Reply to thread

Message: [QUOTE="Bombastus, post: 92006, member: 2178"] The Lexicon 224 gives diffusion an arbitrary 0 to 63 range, while newer reverbs can map the same label to very different behavior. A matching number across two plugins therefore tells you almost nothing about whether their reflection patterns will match. A useful baseline is [B][URL='https://goldmidi.com/community/resources/reverb-diffusion-explained.72/']reverb diffusion and reflection buildup[/URL][/B], because diffusion usually concerns how quickly echoes become crowded enough to lose their individual edges. The trouble starts when designers decide which part of the reverb should diffuse, how the control should scale, and whether several internal variables should move together. [HEADING=2]Diffusion is a job description rather than a standard[/HEADING] Early digital reverbs often used short delay structures to increase echo density without simply making the wet signal louder. The underlying idea appears in [B][URL='https://aes2.org/publications/elibrary-page/?id=849']classic all-pass reverberator design[/URL][/B], where cascaded all-pass sections can multiply audible reflections while keeping the overall amplitude response comparatively flat. Modern algorithms have many more options, so a control called diffusion may sit at the input, inside an early-reflection stage, deeper in a feedback network, or across more than one stage. Lexicon's 224 is a good warning against reading a diffusion knob as a simple percentage scale. Its documented diffusion range runs from 0 to 63, yet the fastest density buildup occurs around 32 to 37, and values above 40 can actually sound less dense. More is not automatically more. Other reverbs behave in a cleaner direction because their designers chose different mappings. Eventide's SP2016 model describes diffusion as thickening or thinning density without changing decay time, while its Plate and Room algorithms connect the same control to reverb buildup and tail density. The shared label survives, but the audible target shifts with the algorithm. [HEADING=2]Early and late diffusion can be separate decisions[/HEADING] Some reverbs split the problem in two. Valhalla VintageVerb's Palace mode gives you early diffusion and late diffusion separately, with the first affecting the level and density of early reflections while the second changes later echo density and the perceived spacing before the tail settles. Moving one knob cannot be assumed to reproduce what a single diffusion control does elsewhere. A similar split appears in reverbs that expose independent early and late engines. One stage can soften the entrance while another determines how the sustained field blooms, which means a smooth attack and a dense tail no longer have to arrive as one package. Copying a diffusion value from a single-stage design into this kind of architecture is mostly guesswork. Size can complicate the comparison further. Changing delay lengths inside an algorithm can alter how quickly reflections recur, so a diffusion setting that feels smooth in a small space may become obviously grainy after the virtual size changes. Some reverbs compensate internally, while others leave the interaction exposed for you to hear and adjust. A missing diffusion knob does not mean diffusion vanished from the algorithm. Designers can fix the relevant coefficients, tie them to room size, or hide several related changes behind a broader character control. You may be hearing substantial internal diffusion without ever seeing its name on the interface. [HEADING=2]Matching diffusion requires listening to behavior, not numbers[/HEADING] Preset translation works better when you ignore the printed value and listen for the point where the reverb changes character. Start with a short, exposed sound, keep decay and wet level stable, then sweep diffusion slowly. Notice when separate reflections stop chattering, when the tail thickens, and whether the front of the effect becomes softer at the same time. Repeat the test after changing the algorithm rather than assuming the same knob position still means the same thing. A plate, room, chamber, or vintage hall can use different internal structures even inside one plugin, so identical settings may produce very different buildup rates. The percentage display is only a user-interface coordinate. Pay attention to non-monotonic controls as well. If a knob becomes smoother through its middle range and then turns thinner or more colored near the top, forcing it to 100 percent because you want maximum diffusion can move you away from the densest result. The Lexicon behavior proves this is not merely theoretical. The practical goal is to match three audible events rather than one number. Listen for how quickly the first reflections merge, how dense the middle of the tail becomes, and whether the texture changes when you alter room size or algorithm type. Once those three behaviors line up, the two reverbs can feel surprisingly close even when their diffusion controls show completely different values. [/QUOTE]

Name