So I tuned a system with an SM81….

So I tuned a system with an SM81….

Since I’m a doofus, I grabbed the wrong toolkit on the way out the door and arrived at the site without a reference mic. Being as that the primary goal in system optimization is uniformity, rather than absolute response, I figured I’d use one of the venue’s SM81 small-diaphragm omnidirectional condensers and see how well I made out. LR system of four boxes per side + front fills. The upcoming events were speech reinforcement so I’m tuning for a relatively flat response.

Here is Main L, all boxes on, measured at ONAX A,B,C,D. HF phase coherency is surprisingly good for such a reverberant space.


Averaging is always a controversial topic when it comes to optimization, and in this case it doesn’t make much sense, because we’re shooting for uniformity so we want to see variance over the space, however since I’ve only got one channel of DSP per side, I have to decide on a ‘best fit’ EQ solution, and so let’s see what the average gives us:

The average seems to agree well with the two issues I see optically: bumps around 300 and 1k.

For speed’s sake I don’t always pick the signal off post-EQ into the analyzer but I did this time, and here’s the two filters I applied along with the data as I viewed it at the time:

And here it is with 1/EQ:


And here is how the 1/EQ fits with the averaged traces:

I chose to leave the bump around 160 for now, as music playback sounds completely anemic without it. (I also run the sub feed a little hotter than unity in this room.)

Now let’s revisit OnAX A, Post-EQ and compare the before and after:

Here’s all four ONAX locations Post-EQ:

The weird loss around 7k comes from box C, and there’s unfortunately not much to be done about it short of pull all the boxes down and test the drivers.

And let’s see how that compares to the averaged traces:


Returning to the venue the next day, I brought an actual measurement mic and took another set of measurements. Here is how the ONAX L at ABCD compares between SM81 (purple) and a MicW M215 (blue):



If I had to guess, the slight HF rolloff of the measurement mic is due to the smaller capsule creating less of a pressure zone at wavelengths approaching the capsule size. The 81 is also a cardioid mic, so it’s potentially less affected by reverberant energy, but I have to say I would have expected that to show up as less HF, not more.

This trend shows up in the single traces as well. For example, ONAX A, and again the blue is the measurement mic:

Note the phase response match as well.

What I learned from this is that we can still meet the goal of uniformity over the space by using a “non-measurement” mic. As we see, the absolute curve may not be represented as we’re used to it, but this may not be a question of which is ‘more accurate’ as much as the different mics provide different lensing effects on the data. After all, neither mic is a great model for human perception, but I think the omni is probably closer in this case.

One Reply to “So I tuned a system with an SM81….”

Leave a Reply

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