Hi, below comment is for clarification;
1. Does the funny noise occur during start-up?
2. Does your attached spectrum is taken during your so called funny noise?
3. Does the funny noise didn't disappear during operation?
4. Whats is the recorded enveloping acceleration for motor drive side and non-drive side? also the timewaveform? any significance changes in trend history?
For me, when i encountered abnormal noise like a sound pressure noise at motor, i do phase analysis for motor base bolt, for me to check if there is a soft foot or looseness which affect the airgap of motor resulting the abnormal noise. also, abnormal noise will generate if there is a clearance a outerace of bearing and housing, but the noise will not last until the bearing meet the thermal expansion, also when the bearing is cocked, you can confirm it through phase reading. if bearing noise through metal to metal contact, you can check it in the TWF and you can perform greasing, anyhow, with your limited data and problem state, all this i can say.
Don't depend on velocity spectrum for bearing fault, use envelop acceleration, TWF and correlate them both.