Rational Acoustics



Jivey1303
August 28th, 2011, 06:26 PM
Would anyone care to share some tips on blanking thresholds? Is the default of 20% a good starting point or should it be adjusted stronger or lighter?

Thanks

lundbergsound
August 29th, 2011, 10:58 AM
Like with most of what we do, it really depends on what you're measuring / for what you're looking. If you are measuring a processor or console (i.e., electrical signals only; not acoustic), any part of a measurement with coherence less than 95-100% is suspect. If you're in a reverberant concert hall or arena, 20% may be the best data you can get.

I tend not to pay much attention to the blanking threshold per se, because I am always watching the coherence trace and making decisions about how much to trust the frequency response measurements based on the coherence. Blanking is helpful to an extent, but if most of my measurement is 80% coherence and one frequency range is 40%, I am going to look for more information about what is happening in that frequency range even if it's not blanking.

Having said that, in direct answer to your question, my blanking threshold is currently at 50%, although I don't recall for what reason I set it to that.

Best,
Daniel

Arthur Skudra
August 29th, 2011, 09:50 PM
Depends on what you're measuring. For live sound, I usually set it 15 to 20% and leave it be. If I'm measuring an electronic circuit, then a lot less. You will find that when you live average several measurement engines together, your coherence improves dramatically, and you'll find blanking to be less of an issue.

Jivey1303
August 30th, 2011, 09:25 AM
@Arthur - What do you mean by live average several measurement engines? I know how to average captured traces, but how do you do it to a live trace?

Arthur Skudra
August 30th, 2011, 10:32 AM
@Arthur - What do you mean by live average several measurement engines? I know how to average captured traces, but how do you do it to a live trace?
It's actually quite easy. After you set up your measurement engines for all of your mic inputs in the Measurement Config window, click on the "New Average" button, then click on the newly created average engine, and check the boxes to the left of the individual mic engines that you want included in the live average engine. You can have as many live average engines running simultaneously as you want, each a different combination of several mics. This is where having 8 mics really speeds things up when measuring a line array. Just as an example I have one average engine including all 8 mics, and perhaps another for the front 4, and another for the rear 4.