User Question: Smaart Practice

User Question: Smaart Practice

I received a message from Jedidia, a church audio engineer in Africa, asking if I could look over his measurement data and talk about time alignment between the two sources.

Here is Jedidia’s measurement setup:

And here are his traces for his main and his sub:

Thank you to Jedidia for allowing me to share his photos as well as my response:

It definitely looks like we are on the right track here. The measurement data looks reasonable for your situation. The measurement of the full-range loudspeaker has pretty low coherence – this could be greatly improved by using longer time averaging – generally I would recommend 2 or 3 seconds of averaging for acoustic measurements, and 4 or 5 if you’re measuring in a large reverberant space like a church. Averaging is a great way to suppress the noise in a measurement and get cleaner, high-coherence data.

The time/phase alignment point should be done at the frequency range where the two sources are equal in level. On your traces, that’s just above 125 Hz (where the magnitude traces cross). That seems like it will leave quite a dip in the response, so if you can decrease the frequency of the high pass filter on the mains, or remove it completely, I would recommend doing that. Ideally we don’t want any gaps. 

Notice that around 125 Hz – again, where both sources meet in level – the phase traces have the same slope and are within about a third of a cycle of each other, so I would consider them aligned-sh, or “close enough for show business.” However, that’s going to change once you play with the crossover filters to fill in that hole, so you should repeat the process, matching the phase slopes at the new crossover point.

 

One Reply to “User Question: Smaart Practice”

Leave a Reply

Your email address will not be published. Required fields are marked *