t really do that anymore either, but what you call “maximization” is just limiting with a soft knee
What you're describing is reducing the dynamic range. Maximization or bringing up quieter details reduces the dynamic range - limiters do this as well. You can decrease the dynamic range without attenuating peaks - that's the concept I'm trying to demonstrate here.
The regular limiters you're describing here would have to operate as a clipper. A limiter, by its nature, utilizes lookahead to attenuate - otherwise, it would clip. There is no place in a waveform that a limiter can choose to start attenuating. To be a brick-wall limiter, it must attenuate the full "over," which necessitates the use of lookahead.
Saturation and clipping are the same thing - waveshaping to peaks. Clipping is usually associated with hard clipping or a complete squaring of the waveform. Saturation is usually associated with a gradual squaring of the waveform (unless asymmetrical waveshaping is used like with tube saturation), but they're 2 sides of the same coin, so to speak. I bring this up because I don't understand what you mean by turning down peaks and saturation being functionally the same thing. The purpose of a limiter is to not saturate or clip
@ thanks for the reply. I’m actually glad to be discussing with you, you very well may know things I don’t about this, but here is my understanding:
1. Increasing the volume of everything but the peaks is functionally equivalent to decreasing the volume of the peaks, save for some makeup gain. Whether we call this limiting or not is a matter of convention, but a unique process is not occurring. Now, it could very well be that your method allows for bringing up quieter parts of the signal in a particular way that is less intrusive to the transients in your signal, but, in theory, you could be bringing up the quiet parts of the signal that results in nearly identical results to decreasing the peaks of the signal. If so, you would see the same decrease in transient “impact.” Transients don’t exist in a vacuum - if the samples surrounding a transient are quieter, that transient will have more “impact” than if the surrounding samples are approaching the same amplitude as the transient itself. Again, maybe you can do this in a cleverer way or have more control with certain methods than with others, but principally, depending on how you do it and how a given limiter is configured, the operations could be functionally equivalent.
2. Let’s consider clipping vs. limiting vs. lookahead limiting. A simple clipper is just a few lines of code: is the sample in question, regardless of its surroundings, above the threshold? If so, change the value of the sample to be identical to the threshold. The problem with this is that it audibly distorts the signal.
Logic has (if I’m not mistaken) a limiter that can be configured to a 0 ms lookahead time. This limiter has a release time but no attack time. What (I imagine) is happening here, is that the signal is run through a similar circuit to the simple clipper, BUT, when the input signal dips back down below the threshold, the processing doesn’t just leave the signal alone again. It remembers how much it turned down previous samples and then “ramps down” (with some shape that depends on the release time and algorithm used) the gain reduction over some time frame, typically milliseconds. This preserves the transients a bit better, because the samples AFTER the over-threshold samples are also attenuated, preserving a bit of the transient amplitude relative to its surroundings (at least on one side).
Finally, the look-ahead limiter. This requires the most clever processing. The limiter “looks ahead” into the future, detects some over-samples, but then considers the samples before AND after them as well. Depending on the algorithm used, it both “ramps up” AND “ramps down” the gain reduction. The loudest sample would then be at the threshold, and surrounding samples, whether they were over threshold or not at the input, are turned down a bit as well to preserve their amplitudes relative to the loudest sample of the transient.
A limiter with true 0 attack time and 0 release time is a clipper. A non-brickwall compressor with true 0 attack time and 0 release time is a gentler saturator. The attack and release (or, in the case of the lookahead limiter, pre- and post- attenuation) are what makes compressors and limiters not saturators and clippers. If it weren’t for that ramping up and down of the gain reduction, it would just be clipping/saturating, and may sound like audible distortion.
@@sageaudio not sure if I was able to tag you in my reply… but see above :)
I really liked this one! I'll definitely use the parallel compression method and the Oxford plugin too, sometimes limiters can change your sound not like you wanted, and these techniques are great options
Just 'wow'. And in addition to 'wow', also, thanks for sharing this rebellious info. Might've never the gall to consider limiterlessness. To '25 and new territory!
And inflators are free to all in many forms, for the yet-uninformed, hopeful inflators out there. Keep on inflatin'!
i only keep limiters on my master so i don’t go deaf by 25
Very ear opening content. Thank you.
What about upward compression?
You've started with an explanation of how does the limiter operates - this is immediate like and respect from me.
Really interesting! I'm going to try to do this and compare how it sounds against the typical Pro L 2 on master, but I'm excited for the Dynamics processing technique since it'll keep the transients intact haha
Great video and solid explanation, thank you as always! 🙌🏼
Thanks for watching! Awesome to hear that you'll give this a try 👍
This is… revolutionary. Loudness without ruining transients.
PSP is the King 👑 THANKS 🙏🏼
What about vintage style limiters - Abbey Road TGI or Fairchild, stuff with a tone and feel?
The kind without look ahead? I’ll never make a commercial recording - but I want my songs to sound like tunes from 1955-1975. Any tips if I’m trying to emulate the sounds from that era?
I reckon it'll be just status quo for you assuming you're emulating the mixing and mastering chain of that time period (vintage analog mixing + mastering for vinyl)
I just tried this on a song I've been messing with, and i do feel like the overall mix felt much more alive. Going to be considering this every time I master!
I use clipping + limiting on my masters. Fat, punchy, and full. Limiters help pull the sound into a prettier image when utilized properly - similar to framing a picture.
What chain would you use for this? I'm getting ready to master my first track and limiting was the centerpiece of the whole thing I thought. Please lmk how I should approach a limitless master because honestly I really hate the squashed transient sound but want to bring up my low and increase clarity of the overall track.
Is Ozone Maximizer okay to use for Post-Punk/ Darkwave type music with low bass and higher treble sound?
Currently using an Ableton stock limiter on master just to tame what otherwise could cause distortion, mainly on vocals that I push to the limits.
Only .1-.2 db.
You think i should just soft clip on the channels and glue or parallel compress on master? Idk i will test it out.
Any tips for my mixes that are a little quiet compared to other songs on Spotify? LoudnessPenalty is saying it's at -4.5 reduction on Spotify so I thought I had that part right.
Have you tried the Loudener plu-in by Techivation? I’m wondering what method it uses.
Make a song without a limiter challenge?
Awesome I was just trying to think of how to accomplish this.
For the opposite outcome tho- i want to bring my mix (-3db peak and -12 Lufs) to streaming levels for the digital publishing.
I want to make this mix of (-3db peak, -12 Lufs) to (-1db peak, and -14 Lufs).
Then for the masters that I want to sell and distribute, I don’t mind the louder sounds.
I did notice that my pre-mix conscious songs ended up sounding a bit clearer albeit around -9 to -12 LUFS compared to clipped and limited songs around -6 to -7 LUFS. Honestly, the limiting might be more harm than good unless it's a distortion heavy genre, or a Rick Rubin production.
Thanks for the insight I’ll implement this to my next session 🤞🏽🔥
Well, what If I don't have Pro-C2, how can I compress like that?
you don't need pro c, you just need any compressor with similar features.
your videos have sped up my mixing so much i nail shit in 3 mixes and some quick acoustic mixes in 2 or less now :)))) tysm!
What is this measurement plugin that you’re using?
Plugin doctor
Is it okay to use 100% effect on the Sonnox Inflator while mastering? I just need to be sure
So many comments to refute this but Andrew Schultz has been using similar technique for years. He chooses not to parallel compressed drums during mastering. Worth a try anyway
What if the limiter has the option of bringing back those transients
Also I usually use a limiter paired with a maximizer so I keep the dynamics
I personally don’t think the vocals sound better without the limiter, but the rest of the mix sounds better
The biggest takeaway that I got from Bob Katz' "Mastering Audio" book was that the best master is one where the ME has to do very little because the mix is so good. That sorta opened my bedroom musician eyes to the importance of "fixing it in the mix" before it even gets to the master bus, let alone the ME. So if I think I need to strap some low end transient beeferizer on the master bus to fix a kick that doesn't kick... then maybe I should go figure out the kick drum first. So in your example, maybe the vocals are overcompressed... OR the vocals are just right, but the rest of the track is limp.
This might be common sense to someone with your experience
But how would fab c, or oxford inflator keep you from digital clipping
I agree that it sounds much better
But As far as i know, these plugins dont prevent you from going above 0
So while i preserve my transients better with these plugins, how do i maximize loudness while keeping my master fader from hitting red
iirc any fabfilter plugin will never have output above 0, they have built in limiter/clipper
@@SToXC_. I Just tested it and you are right. Thank you for the response my friend
Ima test my luck again lmao, but what about inflator
Any way to prevent peaks
Tbh i might just use pro c as limit and push into it with an inflator, but id like to hear your opinion as well
Thanks again
@@ITRiBUTEI Inflator is a waveshaper. It will hard clip when it hits 0.
@@EversonBernardes You know what you're right. Idk why, i havent used it in a long time and i only used it on my drum bus so i had no experience. Thank you both
Ill leave my comment up in case someone inexperienced like myself is wondering.
@@EversonBernardes ahh and to add as im testing it rn, when activating band split it doesnt prevent it but im sure there's a reason for that as well. Regardless thank you
What would you say about limiters on the mix bus during the mixing stage?
I assume it’s a similar deal but are there any benefits in the mixing stage absent in the mastering stage?
it's not inherently wrong but mastering engineers generally don't like this, & if you're using a limiter on your mix bus, you are likely using it to "fix" things that are better off tweaked in the mix. if you take off the limiter and your mix falls apart, your mix needs work
@ Of course to your last point lol. But appreciate the feedback. Yea, I usually use the limiter for one particular section that seems a little out of control. But I was thinking that same thought process this morning (about just needing to tweak the mix in those areas anyway)
I would definitely check out limitless by dmg cause I think you'll see that just a better limiter is needed.
Actually why not just use both? Or use the tool that sounds better for the use case
Interested to hear Panorama’s take on this. Thanks for the vid!
he will agree & disagree with different points but overall disagree with the idea of avoiding limiters. that's my prediction
@@Fire-Toolzbecause u can’t get super loud this way and that’s his mo/badge of honour.
As a dubstep producer, i just clip my master and end up between -3 to -1 lufs. I never touch a limiter ever 😂
This really opened my eyes
💥 The thing is, Skrillex`s masters look like a -1LUFS SAUSAGE, but the kicks sound clean and crisp. If THIS is the way how to achieve it, i`m in. But is this the way?💥
Looooool you hungry?😊
@@ChibindaChipawa FO SURE I AM LADY! 🤩
@@ILOVEDRAGO im a dude looool
@@ILOVEDRAGO I was making a joke about you mentioning sausages
No
Great video! What tool do you use for the analysis of the dry and effected signal in these videos please? Ie the graphing tool
DDMF Plugin Doctor
Bro Sage this is Gamechanger (what) the sound is much more pleasant . I Bought Limiter Elephant last time because like your video about him, but this is next level on the finally effect track !
damn. the parallel compression tip slaps, how did i never know about this
0:30 What is this plugin on the right?
That is DDMF Plugindoctor
Very useful
eq analyzer by bertom is a free alt
What's that VST your using to check the peaks? the blue one?
Plugin Doctor?
Now would you use this method also on drum busses so you don't destroy the transients?
You should definitely give that a try! Parallel compression has long been an industry standard for drum processing, but you can definitely try out some of the other techniques/tools we mentioned on your drum busses too!
Interesting concept and proof that perhaps a little bit of every thing (saturation, compression) can go a long way to achieving a better result than whacking a limiter straight on. Just one issue. At 0:59 I think you're doing the excellent Newfangled Audio Saturate plugin a disservice here by having it set to Clipper Shape 'Hard 100%' which is always going to produce that square wave. Setting it to 'Soft 0%' produces a lovely sine wave and in the other view a very gentle knee even when driven fairly hard. It's a great saturator!
Thanks for watching! I definitely agree, that's one of my favorite plugins. But wanted to show what squaring a wave looked like.
@sageaudio Ah! No worries if it was by design! 👍
Last time I've used a limiter on a master was probably 2007. A limiter on a master channel simply does not achieve anything that is not better achieved without.
But all this could have been done years ago, why is it suddenly the thing to do? Has something in the industry changed? (genuine questions not rhetorical) Thanks
I spent some time experimenting and realized I was overlooking this. Limiters are quicker/easier to use, but this method definitely has a big advantage with the retention of transient detail.
I think if you have to slap the Limiter at mastering with more than 3db it's better to fix the mix 😂😂😂
Heyyy, how'd you know that's exactly how I end up naming my projects near the end? 🤣 Been snooping? lol
🤣
I tried it and it sounds good. But it doesn't fit my track (Orchestral liquid drum and bass).
I think it depends on the situation.
->My track: Warm and soft with my Softclipper->limiter master.
->My track: Hard and I'll destroy your speakers with the compressor master.
Good points, definitely should be a situational decision!
-14 is loud? I don't think so. I would love to see these examples while reaching -9 or -8, as most clients expect, or even -6 for some styles, without limiting. All in all, the video is still about loudness war, you just didn't approach the problem raised by it.
Part of the sound of someone like Skrillex has been heavily reliant on proL2. I wonder what guys like that would say about this video. Interesting idea though.
Thanks for watching and commenting!
completely right!
Ive been following this channel for a while now and id say i agree for about 80% of the time, considering im a dnb producer. But in this case i have to disagree. In the age where every label tells me tp get a song to at least -5 lufs integrated it doesnt work without a limiter.
I'd love to send you a premaster of mine and you try and master it however you see fit and i only use an instance of standardclip and a pro L.
Feel free to reach out to us via email at admin@sageaudio.com about this!
There's also the imprinting to the brains of listeners you would need to factor in. If they got used to the squashed masters they'd expect it. It's not good, and maybe we could get used to cleaner masters.
I Love you guys!
You're still limiting though because a compressor is a type of Limiter.
Thanks for watching! In the traditional sense a limiter is anything with a 8:1 ratio or higher. But brickwall limiters behave much differently than compression.
Whereas compressors have an attack time, which represents how long the compressor takes before the maximum amount of attenuation occurs, brickwall limiters have no attack time. They utilize lookahead to measure the peaks ahead of time, then attenuates accordingly (in a way that introduces significantly less distortion and waveshaping than compression).
This results in significantly different treatment of transients.
I was always told to limit my master. When I was working with the Pro-L², I hated looking at the red crushing my transients. No matter how many times I saw people crushing them, I couldn't do that to mine. I always used the limiter in a subtle way, which kinda defeats the purpose of limiting. I think I'm just going to stick with compression & my Softube Bus Processor for the final stages.
i think one of the problems is that you were judging the way a meter looks. a limiter is meant to do what you're afraid of it doing. i think the problem is this anxiety of dynamics reduction. you know your compressor and softube stuff is reducing dynamics too. it's just not showing you that in a scary red color. don't be afraid of dynamics reduction! so many masters out there sound weak because someone was afraid of making the needle move too much, as if it's bad to do that. if it sounds bad then of course tweak it.
isnt this sort of maximizing? i usually use FL studios plugin Maximus to shape the compression to my taste
It definitely can be! We mention maximizing in the video, specifically in reference to iZotope's maximizer!
Each of those plug-ins is over $150 dang
It is a bit steep, but sometimes that's the price you pay for quality processing, especially when it comes to mastering-grade needs!
Regarding the last segment the Inflator easily sounds the worst out of the 3, a bit more of a tie between Saturn and PSP
inflator really kills your kicks by amplifying mids
Thanks for sharing your thoughts!
At the end of the day no listener will care or know why it's loud, as long as it sounds good
Hmmm. This is so simple and kinda brilliant.
Thanks! It seems like the next logical step if there are plugins that make processing this way possible
I really appreciate your content and knowledge I gained through years of watching this channel but I feel I need to disagree on this matter. Modern music standards requires you to use not only one, but sometimes two or even more different limiters in the chain. Music nowadays is 6 LUFS and streaming services requires 2 db of TP headroom with that loudness. Of course we still can master at by the book level of 14 LUFS but that will sound weak and practically no one talks about that. Every newbie mastering tutorial is about 14 LUFS and that’s so wrong. And even this level of loudness requires 1 db of TP headroom.
I think this mindeset is very genre/style dependent, but you definitely make some good points here. Thanks for watching and commenting!
Yea sorry man, every actual mastering engineer out there uses limiters for a reason. If Ted Jensen is limiting masters, it's fine to limit masters. Get better balances, clip before limiting, manage frequency balance better.
Thanks for watching! I first learned this from John Mayfield about 10 years ago (he's mastered some of the most successful country albums over the past few decades). I sat in on a few of his sessions and asked "why aren't you using a limiter?" and he just laughed at me haha.
@@sageaudio Yea I'd say that's genre dependent, and engineer dependent. You're not going to hit -5 LUFS on a metal track without using clipping and limiting.
great lesson, gracias!
Some ASIO drivers have built in limiters, be sure to check.
I'm an artist. I found the video very inspiring. I however think you could do with a little more assertiveness in your presentation. I loved this one
Thank you!
am i the only one that liked the limiter way more at all of the examples?
Yes.
Yes
Really good video 🎉
Thanks for watching!
Gread video, I've been saying this for years!
Now that's an idea that catches the eye, we'll see what happens in practice. But the concept is interesting.
2nd kooky video I've seen from this channel, after the "EQ the fundamental" thing. Avoiding from now on.
PS: Being aware of a sound's fundmental does help, but the thrust of the video was to move away from thinking of the EQ spectrum as a spectrum at all, and more toward looking at the fundamental. It was just too much "mix following an idea" rather than mix following your ears.
Thanks for watching! Sounds like the video made you consider a different approach for equalization, so it's genuinely cool to see it help another engineer. Kooky maybe, but calling a frequency range "warm" is also kind of kooky lol.
Amazing content as always! Thank you!
Thanks for watching!
Good video 😎🔊🔊
Appreciate your comment and viewership!
I just make it way way louder - DJ carnage