Resample

Resample hands back the same window at a new sample rate. Set Sample Rate in Hz, or leave it at 0 and connect a Reference Signal to match its rate — which is how you get two Signals onto a common grid before comparing them.

Method picks linear interpolation, sample-and-hold, or box averaging when going down, which is properly anti-aliased rather than just dropping samples.

The duration is preserved, and the new samples sit on a fixed time grid — the Reference's own sample times when one is connected — so the result does not shimmer as the window slides. That detail is the difference between a usable resampled trace and one that crawls by a fraction of a sample every frame.

• Prepare for Compare — Two sources at different rates onto one grid so their difference means something.
• Cheapen an analysis — Downsample a 48 kHz audio window before a slow analyser that does not need the bandwidth.
• Align for export — One rate for every channel before Signal Export writes the CSV.
• Match a plot — Put a fast and a slow signal on the same grid so a Heatmap's rows line up.
• Upsample for smoothness — Linear interpolation to a higher rate before a visual consumer.

• Use box averaging when going down. Linear interpolation while downsampling aliases.
• Sample-and-hold is right for stepped data (an integer selector, a gate) where interpolation would invent values.
• Resampling cannot add information — upsampling a 10 Hz source to 1 kHz gives a smooth curve, not detail.
• Every Signal carries the wall time of its newest sample, and Plot places traces by it, so a 10 Hz and a 100 Hz window already line up on one chart without resampling.