Is this clipping, what causes it?

[Matthew J Poes]

everything over a 6th harmonic looks clipped. Why is that?

 

[John Mulcahy]

Are you referring to the flattening of the harmonics as they approach their frequency limit? That is due to limiting the boost applied by normalisation to 30 dB in the distortion controls, the rise is due to boosting of the noise floor as the level of the fundamental drops.

 

[Matthew J Poes]

Are you referring to the flattening of the harmonics as they approach their frequency limit? That is due to limiting the boost applied by normalisation to 30 dB in the distortion controls, the rise is due to boosting of the noise floor as the level of the fundamental drops.

Thanks John. I didn't ever pay much attention to that.

Ok so I can completely see value in measuring distortion of a speaker in near or free field where direct energy is dominant. It seems like for in room measurements it's not so helpful. Distortion is largely below the noise floor in that measurement right?

I've used it to try and identify rattles or objects resonating. Like if I take a measurement I might scan it for a sudden spike at a certain frequency. Then I might test just those frequencies to identify a problem. I know you can't use normalized for this. Does that seem reasonable? I'm sure I've come up with lots of crazy ideas trying to assess a room with your tools.

 

[John Mulcahy]

An easier way to narrow in on rattles or resonances is perhaps to use the sine generator's 'Frequency tracks cursor' option and slide the cursor about on the graph as you listen for them.

Listening position distortion measurements can show up clipping somewhere in the signal chain, or over-driven subwoofers or faulty drivers and the like, so it is worth scanning the distortion plot when a measurement is made. All would show up more clearly on a near field measurement if there is an issue to investigate.

 

[Matthew J Poes]

An easier way to narrow in on rattles or resonances is perhaps to use the sine generator's 'Frequency tracks cursor' option and slide the cursor about on the graph as you listen for them.

Listening position distortion measurements can show up clipping somewhere in the signal chain, or over-driven subwoofers or faulty drivers and the like, so it is worth scanning the distortion plot when a measurement is made. All would show up more clearly on a near field measurement if there is an issue to investigate.

I see. Ok that makes sense. I do typically scan it but have never spent a lot of time looking at it. I guess I subconsciously did what you suggest in some ways (does everything look normal here?). Iny YouTube video of my room assessment primer I did address this too, just as a quick check.