Reply to thread

Message: [QUOTE="Bombastus, post: 91962, member: 2178"] Ryan Schwabe places HiFAL after clipping on most mix buses, either early in the chain or directly before the final limiter. Placement matters because each processor changes the signal handed to the next one, and high-frequency dynamics can shift after loudness processing has already done its work. A clipper can remove short peaks before a limiter has to react to them, but the useful point here is not the usual clipper-versus-limiter debate. High-frequency limiting has a different job. It can sit after peak reduction and listen to the brighter, denser signal those earlier stages actually created. Once the chain gets loud enough to alter its own tonal balance, [B][URL='https://goldmidi.com/community/threads/ryan-schwabes-hifal-plug-in-kills-harsh-peaks-without-murder.48966/']the high-frequency acceleration limiting approach[/URL][/B] can make more sense after the loudness stage than before it. You are controlling the version of the mix listeners will actually hear, not a cleaner version that existed several processors earlier. [HEADING=2]Peak control before the limiter changes what follows[/HEADING] Putting a clipper before high-frequency control can be useful when drums or percussion carry isolated peaks that eat headroom. The clipper handles those brief waveform tops first, so the high-frequency processor is not forced to react to a signal whose overall peak structure is still being dominated by a few huge attacks. Keep the distinction clear. Clipping changes waveform shape and can create distortion, while a limiter changes gain over time. If the clipper is pushed hard enough to make cymbals brittle or consonants grainy, placing a high-frequency limiter afterward may hide part of the symptom without fixing the damage that caused it. A cleaner sequence starts with modest peak control, then checks whether the upper band became harder, denser, or more forward. HiFAL can work on those upper-frequency events before the final broadband limiter decides how much gain reduction the whole mix needs. Each processor gets a narrower problem. Processing order also changes gain staging. A high-frequency limiter placed too early may behave beautifully, only for later clipping, saturation, EQ, or broadband limiting to generate fresh upper-frequency peaks. Moving it later lets the processor respond to more of the tonal shift produced by the chain itself. [HEADING=2]Full-band limiting can make the top end feel brighter[/HEADING] A broadband limiter reacts to whatever pushes its detector hardest, and kick or bass energy can dominate that decision. When a heavy low-frequency hit drives gain reduction across the entire mix, mids and highs drop with it. As the limiter releases after the low end falls away, the upper spectrum can move forward again. No extra treble has to be generated for the mix to feel brighter. Repeated broadband gain reduction changes the relationship between low-frequency hits and everything surrounding them, especially on loud material with dense cymbals, vocal consonants, or bright synth layers. The tonal shift is dynamic rather than a simple EQ boost. Placement gets interesting only after loudness processing starts changing the tone. High-frequency control before the main loudness limiter cannot respond to brightness movement created by that limiter later. Put the processor after the loudness stage, or close enough to the final limiter that little tonal processing follows, and it can catch the upper-frequency behavior produced by the chain. A [B][URL='https://patents.google.com/patent/US8488811']multi-stage audio peak limiting design[/URL][/B] uses separate slow and fast gain-reduction stages, a useful reminder that peak control does not have to happen in one indivisible block. Mastering chains can divide jobs too. One stage can create loudness while another catches smaller problems created after it. [HEADING=2]A safety limiter protects processing placed after loudness[/HEADING] Schwabe also uses a two-limiter arrangement in some mastering work. The first limiter performs the loudness gain reduction with a slightly lower ceiling, while a second limiter sits later with a higher ceiling and no added gain. Processing can live in the small headroom window between them. High-frequency limiting fits neatly into that window because the first limiter has already exposed the tonal behavior of the loud master. HiFAL can control any resulting brightness, and the later limiter can catch new peaks created by processing between the two stages. The second limiter is protection, not another loudness push. Anything placed after a limiter can change peak level again. EQ boosts, saturation, clipping, high-frequency trim, stereo processing, or another dynamics stage can create samples that exceed the earlier ceiling. Calling the first limiter “final” stops making sense once signal-changing processing follows it. Use the later limiter lightly enough that it is mostly idle. Its job is to catch unexpected overshoot, not flatten the transients rebuilt or preserved between stages. If it shows constant reduction, the earlier stages have probably left too little headroom or the processing window is being driven too aggressively. [/QUOTE]

Name