256 k 4 times choose. noise floor in distortion measure is much higher as 5. and 7. harmonic but when do sine and RTA noise floor is much lower

[user44455555]

I understand that in RTA with constant frequency noise floor is lower as a single sweep.. but i use 4 sweeps(set in measure dialog). it reduce noise floor not much. see screenshots for compare
here can see clear 5 and 7 harmonic

here noise floor is much too heigh. how can get this less noise floor here so can trust the values ?

this are settings i use for the RTA

 

[John Mulcahy]

When you are using the RTA all the stimulus energy is in the tone you are measuring. When a sweep measurement is made the stimulus energy is spread over the whole sweep, it cannot approach the signal-to-noise of an RTA capture. Each doubling the sweep length, or each doubling of the number of sweeps, gives about a 3 dB improvement in signal-to-noise, which can help bring harmonics above the noise limits of the measurement chain. If you need to resolve low distortion levels you need to either use very long sweeps or make a stepped sine measurement with the RTA.

The "noise floor" trace of a sweep measurement is the spectrum of the signal captured from the microphone before the measurement is made. That shows what else is present before you start measuring, it isn't affected by sweep length or number of sweeps.

 

[user44455555]

so when i do 4 measure sweeps i get 6 db less noise floor which mean half. when noise floor is at a freuency 0.1% then i get on this frequency only 0.05% noise floor level ?. high harmonic 7. th and 9. th sound really bad even if it is only 0.1% 0.05% is then good for my ears.