Sudden "High measurement distortion" warning - cause?

[Spennis]

Hi,

I was running individual speaker sweeps of my 7.2.4 system. I completed the sweeps for all speakers at two different microphone positions. In the third microphone position I suddenly got a "High measurement distortion" warning on one of the speakers. I have never run into this before, and I have measured the exact same system at the exact same microphone position and SPL level very recently.

What might be the cause of this sudden distortion warning? Is it only the signal chain that can cause issues like this, or can external factors (such as stuff rattling in the room) also cause this? If it is only the signal chain than can cause this, can it be both the input (e.g. HDMI to AVR and USB to mic) and the output path (e.g. amp to speaker)?

Regards
Isak

 

[John Mulcahy]

Could be clipping somewhere along the signal path. If you didn't get a clipping warning when making the measurement the replay path might be the culprit, but it could also be analog saturation along the capture path before conversion. You can look at the distortion graph to see where the distortion occurs and what harmonics there are, clipping would be predominantly odd orders.

 

[Spennis]

I'm a novice when it comes to this. Could a loose USB cable connection between computer and mic, or a loose HDMI cable connection between computer and AVR cause this kind of issue? I'm using a UMIK-1, a Dell XPS 7590 laptop and a Denon AVR-X4500H receiver.

 

[John Mulcahy]

Loose cables could cause dropouts which could cause a range of problems.

 

[Spennis]

Thanks. I measured today again and managed to take a measurement of that speaker (surround back left, SBL) without any distortion warning. Here's the distortion plot which I am not sure how to read, along with the distortion warning message. The measurement is taken at the MLP, about 3-4 meters away from the speaker. Any cause for concern, either with regards to speaker performance or measurement quality? The reason I was taking measurement in the first place was to time align and create EQ filters.

Edit: I did a measurement of another speaker and got the same warning again. I do not experience any popping sounds or image issues on the TV while doing the sweeps, but I have ordered new HDMI and USB cables in case there is an issue there.

Edit #2: I tried different HDMI cables, I tried removing the USB extension cable I was using for the UMIK-1, I tried a different HDMI port on the AVR, I checked that the speaker connections were not loose. Finally, I moved the mic around the room and took measurements, and I noticed that I did not get any "High measurement distortion" warning when measuring closer to the speaker. However, one I moved back to the main listening position, same warning appeared. I tried back and forth several times, same thing. In the end, I only moved the mic (and it's USB cable), not the laptop or HDMI, to rule out any possibility of a poor connection with the laptop. I also measured both speakers that were causing issue (SBL and SBR), same thing.

Why am I getting a distortion warning at the MLP, but not closer to the speakers? Is REW reacting to the room acoustics, or what is going on? And do I need to be concerned with regards to using the frequency response and impulse response measurements for manual EQ purposes?

 

[John Mulcahy]

Below 40 Hz the measurement is pretty much all distortion, since the speaker doesn't have any output to speak of down there. You could try reducing the measurement level, or just ignore the warnings.

 

[Spennis]

Is that the cause of the warning you think? Here's the MLP vs. nearfield (quick and dirty, same 90 degree mic cal file) measurements.

 

[robbnj]

I've had some pretty good successes with REW, but am still definitely a neophyte.
That said, I think John is pointing to the capabilities of your speakers as the possible cause. It looks like your speakers have a pretty dramatic output decline starting at 80Hz (that magic crossover number).
Maybe try setting your frequency sweep to start at 50 or even 60Hz, and measure from there up?

 

[Spennis]

I've had some pretty good successes with REW, but am still definitely a neophyte.
That said, I think John is pointing to the capabilities of your speakers as the possible cause. It looks like your speakers have a pretty dramatic output decline starting at 80Hz (that magic crossover number).
Maybe try setting your frequency sweep to start at 50 or even 60Hz, and measure from there up?

Thank you. I will definitely try that. Still wondering whether I can use the full sweep measurements though, or whether the measurement distortion makes them unreliable to use as a basis for EQ filter calculations etc.

 

[John Mulcahy]

It's fine to use the measurements, REW is just warning you there is high distortion. The Distortion graph shows you the frequency ranges where that distortion is happening. In yous case it is where the speakers roll off at the low end and have no output to speak of, so it doesn't really matter. You shouldn't try using EQ to raise the response in that region as it would only make things worse.

 

[Spennis]

It's fine to use the measurements, REW is just warning you there is high distortion. The Distortion graph shows you the frequency ranges where that distortion is happening. In yous case it is where the speakers roll off at the low end and have no output to speak of, so it doesn't really matter. You shouldn't try using EQ to raise the response in that region as it would only make things worse.

Thank you. That's comforting to hear.

How do I determine what an acceptable level of distortion is in these graphs? What am I looking for?

Furthermore, would you recommend using distortion graphs to determine what a sensible crossover point between the speaker and subs would be (to avoid unnecessary distortion), or is it fine too just look at the frequency responses and choose the point where they intersect?

 

[John Mulcahy]

Use the frequency response to decide crossover points. Distortion will depend on level, the louder you try to run, the higher the distortion will become, very rapidly so once you exceed the capabilities of the drive unit(s).

 

[Spennis]

Use the frequency response to decide crossover points. Distortion will depend on level, the louder you try to run, the higher the distortion will become, very rapidly so once you exceed the capabilities of the drive unit(s).

OK. Thank you very much.