Say I’m looking at a .WAV that someone else has recorded… it’s supposed to be an archival backup of a cassette with speech. Zoomed out, the waveform looks like a giant block of red due to clipping. But zoomed in, I see that it’s really just one peak getting chopped every second or two. Mostly not something I can hear. Would you consider this no big deal, or a problem that merits re-recording? (I’m inclined to re-record, but I’d like a third party opinion because I have some financial stake in the answer.)
Is it actually being “chopped” or is it just “touching” 0dB (plus or minus 1 on the vertical scale).
If a recording has been normalized to 0dB then you will get a red “clipping” indicator at the highest peak.
If a recording has been “hard limited” to 0dB, then you will get many red “clipping” indicators.
In either of these two cases, since the audio has not actually been “chopped” there is probably no need to re-record.
On the other hand, if the audio has been “chopped” (clipped), then it is likely to be audible if listened to closely on high quality monitoring equipment, and I would re-record if possible.
That is bound to be a contributing factor in your decision, but being “reasonable” should be another influence.
The kind of overload counts, too. People pay handsome sums of money and put up with technical inconveniences to get the overload characteristics of a vacuum tube.
Digital clipping is pretty unforgiving. Instant third order harmonics as far as the eye can see. That’s technical talk for scratching on a blackboard or a baby screaming. Tape saturation can sound OK as long as you don’t overdo it. Tape tends to overload “soft” because of the interaction with the bias system.
You are strongly urged to listen to the work on something other than a computer sound system. I tested one mini speaker system at work that has no response at all in the middle of the voice range. Zero. One small amplified speaker system has 27% harmonic distortion right out of the blister pack.