Im at the point where I want to start measuring FR of my projects for making adjustments to passive xovers and enclosure design.
I have a Omnimic V2 and dedicated PC to use, I know I can just plug it in a get a FR in real time but what Id like to find out is what matters and what doesnt in the set up.
My situation is a large echoing room(tiled floor and concrete walls), I dont think SWMBO would like the room to be lined with spikey acoustic treatments 😱
Would it be better off starting with a small area maybe a corner with some temporary treatment in place on sheets of ply? Do I surround the speaker completely or can firing into a corner with the temporary treatment work.
Im not looking to quantify to the umpteenth degree and build a chamber, just good enough to improve my projects without to much guesswork.
Any good links or YT videos I can study?
I have a Omnimic V2 and dedicated PC to use, I know I can just plug it in a get a FR in real time but what Id like to find out is what matters and what doesnt in the set up.
My situation is a large echoing room(tiled floor and concrete walls), I dont think SWMBO would like the room to be lined with spikey acoustic treatments 😱
Would it be better off starting with a small area maybe a corner with some temporary treatment in place on sheets of ply? Do I surround the speaker completely or can firing into a corner with the temporary treatment work.
Im not looking to quantify to the umpteenth degree and build a chamber, just good enough to improve my projects without to much guesswork.
Any good links or YT videos I can study?
You're better off using the middle of the room. Branch the speaker and mic a little toward opposite corners.
You seem to have a USB mic. Not as easy to do timed measurements.
You seem to have a USB mic. Not as easy to do timed measurements.
You just set the gate in omnimic to remove reflections from the measurements. Lift the speakers off the ground, and as far away from boundaries as possible. There's no need to acoustically treat the walls.
Thanks guys.
So the middle of the room(its 12x6m)set about middle height of 1.35m and use the gate settings... almost sounds to simple 🙂
Some real noob questions
White or pink noise or a sweep, how loud does it need to be?
When measuring individual drivers how do I make a .frd file to use in Xsim, can you just select the file type when saving?
So the middle of the room(its 12x6m)set about middle height of 1.35m and use the gate settings... almost sounds to simple 🙂
Some real noob questions
White or pink noise or a sweep, how loud does it need to be?
When measuring individual drivers how do I make a .frd file to use in Xsim, can you just select the file type when saving?
For measuring, this might be helpful: How to Achieve Accurate In-Room Quasi-Anechoic Free-Field Frequency Measurements by Jeff Bagby.
To work with the files in XSim, it is necessary to also measure the relative acoustic centers of the drivers if you are not including timing in your measurements.
Box with Passive Crossover Designer (PCD) or
Speaker acoustic center - How to find it | Audio Judgement with XSim.
To work with the files in XSim, it is necessary to also measure the relative acoustic centers of the drivers if you are not including timing in your measurements.
Box with Passive Crossover Designer (PCD) or
Speaker acoustic center - How to find it | Audio Judgement with XSim.
I use OmniMic, as well as the 2-channel method with REW.
Allen points out that a USB mic will not directly capture the timing differences among drivers. The timing differences give you the z-offset. This is the disadvantage (perhaps only disadvantage) to omnimic. However, there is a way to accurately calculate the timing differences using XSim. There is also a method described in the Omnimic documentation, but I must be dumb because I could not understand how to make it work. It works for other people.
The Jeff Bagby white paper that jReave linked is excellent... basically all you need to know. The other links show how to us XSim (or PCD) to calculate the z-offset.
The omnimic psuedo noise sequence is fine to use up to about 2 khz. I usually use this for near field or ground plane testing (it is easier on the ears). Above 2k, use the swept sine wave.
If your sound source is a PC or laptop, be sure you have ASIO drivers loaded. For a while I was using Windows Media Player, and the frequency response of the test signal WAS NOT FLAT. Now I use Foobar with ASIO drivers. In order to be confident I had a good test signal, I loaded the omnimic test tracks to a CD-ROM and played them through a CD player. I measured a driver using both the CD ROM and Foobar / ASIO... Only then did I trust that my laptop was accurate.
Use a microphone stand when making gated far field measurements. see
Loudspeaker Measurements
In general, you want your first reflection to be from the floor. If it is 35 inches from the driver being tested, you want a 35 inch radius sphere around the speaker to be clear of any reflective surface. You want a 35 inch radius sphere around the microphone to be clear of any reflective surface. You want a 35 inch radius cylinder of clear space between the speaker and the microphone.
If I think of anything else, I will post more...
j.
Allen points out that a USB mic will not directly capture the timing differences among drivers. The timing differences give you the z-offset. This is the disadvantage (perhaps only disadvantage) to omnimic. However, there is a way to accurately calculate the timing differences using XSim. There is also a method described in the Omnimic documentation, but I must be dumb because I could not understand how to make it work. It works for other people.
The Jeff Bagby white paper that jReave linked is excellent... basically all you need to know. The other links show how to us XSim (or PCD) to calculate the z-offset.
The omnimic psuedo noise sequence is fine to use up to about 2 khz. I usually use this for near field or ground plane testing (it is easier on the ears). Above 2k, use the swept sine wave.
If your sound source is a PC or laptop, be sure you have ASIO drivers loaded. For a while I was using Windows Media Player, and the frequency response of the test signal WAS NOT FLAT. Now I use Foobar with ASIO drivers. In order to be confident I had a good test signal, I loaded the omnimic test tracks to a CD-ROM and played them through a CD player. I measured a driver using both the CD ROM and Foobar / ASIO... Only then did I trust that my laptop was accurate.
Use a microphone stand when making gated far field measurements. see
Loudspeaker Measurements
In general, you want your first reflection to be from the floor. If it is 35 inches from the driver being tested, you want a 35 inch radius sphere around the speaker to be clear of any reflective surface. You want a 35 inch radius sphere around the microphone to be clear of any reflective surface. You want a 35 inch radius cylinder of clear space between the speaker and the microphone.
If I think of anything else, I will post more...
j.
You shouldn't need to go to that length, REW will do it using a technique (an available option) called 'acoustic timing'. It sends a second, known signal which it can refer to in each measurement. Just avoid USB mics in the future. 🙂
I use UMIK-1 a USB mic calibrated by third party Spectrum Labs for a nominal charge. I use the acoustic interferometer method to determine the acoustic offsets between the two drivers. Works quite well.