My favourite mastering limiters

People are always asking me - what limiters do you recommend for mastering?
And often, I sidestep the question.
But not today.
In this video, I show you two of my favourite mastering limiters, explain why I like them, and demonstrate the differences with a challenging example.
(I also explain why I don't love answering this question, along the way, plus digging into the topic of clipping versus limiting, the benefits and problems with each, and quickly test the new Slate FG-X 2 plugin.)
Music: "Beat 1v3" by Leon
Mix: Sebastian Demydczuk
Web: soundbalancest...
FB: / mixmasterdrumsrecordin...
IG: / soundbalancestudio

[@DanWorrall]

Did you try increasing lookahead in the FabFilter? With zero lookahead you're literally clipping the leading edge of transients. The larger the lookahead the cleaner it gets, at the risk of softening transients too much.

[@ProductionAdvice]

Yes, you're absolutely right that longer lookahead reduces the clipping, but I waned to use the default settings as much as possible for this video since that's what most people will try, and may not be aware that hard clipping is part of the sound if they start with something where it's being masked by high frequency content, noise-like signals etc. TBH I find it a little strange that the preset has such a short lookahead, I guess it's to minimise latency...?

[@DanWorrall]

@@ProductionAdvice I'm not sure if it affects the latency? Anyway that becomes irrelevant with oversampling enabled. I think they probably judged that in most cases that kind of clipping is preferable to a loss of punch and impact.

[@ProductionAdvice]

@@DanWorrall Right, that was my impression too, which is why I felt it was a fair comparison to make without diving into a discussion of lookahead

[@russellcameronthomas2116]

@@ProductionAdvice I understand your evaluation choice for this comparison video, but many viewers will get the impression that the Pro-L2 is unable to handle this material at all, and that they should buy one of the alternatives. IMO there is entirely too much attention in the music production industry on "oooh... buy this new thing" and not enough on getting good at what we already have. The FabFilter products, in particular, are very deep and capable, and it is very worthwhile for people to go beyond defaults and presets.

[@mixphantom0101]

Fair and professional observations as always! A follow-up video showing how you would use the same limiters and finding the "sweet spot" for each would be very interesting! I understand the "default" thing... but punchy? 😉

[@ProductionAdvice]

Interesting - you mean you don’t like the “Punchy” preset ?

[@mixphantom0101]

@@ProductionAdvice Punchy and Safe I have not found a use for. I actually made a couple of videos investigating "clean bass limiting" on my new channel and became a complete advocate for the Modern setting. Pro-L2 controls are just not like any other limiter I've tried, so the "simplicity" in the GUI is deceiving - and probably why FabFilter recommend you use their presets!

[@DomSigalas]

Very interesting comparison Ian, thank you! It was very interesting to hear Pro-L2 getting so crunchy- I wonder what is the combination of elements in the program material that's triggering this behaviour. My experienc with Pro-L2 compared to other limiters is that after a point it just sounds squashed and lifeless but it resists the temptation to distort (which can be good or bad thing :) ) I am very intrigued by your examples here, could it be the midrange of the belly synth with the big 808 that creates some sort of intermodulation? Thank you for this amazing video!

[@briancase6180]

Dom, it was you who turned me on to MD4. I waited for the summer sale and picked up the whole shebang for like $700 (? I forget exactly, but I really only wanted MD4 and a couple of the beautiful reverbs...). Thanks to you and Ian for helpful, fun content.

[@ProductionAdvice]

Actually @Dom I think the cause is the high level bass content with no high-frequency stuff to mask the distortion

[@DomSigalas]

@@ProductionAdvice aaah this makes total sense. Proves that point that there is no “best” limiter, it all depends on the program material. Great work Ian!

[@jbenoit1962]

I didn't recognize the channel name, was questioning the validity of the content then rolled over the thumbnail saw your face and said to my wife "Holy Shit, it's Ian!" Question resolved with astounding authority.

[@ProductionAdvice]

😂 Well that's very kind of you, but I also get things wrong and make mistakes, so please don't trust what I say just because it's me !

[@jbenoit1962]

@@ProductionAdvice For me it's to do with the process and the perspective. There's testing, critique, methodology applied. And an actively developing approach song to song, year to year, hardware and software release to release etc. You're a real Mastering engineer who contributed to my understanding of how to work and learn like a real mastering engineer. It's a different thing that I value immensely in a landscape now populated by content creators and salespeople serving audiences who often don't have the opportunity to learn how to be discerning in their sources. Dan Worrall's always amazing as we know and Panorama Mix & Mastering is probably my favorite Mastering content creator of 2022 who is only getting better. It's stuff that used to be secret dark knowledge now widely available and even something as simple as seeing your more conservative approach to LUFS and meticulousness regarding acceptable levels of distortion is a point of reference that allows me to better consider and contextualize my own work. :) One day it'll be my responsibility to serve the next generation the way you guys have supported me so I'm working hard to carry the torch. Big part of that is learning the culture of generosity you embody as well.

[@ProductionAdvice]

@@jbenoit1962 Thanks, I really appreciate the kind words !

[@amendippanesar]

I'm really loving that FGX2 it's my fave at the moment

[@ProductionAdvice]

I need to test it some more, but it looks promising

[@rolfmollvik]

Great video! I haven't used Pro L2 or any of the other limiters you are presenting in the video, but I have now done the same kind of test with two of the limiters I have: Nugen Audio's ISL, and Sonible's Smart Limit. Everytime I have got distortion, I have tried to increase the lookout, but it has not always been a success.

[@ProductionAdvice]

Increasing the lookahead helps with Pro-L - how much gain reduction are you trying to get ?

[@touc2007]

Thanks Ian. Really helpful and useful presentation.

[@ProductionAdvice]

Glad it was helpful!

[@mcdebugger]

Would be interesting to hear TDR Limiter in this particular example. It has separate clipper and limiter modules (actually, two limiters - one for regular peaks and one for TP inter-sample peaks). I've heard about it in one of episodes of your podcast and really like it.

[@acrosonusmastering]

fabfilter increase the lookhead, cos it work in 2 stage , first stage like a clipping and the second stage of compressor you set with the attack time, for example attack 30ms , for fB L2 it will act like if the limiting pass 30ms it will compress than just "clipping". Tonebooster barricade work in the same way and as good as fabfilter, good thing you get preset for the limiter part and in safe it will not clip (good for play security), for reducing the pumping place 2 instead 1. Prefer to use OVC 128 voxengo in softclipping knob 12 o clock before limiter, cos some limiter get hardclipping and do this bad clipping sound like AOM, it permits to go louder than most limiter and sound "transparent" but sometime it will do this hardclipping sound, so this why i use or FBL2 tonebooster as backup After it a balance, after the gain reduction meter it depends on you can get -5 but mostly a peak and a -1 who sound more obviously cos it more a longer part and "RMS level signal" And try to fight against the loudness wars, but some clients don't want to understand, cos more you crush the sound more it hade the default of mixing balance too. and now LUFS some ask the RMS value as LUFS but both are different..... it becomes a mess

[@saardean4481]

Interesting test. Thank you for that. On music of this style I almost always end up liking multi band working limiter such as Limitless, Elevate etc . Very often there will be rumble below 20-30 hz which overwhelms Pro L2 imho. If this type of bass music is not properly side chained you get swollen passages that limit (pun intended) how much you can push it. While these limiters sound more aggressive at times I find it suiting if not overdone. Having written all these thoughts of the moment, I live your channel Ian. Hope you are doing superb.

[@briancase6180]

I LOVE the MD4. I just choose my favorite preset and don't touch it. Someday, I'll try to actually use its capabilities, but for now, it's a sound-good-izer. Still, my limiter is Elevate (used after MD4 most of the time). I have yet to be happier with anything else. This is a really eye-opening comparison; thanks very much!

[@jbenoit1962]

That's me with Soothe2. I know how it works but I've found best application for me has been slapping on the same gentle preset when appropriate and building my chain from there.

[@GunterBrenner]

Is it just a display setting or the Pro-L2 is being fed an already clipping signal at 11:47?

[@nonamegrs]

Very interesting comparison. Thanks Ian! I usually use the L2 with lookahead at 70-80% in modern style, attack 1-2 o'clock, release about 10 and TPL always on. I would be curious if with this settings the L2 still clips that hard.

[@ProductionAdvice]

Great point, increasing the lookahead can reduce the clipping, yes, but for this comparison I wanted to stick with default settings since that's what most people use.

[@audeption]

@@ProductionAdvice Hi Ian! Increasing the lookahead was exaclty what comes to my mind when i saw this. This is what i try first, when i run in such an issue. Thank you very much, for this video!

[@mixey01]

Would like to hear this with the Sonnox Limiter

[@darrellroseborough7275]

Ian, you're a god sent. Been following you since your waves heydays. Therefore does this end the loudness war? I myself own both slate limiters and the waves L2. I don't have the pro 2 fabfilter. I tried the same settings and I do notice using the different presets opposed to the clear style setting that everything I throw at it is distorted. punchy, loud, tight settings. What about true peak is that important? Ian another quick question what do you think about Massy L2007 limiter? I've used that and gotten good results without distortion. very simplistic limiter plugin but giant results. Thanks for all you do for the recording Community.

[@ProductionAdvice]

Glad you liked the video ! The Pro-L clipping can be reduced by increasing the Lookahead value, but in this case I wanted to stick to default settings since that's where most people start. I don't think this really affect the loudness war either way, to be honest - if people want loud masters, there are various ways to achieve it - this video just shows some of the things you need to be aware of in that case. The only thing that will end the war is people deciding not to master so loud... I haven't tried the Massey Limiter yet, so I can'r comment, sorry.

[@ReeseWitherknife]

Great video, thanks a lot. Do you also happen to know why limiters seems to not catch all peaks completely, even with intersample peak detection and lookahead turned on?

[@ProductionAdvice]

Can you give specific examples ? Most that I use work pretty well in this respect

[@jacobsmith1877]

I use Submission Audio Flatline to clip a few db of transients, then a clean limiter for final level

[@esahm373]

Submission Audio Flatline is a generic clipper. Nothing special about it. Plenty of freeware can do the exact same and better!

[@toprekmusic]

true PeakeLimiting is Off in Pro-L 2? Why? I mran the rest of them have it on from what i can see. TPL in fab Pro-L is clipping as far as I'm aware?

[@ProductionAdvice]

You're right, I forgot to enable TP limiting in the Pro-L, but the peaks were only reaching -1.01 dBTP so this isn't an issue. I just checked by enabling it and the results are the same

[@toprekmusic]

@@ProductionAdvice Thank you for the video!

[@todezwis]

Have you done any work with DMG products: Limitless, Equilibrium, Multiplicity?

[@ProductionAdvice]

Not yet, but they're on the list. I think I've read that they need quite a bit of tweaking to get the best results, I'm not sure if I'd have the patience...!

[@BrentIraEnman]

Dmg limitless is the bomb!

[@587583922]

Just wanted to say....the Pro-L2 settings are a bit odd. It does include clipping for short signals, but the "Attack" knob defines what it considers short. By default (which looks like what you have set), anything above the threshold that lasts less than 275ms is considered short and hard clipped instead of letting its release behavior reduce the distortion. You get a cleaner sound by turning the attack knob down.

[@ProductionAdvice]

Yes, fair point. I deliberately stuck with the default settings for each limiter though, since these are what many people will be using

[@Vospi]

Not really "clipped" but "processed in a way that this mode applies". Which could include straight clipping.

[@Vospi]

Don't see anything good about sticking to default settings. Would you do that with a compressor? Lookahead and attack alone would've solved the problem in seconds, not 22 minutes.

[@ProductionAdvice]

​@@Vospi True, either lookahead *or* attack alone would have solved it, as I said in my comment above. And no, I don't often use default settings, but a limiter is a different beast than a compressor. In a floating-point DAW you don't even need to adjust the input gain if you don't want to. Lookahead can change the way a limiter sounds quite dramatically so I think it's reasonable to expect the default settings to give good results on a wide range of material. And they do, for the Pro-L 2, this was just an interesting example where they didn't. None of that was the point of the video, though. I took 20 minutes to: demonstrate what clipping sounds like, compare 4 different limiters, show to what extent they will clip the signal, plus explaining some of my ideas about using them & how to get better results. Sorry you didn't like it !

[@stevegalante]

I really like the FG-X2. Now the question is, can you get to the same LUFS using a "clean" limiter compared to a "clipping" process ?

[@ProductionAdvice]

At the kinds of levels I typically master to (-10 LUFS at the loudest moments) the answer is absolutely, yes. Up at -8 and higher, it gets much harder - which is why there’s almost always a degree of distortion involved in super-loud masters, no matter how that level was achieved.

[@darrellroseborough7275]

Oh sorry its me again. by the way the FGX2 is waaaayyyyy better than its original counterpart. I also have the sonible smart limiter. I must say when dialing in it has these messages to let you know exactly what all the media platforms will do to the audio Spotify, apple music, KZbin. the limiter tells you how your volume and dynamics looks. sometimes mind will say everything looks good and checks out but on my small set of reference speakers the distortion tells me different. but on my regular size 5"- 10" nearfield things sound good. The music should sound good if the limiter say volume, dynamics and the output say looks good right? when I bring the threshold way down then there's no distortion.

[@ProductionAdvice]

I haven't tried the Sonible limiter, but I looked at a review of their metering package which also makes suggestions for loudness & dynamics. The catch is that I think these are based on data, and of course the data will include lots of really loud masters that are out there. Just because "it's out there" doesn't mean it's good to emulate, in my opinion...

[@slashzerorecords]

Wow, thanks Ian! A really great example, even after the first round it was clear that TC wins by far imho! I tried a lot of limiters myself in shootouts, but haven't tried the TC yet. Have you also tried that same example with Ozone 10's limiter in IRC3 or 4 modes? Would be interesting to know if that can compete with the TC...

[@ProductionAdvice]

I can't use Ozone 10 without upgrading past Mojave, which I don't want to do, so I haven't tried it yet...

[@gnprice]

@@ProductionAdvice I'm also reluctant to give up my Mojave install - newer is not always better 🙂

[@billpodolak7754]

@@ProductionAdvice Ozone 9 and 10 have the same IRC modes, fyi. Was surprised Ozone wasn't included here! If you do end up trying it, definitely be sure to adjust the character slider as that is what controls the speed and as a result, the amount of distortion. Slow, larger character is cleaner, but more likely to pump. Fast character is more likely to clip.

[@ProductionAdvice]

Thanks Bill, I don’t currently have Ozone installed, so I just used the examples I’m familiar with. I emailed you about getting hold of a copy of Ozone 9 though 🙂

[@billpodolak7754]

@@ProductionAdvice I replied a couple days ago! Let me know if it didn't come through 😄

[@Crossfire2003]

I think that it's less about the limiter itself, as it is about the specific preset that has been used.

[@ProductionAdvice]

Different presets in the L2 use less or more clipping, but some limiters don’t clip at all

[@TylerDarlington]

Did you try higher levels of Oversampling (16x or 32x) on the FabFilter Pro-L 2?

[@TylerDarlington]

Also, did the distortion still exist once the track was bounced?

[@ProductionAdvice]

@@TylerDarlington I didn't try either of those things because I wouldn't expect them to make a difference, but just to be sure I just did, and sure enough the clipping was still there.

[@TylerDarlington]

@@ProductionAdvice Spectra 1964 has several analog limiters that boast attack and release times in in the nanosecond range. The claim is that nothing would actually clip or distort, and they claim that because it is so fast deesing isn’t needed (on electric guitar) either. I don’t believe it is possible at this time to make a plug-in that fast, but I was curious what your thoughts were on that, and maybe such a plug-in would solve this problem?

[@tonytemple8195]

Clipper into a limiter or two limiters can help 😉

[@ProductionAdvice]

I'm not a big fan of double limiting, personally. May do a video on this actually

[@tonytemple8195]

@@ProductionAdvice those small track limiters like DMG's track limiter or Kilahearts track limiter, two of the best in my opinion. 😉

[@tonytemple8195]

@@ProductionAdvice a video would be greatly appreciated too.

[@theCloneman5]

Too bad the System 6000 Native is practically abandoned as of now, I can’t even run it on my new machine. Behringer strikes again

[@ProductionAdvice]

That seems unlikely, they only released it fairly recently. What’s your machine ?

[@theCloneman5]

@@ProductionAdvice an M1 Mac system, it‘s been out for two years and still isn‘t properly supported natively. Even Pro Tools is getting around to it now.

[@briancase6180]

I'm running the TC suite (MD4, VSS4, TC8210, etc, etc) on my M2 Macbook running the very latest OS (13.3). No problems at all. Couldn't be happier 'cause I bought the suite on sale. :)

[@tomcarmichael100]

Was the Pro L2 still clipping as heavily if you increased the lookahead time?

[@ProductionAdvice]

No, increasing the lookahead reduces the clipping, but I wanted to compare the default settings as much as possible since these are what many people will use

[@tomcarmichael100]

@@ProductionAdvice Cool, thanks for clarifying Ian!

[@michaelrasbury]

Just to verify- the audio is preserved at lower bit depths with dithering, but in order to do this noise shaping in order to mitigate quantization errors, aren’t you permanently degrading the quality of the music by permanently adding noise that in some cases is quite audible?

[@ProductionAdvice]

Actually the opposite is true - check this video: kzbin.info/www/bejne/iJTCXpiXq7Cjqrssi=-lOZPVfVMOPQ1mZ1 (Short version, the added noise is less detrimental than the truncation distortion)

[@brooks1700]

sonible’s limiter has won me over from pro l2

[@kye903]

By the looks of it, the TC and the new FG-X one aren't limiting the true peaks...? Are you not bothered about doing that these days then? I know Streaky recently said he doesn't do it. You definitely lose something if you do limit them...

[@ProductionAdvice]

Both of them limit True Peaks, yes - in fact I forgot to enable TP limiting in the Pro-L, but it doesn't make a huge difference to this example. I disagree with Streaky, I always pay attention to True Peak levels and keep them around -1, but I don't obsess about it. I also disagree that TP limiters are any different to normal limiters in terms of the result, I think I need to do a video on that...

[@benshepherd2419]

Tbh I think putting any limiter on your mastering is really reductive- you should strive to push past all your limits to get the best master you can

[@normalizedaudio2481]

I don't like to use a limiter that much.

[@themagicianofsound]

It was the same for me, I didn't like to use a limiter until I got a song to master for a major artist / producer who forced me to make his master super loud, then I knew I had to improvise and I was in deep trouble haha ... Finally after some sleepless nights I got it sounding fine but I needed to "learn" at an accelerated pace how a limiter can be used to make music loud and good without perceiving any distortion...

[@colonelporridge8974]

I can hear the distortion on an iphone speaker

[@ProductionAdvice]

Me too !

[@benshepherd2419]

Slayyy