Reply to thread

Message: [QUOTE="Shamiso, post: 92670, member: 160"] Mid-side bitcrushing lets you degrade the shared and difference parts of a stereo signal separately instead of crushing both speaker channels equally. BITLAYER can route each frequency band to the mid, the side, or both, which makes the feature more useful than a normal stereo width switch. The [B][URL='https://goldmidi.com/community/threads/riverflux-audio-debuted-a-free-bitlayer-multiband-bitcrusher.78061/']BITLAYER per-band mid-side routing[/URL][/B] can leave a solid center relatively clean while the wider material gets bit reduction, downsampling, jitter, or some combination of the three. The catch is that mid and side are not tidy instrument stems. Treating Mid as “the vocal and kick” and Side as “the reverbs and hard-panned stuff” works only as a rough shortcut, and it gets misleading fast. [HEADING=2]Mid and side are signals, not stereo stems[/HEADING] A mid-side encoder builds Mid from the sum of left and right, while Side comes from their difference. Decode them and the two signals reconstruct the original left and right channels. [B][URL='https://patents.google.com/patent/GB394325A/en']Blumlein’s stereophonic patent[/URL][/B] sits behind the same sum-and-difference idea that modern plugins still use. A sound panned dead center contributes strongly to Mid because left and right carry the same thing. A hard-panned guitar is different. It contributes to both Mid and Side, so crushing only Side does not mean you have somehow isolated the guitar from everything else. This matters when you use mid-side distortion on a busy mix. Side-only processing changes whatever differs between left and right, which can include room sound, doubled parts, stereo synth movement, chorus, hard panning, and timing differences. The Mid still carries some of many of those sounds. Goldmidi’s [URL='https://goldmidi.com/community/threads/mid-side-compression-splits-a-mix-down-the-middle.75885/']mid-side compression breakdown[/URL] goes deeper into the matrix and why the Side disappears during a mono fold. The same routing math applies before you swap compression for a bitcrusher. [HEADING=2]Side-only crushing can disappear in mono[/HEADING] Side-only bitcrushing has a weirdly useful property. Process the Side signal, decode it back to left and right, then sum the result to mono, and the processed Side cancels out if the matrix is behaving cleanly. The stereo version can therefore sound much rougher without forcing the same damage into the mono signal. Downsample the upper sideband and cymbals, room tails, wide pads, or doubled guitars can pick up extra grit around the edges while the mono fold keeps only the Mid contribution. This does not mean the stereo result is automatically safe. Push the sides hard and the balance between center and width changes. A lead vocal can feel smaller simply because the degraded edges became louder, brighter, or more attention-grabbing around it. Mid-only crushing behaves differently. Any degradation applied to Mid survives the mono fold because Mid is the part left and right share. Crushing the center can be useful when the vocal, snare, kick, bass, or other central energy needs texture, but the effect is no longer disposable stereo decoration. The distinction also explains why a correlation meter cannot decide for you. Side processing can remain mathematically cancelable in mono while still making the stereo mix feel lopsided, abrasive, or hollow. Listen to both presentations instead of chasing one meter value. [HEADING=2]Per-band routing keeps the damage where it belongs[/HEADING] BITLAYER’s eight bands make the routing more surgical. You can keep low frequencies on Mid or full stereo while sending only an upper band through Side processing, which is a cleaner move than crushing the entire difference signal from sub bass to air. Keeping the bottom stable is especially useful when a stereo bass patch carries width mostly in its harmonics. Goldmidi’s [URL='https://goldmidi.com/community/threads/keep-the-low-end-mono-without-shrinking-your-mix.77373/']low-end mono guide[/URL] explains why the useful cutoff depends on the source rather than a magic frequency. Start with the sideband cleaner than you think it needs to be. Lower bit depth until the width gets grain, then try sample-rate reduction if you want harder aliasing or a more obvious digital edge. Jitter can make the sides less repetitive, but it can also turn bright stereo material into fizz surprisingly quickly. Keep the Mid path unchanged while finding the Side texture first. Once the width has the right amount of damage, switch to mono and check what actually remains. If the groove, lead, and low end still make sense without the extra stereo dirt, the routing is doing something useful instead of just making the headphones version louder. A final level match matters here more than it seems. Added aliasing and quantization noise can make the sides feel larger even when their measured level barely moves. Pull the processed version back until bypass feels equally loud, then decide whether the extra width is real separation or just brighter damage. [/QUOTE]

Name