Hi,
I'm setting up an audio lab to test speakers and speak enclosures. Any suggestions on what microphone to get to test speaker audio frequency response. Also, recommenations on software to plot frequency response, thanks!
I'm setting up an audio lab to test speakers and speak enclosures. Any suggestions on what microphone to get to test speaker audio frequency response. Also, recommenations on software to plot frequency response, thanks!
If you're measuring indoors, using gated measurements, the mike has to be omnidirectional right in front of the driver. This is because bass fades away with a cardiod mike under 1/4 wavelength away. Gating terminates the measurement before the room reflections come back, theoretically.
If you live in US lots of people use this omni mike. https://www.parts-express.com/miniD...USB-Measurement-Calibrated-Microphone-230-332 This mike is USB so you have to own an up to date PC or laptop. USB is fairly useless for recording music because of variable latency. What PC op system do you have?
Outside the US some people recommend behringer omni mike.
Lots of people recommend software roomequalizerwizard roomeqwizard.com which is a free download. I didn't get it to install until I'd taken my measurements with audacity and a white noise file. I got the audacity from the ubuntu software center (for linux) and the white noise file off the internet.
If you can measure out of doors, omni mikes set so closely ignore the blending of 2 or more drivers. I use a 1" cardioid mike from 2 m away at ear level to include the blending ( and phase cancelation) in my measurement, I bought my cardioid mikes from newark (farnell USA) for $36 each. They require phantom power from a mixer, and your PC has to have a line level 1/8" TRS input. The farnell mikes roll off at 15 khz, which is fine for me since my ears stop at 14 khz. If you need 14 khz to 20 khz, I used to own Shure KSM27 but they were stolen in 2018. Line level condensor mike saves $150 for a USB sound card, the $99 omni mike, all that garbage that is useless for recording live music. I paid $45 for my PV8 mixer and repaired it by buying a used 12vac wall transformer for $2 at a resale shop. Some mixers "for parts or repair" have bad master volume pots or other problems. If ordering from farnell etc you'll also need male to female XLR cable, and a 1/4 phone plug (mixer master out) to 1/8" stereo phon plug (PC input) cable. You need a headphone with 1/4" stereo phone plug or 1/8" to 1/4" adapter to check your setup with the mixer. Shipping is $10, order all the stuff at once.
If you live in US lots of people use this omni mike. https://www.parts-express.com/miniD...USB-Measurement-Calibrated-Microphone-230-332 This mike is USB so you have to own an up to date PC or laptop. USB is fairly useless for recording music because of variable latency. What PC op system do you have?
Outside the US some people recommend behringer omni mike.
Lots of people recommend software roomequalizerwizard roomeqwizard.com which is a free download. I didn't get it to install until I'd taken my measurements with audacity and a white noise file. I got the audacity from the ubuntu software center (for linux) and the white noise file off the internet.
If you can measure out of doors, omni mikes set so closely ignore the blending of 2 or more drivers. I use a 1" cardioid mike from 2 m away at ear level to include the blending ( and phase cancelation) in my measurement, I bought my cardioid mikes from newark (farnell USA) for $36 each. They require phantom power from a mixer, and your PC has to have a line level 1/8" TRS input. The farnell mikes roll off at 15 khz, which is fine for me since my ears stop at 14 khz. If you need 14 khz to 20 khz, I used to own Shure KSM27 but they were stolen in 2018. Line level condensor mike saves $150 for a USB sound card, the $99 omni mike, all that garbage that is useless for recording live music. I paid $45 for my PV8 mixer and repaired it by buying a used 12vac wall transformer for $2 at a resale shop. Some mixers "for parts or repair" have bad master volume pots or other problems. If ordering from farnell etc you'll also need male to female XLR cable, and a 1/4 phone plug (mixer master out) to 1/8" stereo phon plug (PC input) cable. You need a headphone with 1/4" stereo phone plug or 1/8" to 1/4" adapter to check your setup with the mixer. Shipping is $10, order all the stuff at once.
Last edited:
Hi,
Microphones: behringer ECM800, Dayton, Sonarworks,... whatever is omni and include calibration files ( this is important).
For software REW is free and what most peoples use this days so easier to have help in case you need.
Microphones: behringer ECM800, Dayton, Sonarworks,... whatever is omni and include calibration files ( this is important).
For software REW is free and what most peoples use this days so easier to have help in case you need.
Thanks guys, I really appreciate your suggestions. I just placed an order for a miniDSP UMIK-1 microphone and have installed REW software. I’m new at this but excited to start experimenting with speakers and speaker enclosures
The UMIK makes it difficult to measure relative phase. The USB interface makes it easy to plug in but impossible to lock onto the timing.
The USB driver would affect phase measurements between two or more signals, I suspect there is some time delay with a certain +/- tolerance depending on how busy the CPU is at any time. If this is true, I wonder if a real-time OS might improve measurement accuracy with a USB based mic?
Don't bother, a soundcard with two chanels, one dedicated to loopback, the other to your regular mic, done.
Hi, and welcome!
If you want to time align drivers which have offset between their emmission point: horn system where the compression driver sometimes located 10cm or more behind the direct radiator is an (extreme) example.
If you plan to use Linkwitz Riley filters it is needed to time align too.
This is what AllenB message is related.
If you want to time align drivers which have offset between their emmission point: horn system where the compression driver sometimes located 10cm or more behind the direct radiator is an (extreme) example.
If you plan to use Linkwitz Riley filters it is needed to time align too.
This is what AllenB message is related.
To put things in perspective, many computers already have sound in/out without adding something like this. That just leaves the question of powering the mic. Phantom power can be made with some 9v batteries and a few components, maybe put it in a case, some might make up a cable. So there is a DIY element in that. A person has a choice.
The Solo is OK for measuring loudspeakers. Some advanced tricks could be made with a second channel for loudspeaker. There is a possibility for using the second one in a loopback configuration which a clever program can use to improved the accuracy and null out latency "errors" - but I'd say you will be fine with the Solo - I use one and havent missed another channel.
//
//
I agree flat fr is not a great idea for final outcome imho, BUT it's a very convenient intermediate step within 'voicing' of loudspeakers.
In my own process it is the last one before applying house curve ( a downward pointing linear attenuation from 1khz to 24khz @-6db). Other modifications can take place depending on spl level expected at listening position.
I agree too accelerometer ( and taking care of box issues) are nice to have too. Once you start measuring things there is many one can make.
The number of input/output of a soundcard will depend of what you plan to do and how you'll evolve.
Basicaly you need one input for your mic.
If you start to want to play with time referenced measurements then a second chanel is mandatory.
Then if you want to measure at multiple different locations for averaging or other reason ( like doing an MMM technique at same time) then 4x input (or even more) can be interesting.
Another things most people forget: if you have multiple input, you usually have multiple output availlable too. At least this is true in semipro gear.
This open horizon to dsp processing and multiamp approach (within your computer) if this is your thing ( it's mine).
In my own process it is the last one before applying house curve ( a downward pointing linear attenuation from 1khz to 24khz @-6db). Other modifications can take place depending on spl level expected at listening position.
I agree too accelerometer ( and taking care of box issues) are nice to have too. Once you start measuring things there is many one can make.
The number of input/output of a soundcard will depend of what you plan to do and how you'll evolve.
Basicaly you need one input for your mic.
If you start to want to play with time referenced measurements then a second chanel is mandatory.
Then if you want to measure at multiple different locations for averaging or other reason ( like doing an MMM technique at same time) then 4x input (or even more) can be interesting.
Another things most people forget: if you have multiple input, you usually have multiple output availlable too. At least this is true in semipro gear.
This open horizon to dsp processing and multiamp approach (within your computer) if this is your thing ( it's mine).