Reply to thread

Message: [QUOTE="Shamiso, post: 92673, member: 160"] BITLAYER’s Transient mode targets attacks, while Tonal mode targets sustained material instead of sending the same bitcrushing across the whole signal. A snare crack and the ringing body behind it are not interchangeable pieces of audio. The attack changes quickly and often spreads energy across more of the spectrum, while the later portion can settle into steadier resonances. In the [B][URL='https://goldmidi.com/community/threads/riverflux-audio-debuted-a-free-bitlayer-multiband-bitcrusher.78061/']BITLAYER transient and tonal processing[/URL][/B], Normal handles the selected content as a whole, while the other two modes separate the sharp and sustained parts. The public descriptions available for BITLAYER do not publish the detector math, so the useful move is judging what each mode actually leaves exposed rather than assuming a fixed millisecond boundary. [HEADING=2]Transient and tonal regions carry different information[/HEADING] A transient is not simply the loudest sample in a sound. It is a short region where the signal changes rapidly, and those fast changes can contain broad spectral energy that is very different from the steadier part that follows. The [B][URL='https://doi.org/10.1186/s13636-014-0026-5']percussive and harmonic separation model[/URL][/B] treats percussive material as sparse in time and relatively broad in frequency, while harmonic material is smoother over time and more concentrated spectrally. A bitcrusher does not become a transient shaper just because it can isolate those regions, but the split changes which part receives the damage. Lower bit depth on an attack can roughen the snap, click, or stick noise without forcing the same quantization texture through the whole tail. Sample-rate reduction can make the leading edge feel more stepped or brittle, especially when the attack contains plenty of upper-frequency energy. The result can be useful on drums because the attack often carries the cue that tells your ear where the hit starts. Crush only that region and the body can stay comparatively clean, so a snare can keep its ring while the crack gets nastier, or a kick can keep its low body while the upper click gets chewed up. Transient mode does not guarantee more punch. Too much degradation can blur the very edge that made the hit easy to locate, especially once downsampling or jitter becomes aggressive. Dirtier is not automatically harder. [HEADING=2]Tonal crushing can leave the front edge readable[/HEADING] Tonal processing flips the emphasis. Instead of concentrating degradation around the attack, it aims at the sustained material, which can leave the initial hit comparatively intact while the body gets rougher. This is useful on basses, plucks, and drums where the onset already works, but the sustain feels too polite. A clean front edge can keep timing obvious, while the later portion picks up quantization grain, aliasing, or a more broken digital texture. Fixed sample-rate reduction behaves especially differently on sustained pitched material. A held note gives repeatable alias components time to remain audible, so the result can sound metallic or pitched rather than appearing as one short burst around the attack. Bit-depth reduction has another wrinkle on decays. As a tail falls toward a coarse set of available amplitude steps, low-level detail has fewer useful values to occupy, so the end of the sound can become grainy or stair-stepped before the louder portion feels equally wrecked. The same hit-to-tail trade-off shows up in [URL='https://goldmidi.com/community/threads/maximillion-can-crush-drums-without-erasing-the-hit.77711/']drum processing that preserves the attack[/URL]. The processing method there is different, but the listening problem is the same because a dense tail and a clear onset affect groove in different ways. [HEADING=2]Pick the target before choosing the damage[/HEADING] Start with one frequency band and one source rather than a whole bus. Use Normal first to find a bit-depth or sample-rate setting whose character you like, then switch between Transient and Tonal without touching the degradation amount. On a snare, listen separately to the first crack and the ring behind it. On a plucked bass, listen to the finger or pick noise first, then the pitch and decay. The better mode is the one that damages the part you actually wanted to change without making the other part pay for it. Jitter needs extra restraint on transient material because timing errors become easier to notice when the waveform changes quickly. The deeper reason is covered in [URL='https://goldmidi.com/community/threads/bitcrusher-jitter-is-not-sample-rate-reduction.78081/']why bitcrusher jitter hits fast-changing audio harder[/URL], and the practical symptom is usually a sharper move from texture into fizz. Keep the dry or per-processor mix available while you set the mode. If Transient makes the attack brittle enough to lose definition, back off the crusher or blend clean attack back in. If Tonal turns the sustain into a flat digital haze, raise the bit depth or sample rate until the decay still carries the note instead of only the artifact. [/QUOTE]

Name