Along with what other posters have stated, it's important to realize that while we're trying to use barometric pressure sensors to just detect altitude, they also are subjected to other causes of pressure disturbance.
1. Ram pressure (will cause altitude graph to dip DOWN at high speed). Common when vents are on the curved nosecone surface.
2. Bernoulli effect (graph will jump UP at high speed). Common when vents are right behind a curved surface, like on the base of a nosecone.
3. Shock effect (pressure sensors are mechanical devices, and can be affected by shock; some are more rigid than others).
4. Base drag suction (the low pressure at the tail of the rocket causes low pressure during high speed which isn't relieved by venting, and is a form of Bernoulli effect)
5. Pressures during ejection (for those who use their altimeter in the fuselage, which is fine if vented).
6. Mach effects (as standing pressure waves migrate down the fuselage and burble over vents during strong boosts)
Some of these we try and filter when we can. When you see an obvious anomaly, try to see if you can correlate it with high speed (the boost phase) or ejection. We see these effects a lot on water rockets due to their much-higher G accelerations and frequently mounting on the outside of fuselages.