Why re-encoding a video is almost always a mistake
Every re-encode throws away detail permanently. Since the file you download has already been compressed once by the platform, a second pass is a second loss.
- Lossy compression is not reversible — quality only goes down
- A direct copy is always preferable to a conversion
- Convert only for compatibility or a hard size limit
What generational loss means
Lossy codecs discard information the encoder judges you won't notice. Re-encoding the result makes those judgements again on already-degraded material, and the artefacts compound. Do it three or four times and it becomes obvious even to an untrained eye.
The platform already did one pass
Whatever you download has been compressed by YouTube or Instagram at upload. That's generation one, unavoidable. Everything you do afterwards is generation two or worse.
When converting is justified
Two cases: a device that genuinely cannot play the source codec, and a hard file-size ceiling you must meet. Both are compatibility problems, not quality improvements.
The better habit
Pick the right format at download time. Taking M4A instead of MP3, or MP4 instead of WebM, costs nothing at that moment and saves a conversion later.