Redux: separate bit-depth grit from low-rate aliasing

Experiment A: fewer amplitude steps

The first source is an original eight-second synthetic bass pattern with strong accents and quieter notes. Rate starts at 48 kHz and Post is off, isolating the low Bits setting from intentional rate reduction. The source itself was generated at 44.1 kHz; selecting 48 kHz does not add detail that was never present.

Compare five bits with eight, then sixteen. Follow the end of a quiet note rather than only the strongest attack. Quantization rounds amplitude to fewer available levels, so the error can become especially noticeable as the original waveform shrinks. A decay may become grainy, buzzy or abruptly sparse instead of fading smoothly. Lower bits does not mean fewer musical notes or a slower tempo.

The louder accents and quieter notes will not necessarily acquire the same apparent amount of distortion. That is useful diagnostic information, not a reason to keep reducing Bits until every note sounds equally broken. If quiet notes lose their identity, return to eight bits or bring Dry/Wet down to about 40%. The dry component preserves the line while the quantized component supplies the edge.

Experiment B: fewer updates in time

This separate original eight-second pluck uses brighter harmonics. Bits stays at sixteen while Rate begins at 6 kHz, with Post off. Listen for new metallic or inharmonic detail around each attack. Sample holding keeps a value for several engine samples before updating it; this differs from reducing the precision of every amplitude value.

Move Rate upward toward 12 kHz, then downward toward 3 kHz. Judge whether the added tones support the pitch or compete with it. Rate reduction without a pre-filter can fold high-frequency information into lower frequencies. Consequently, a lower Rate need not behave like a gentler low-pass filter: it can produce extra audible material as well as remove a sense of clarity.

The actual hold length is an integer number of audio-engine samples. TuttiDAW rounds the ratio between the engine sample rate and the requested Rate, so nearby knob positions can share the same hold length. The effective rate is stepped and depends on the engine's sample rate; do not expect every small numerical change to yield a new sound. The control never upsamples when set above the engine rate.

What the Post button can and cannot repair

Return near 6 kHz and enable Post. Its low-pass cutoff follows the requested Rate, at roughly 45% of that value, also limited by the engine rate. It softens high-frequency sample-hold images after the reduction. It cannot unfold alias components already created below that cutoff, and there is no pre-filter switch in this panel.

If Post makes the pluck too dull, compare Post off with less wet signal instead of immediately increasing Bits. Bits cannot undo rate artefacts. Conversely, raising Rate will not restore a smooth quiet decay if you have left Bits very low. Keep identifying which dimension caused the problem before choosing the remedy.

Combine only after you can name the difference

Try eight bits and about 12 kHz as a combined starting point on the pluck, then reduce Dry/Wet until the note still leads the texture. Compare at a similar perceived level. A bass that needs stable pitch may tolerate a different compromise from a decorative percussion sound. These settings demonstrate two mechanisms; they do not certify a vintage converter, console or sampler emulation.