✅ Support Me in Making More Free Videos: warpacademy.com/membership-account/membership-plans/ Video Correction: When I spoke about Pro-L2, I reversed the Attack and Release settings that produce clipping. It's maximum long attack and minimum short release with 0 lookahead that will produce clipping. Sorry for the confusion. ➤ ➤ ➤ This is a Series - Watch More Here: th-cam.com/play/PLsyXZjufQdPzmm46ecm0HBgSpLNqdeb9_.html ✅ Get Our Free Studio Design & Acoustics Course: warpacademy.com/free-studio-design-course/ We also have an epic Discord server going on, very active with great discussion threads and tons of amazing people in the industry. ➤ ➤ ➤ Join Up! Hop on Our Discord Server ➤ discord.gg/ZwNgZteGKw
Great Video! Just wanted to point out that in order to use FF PRO-L2 as a hard clipper, the attack and release settings need to be the exact opposite of the setting you showed at 3:02. Attack needs to be at 10s and Release at 0s in order do bypass the dynamics module.
Hey hey. Thanks for the comment. Yes, you're 100% right. I was moving quickly in the video and initially got it backwards, ha! I already pinned a comment about this, and if you watch further into the video where I actually set Attack and Release, I explain it correctly. Good catch!
All good, I appreciate people pointing things out. I'm not perfect and sometimes I mis-explain stuff with all the tech lingo. Thanks for the support! Join up on our Discord server and stay in touch with the whole crew: discord.gg/ZwNgZteGKw
03:03 from what i understood both using it and reading the manual to get it working only as clipper you should leave the attack the longest. As per the manual " The Attack and Release knobs control how quickly and heavily the release stage sets in. Shorter attack times will allow the release stage to set in sooner; longer release times will cause it to have more effect."
Yes, that is 100% correct. I initially explained it backwards (you can see in my pinned comment I addressed this, as have several other comments). Later in the video, I explain it the right way around. Oops! And good catch. Nice to see people are paying attention.
Man i'm so so glad i found your channel a few weeks ago! Your videos are super high quality content and i'm extremely grateful that you give us all this useful information for free. You are much appreciated and respected. Looking forward to keep learning from you! Thank you so much! all the best from germany
Thanks for this! Limiters have been a mystical plugin for me, but now I understand it a bit better! I mix my own orchestral music, which is very dynamic, and I've not understood how to use limiters very well.
That's a great challenge! Orchestral music has such a wide dynamic range, so there's always some fine tuning involved. Note that my settings here in these limiters would be totally off for gentle orchestral music. Try this: -Use styles like "Transparent" "Safe" "Dynamic" or the Soft vs. Hard style in smart:limit -Use long lookahead times -Use shorter dynamics attack and longer release times -Use the limiter knee if it has one -Use oversampling at 4X -Monitor for True Peaks and leave enough margin where your TP level is well below 0 dB Hope that helps. Cheers!
@@warpacademy I was talking to engineers at the Cleveland museum of arts black box theater when I was a student. I asked one of them "Do you use limiters when you're mixing for orchestras?" He said "Well if you do that, you might not get a call back for another gig!" 🤣 For sure the music is too dynamic for any sort of aggressive processing. But also, different music is experienced in different ways. Our mixes don't always have to accommodate one particular way of listening
Love to know the nuts and bolts. Thank you. The Sonible unit is fantastic. Fast, easy and it works. Actually the whole line is great. Made for us mastering newbies and gives Ozone a run for it's money.
Fantastic content as always ! Limitless is still king, right ? :) No matter which limiter I genuinely test against it, I still come back to Limitless. Cheers from a fellow CT & mastering engineer.
Hey Nil. Nice to connect with a fellow CT and ME! Where are you base out of? Come join our Warp Academy Discord server, I'd love to see you there: discord.gg/ZwNgZteGKw. To answer your question, Limitless is the limiter I end up using on the majority of my masters, but not all of them. I always test it and A/B against other tools (usually Pro-L2, smart:limit, iZo Ozone 11 Maximizer, and TDR Limiter No. 6 GE). The reason I go with Limitless in most cases is because it sounds very clean at high loudness levels and more open on the top end. Very occasionally I prefer the sound of Pro-L2 with certain kick drums, and when I want a more contained, tighter sound. Sometimes it works better, but only every now and then. I'd not been using smart:limit that much, but now that I've gotten to know it better and talked lots with their team, I'm using it more and more. I'm not a TPL fan and in the earlier versions of smart:limit you could not turn TPL off. They released an update where you now can, and hence, I'm using it more! Stay tuned for upcoming videos as I did 2 big head to head comparisons of Pro-L2, iZo Ozone 11 Maximizer, smart:limit, and Limitless in loud drum n bass, and a hip hop session. Coming out soon!
Great video! I flip-flop between Ozone 11, smart limit and Acustica Lace. I have to say smart limit is super easy to set and gets great results. I use Lace if I'm feeling bold and want to make more changes. Ozone 11 has some great modules in general but I stopped using its AI settings. I like using the Clarity module along with Maximizer. Depends on the material too.
Isn't it a bad decision to use a plugin on the master that supports multiband (in this case I'm talking about DMG Limitless)? As far as I know, the respective bands of Limitless can only be processed individually because the plugin internally sets high and low cuts, which in turn lead to resonances and phase changes at the intersection points. Consequently, in terms of a homogeneous sounding stereo track, a limiter with one band should sound best and cause the fewest problems. Please correct me if I am wrong. Thank you very much 😊
Hey hey. Yeah, so let me set the record straight on a few things here. First, any well-coded multi-band plugin is going to use Linkwitz-Riley crossovers, which are completely transparent in the frequency domain. Meaning there are no "resonances" that affect the frequency response. I posted a link below. Also, the filters comprising the crossovers are linear phase, which means no phase rotation as you'd expect from minimum phase crossovers. In fact, it's entirely possible to design crossovers that are minimum phase that are perceptually transparent as well - Soundtheory accomplished this in their Kraftur plugin. The human auditory system is relatively insensitive to phase shifts. Did you know that the speakers you're listening on - if they're 2-way or 3-way - have crossovers that are rotating phase? Do you hear that as bad? Also, all analog mastering gear uses filters that cause phase shifts. And the audio you're hearing in your room all has phase shifts happening at various frequencies due to the reflected energy off your room boundaries. Frequencies even drift in and out of phase relative to each other as a wavefront travels. It's only in the digital domain that things like linear-phase filters can exist by delaying audio. It's also not bad to use multi-band processing on your master. This is part of the magic of Limitless. By splitting the bands, it's able to use less gain reduction than single band limiters because it's only inducing gain reduction exactly where needed (kind of like a parametric dynamic EQ). The gain changes on a sample by sample basis can also be optimized for less distortion better in a multi-band limiter. It doesn't limit if it doesn't need to, which means a more open sound, less distortion, and less overall limiting. Limit-LESS, it's in the plugin name ;) In this case, I would say to stop thinking about all the theory and things people have said about filters and listen closely. How does Limitless sound to you? Watch this video for a careful and extensive comparison between single band limiters, and Limitless: th-cam.com/video/Yp3Q4g6Vtjw/w-d-xo.html Learn about Linkwitz-Riley crossovers here: en.wikipedia.org/wiki/Linkwitz%E2%80%93Riley_filter Read this post from Dave Gamble (from DMG) about why multi-band limiters can solve problems that single band limiters can't: gearspace.com/board/new-product-alert-2-older-threads/1058709-new-dmg-audio-limiter-limitless.html
Thanks for the video. I almost always use Smart limiter - great limiter.But lately I started (sometimes) to use Bus PEAK from Arturia. Bus PEAK from Arturia is a good limiter, but the problem is that it consumes a lot of memory (DSP).
Thanks for sharing! Yeah mastering limiters are some of the most CPU intensive plugins. But even still they should be coded in a way that they don’t hog resources. Pro-L2 is a great example. I was able to load up 40 instances of it at only 15% CPU usage. Pretty amazing. I’ve never used PEAK. Sounds like it’s aimed more at bus limiting than master limiting perhaps. Cheers!
Hello sir could you please recommend plugin that works as limiter and compressed at the same time. For example we raised the loudness and it will compress the loudest part to make sure it won't clip.
Hey hey. Sure thing. I just want to make sure you fully get what I'm saying in the video. These modern limiters have exactly what you need. They have the fast transient stage that will ensure no peaks escape (this ensures you don't have any samples going over the output / 0 dBFS). And the slower dynamics stage compresses the general dynamics of the song. For this reason, a lot of mastering engineers I know (myself included) rarely or never use compressors on their master. These modern limiters do both jobs. Let me know if that doesn't make sense and I can elaborate. Cheers!
Hey hey. If you’re looking for a second limiter and you already have smart:limit then I would get Limitless. It’s just so good. And it’s very different than Pro-L2 because it’s multi-band. Watch this video for a comparison between Pro-L2 and Limitless: th-cam.com/video/Yp3Q4g6Vtjw/w-d-xo.htmlsi=4XOTh2EKtGgU-Muc
Oh, right, I misread ya. I would prefer smart:limit over Pro-L2 personally. Less distortion, more features - especially with the Saturation and Bass parameters. Cheers.
Oversampling is more complex than what you explain. The higher the oversampling the more "sampling points" so the less chances you have to get inter-sample peaks so the limiter will handle better those peaks (and obviously their difference in level will be smaller) BUT in fact while downsampling your can create back overshoots BUT these are not taken care of in any way UNLESS you turn on "true peak limiting" which is the button to enable this 3rd level of dynamic control. This being said TPL has a color and has a tendency to "blur" and remove "sparkle" in the transients. In most of the cases ideal scenario would be lookahead of 0,1, highest possible oversampling and setting the output somewhere around -0,3 to -0,5 of the desired output ceiling to be safe that ultimately no ISP will peak above without needing this 3rd level of dynamic control induced by the TPL
Oversampling is much more complex than almost anyone explains in a single video. That said, both myself and some other mastering engineers have covered the topic very well with several videos. I'll unpack your comments one by one: "The higher the oversampling the more sampling points" - yes, correct "So the less chances you get intersample peaks" - Kind of - I get you here. By using OS you cause the limiters detection circuit to read higher peaks (ISPs) then if you were not using OS, and then the end result of the limiter output should have a lower TP value. The thing is, you now get more limiting due to it reading higher peak level. Watch this video I just did on using oversampling in limiters: th-cam.com/video/Xf9gVOKa_U4/w-d-xo.html "While downsampling you can create back overshoots BUT these are not taken care of in any way UNLESS you turn on TPL" - Incorrect. A limiter must create a hard ceiling at project sample rate without overshoots. This is completely independent of the TPL stage. If you set the limiter to oversample, it will downsample, it will create overshoots, and it must 100% have another limiter to catch overshoots. The TPL stage, by virtue of oversampling and downsampling again, must also have a final limiter that runs at the project sample rate to catch downsampling overshoots. "TPL has a color...." - I mean, yeah, but only because limiting has a color. All TPL does is engage more limiting. There's nothing special about it. It's just another limiter. If you perform more limiting in the main limiter it will also have all the same downsides you describe. "Highest possible oversampling setting" - As it turns out, no. There is no benefit to more than 4X oversampling in any of the limiters I use if you're running a base rate of 48 kHz in your session. The ITU for TPL specifies 4X, not more, for this reason. Sonible's team also did a ton of research into this and confirmed that anything more than 4X was overkill. And finally, Nicholas Di Lorenzo @Panorama_mastering did a video about “The Best Oversampling Amount” and also confirmed 4X is where you should max out and that was for a clipper, not a limiter: th-cam.com/video/6Ss3cHXGjl4/w-d-xo.html& You know your stuff, that's clear. Great questions and discussion. Follow up on those videos I posted and let me know if you have any other questions.
@warpacademy overshoot can happen because au phase rotation of the anti aliasing filters fore sure. Yet I can swear there might not be added value of oversampling above 4x according to Nicholas but I did some Delta tests and you hear a clear difference between x4 and x32 on pro L2. Transients stay somehow sharper and snappier, les grungy. I really encourage you to try it out yourself. ITU recoms are based I guess about a minimum viable oversampling amount to be tolerable if ISP occur that they are so little and so short they won't cause any audible artifacts on most convertors. Nice discussion in fact and thanks for the in depth content you share , it's not the usual YT entry level BS ;) keep them coming ! Cheers
@@incidence.studio Hey hey. Yeah, it's a good conversation. Regarding overshoots, the main thing I'd like to point out is what causes them. Overshoots happen from band limiting a signal. So they happen with any type of anti-aliasing LPF - it can be linear phase or minimum phase. So even without any phase rotation, you still get overshoots from the downsampling process because by removing higher frequency sinusoid partials you change the waveform and its peak amplitude. For this reason, anytime you have a last-in-chain mastering-grade plugin that oversamples, you will see another final limiter or clipper after the downsampling LPF. Some manuals for plugins specifically discuss this, such as Sountheory Kraftur. It's interesting to hear about Pro-L2 and transients. I think that may have more to do with their windowing on their anti-aliasing filter. It definitely affects the top end. Also interesting to note is that oversampling is not even implemented in DMG Limitless (except for its clipper and ISP, not for the main limiter) or Ozone 11 Maximizer. They are both sample peak limiters, and to be honest, DMG Limitless sounds cleaner than either Pro-L2 or Maximizer. Sonible's smart:limit maxes out at 4X. So Pro-L2 is the odd one out in that bunch. I don't use Pro-L2 much anymore as I almost never like how it sounds compared to Limitless. Anyways, nice to chat. All the best.
one question, how is the multi band limiter affects the tonal ballance of the track? I know many engineers avoid them because of this, working hard to make fine eq moves to get the perfect ballance, then changing that with the limiter.
Hey Alex. Great question! Limitless is fully customizable in this respect with the Separation control. If you set Separation at 0 it's a single band limiter. Separation of 100% is still great, and will reduce distortion by ensuring bands that don't need to be limited as much are not as affected. More than 100% runs the risk of making the mix sound re-EQed, but can be useful for problem bands. As an ME, we would always keep a dynamic EQ in our tool kit for this. Limitless can do that right in the limiter. So, it's a "use with caution" kind of thing, but you have full control over it. If you don't want Limitless to be multi-band, you can turn down Separation. Cheers!
@@warpacademy thank you very much for the time to answer. Yes, I fully understand the reason why a multi-band limiter would be useful, exactly as you mentioned. But still, provided it is needed, it will definatelly affect the tonal ballance, as the bands which are limited less, will be emphasized. So how do you approach this? do you accept those tonal changes as a tradeoff for limiting separate bands based on needs? or would you then revisit the eq and make some changes? (with this potentially require a second visit to the limiter and so on)
Hey hey. Sure, happy to address this. You're thinking that band-specific compression is a bad thing. It's not. It's just a tool. I mean, think about how many engineers for years have been using multi-band compression on the master? It's nearly the same thing. In a single band limiter, if a huge kick comes in and triggers the limiter, it pulls everything down. The whole song can sound quiet and sucked out from the big kick. In a MB process, that doesn't happen. It's not an advantage or a disadvantage, it's just a thing you can use. If you set Limitless to 200% Separation, you would notice it, perhaps. At 100% or less, like I set it, you really don't notice it affecting the tonal balance in a negative or overt way. It simply sucks down any specific frequencies that may pop out from time to time, rather than making the bands that aren't limited as much sound "emphasized". The other large advantage of an MB setup is that you can scale the release times to be longer in the LF and shorter in the HF, rather than one single release timescale for everything. This was also a good MB compression trick used by engineers. So the big sub hit, the 808, the kick, the bass guitar get a nice gluey release, and hi hats get a more rapid release to recover for the next transient of fast hat passages. Limitless scales the per-band release time in the same way. And finally, fast MB gain limiting processes will produce less intermodulation than a single band process. When using minimal lookahead and fast settings, limiters will definitely produce intermodulation on complex signals, which can and does produce non-musical sum and difference partials that can extend into the extreme LF below the fundamentals of the kick and bass. For proof of that, look at the documentation on Soundtheory Kraftur (it's a MB clipper, but that's very similar to a fast MB limiter set to waveshaping timings). They clearly state that the MB processing in Kraftur produces less intermodulation on a master, and they're right. Do you use Limitless? You should try it. Also, if you have iZotope Ozone, the Maximizer is multiband if you use the IRC IV style, you just don't get any control over it.
@@warpacademy Wow, awesome detailed explanation, thank you. I don't have Limitless, I do have Ozone Maximizer though. I understand what you are saying, I will try to experiment with multiband to see the advantages and disadvantages on the sound :)
Right on Alex. Keep me posted on how that goes! Also, we have a Discord server for Warp Academy. You seem like a good fit for the community there. Come join up! discord.gg/ZwNgZteGKw
Great question. A longer lookahead will cause the limiter transient stage to kick in full gain reduction very early. The upside is no distortion, the downsides are: quieter master and transients that sound dull, suppressed and lacking in impact. Plus more latency, which may not matter to you if you're in a mastering setting. Longer lookahead is good for some material, but I almost never use it for the music I engineer.
Good suggestion. I'll keep that in mind as I have more limiter videos coming up in this series. In this video, I was focusing more on the stages and general intro to workflow rather than a deep dive on the effect of settings in each limiter. If I did that, this video would be 2 hours and not 25 min. My videos are already on the long side so I keep the scope for each one as tight as I can.
@@warpacademy but you did other A/B’s and really talked about how you lean into hop hop mixing, and how important bass is to hip hop and mentioned how the weighing really affects the bass, but no demonstration.
I get you. I'll explain. That's because I already shot a separate full limiter comparison video using a hip hop song (the same one) with Pro-L2, iZo 11 Ozone Maximizer, smart:limit, and Limitless. I show head to head comparisons and go more deep dive into settings. I don't try to cover every aspect of workflow in a single video. I break them up, to make them a bit more bite sized. This video was already 27 minutes right? Most people don't have the attention span for that, so I make series. Stay tuned, I think you'll like the hip hop limiter comparison coming up. Or, if you don't want to wait, you can download the trial of Limitless, and A/B/X to your heart's content with a hip hop track on your end. All the best.
The windowing function for true peak in Pro-L 2 is too aggressively interpolated. no matter the amount of gain reduction i do with it, even just the odd overshoot here and there, i can hear it push the transients further back than it should. if you compare the True Peak limiting in Pro-L 2 to say Tokyo Dawn Labs LimiterNo6 GE it's a night and day difference. there's definitely some softening on the latter if you try and do about 1dB of almost constant limiting on transients (nobody who knows what they are doing though would do that i hope), but using a TP limiter how it is supposed to be used, LimiterNo6 wins out over Pro-L2 for me
Hey John. Thanks for the comments. I wonder if this has to do with the fact that Pro-L2 uses 8X OS in their TPL stage and not 4X like most do. I talked with the Sonible team and several other MEs and they all agree that anything over 4X provides no benefit, but it will likely read and react to higher ISPs and push the transients down more as a result. Thoughts on that? Are you an engineer or DSP guy? Cheers!
@@warpacademy That might explain it, yeah. i was wondering if the oversampling for TP in Pro-L 2 is perhaps minimum phase as it really does sound like it almost slewing the transients. the phase shift might be what i am hearing. OS in most other Fabfilter plugins is minimum phase(except for Saturn when you engage the Linear Phase crossovers), so it's not out of the realm of possibility I am or was rather a an inhouse mix engineer for some subsidiary dance music labels in the UK and Europe for Sony and Virgin Records(waaaay before UMG gobbled them up) before that, but I've pretty much retired now about 4-5 years ago after doing the freelance thing for a bit. still interested in all of it as a hobby though. doubt that will ever change
Ah neat!! Thanks for the background. Yeah it sounded like you really knew your stuff. A lot of plugins give you control over the downsampling LPF and you can use minimum or linear phase (like Schwabe Digital stuff) but some don’t. Linear phase can smear transients. I’m working with the Soundtheory team to do a video on their new Kraftur plugin and they only use minimum phase crossovers for that reason. In general I like the option to choose as a user. I’ll have to ask the FabFilter team more about their LPF configuration. Anyways, nice to see you around the channel and thanks for the contribution to the discussion. Stay in touch!
I will have to watch on replay....But I will take notes for this....I have software that have limiters. I won't touch it because I do not know what I am doing....hahahahaha Thank you for this video....
Yep, it's an eye opener hey? Back in the day, this is why engineers would sometimes use 2 limiters in series. One would have fast settings for shearing peaks (small time scale for managing micro dynamics) and the second limiter would have smoother settings (larger time scale, macro-dynamics). If you have a single stage limiter, then small, short transients that are loud will trigger the full release envelope and cause a big "hole" in the audio.
Longest attack and shortest release, for sure. You're 100% right. I reversed that in the video. Sometimes happens ;). I did discuss it later on and described it correctly that time, explaining how longer attack is louder and lets more transient through and the shorter release is louder and prevents transients from being suppressed as much. Good catch!
Good question. I’ve tested this myself and those settings in Pro-L2 do sound substantially different than a hard clipper to me. I’m not sure why but perhaps that’s why they named the preset “practically” and not just clipping. I expect, if I had to venture a guess, that there may be a limiter knee happening under the hood that makes it into a soft clipper transfer curve.
Hey hey. Yeah good thinking on that one. I bet it does. I mean I can’t imagine that it wouldn’t actually. But seeing as it’s all happening under the hood we may never know for sure. Cheers!
It’s not just a drum beat really. There’s more to it. And also, a huge amount of music consists of drums, synths and an 808. If you want to see different music in a mastering session then I have a whole playlist that has different genres including vocals and guitars. Just check the playlists section of the channel for Mastering Sessions.
loved the video but I agree, the music was not good (very little bass content, no vocal taking up a massive part of the frequency range, etc) for showing what a limiter struggles to fix, but that wasn't really the point of the video. Next time try a dense pop mix with heavy bass, a punchy kick and an in-your-face up front vocal like every mix I work on. Then you have no choice but to hit it at about 3-4db of limiting if the request is -6 or -7 LUFS. That is where the limiters really start to fall apart, but Limitless still stands tall for me. I have used it with great success doing far too much limiting than I'd normally be comfortable with (this is after plenty of clipping, bus limiting, etc...) Another interesting video would be whether or not clipping is pointless before any sort of multiband (non linear phase) processing such as a Pro-MB or EQ on the master.... It literally moves the clipped point to a different part of the waveform leaving it unclipped. I still need to do more testing, but it would essentially mean that everyone who has recommended clipping for peak control at various stages in the project was wrong.
Hey hey. Noted. Thanks for the comment. I’ll keep that in mind. I agree that around 3-4 dB is where limiters fall apart. And at the same time I’m able to turn out mixes that are -9 LUFS or so, so I have been able to reach loudness in the range of -6 in the heavy sections without more than 2 dB on the limiter. The concept is straightforward: you do more of the “mastering” duties on your Sub-Mixes where you have less complex signals and thus less intermodulation. And you do less on the master. And using multiband clipping and multiband limiting on your master, again to reduce intermodulation and get the most minimal amount of GR needed per band. Regarding phase shifts and clipping yes that’s an important point. Ian Stewart and Nicholas Di Lorenzo had a debate about this on his channel. Nicholas tested it and found that as long as you’re not creating extreme phase shifts (IE minimum phase cut filter) after the master clipper than the amplitude change on the sampling points is negligible. IMO using clippers to control peak level at a track / individual sound level is 100% a needed best practice regardless of what is going on downstream. Also interesting to note is how Soundtheory Kraftur is set up. It’s a mastering grade clipper. Yet its crossovers create full phase rotation at the x points. And they have wideband clipper after that. It’s also the loudest and cleanest clipper on the market today. All interesting things to ponder. Come join up on our Discord server and chat more! discord.gg/ZwNgZteGKw
@@warpacademy Anytime I ask you a question on your channel I am reminded of why I consider your channel one of the absolute best on YT. You are a true engineer and an incredible teacher. What a thoughtful and helpful answer. I have asked all over the place for thoughts on this topic and have received some tepid responses. This was so thorough and pointed me in the right direction with references. Do you have a video regarding multiband clipping and when you prefer to use that over a wideband clipper? I have never tried a multiband clipper out of fear of phase rotation undoing my hard work in the mix. Also because I have a hard time wrapping my head around when it would be preferred other than preserving low end while killing some high transients. Most modern mixes are heavier in the lows so it's not really the high freq transients causing peak issues. Maybe I have the use case backward. Regarding Kraftur, do those crossovers not cause problems?? I would think they would negate a lot of the clipping you've used throughout the mix if you're throwing that on the master. Perhaps the phase is not affected enough to cause an issue? Loudest and cleanest is a bold claim, now I've GOT to try it! I've used so many clippers and always find myself returning to standardclip. I joined the discord :)
@@musiclover-r9m Thanks so much for the kind words and the reflection of how I've been showing up. I love detailed conversations like this, and taking the time to explore the many nuances of audio. In terms of multi-band clipping, I've noted your questions and will aim to answer as many of them as I can in my upcoming video on Kraftur. I'm actually about to hop on a call with their product specialist so it's good timing. But to quickly answer what you've asked, I would want to use an MB clipper on my master vs a wideband clipper if it's doing much work. It all comes down to intermodulation. MB plugins have much less intermodulation than wideband ones with distortion. There are use cases where I would want to leave the low end untouched, but clip the mids and highs. Luca Pretolesi often uses K-Clip with a subtractive low shelf before it and an equal / opposite additive low shelf after it to reduce the amount of low end triggering the clipper and to lower the clip point. Solid technique, but now Kraftur can do that without all the EQ. Regarding Kraftur crossovers. They cause full 180 degree phase rotation at the crossover points. I was initially alarmed by this and had a fairly extensive conversation with the Director of Research at Soundtheory. He said they've built the crossovers as minimum phase intentionally to avoid transient smearing from non-causal linear phase filters, and they've developed an in-house perceptual model of the human auditory system to use in design. He told me to stop looking at Plugin Doctor and to go listen to it. So I did. I can throw it on my master, with no clipping, and move the crossover points around and I don't hear anything changing. The crossovers sound completely transparent. This has to do partially with the fact that, and I quote, "Our auditory system is very forgiving when it comes to gentle phase changes, mostly because the signal chain from the source to your inner ear adds a lot of phase changes that the auditory system needs to deal with. Unless you have a phase response whose second derivative with respect to frequency, which is called the dispersion of the response, exceeds a certain threshold, the phase changes will not introduce audible artefacts." More on this soon as Kraftur is one of the next videos in my production schedule. See you on the Discord server!
Hi Warp academy - Good video, but you assume all of us use external paid for 3rd party plugins. Not all of us can afford these paid for bells and whistles.
Hey hey. I totally hear you. When you're starting out, often times you work with the stock plugins in your DAW or free plugins. You don't necessarily need a fancy limiter to get a good master. There are also some decent ones in some DAWs, but I would actually recommend that you look TDR Limiter No 6. It's only 50 euro and it's very good: www.tokyodawn.net/tdr-limiter6-ge/. If you are serious about music, you'll invest in your tools incrementally as you can. You don't need the most expensive plugins, but you would benefit from the mindset of growing your tools. Or find good freeware. There's never been so much good freeware on the market as now.
They did update their limiter. Added some new much needed features. Worth stacking it up against these for sure but I don’t have high hopes. Maybe I’ll be surprised?
Done! I have quite a few videos of singer / songwriter stuff with vocals, guitar, bass, drums in my Mixing and Mastering Playlist: th-cam.com/play/PLsyXZjufQdPzK5ndIgFQiCWIDICCKPaj4.html. This one specifically could be useful for you: th-cam.com/video/xnTpwu0YbU0/w-d-xo.html
![The Best Limiter for Loud, Aggressive Music? [DnB, EDM, Bass Music]](http://i.ytimg.com/vi/Yp3Q4g6Vtjw/mqdefault.jpg)
![The Best Limiter for Loud, Aggressive Music? [DnB, EDM, Bass Music]](/img/tr.png)
✅ Support Me in Making More Free Videos: warpacademy.com/membership-account/membership-plans/
Video Correction: When I spoke about Pro-L2, I reversed the Attack and Release settings that produce clipping. It's maximum long attack and minimum short release with 0 lookahead that will produce clipping. Sorry for the confusion.
➤ ➤ ➤ This is a Series - Watch More Here: th-cam.com/play/PLsyXZjufQdPzmm46ecm0HBgSpLNqdeb9_.html
✅ Get Our Free Studio Design & Acoustics Course: warpacademy.com/free-studio-design-course/
We also have an epic Discord server going on, very active with great discussion threads and tons of amazing people in the industry.
➤ ➤ ➤ Join Up! Hop on Our Discord Server ➤ discord.gg/ZwNgZteGKw
Great Video! Just wanted to point out that in order to use FF PRO-L2 as a hard clipper, the attack and release settings need to be the exact opposite of the setting you showed at 3:02. Attack needs to be at 10s and Release at 0s in order do bypass the dynamics module.
Hey hey. Thanks for the comment. Yes, you're 100% right. I was moving quickly in the video and initially got it backwards, ha! I already pinned a comment about this, and if you watch further into the video where I actually set Attack and Release, I explain it correctly. Good catch!
@@warpacademy Ahh, I should've looked closer 🫡 thanks for clearing that up. Love your Videos, been enjoying silently for a long time. Keep it going 🙏🙏
All good, I appreciate people pointing things out. I'm not perfect and sometimes I mis-explain stuff with all the tech lingo. Thanks for the support! Join up on our Discord server and stay in touch with the whole crew: discord.gg/ZwNgZteGKw
03:03 from what i understood both using it and reading the manual to get it working only as clipper you should leave the attack the longest.
As per the manual " The Attack and Release knobs control how quickly and heavily the release stage sets in.
Shorter attack times will allow the release stage to set in sooner; longer release times will cause it to have more effect."
Yes, that is 100% correct. I initially explained it backwards (you can see in my pinned comment I addressed this, as have several other comments). Later in the video, I explain it the right way around. Oops! And good catch. Nice to see people are paying attention.
Man i'm so so glad i found your channel a few weeks ago! Your videos are super high quality content and i'm extremely grateful that you give us all this useful information for free. You are much appreciated and respected. Looking forward to keep learning from you! Thank you so much!
all the best from germany
Hey mate. Thanks for the kind words. Glad you've been enjoying the channel. Subscribe and keep in touch!
@@warpacademy of course i already subbed instantly back when i discovered you haha ;D cheers!
Right on. Thanks for the support.
Thanks for this! Limiters have been a mystical plugin for me, but now I understand it a bit better! I mix my own orchestral music, which is very dynamic, and I've not understood how to use limiters very well.
That's a great challenge! Orchestral music has such a wide dynamic range, so there's always some fine tuning involved. Note that my settings here in these limiters would be totally off for gentle orchestral music. Try this:
-Use styles like "Transparent" "Safe" "Dynamic" or the Soft vs. Hard style in smart:limit
-Use long lookahead times
-Use shorter dynamics attack and longer release times
-Use the limiter knee if it has one
-Use oversampling at 4X
-Monitor for True Peaks and leave enough margin where your TP level is well below 0 dB
Hope that helps. Cheers!
@@warpacademy I was talking to engineers at the Cleveland museum of arts black box theater when I was a student. I asked one of them "Do you use limiters when you're mixing for orchestras?" He said "Well if you do that, you might not get a call back for another gig!" 🤣 For sure the music is too dynamic for any sort of aggressive processing. But also, different music is experienced in different ways. Our mixes don't always have to accommodate one particular way of listening
Well said. Great insights.
Love to know the nuts and bolts. Thank you. The Sonible unit is fantastic. Fast, easy and it works. Actually the whole line is great. Made for us mastering newbies and gives Ozone a run for it's money.
Right on. Glad to hear that. And yes the full Sonible lineup is fantastic. Cheers!
limitless does sound the best IMO. more transparent, transient preserved, and spacious.
Agreed!
Great video thank you! Would you ever do an Ozone 11 video?
Great suggestion!
Fantastic content as always ! Limitless is still king, right ? :) No matter which limiter I genuinely test against it, I still come back to Limitless. Cheers from a fellow CT & mastering engineer.
Hey Nil. Nice to connect with a fellow CT and ME! Where are you base out of? Come join our Warp Academy Discord server, I'd love to see you there: discord.gg/ZwNgZteGKw.
To answer your question, Limitless is the limiter I end up using on the majority of my masters, but not all of them. I always test it and A/B against other tools (usually Pro-L2, smart:limit, iZo Ozone 11 Maximizer, and TDR Limiter No. 6 GE). The reason I go with Limitless in most cases is because it sounds very clean at high loudness levels and more open on the top end.
Very occasionally I prefer the sound of Pro-L2 with certain kick drums, and when I want a more contained, tighter sound. Sometimes it works better, but only every now and then.
I'd not been using smart:limit that much, but now that I've gotten to know it better and talked lots with their team, I'm using it more and more. I'm not a TPL fan and in the earlier versions of smart:limit you could not turn TPL off. They released an update where you now can, and hence, I'm using it more!
Stay tuned for upcoming videos as I did 2 big head to head comparisons of Pro-L2, iZo Ozone 11 Maximizer, smart:limit, and Limitless in loud drum n bass, and a hip hop session. Coming out soon!
Great video! I flip-flop between Ozone 11, smart limit and Acustica Lace. I have to say smart limit is super easy to set and gets great results. I use Lace if I'm feeling bold and want to make more changes. Ozone 11 has some great modules in general but I stopped using its AI settings. I like using the Clarity module along with Maximizer. Depends on the material too.
Right on! I like using smart:limit too - I know it’s a favorite. Thanks for sharing your experience.
Isn't it a bad decision to use a plugin on the master that supports multiband (in this case I'm talking about DMG Limitless)? As far as I know, the respective bands of Limitless can only be processed individually because the plugin internally sets high and low cuts, which in turn lead to resonances and phase changes at the intersection points. Consequently, in terms of a homogeneous sounding stereo track, a limiter with one band should sound best and cause the fewest problems. Please correct me if I am wrong. Thank you very much 😊
Hey hey. Yeah, so let me set the record straight on a few things here. First, any well-coded multi-band plugin is going to use Linkwitz-Riley crossovers, which are completely transparent in the frequency domain. Meaning there are no "resonances" that affect the frequency response. I posted a link below.
Also, the filters comprising the crossovers are linear phase, which means no phase rotation as you'd expect from minimum phase crossovers. In fact, it's entirely possible to design crossovers that are minimum phase that are perceptually transparent as well - Soundtheory accomplished this in their Kraftur plugin. The human auditory system is relatively insensitive to phase shifts. Did you know that the speakers you're listening on - if they're 2-way or 3-way - have crossovers that are rotating phase? Do you hear that as bad? Also, all analog mastering gear uses filters that cause phase shifts. And the audio you're hearing in your room all has phase shifts happening at various frequencies due to the reflected energy off your room boundaries. Frequencies even drift in and out of phase relative to each other as a wavefront travels. It's only in the digital domain that things like linear-phase filters can exist by delaying audio.
It's also not bad to use multi-band processing on your master. This is part of the magic of Limitless. By splitting the bands, it's able to use less gain reduction than single band limiters because it's only inducing gain reduction exactly where needed (kind of like a parametric dynamic EQ). The gain changes on a sample by sample basis can also be optimized for less distortion better in a multi-band limiter. It doesn't limit if it doesn't need to, which means a more open sound, less distortion, and less overall limiting. Limit-LESS, it's in the plugin name ;)
In this case, I would say to stop thinking about all the theory and things people have said about filters and listen closely. How does Limitless sound to you?
Watch this video for a careful and extensive comparison between single band limiters, and Limitless: th-cam.com/video/Yp3Q4g6Vtjw/w-d-xo.html
Learn about Linkwitz-Riley crossovers here: en.wikipedia.org/wiki/Linkwitz%E2%80%93Riley_filter
Read this post from Dave Gamble (from DMG) about why multi-band limiters can solve problems that single band limiters can't: gearspace.com/board/new-product-alert-2-older-threads/1058709-new-dmg-audio-limiter-limitless.html
@@warpacademy Wow! Thanks for this huge answer. I will have a look at the links :) Greetings from Germany
You’re very welcome. I hope that all makes sense. Cheers!
Great tutorial, thank you!
My pleasure. Glad you enjoyed it.
Thanks brotha, great vid.
My pleasure! Glad you enjoyed it. Thanks for stopping by the channel. Subscribe and keep in touch!
Thank you so much warp academy, very helpful and insightful information
Thanks for watching and subscribing!
Thanks for the video. I almost always use Smart limiter - great limiter.But lately I started (sometimes) to use Bus PEAK from Arturia. Bus PEAK from Arturia is a good limiter, but the problem is that it consumes a lot of memory (DSP).
Thanks for sharing! Yeah mastering limiters are some of the most CPU intensive plugins. But even still they should be coded in a way that they don’t hog resources. Pro-L2 is a great example. I was able to load up 40 instances of it at only 15% CPU usage. Pretty amazing.
I’ve never used PEAK. Sounds like it’s aimed more at bus limiting than master limiting perhaps. Cheers!
Fantastic!!!Thanks!!!!
My pleasure. Cheers!
I have the Sonible Smart Limit, so I guess that's what I'll be using.
Nice! It’s a great plugin. Stay tuned on the channel as I have several more videos coming up with it.
Nice one thx again
No problem 👍
Hello sir could you please recommend plugin that works as limiter and compressed at the same time. For example we raised the loudness and it will compress the loudest part to make sure it won't clip.
you just described a limiter
Hey hey. Sure thing. I just want to make sure you fully get what I'm saying in the video. These modern limiters have exactly what you need. They have the fast transient stage that will ensure no peaks escape (this ensures you don't have any samples going over the output / 0 dBFS). And the slower dynamics stage compresses the general dynamics of the song.
For this reason, a lot of mastering engineers I know (myself included) rarely or never use compressors on their master. These modern limiters do both jobs. Let me know if that doesn't make sense and I can elaborate. Cheers!
@mttlsa686 ^^^^ indeed!
Toneboosters Barricade has a compressor stage if that's what you mean. Ozone 11 maximizer also has a parallel compressor stage, and a soft clipper
@@JazzyFizzleDrummers good comments. Thanks for joining in.
do you prefer sonible smart limit over l2 (transparent etc) i have limitless but want one more
Hey hey. If you’re looking for a second limiter and you already have smart:limit then I would get Limitless. It’s just so good. And it’s very different than Pro-L2 because it’s multi-band.
Watch this video for a comparison between Pro-L2 and Limitless: th-cam.com/video/Yp3Q4g6Vtjw/w-d-xo.htmlsi=4XOTh2EKtGgU-Muc
@@warpacademy no no i have limitless and want L2 or sonible as second
@@warpacademy do you feel sonible is louder, more transparent than L2 for mastering
Oh, right, I misread ya. I would prefer smart:limit over Pro-L2 personally. Less distortion, more features - especially with the Saturation and Bass parameters. Cheers.
More transparent yes. Both can be loud. smart:limit can be cleaner and more free of distortion.
Nice 🙏🏻
Thank you 🙌
Thanks! Best/Mathias
My pleasure. Thanks for stopping by the channel!
Oversampling is more complex than what you explain. The higher the oversampling the more "sampling points" so the less chances you have to get inter-sample peaks so the limiter will handle better those peaks (and obviously their difference in level will be smaller) BUT in fact while downsampling your can create back overshoots BUT these are not taken care of in any way UNLESS you turn on "true peak limiting" which is the button to enable this 3rd level of dynamic control.
This being said TPL has a color and has a tendency to "blur" and remove "sparkle" in the transients. In most of the cases ideal scenario would be lookahead of 0,1, highest possible oversampling and setting the output somewhere around -0,3 to -0,5 of the desired output ceiling to be safe that ultimately no ISP will peak above without needing this 3rd level of dynamic control induced by the TPL
Oversampling is much more complex than almost anyone explains in a single video. That said, both myself and some other mastering engineers have covered the topic very well with several videos. I'll unpack your comments one by one:
"The higher the oversampling the more sampling points" - yes, correct
"So the less chances you get intersample peaks" - Kind of - I get you here. By using OS you cause the limiters detection circuit to read higher peaks (ISPs) then if you were not using OS, and then the end result of the limiter output should have a lower TP value. The thing is, you now get more limiting due to it reading higher peak level. Watch this video I just did on using oversampling in limiters: th-cam.com/video/Xf9gVOKa_U4/w-d-xo.html
"While downsampling you can create back overshoots BUT these are not taken care of in any way UNLESS you turn on TPL" - Incorrect. A limiter must create a hard ceiling at project sample rate without overshoots. This is completely independent of the TPL stage. If you set the limiter to oversample, it will downsample, it will create overshoots, and it must 100% have another limiter to catch overshoots. The TPL stage, by virtue of oversampling and downsampling again, must also have a final limiter that runs at the project sample rate to catch downsampling overshoots.
"TPL has a color...." - I mean, yeah, but only because limiting has a color. All TPL does is engage more limiting. There's nothing special about it. It's just another limiter. If you perform more limiting in the main limiter it will also have all the same downsides you describe.
"Highest possible oversampling setting" - As it turns out, no. There is no benefit to more than 4X oversampling in any of the limiters I use if you're running a base rate of 48 kHz in your session. The ITU for TPL specifies 4X, not more, for this reason. Sonible's team also did a ton of research into this and confirmed that anything more than 4X was overkill. And finally, Nicholas Di Lorenzo @Panorama_mastering did a video about “The Best Oversampling Amount” and also confirmed 4X is where you should max out and that was for a clipper, not a limiter: th-cam.com/video/6Ss3cHXGjl4/w-d-xo.html&
You know your stuff, that's clear. Great questions and discussion. Follow up on those videos I posted and let me know if you have any other questions.
@warpacademy overshoot can happen because au phase rotation of the anti aliasing filters fore sure. Yet I can swear there might not be added value of oversampling above 4x according to Nicholas but I did some Delta tests and you hear a clear difference between x4 and x32 on pro L2. Transients stay somehow sharper and snappier, les grungy. I really encourage you to try it out yourself. ITU recoms are based I guess about a minimum viable oversampling amount to be tolerable if ISP occur that they are so little and so short they won't cause any audible artifacts on most convertors. Nice discussion in fact and thanks for the in depth content you share , it's not the usual YT entry level BS ;) keep them coming ! Cheers
@@incidence.studio Hey hey. Yeah, it's a good conversation.
Regarding overshoots, the main thing I'd like to point out is what causes them. Overshoots happen from band limiting a signal. So they happen with any type of anti-aliasing LPF - it can be linear phase or minimum phase. So even without any phase rotation, you still get overshoots from the downsampling process because by removing higher frequency sinusoid partials you change the waveform and its peak amplitude.
For this reason, anytime you have a last-in-chain mastering-grade plugin that oversamples, you will see another final limiter or clipper after the downsampling LPF. Some manuals for plugins specifically discuss this, such as Sountheory Kraftur.
It's interesting to hear about Pro-L2 and transients. I think that may have more to do with their windowing on their anti-aliasing filter. It definitely affects the top end.
Also interesting to note is that oversampling is not even implemented in DMG Limitless (except for its clipper and ISP, not for the main limiter) or Ozone 11 Maximizer. They are both sample peak limiters, and to be honest, DMG Limitless sounds cleaner than either Pro-L2 or Maximizer. Sonible's smart:limit maxes out at 4X.
So Pro-L2 is the odd one out in that bunch. I don't use Pro-L2 much anymore as I almost never like how it sounds compared to Limitless.
Anyways, nice to chat. All the best.
one question, how is the multi band limiter affects the tonal ballance of the track? I know many engineers avoid them because of this, working hard to make fine eq moves to get the perfect ballance, then changing that with the limiter.
Hey Alex. Great question! Limitless is fully customizable in this respect with the Separation control. If you set Separation at 0 it's a single band limiter. Separation of 100% is still great, and will reduce distortion by ensuring bands that don't need to be limited as much are not as affected. More than 100% runs the risk of making the mix sound re-EQed, but can be useful for problem bands. As an ME, we would always keep a dynamic EQ in our tool kit for this. Limitless can do that right in the limiter.
So, it's a "use with caution" kind of thing, but you have full control over it. If you don't want Limitless to be multi-band, you can turn down Separation. Cheers!
@@warpacademy thank you very much for the time to answer. Yes, I fully understand the reason why a multi-band limiter would be useful, exactly as you mentioned.
But still, provided it is needed, it will definatelly affect the tonal ballance, as the bands which are limited less, will be emphasized. So how do you approach this? do you accept those tonal changes as a tradeoff for limiting separate bands based on needs? or would you then revisit the eq and make some changes? (with this potentially require a second visit to the limiter and so on)
Hey hey. Sure, happy to address this. You're thinking that band-specific compression is a bad thing. It's not. It's just a tool. I mean, think about how many engineers for years have been using multi-band compression on the master? It's nearly the same thing.
In a single band limiter, if a huge kick comes in and triggers the limiter, it pulls everything down. The whole song can sound quiet and sucked out from the big kick. In a MB process, that doesn't happen. It's not an advantage or a disadvantage, it's just a thing you can use.
If you set Limitless to 200% Separation, you would notice it, perhaps. At 100% or less, like I set it, you really don't notice it affecting the tonal balance in a negative or overt way. It simply sucks down any specific frequencies that may pop out from time to time, rather than making the bands that aren't limited as much sound "emphasized".
The other large advantage of an MB setup is that you can scale the release times to be longer in the LF and shorter in the HF, rather than one single release timescale for everything. This was also a good MB compression trick used by engineers. So the big sub hit, the 808, the kick, the bass guitar get a nice gluey release, and hi hats get a more rapid release to recover for the next transient of fast hat passages. Limitless scales the per-band release time in the same way.
And finally, fast MB gain limiting processes will produce less intermodulation than a single band process. When using minimal lookahead and fast settings, limiters will definitely produce intermodulation on complex signals, which can and does produce non-musical sum and difference partials that can extend into the extreme LF below the fundamentals of the kick and bass. For proof of that, look at the documentation on Soundtheory Kraftur (it's a MB clipper, but that's very similar to a fast MB limiter set to waveshaping timings). They clearly state that the MB processing in Kraftur produces less intermodulation on a master, and they're right.
Do you use Limitless? You should try it. Also, if you have iZotope Ozone, the Maximizer is multiband if you use the IRC IV style, you just don't get any control over it.
@@warpacademy Wow, awesome detailed explanation, thank you. I don't have Limitless, I do have Ozone Maximizer though. I understand what you are saying, I will try to experiment with multiband to see the advantages and disadvantages on the sound :)
Right on Alex. Keep me posted on how that goes! Also, we have a Discord server for Warp Academy. You seem like a good fit for the community there. Come join up! discord.gg/ZwNgZteGKw
Why not set the lookahead as long as we can? What is the downside?
Great question. A longer lookahead will cause the limiter transient stage to kick in full gain reduction very early. The upside is no distortion, the downsides are: quieter master and transients that sound dull, suppressed and lacking in impact. Plus more latency, which may not matter to you if you're in a mastering setting. Longer lookahead is good for some material, but I almost never use it for the music I engineer.
Why no limitless weighting A/B? , you did it with other parameters.
Good suggestion. I'll keep that in mind as I have more limiter videos coming up in this series. In this video, I was focusing more on the stages and general intro to workflow rather than a deep dive on the effect of settings in each limiter. If I did that, this video would be 2 hours and not 25 min. My videos are already on the long side so I keep the scope for each one as tight as I can.
@@warpacademy but you did other A/B’s and really talked about how you lean into hop hop mixing, and how important bass is to hip hop and mentioned how the weighing really affects the bass, but no demonstration.
I get you. I'll explain. That's because I already shot a separate full limiter comparison video using a hip hop song (the same one) with Pro-L2, iZo 11 Ozone Maximizer, smart:limit, and Limitless. I show head to head comparisons and go more deep dive into settings.
I don't try to cover every aspect of workflow in a single video. I break them up, to make them a bit more bite sized. This video was already 27 minutes right? Most people don't have the attention span for that, so I make series.
Stay tuned, I think you'll like the hip hop limiter comparison coming up.
Or, if you don't want to wait, you can download the trial of Limitless, and A/B/X to your heart's content with a hip hop track on your end. All the best.
@@warpacademy awesome, I’ll checking out your vids today, great content !
Thanks!! The limiter videos I mentioned will be released soon. I’m just editing them now. Stay tuned.
The windowing function for true peak in Pro-L 2 is too aggressively interpolated. no matter the amount of gain reduction i do with it, even just the odd overshoot here and there, i can hear it push the transients further back than it should. if you compare the True Peak limiting in Pro-L 2 to say Tokyo Dawn Labs LimiterNo6 GE it's a night and day difference. there's definitely some softening on the latter if you try and do about 1dB of almost constant limiting on transients (nobody who knows what they are doing though would do that i hope), but using a TP limiter how it is supposed to be used, LimiterNo6 wins out over Pro-L2 for me
Hey John. Thanks for the comments. I wonder if this has to do with the fact that Pro-L2 uses 8X OS in their TPL stage and not 4X like most do. I talked with the Sonible team and several other MEs and they all agree that anything over 4X provides no benefit, but it will likely read and react to higher ISPs and push the transients down more as a result. Thoughts on that?
Are you an engineer or DSP guy? Cheers!
@@warpacademy That might explain it, yeah. i was wondering if the oversampling for TP in Pro-L 2 is perhaps minimum phase as it really does sound like it almost slewing the transients. the phase shift might be what i am hearing. OS in most other Fabfilter plugins is minimum phase(except for Saturn when you engage the Linear Phase crossovers), so it's not out of the realm of possibility
I am or was rather a an inhouse mix engineer for some subsidiary dance music labels in the UK and Europe for Sony and Virgin Records(waaaay before UMG gobbled them up) before that, but I've pretty much retired now about 4-5 years ago after doing the freelance thing for a bit. still interested in all of it as a hobby though. doubt that will ever change
Ah neat!! Thanks for the background. Yeah it sounded like you really knew your stuff.
A lot of plugins give you control over the downsampling LPF and you can use minimum or linear phase (like Schwabe Digital stuff) but some don’t.
Linear phase can smear transients. I’m working with the Soundtheory team to do a video on their new Kraftur plugin and they only use minimum phase crossovers for that reason.
In general I like the option to choose as a user. I’ll have to ask the FabFilter team more about their LPF configuration.
Anyways, nice to see you around the channel and thanks for the contribution to the discussion. Stay in touch!
I will have to watch on replay....But I will take notes for this....I have software that have limiters. I won't touch it because I do not know what I am doing....hahahahaha Thank you for this video....
You got it! That's what this video is all about. Good luck.
1:26 REALLY??!? Did NOT know that.
Yep, it's an eye opener hey? Back in the day, this is why engineers would sometimes use 2 limiters in series. One would have fast settings for shearing peaks (small time scale for managing micro dynamics) and the second limiter would have smoother settings (larger time scale, macro-dynamics). If you have a single stage limiter, then small, short transients that are loud will trigger the full release envelope and cause a big "hole" in the audio.
Practically clipping with the L2 should be the fastest/shortest release, and the slowest/longest attack ;)
Longest attack and shortest release, for sure. You're 100% right. I reversed that in the video. Sometimes happens ;). I did discuss it later on and described it correctly that time, explaining how longer attack is louder and lets more transient through and the shorter release is louder and prevents transients from being suppressed as much. Good catch!
I heard this isn't "practically" clipping, it is actually 100% hard clipping as long as there is zero lookahead. Maybe that info is incorrect?
Good question. I’ve tested this myself and those settings in Pro-L2 do sound substantially different than a hard clipper to me. I’m not sure why but perhaps that’s why they named the preset “practically” and not just clipping.
I expect, if I had to venture a guess, that there may be a limiter knee happening under the hood that makes it into a soft clipper transfer curve.
@@warpacademy thanks for sharing your observations. That makes a lot of sense, I wonder if the selected algorithm affects that knee.
Hey hey. Yeah good thinking on that one. I bet it does. I mean I can’t imagine that it wouldn’t actually. But seeing as it’s all happening under the hood we may never know for sure. Cheers!
Idk if a drum beat is a good way to demonstrate mastering processes
It’s not just a drum beat really. There’s more to it. And also, a huge amount of music consists of drums, synths and an 808. If you want to see different music in a mastering session then I have a whole playlist that has different genres including vocals and guitars. Just check the playlists section of the channel for Mastering Sessions.
loved the video but I agree, the music was not good (very little bass content, no vocal taking up a massive part of the frequency range, etc) for showing what a limiter struggles to fix, but that wasn't really the point of the video.
Next time try a dense pop mix with heavy bass, a punchy kick and an in-your-face up front vocal like every mix I work on. Then you have no choice but to hit it at about 3-4db of limiting if the request is -6 or -7 LUFS. That is where the limiters really start to fall apart, but Limitless still stands tall for me. I have used it with great success doing far too much limiting than I'd normally be comfortable with (this is after plenty of clipping, bus limiting, etc...)
Another interesting video would be whether or not clipping is pointless before any sort of multiband (non linear phase) processing such as a Pro-MB or EQ on the master.... It literally moves the clipped point to a different part of the waveform leaving it unclipped. I still need to do more testing, but it would essentially mean that everyone who has recommended clipping for peak control at various stages in the project was wrong.
Hey hey. Noted. Thanks for the comment. I’ll keep that in mind.
I agree that around 3-4 dB is where limiters fall apart. And at the same time I’m able to turn out mixes that are -9 LUFS or so, so I have been able to reach loudness in the range of -6 in the heavy sections without more than 2 dB on the limiter. The concept is straightforward: you do more of the “mastering” duties on your Sub-Mixes where you have less complex signals and thus less intermodulation. And you do less on the master.
And using multiband clipping and multiband limiting on your master, again to reduce intermodulation and get the most minimal amount of GR needed per band.
Regarding phase shifts and clipping yes that’s an important point. Ian Stewart and Nicholas Di Lorenzo had a debate about this on his channel. Nicholas tested it and found that as long as you’re not creating extreme phase shifts (IE minimum phase cut filter) after the master clipper than the amplitude change on the sampling points is negligible.
IMO using clippers to control peak level at a track / individual sound level is 100% a needed best practice regardless of what is going on downstream.
Also interesting to note is how Soundtheory Kraftur is set up. It’s a mastering grade clipper. Yet its crossovers create full phase rotation at the x points. And they have wideband clipper after that. It’s also the loudest and cleanest clipper on the market today.
All interesting things to ponder.
Come join up on our Discord server and chat more! discord.gg/ZwNgZteGKw
@@warpacademy Anytime I ask you a question on your channel I am reminded of why I consider your channel one of the absolute best on YT. You are a true engineer and an incredible teacher. What a thoughtful and helpful answer. I have asked all over the place for thoughts on this topic and have received some tepid responses. This was so thorough and pointed me in the right direction with references.
Do you have a video regarding multiband clipping and when you prefer to use that over a wideband clipper? I have never tried a multiband clipper out of fear of phase rotation undoing my hard work in the mix. Also because I have a hard time wrapping my head around when it would be preferred other than preserving low end while killing some high transients. Most modern mixes are heavier in the lows so it's not really the high freq transients causing peak issues. Maybe I have the use case backward.
Regarding Kraftur, do those crossovers not cause problems?? I would think they would negate a lot of the clipping you've used throughout the mix if you're throwing that on the master. Perhaps the phase is not affected enough to cause an issue?
Loudest and cleanest is a bold claim, now I've GOT to try it! I've used so many clippers and always find myself returning to standardclip.
I joined the discord :)
@@musiclover-r9m Thanks so much for the kind words and the reflection of how I've been showing up. I love detailed conversations like this, and taking the time to explore the many nuances of audio.
In terms of multi-band clipping, I've noted your questions and will aim to answer as many of them as I can in my upcoming video on Kraftur. I'm actually about to hop on a call with their product specialist so it's good timing.
But to quickly answer what you've asked, I would want to use an MB clipper on my master vs a wideband clipper if it's doing much work. It all comes down to intermodulation. MB plugins have much less intermodulation than wideband ones with distortion.
There are use cases where I would want to leave the low end untouched, but clip the mids and highs. Luca Pretolesi often uses K-Clip with a subtractive low shelf before it and an equal / opposite additive low shelf after it to reduce the amount of low end triggering the clipper and to lower the clip point. Solid technique, but now Kraftur can do that without all the EQ.
Regarding Kraftur crossovers. They cause full 180 degree phase rotation at the crossover points. I was initially alarmed by this and had a fairly extensive conversation with the Director of Research at Soundtheory. He said they've built the crossovers as minimum phase intentionally to avoid transient smearing from non-causal linear phase filters, and they've developed an in-house perceptual model of the human auditory system to use in design. He told me to stop looking at Plugin Doctor and to go listen to it. So I did.
I can throw it on my master, with no clipping, and move the crossover points around and I don't hear anything changing. The crossovers sound completely transparent. This has to do partially with the fact that, and I quote, "Our auditory system is very forgiving when it comes to gentle phase changes, mostly because the signal chain from the source to your inner ear adds a lot of phase changes that the auditory system needs to deal with. Unless you have a phase response whose second derivative with respect to frequency, which is called the dispersion of the response, exceeds a certain threshold, the phase changes will not introduce audible artefacts."
More on this soon as Kraftur is one of the next videos in my production schedule. See you on the Discord server!
Hi Warp academy - Good video, but you assume all of us use external paid for 3rd party plugins. Not all of us can afford these paid for bells and whistles.
Hey hey. I totally hear you. When you're starting out, often times you work with the stock plugins in your DAW or free plugins. You don't necessarily need a fancy limiter to get a good master. There are also some decent ones in some DAWs, but I would actually recommend that you look TDR Limiter No 6. It's only 50 euro and it's very good: www.tokyodawn.net/tdr-limiter6-ge/.
If you are serious about music, you'll invest in your tools incrementally as you can. You don't need the most expensive plugins, but you would benefit from the mindset of growing your tools. Or find good freeware. There's never been so much good freeware on the market as now.
the point is to know what is happening, then you would be able to use any limiter in an efficient way
^^^^ also this. Yes.
Ableton Live just updated their limiter and it looks wayyyyyyy better than the previous version - potentially could rival some of these in mastering.
They did update their limiter. Added some new much needed features. Worth stacking it up against these for sure but I don’t have high hopes. Maybe I’ll be surprised?
Music keeps playing. It’s distracting. I skip videos if music keeps playing
Skip ahead to your heart's content then
I would love to see a version of this video with a more cleaner, natural music style with vocals.
Done! I have quite a few videos of singer / songwriter stuff with vocals, guitar, bass, drums in my Mixing and Mastering Playlist: th-cam.com/play/PLsyXZjufQdPzK5ndIgFQiCWIDICCKPaj4.html.
This one specifically could be useful for you: th-cam.com/video/xnTpwu0YbU0/w-d-xo.html