I'm with you on the speaker distortion 'myth', could be due to bad mics, but the general consensus seems to be that speaker drivers make 1% distortion or so, but that is definitely not the case.
What leads you to believe the 7052 has lower distortion? I see nothing else than <1% mentioned? From what I could find by googling a bit, it seems condenser mics usually have higher distortion, and that's why I'm doubting the B&K I borrowed from work, even if it's an expensive (condenser) mic.
What leads you to believe the 7052 has lower distortion? I see nothing else than <1% mentioned? From what I could find by googling a bit, it seems condenser mics usually have higher distortion, and that's why I'm doubting the B&K I borrowed from work, even if it's an expensive (condenser) mic.
I'm basing it on a recommendation from a colleague that's been using it professionally for about ten years. ACO Pacific does not have IMD data available for their mics, and the data provided does not indicate anything relevant to IMD, so I am left with trial and error. The SM58 can be used for measurement if I use a correction file. ARTA has a feature where you can create your own correction file by using another microphone as a reference and it calculates the correction file based on the difference between the two mics. Alternatively there is the SM81 which has a much flatter response than the SM57, however it is $400 USD.
Check these open source alternatives. Octave is intended to be compatible.
https://opensource.com/alternatives/matlab
Ok, looking forward to you breaking new distortion records! 
Yes, I know about the calibration file, I was just thinking that FR would not matter much for multitone IMD measurements, more the signal to distortion. FR gets pretty wobbly anyway when doing these ungated multitone measurements, so I thought the mic's contribution to that would not matter much.
Yes, I know about the calibration file, I was just thinking that FR would not matter much for multitone IMD measurements, more the signal to distortion. FR gets pretty wobbly anyway when doing these ungated multitone measurements, so I thought the mic's contribution to that would not matter much.
What model of Bruel are you using? You’re getting excellent results!Ok, looking forward to you breaking new distortion records!
Yes, I know about the calibration file, I was just thinking that FR would not matter much for multitone IMD measurements, more the signal to distortion. FR gets pretty wobbly anyway when doing these ungated multitone measurements, so I thought the mic's contribution to that would not matter much.
There’s a “no return” policy on the Aco Pacific and so I’m not buying it. It’s too risky if it doesn’t perform.
I have to check the mic model later, but don't look at that graph for distortion. That's the ITU... multitone we discussed just before you joined (on the previous page) where all the harmonics are hidden by the test tone itself, starts at 100Hz, and tones are with 100Hz intervals so all harmonics are hidden.
The only visible distortion components are from 4k and up where there is no input signal. I'm thinking this could be used to reveal the actual noise floor of a running measurement (reflections with altered frequency that you mentioned in your video, background noise, noise from measurement gear etc)
The best actual measurement has been around -60dB distortion on tweeters so far, and I don't know if its the mic or the drivers limit. I'm also not sure if the mic distortion rises when frequency goes down, and that could be a reason mids show significantly higher distortion. but, again, I don't have any 'low distortion' drivers to test. This is why I'm curious about the SM mic, since you got around -80dB with that if I remember correctly? Easiest would be if I could just borrow a SM58 from a friend and compare, saves me buying one.
The only visible distortion components are from 4k and up where there is no input signal. I'm thinking this could be used to reveal the actual noise floor of a running measurement (reflections with altered frequency that you mentioned in your video, background noise, noise from measurement gear etc)
The best actual measurement has been around -60dB distortion on tweeters so far, and I don't know if its the mic or the drivers limit. I'm also not sure if the mic distortion rises when frequency goes down, and that could be a reason mids show significantly higher distortion. but, again, I don't have any 'low distortion' drivers to test. This is why I'm curious about the SM mic, since you got around -80dB with that if I remember correctly? Easiest would be if I could just borrow a SM58 from a friend and compare, saves me buying one.
They are. A world class speaker will barely maintain less than 0.1% THD across all frequencies at a modest listening level (90-95dB/1m). Even a mediocre amplifier should easily do <0.01% THD at that level.
An average speaker will get into the 0.1-1% range at certain frequencies. >1% at anything other than very low bass frequencies or frequencies where the distortion products are ultrasonic, is trash.
Also good speaker drivers typically make only low order (2nd and 3rd) harmonics and little higher order harmonics. That makes them sound better at the same numerical THD.
I’m quite surprised on the topic of measurement mics. From everything I’ve read over the years condenser mics are very low distortion at reasonable SPLs. But I personally have no way to measure that.
Using a dynamic vocal mic like the SM58 or SM57 seems a strange choice. They are anything but flat or wide bandwidth. Of course a calibration file can compensate for that, but wouldn’t S/N suffer greatly?
Is there any indication in the literature that a dynamic mic is lower distortion than a condenser? This would be a big surprise to me.
Using a dynamic vocal mic like the SM58 or SM57 seems a strange choice. They are anything but flat or wide bandwidth. Of course a calibration file can compensate for that, but wouldn’t S/N suffer greatly?
Is there any indication in the literature that a dynamic mic is lower distortion than a condenser? This would be a big surprise to me.
At this point there is nothing conclusive as to why the SM57 is better than the Dayton UMM6 USB measurement microphone. All I know for sure is that it has substantially better IMD performance. These are the only two mics that I've tested. I am hoping that others will do more testing of their own mics. Testing a mic with two sound sources it a bit time consuming to setup, and requires one speaker to create the modulating tone, while the other creates the higher tone to observe the side band products of the mic alone.
Sorry but what a crap. A Dayton USB microphone has more noise as a preamp ... no way Sherlock
He is measureing the noise of the microphone. ALWAYS start with a proper noise floor measurement - unplug your speaker, make a measurement, replug and have a look. Then you know if it's a signal from the speaker or from your mic/environment/preamp/....
Dayton USB mic is one of the worst microphones when you need low noise. But there is a reason why NO CARDIOD microphones are used for measurements, ESPECIALLY in close distance! They change their frequency response A LOT in the nearfield cause they record sound particle velocity and not sound pressure.
Not the IMD performance of an SM57 is better - it has less noise. That's a completely different thing.
If you wand to measure distortion without noise - use the right measurement method. You CAN measure THD (summ of overtones) instead of THD+N (sum of everything minus the stimulus).
There actually ARE omni dynamic microphones (MD42, M58, EV 635A...) and even when their linearity doesn't qualify them as measurement mics (but all these cheap USB mics or Behringer mics arn't good enough either without calibration) at least they have a stable behaviour. Use these if you have to.
In these posts so often noise is mistaken with actual IMD or overtones ... be SURE what you measure and when it's not working use a different methode.
I suspect that is the case, being a passive mic. As you say it’s easy enough to test.
Yes the USB mic has distortion producing *elements in addition to the point of **clipping for the electret condensor mic. itself.
*amplifer and Dac. You can get rid of both sources of distortion with a non-USB microphone in a "full dual channel" setup.
**clipping is related to self-noise through the dynamic range for the microphone.
*amplifer and Dac. You can get rid of both sources of distortion with a non-USB microphone in a "full dual channel" setup.
**clipping is related to self-noise through the dynamic range for the microphone.