Good to know!
I mostly use the M50 cause I have to measure high frequency stuff - the nice thing it's very linear up to 20kHz, I don't need a correction file. When I need even more there are G.R.A.S. 1/4" mics, but these are very noisy and THD is already compromised at lower volumes.
I bought an Audix TM1 at the suggestion of profiguy, and I have been very satisfied.
The TM1 is good enough to record with thanks to its lower than usual noise floor for a measurement mic. It also goes high enough to pick up the UHF breakup of metal domes. Thats far enough to identify any weird issues past 20k which may affect phase or IMD in the upper pass band.
Looking at the cal file on this mic, it doesn't need much correction compared to other cheaper mics, so its not just some Chinese rebranded mic stuck in a fancy box and marked up in price.
Looking at the cal file on this mic, it doesn't need much correction compared to other cheaper mics, so its not just some Chinese rebranded mic stuck in a fancy box and marked up in price.
Condenser mics usually have quite high distortion and self noise, TM1 SNR is only 66 dB. That makes them not-ideal for measuring low levels of distortion. It's probably very good for other loudspeaker measurements though.
It's not a bad idea to use a dynamic mic for distortion measurements.
It's not a bad idea to use a dynamic mic for distortion measurements.
Do you know an omnidirectional dynamic mic with good enough linearity? I don't.
Just use a proper measurement mic! There are special microphones with self noise as low as 6.5dB(A)Spl (we use them for sound isolation measurements). Even the cheap MicW M215 is specified with self noise level <18dBA and max level (1% THD) of 135dBSpl.
Proper 1/2" measurement mics have a range from 14-15 dB(A) to 145-150dBSpl. That's fine for all measurements you need to do. (except super low noise stuff - but when you can afford the room for that the mic is not the problem ;-))
https://www.grasacoustics.com/products/measurement-microphone-cartridge/product/163-40af
Last edited:
Mine measures 88dB SNR on my motu interface. Thats why I said its good enough for recording. Not sure why it would be advertised as low as 66dB
Sorry - but that's for sure a mistake. How do you measure that? It's not S/N, it's noisefloor relative to 94dBSpl, 1kHz or 400Hz.
Its not a mistake but I used 100 dB reference to measure (this is all I had available). That translates into a self noise of 12 dB - much lower than the 28 dB advertised, which is still a decent figure for a calibrated measurement mic of this price range.
The better specs are why I purchased a pair of TM1s to begin with. Both mics measure close to each other all around. The high spl/distortion is also lower than stated - less than .5% at 140 dB for both mics. I had that verified by the Audix tech up in Oregon after I purchased them (I know the rep). I was also skeptical about the better than advertised specs but the measurements and listening tests I did convinced me. They're just very good mics for the money.
I actually tried to get a TM1 to distort by close micing a snare drum with it and I did this with some other cheaper measurement mics (Behringer, DBX, etc). The TM1 stayed very clean and clear at very high spls unlike the cheap black DBX and silver Behringer mic everyone uses - they both sounded dirty and awful in comparison. Compared to my Neumann TLM103 (about 4 dB self noise) at the same gain setting, the TM1 was marginally noisier when taking the noise distribution into account.
For me, anything under 20 dB self noise is fine for measuring speakers, unless you want to measure very low levels of distortion. The self noise spectrum isn't linear anyways and it can also depend on how the mic is loaded by the preamp, so there are other variables involved.
The better specs are why I purchased a pair of TM1s to begin with. Both mics measure close to each other all around. The high spl/distortion is also lower than stated - less than .5% at 140 dB for both mics. I had that verified by the Audix tech up in Oregon after I purchased them (I know the rep). I was also skeptical about the better than advertised specs but the measurements and listening tests I did convinced me. They're just very good mics for the money.
I actually tried to get a TM1 to distort by close micing a snare drum with it and I did this with some other cheaper measurement mics (Behringer, DBX, etc). The TM1 stayed very clean and clear at very high spls unlike the cheap black DBX and silver Behringer mic everyone uses - they both sounded dirty and awful in comparison. Compared to my Neumann TLM103 (about 4 dB self noise) at the same gain setting, the TM1 was marginally noisier when taking the noise distribution into account.
For me, anything under 20 dB self noise is fine for measuring speakers, unless you want to measure very low levels of distortion. The self noise spectrum isn't linear anyways and it can also depend on how the mic is loaded by the preamp, so there are other variables involved.
Well, again I'm no expert, but the popular UMIK-1 has quite similar specs for distortion and it's considered too much for measuring distortion. Please remember that you can use different mics for different tasks. If TM1 real world performance is better than specsheet, then it's all the better.
In context of this thread, if you want more than 1% HD out of, say 15NMB420, you probably need more than 100 dB and things can get quite uncomfortable.
PS. Just a thought: why do speaker measurerers use omni-pattern mics while building anechoid chambers to capture only direct sound? If all the money went into making a directional measurement mic, wouldn't that be easier?
In context of this thread, if you want more than 1% HD out of, say 15NMB420, you probably need more than 100 dB and things can get quite uncomfortable.
PS. Just a thought: why do speaker measurerers use omni-pattern mics while building anechoid chambers to capture only direct sound? If all the money went into making a directional measurement mic, wouldn't that be easier?
What you write here is technically not possible. (I developed electronics for AKG - I know how to measure and calculate that stuff)
The TM1 has 6mV sensitivity and a 6mm capsule - that's noisy. For 12dB of self noise it would need 0,5uV(A)rms noise voltage, that's not possible for such a small capsule + buffer. The lowest noise 1" studio condensers have around 1uV(A)rms self noise (from 20-20kHz).
When you measure with 100dBSpl reference you have of course to recalculate to 94dB! So when you have 88dB difference between 100dBSpl and noise floor - you get 18dB(A) self noise - still very good! Too good I would say but I don't have a mic here for a valid measurement. (1uVrms(A) is actually possible but for such a small capsule it's not done by manufacturers cause you would need an incredible high impedance input buffer (normally a few GOhm, here you would probably need 100GOhm). Needs ages to settle and hard to manufacture reliable etc.)
The Earthworks M50 has 30mV sensitivity, a little larger capsule and get's 22dB(A) self noise (which is actually not that good, could be better. And mine es even 1,5dB worse). The TM1 only has 6mV sensitivity ... to get 28dB(A) self noise is actually quite ok.
When you hear very little difference to your TLM103 - your room is noisier as the mic self noise. Btw - whe your TLM103 really has 4dB(A) self noise you must got a very high sensitive one.
Behringer and DBX and all the other electret mics are very bad - I would not say measurement microphones to them. They have HUGE tolerances and are totally useless without calibration. You CAN be lucky to get a good one ... but you never know. The MEMS microphone in your Iphone is actually better.
Here are calibration measurements of these cheap mics, but only in German - up to +-5dB tolerance at 20kHz: https://www.hifi-selbstbau.de/grund...0-1000-mikrofonkalibrierungen-eine-uebersicht
MY 4 MicW are performing to spec. But I can imagine that quality control is not to strickt for these cheap models.
p.s.: We should probably make a new topic about measurement microphones and not hack this speaker project ;-)
Thanks for the explanation. Its possible my interface wasn't set up right on the software side ie. preamp settings didn't add up. That would explain the optimistic numbers. The only reason I defend the TM1 is the verification I got from Audix, and it favors that reasoning when I compare it to the EW M40 I typically use - don't remember the specs on that one, but my specific one is close to both the TM1s. The TLM103 was a cherry picked one from a former commercial studio, but those generally had very low self noise to start with like the Rodes NT1 did. Maybe I just got very lucky with my TM1s. They did come from Audix directly in Oregon, personally set aside by the rep.
Good info on the mics! I think the B&K I'm using should be fairly good, but if there was a reasonably prized dynamic mic that performs better distortion-wise, I might be interested in using that for distortion only. I'm thinking FR is not critical for that.
This project has stalled a bit. Main reason is the missing mid driver. They received it, but it turns out I returned it in the wrong box (box for the 8ohm driver) so I had to send them the correct box yesterday, so it will probably be a couple of weeks before I have the other mid..
I also received the minidsp flex eight, so I have been playing around with it a bit, just installing the config tool, connecting, trying bluetooth to phone, and USB to my Daphile PC. All seems fine so far. I should start looking at amps too. I'm thinking I'll use my Alpha Nirvana for mid, possibly JLH-69 for treble, and thinking I should build a AB amp with high loop gain (low output impedance) for bass, since I have found I like that for well controlled bass. I could probably throw something together from stuff I have lying around, maybe maximize loop gain and re-compensate some chinese LJM kit. Want to fit it in a small case, since space can become an issue going from one to three amps..
This project has stalled a bit. Main reason is the missing mid driver. They received it, but it turns out I returned it in the wrong box (box for the 8ohm driver) so I had to send them the correct box yesterday, so it will probably be a couple of weeks before I have the other mid..
I also received the minidsp flex eight, so I have been playing around with it a bit, just installing the config tool, connecting, trying bluetooth to phone, and USB to my Daphile PC. All seems fine so far. I should start looking at amps too. I'm thinking I'll use my Alpha Nirvana for mid, possibly JLH-69 for treble, and thinking I should build a AB amp with high loop gain (low output impedance) for bass, since I have found I like that for well controlled bass. I could probably throw something together from stuff I have lying around, maybe maximize loop gain and re-compensate some chinese LJM kit. Want to fit it in a small case, since space can become an issue going from one to three amps..
I feel your pain as far as the 16 ohm mid mistake goes. I had a similar problem ordering a compression driver that should have come with bolts, but it didn't and took nearly 11 months to arrive. I like your build. I think the upper cabinet should work well once all the trouble shooting is done. Regarding that woofer frame resonance, you could fix some slim pieces of lead sheeting with liquid nails , to the basket frame. That will quieten it down a lotDid some initial FR measurements today. At first it all looked a bit odd, and I could not really understand what was going on. The tweeter did not show the typical WG response with boosted mids and falling towards higher frequencies. I thought it was a tweeter issue first, but when I tried another mic it all looked like what I expected. An expensive B&K mic bites the dust?
I see two things in the measurements I don't like so far. The tweeter output is a bit low compared to the mid. Ok, not a big issue since the overall response could be EQ:ed with DSP even if it was falling towards the high end. The other issue is the acoustic centers of the drivers is off. The mid is around 1.5cm behind the tweeter, I expected them to be a lot closer. This is not to my liking if I am to stay with passive XO between mid and tweeter. I don't like 'tricking' the phase to align in the passive XO, so running DSP for mid/tweeter XO is more tempting now than before.
I forgot to thank you for the input mattstat, I appreciate the input!
Hopefully I will have some 'quality time' to play around with these more in the weekend. For now focus is more on the technical matters (not design), need to make some angle indicator on the stand so I can make some off axis measurements and start learning to use vcad for the XO sims.
The 16ohm mid is on it's way back, THLP customer support has been good and quick so far!
I hope you are right, the off axis looks decent at least, I think I have to find out what compromise I like best between smooth overall, and smooth on-axis. Also try some different slopes to try how horisontal directivity affects the sound.
I think I will try supporting the magnet with the bracing first, in case I want to try other woofers, they are easier to sell unmodified.
I think I will try supporting the magnet with the bracing first, in case I want to try other woofers, they are easier to sell unmodified.
After 21y of working with various DSP options, I find a DAW like PC based system to be by far the best.
As a small update, this is what the back of the speakers will look like. I'm planning some room treatments in the near future, so I figured, why not use the back of the speaker for some treatment too, it's not going to visible anyway. If I'm being optimistic, it might even dampen some of the sound radiated from the back of the box too.
The wheels and handle are there to make it easier to move the speakers around. Thay can also be laid on their backs, resting on the wheels and the handle, for transport (need a van for that) or if I do some work on them.
Working on finishing the cosmetics on the front too. All vinyl wrap, so no advanced woodworking here, I just want them to blend in in the living room, and not irritate my eyes (or HER's) from listening distance
The wheels and handle are there to make it easier to move the speakers around. Thay can also be laid on their backs, resting on the wheels and the handle, for transport (need a van for that) or if I do some work on them.
Working on finishing the cosmetics on the front too. All vinyl wrap, so no advanced woodworking here, I just want them to blend in in the living room, and not irritate my eyes (or HER's) from listening distance
Have been making some speaker cables, using 4-pole speakon's for the top cabinets, and banana plugs for the bass. Makes it easier to 'grab and go' the upper cabs, and reduces the risk of swapping mid/tweeter cables. Tricky to fool-proof the whole signal chain when running a fully active setup, and mistakes can be costly..
Today I took a picture of the front too. Some minor finishing work to do on that speaker still, then finish the other one, and wait for the mid driver..
Today I took a picture of the front too. Some minor finishing work to do on that speaker still, then finish the other one, and wait for the mid driver..