AFMG Network Forum

EricC · Member

I was working with an experienced Band Engineer the last couple of days and he was giving me some pointers on measurement. I was looking at a magnitude transfer function and Impulse response during a live show. In the IR I could see 2 peaks a few ms apart of approximately the same level. In the magnitude I saw what looked like comb filtering. I'm assuming the 2 peaks where the left and right side of the PA and the comb filtering was the interaction between them.

He said that I should try to position the mic either exactly between the 2 PA stacks so that I get a single well defined peak in the IR. The other option he said was to get a plastic disk and put it on one side of the mic to essentially block one side of the PA from being picked up by the mic if the mic had to be offset to one side of the room.

Do you use any either of these methods or have a better way to get accurate transfer functions during a live show when both sides of the PA need to be turned on?

Thanks,
Eric

Woodstock · Member

Hi Eric,
of course you get best results when you are turning off one PA side Wink

. But I think the audience will not like this.

Yes, when your microphone is a little bit out of center you will see two peaks in the IR and this is causing comb filter in the transfer function.

Try to place your microphone during a show as far as possible in front of one PA side and as far as possible from the other side. The idea of this method is to get a higher level of the PA side where your mike is placed.

I don't think a plastic disk helps you very much. First you will get a pefect reflection and this is causing comb filter in higher frequencies and you will still have comb filter in lower frequencies.

Woodstock.

gluis · Member

I totally agree. If it is possible, try to align your mic to one side of the PA during the show. If you got to have it at FOH (I wouldn't dare to leave a measurement mic out in the audience at most shows), you can try a couple of things. It is possible to have it so perfectly centered that you get one big impulse from both sides. I have done it, and I even remember a post on another forum from the late Don Pearson (if I remember correctly) with his method for achieving this. The other one is a compromise, but a workable one and better than nothing. Measure one side of the PA with the mic in the final position (the one that is going to have at the show), and store the resulting TF curve. Then open both channels and store again. After storing it raise smoothing enough to stop viewing the comb filtering at TF and compare the 3 curves. Store that one too. At show time you can compare the live curve, with the stored ones and make decisions taking into account the comb filter effect, the low frequency build up of having both channels on, and the effect of smoothing. It is not a perfect solution but at least you have a lot of data to make better decisions.

Best regards,

Guillermo Sánchez