• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Motu M4 - Tear down, bit of internals analysis and few in-house measurements

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,317
Likes
1,331
Location
Iasi, RO
#1
After @amirm published Motu M4's measurements review, several ASR members became interested in this audio interface, me included. I was able to get this from Thomann, after I almost decided to cancel my order due to manufacturer delay, but in the end I was able to get it after few weeks of waiting.

Details and specs about the M4 interface could be found on the manufacturer website: https://cdn-data.motu.com/manuals/usb-c-audio/M_Series_User_Guide.pdf.

The M4 was assembled in the U.S., it is actually written on the board VIRTEX - Austin, TX: https://www.virtex.us/. However, “Assembled in” is different than “Made in”, based on https://www.themadeinamericamovement.com/made-in-usa-certified/difference-between-assembled-made-in-usa/#:~:text=The product is assembled in,of Columbia, and U.S. territories., so I guess the board and soldering, but also the assembly of the parts are all done in the U.S. Probably the case and most of the electronic components are still manufactured in ASIA, perhaps in China, but not sure this matters much given that the final product is really good.


IMG_0610_.jpg

Motu M4 board from top

image3.jpeg

Motu M4 board from top

Soldering looks good to me, the board is clean on top and also on the bottom, just a bit of flux residue around the output plugs. Board is multi-layer and with a huge visible ground plane on top and bottom layers to minimize noise and interferences. Plugs are soldered very well and seem sturdy and not moving when I try to bend or pull them. Also, the RCA plugs from back are kept in place with a couple of dedicated screws, so it's very unlikely these plugs to move or become loose with time due to plugging/unplugging the cables.

All knobs, but all of them, even the small one from the headphones amplifier volume, are made of aluminum, not regular plastic like most interfaces in this price range. The black anodized aluminum case is not easy to get scratched, it seems robust and should last long.

The front display is fancy and eye catchy, a very good tool to get an idea about the input/output levels, although the OLED VU-meters are kind of slow, much slower than regular analogue lights from, let’s say, Scarlett interfaces. They’re displaying more of a top-averaging of the SPL of the sound, instead of rapid peaks, but if clipping occurs then the red light stays lit on top of the display for few seconds, so there should be no problem in getting a good recording after all.

I tend to press the 48V buttons when I unplug the XLR mic plugs with my fingers, so unbalanced dynamic mics might get destroyed by this (https://www.shure.com/es-CO/desempeno-y-produccion/louder/top-8-microphone-myths-exposed) or the phantom power itself might get damaged, depends. Some say (https://royerlabs.com/ribbon-mics-and-phantom-power/#:~:text=The ribbon elements in some,stretched or completely blown ribbon) that older ribbon mics could also get damaged by the 48V too, so take good care when unplugging such mics from the M2/M4 or simply power it off before doing this, as it is also recommended, due to the inherent noise that pops into the speakers anyway. However, I'm using balanced dynamic and condenser mics, so I don't are much about this.


IMG_0616_.jpg

OPA1678 operational amplifiers buffers DAC outputs

Six OPA1678 (https://www.ti.com/lit/ds/symlink/opa1678.pdf), marked as TI 02 1AW7, are used as output buffers for the four TRS and RCA outputs from the backside of the case. These are connected between the DAC-outputs and the interface balanced output plugs, so the TRS and RCA outputs are direct-coupled, without any capacitors in between.


OPA1678.png

OPA1678 THD+N from TI datasheet

One OPA1688 is the driver for the headphones. This is a dedicated op amp for use as output buffer for headphones output and is able to deliver 50mA @32Ohms with a very low distortion and noise. Despite it’s relatively low output power, I was able to listen to comfortable levels to headphones like AKG K701 (62 Ohms and relatively hard to drive) and LCD2-F. However, when switching to Hifiman HE-560 the OPA1688 powered by the 10V rails was not able to deliver enough power to put the joy on my face, so for hard to drive cans an external headamp might be required. When headphones volume knob passes 3 o’clock, with 0dB recorded music, distortions become clearly audible, but this output volume is too much anyway for most studio monitoring headphones, so keeping the volume below 3 o’clock seems reasonable enough.

OPA1688.png

OPA1688 THD+N from TI datasheet


image1.jpeg

OPA1688 headphones output buffer

The USB interfacing is handled by the XMOS XU216-256-TQ128-C20 microcontroller chip. Nearby there is a SO8 chip that might be ST922 op amp (https://www.st.com/resource/en/datasheet/ts922.pdf); given it’s 0.005% THD+N and its location I’m sure it was not placed in the audio signal path, so it’s probably there for a different reason (comparator?, protection?). A flash memory (http://www.issi.com/WW/pdf/25LP-WP080D-040D-020D.pdf) was probably used to store the firmware, so future upgrades can be easily done through the dedicated MOTU application.


IMG_8857 copy_.jpg

XMOS XU216-256-TQ128-C20 microcontroller chip

To combat ripple and noise MOTU installed eight aluminum polymer surface-mount capacitors on the board. Initially I thought that these capacitors were Nichicon CS-series, but on a closer look I find them more likely to be the ones manufactured by a Chinese brand named YTF, more details could be found here: https://www.ytfcapacitor.com/smd-capacitor/Hot-Offer-SMD-470uF-SMD-Electrolytic-Capacitors.html. Also, JB seems that used to manufacture similar looking caps as well: https://www.jbcapacitors.hk/post/2010_7.html. If I am right about the YTF manufacturer, then these caps are rated with a “load life” between 3000-5000 hours, although no datasheet could be found to be downloaded. We could probably press the “wirte us” button from their webpage to ask them about more details about these caps. However, their video presentation
shows us a decent manufactory, so I guess the caps are pretty decent, but time will tell. There are also several smaller capacitors on the boards as well, but unable to tell the manufacturer.

We can also spot seven Panasonic FK-series polymer capacitors, probably used for decoupling purposes.

The device is powered via the USB-C plug, so the 5V coming from computer's USB is split into +/-5V by the 250mA dual converter TPS65133 chip (TI 87I C1NC marking on the board). Inside the chip there is a boost converter that generates the +5V and an inverting buck-boost converter that generates -5V; this way the internal op amps and other active components from M4’s board will be powered by a total of 10V, the headphones driver too.


image0.jpeg

TPS65133 dual converter responsible for getting the +/-5V rails

On the board there are several LDO regulators like: https://www.ti.com/lit/ds/symlink/lp5907.pdf, https://www.ti.com/lit/ds/symlink/lp5912-q1.pdf etc.

The MIDI input, being a serial transmission protocol, is using a H11L1M optocoupler for protection, more details to https://hackaday.com/2018/05/09/opt...microcontroller-midi-and-a-hot-tip-for-speed/ and to https://learn.sparkfun.com/tutorials/midi-tutorial/hardware--electronic-implementation. Per Wikipedia (https://en.wikipedia.org/wiki/MIDI): “Opto-isolators keep MIDI devices electrically separated from their connectors, which prevents the occurrence of ground loops[75]:63 and protects equipment from voltage spikes”.


image2.jpeg

H11L1M optocoupler

Digital to Analog Conversion is done by the -110 dB THD+N ESS ES9016S 8-channel chip. Four channels are used for the four outputs from the backside (two for Monitor Out and two for Line Out), while two seem to be used for driving the headphones, via the two 4580R (https://www.ti.com/lit/ds/slos412d/slos412d.pdf) operational amplifiers and the OPA1688 buffer. The other two channels from the ES9016S DAC chip don't seem to be connected to the board.


IMG_8872 copy_.jpg

ESS ES9016S 8-channel chip

IMG_8885 copy_.jpg

4580R operational amplifiers, placed between DAC and output buffer

Worth mentioning that Headphones-Out is mirroring the 1/2 Monitor outputs only, so when using the 3/4 Line outputs with an external amplifier, pre-amp or mixer you will not be able to use the built-in headphones amplifier. Also, to listen to both 1/2 and 3/4 outputs you will need an application that knows how to work with both outputs at the same time, like a 4-channels player, otherwise the operating system will see the two output as two distinct playback devices and choose only one of them as being primary.

The Analogue to Digital conversion is done by the 4-channel AKM AK554VN chip that is able to provide a THD+N of -106 dB, pretty close to the -104 dB Amir measured on the Line-In 3/4 inputs here: https://www.audiosciencereview.com/forum/index.php?threads/motu-m4-audio-interface-review.15757/.


IMG_8873 copy_.jpg

AKM AK554VN ADC chip

Internal 40MHz quartz clock from around the AKM4554 ADC chip is an American brand: ILSI; not sure these are entirely manufactured in US, but based on https://abracon.com/about they do have headquarter and facilities in Spicewood, TX. I was able to find couple of datasheets too, just not the exact match type for the TCXO installed inside the M4: https://abracon.com/datasheets/ILSI/I533.pdf and https://www.tme.com/Document/5876ac880aecd10660839d4d7584f0de/I547_I747_Series.pdf. Probably the 24MHz TCXO from nearby the XMOS microcontroller is made by ILSI too.​


IMG_0614.jpg

ILSI TCXO nearby the ESS DAC and AKM ADC chips

Analogue inputs from the front panel are buffered by couple of OPA1678 op amps and the gain is adjusted by the dedicated THS4522 and THAT6263 chips. I am guessing that the first one acts as an input buffer and is having a fixed gain, while the second is changing the gain in 3dB, per manufacturer datasheet, although in real-life M4 mic inputs are fine adjusted by 1dB and not by 3dB. Perhaps someone else could jump in here with some thoughts about how the above two chips might be used inside the M2 & M4.


IMG_0613.jpg

THAT6263 dedicated microphone pre-amplifier

Motu M4 is a good looking USB audio interface with a very low latency (<5ms is using 128Kbs buffer) and drivers and control interface are robust under Windows and work flawless, although under MacOS I wasn’t able to find any control panel to do some more settings to the M4. I’m probably spoiled by the Focusrite control panel where several mixing and audio rerouting can be done with ease, but Motu comes in handy with its “loopback audio recording” so everyone can record the audio track that is playing at the same moment (e.g.: recording audio track while playing Tidal or while having a Whatsup call with a friend etc.).

The backside 3/4 balanced Line In TRS plugs are bypassing the input gain controls and can be used for audio measurements with a THD+N of -106 dB, per Amirm’s measurements. However, if you’re not comfortable with the fixed input level where 0 dBFS is a bit over 6 V RMS, then you could do your audio measurements on the 1/2 gain adjustable inputs, but increasing the gain will increase the noise a bit, making the M4 more or less an option for sensitive audio measurements, depending on your needs.

Despite all the THD measurements done to audio interfaces, home microphone recordings done with different modern audio interfaces will usually measure and sound very similar, because today audio interfaces are very capable, while most microphones are having a THD up to 1% @94 dB-SPL. Basically, regarding the THD of the final track, I consider the limiting factor being the microphone itself and not the actual audio interface, due to the transducer (https://www.shure.com/es-CL/desempeno-y-produccion/louder/mic-basics-transducers) and the diaphragm inside. So switching from one interface to another might be considered an upgrade if the EIN of the interface is lower (that means lower noise preamps), more inputs or the internal gain is higher, to better accommodate low sensitivity mics, without worrying much about the THD of the audio interface, which usually is way above the capabilities of the microphone itself. A great reading would be: https://www.neumann.com/homestudio/en/will-a-better-preamp-give-you-lower-noise.


60Hz_SE_X1_S_-1dB_THD_MotuM4_Mic1.png

SE X1S connected to Motu M4 @60Hz sinewave

60Hz_SE_X1_S_-1dB_THD_Solo3_Mic1.png

SE X1S connected to Focusrite Solo Gen3 @60Hz sinewave

Sweep_ECM999_THD_MotuM4_Mic1.png

ECM999 mic connected to M4 - THD sweep

Sweep_ECM999_THD_Solo3_Mic1.png

ECM999 mic connected to Solo Gen3 - THD sweep
As we can se from the above two screenshots, using the same microphone to two different audio interfaces, made by different manufacturers too, didn't changed much the harmonic profile nor the final THD+N, although track's recorded background noise might be different in the final mix, depending on the EIN of each interface.

Motu M4 has a EIN of -129 dBu at maximum gain, measured with a 150Ω resistor, A-weighted . Testing the M4 on my high impedance 600 Ohms and rather low sensitivity AKG D5 S dynamic mic (2.6 mV/Pa or -51.7 dBV/94 dB-SPL) proved a low noise recording and the +60 dB of internal gain was pretty much OK, even for low talk during the night.

Moving on the SE X1S condenser mic, which is a very sensitive one (30 mV/Pa or -30.5 dBV/94 dB-SPL), I didn’t noticed any significant background noise and even the tiniest move of the lips was clearly recorded on the track.

In case you’re wondering if the 150 Ohms 1.12 mV (-59 dBV/94 dB-SPL) sensitivity SHURE SM7B dynamic mic will be a good match for the +60 dB of gain from the M4, I don’t have one SM7B to test it right now, but if you’re a low talker then you will probably be needing an interface having at least +65 dB of gain or add an in-line or an external pre-amp to gain few more extra dB. However, in this thread https://www.audiosciencereview.com/forum/index.php?threads/which-preamp-for-sm7b.14028/#post-427838 there’s lot of talking about the SM7B/M4 compatibility and seems that under normal conditions the match between the SM7B and M4 should be a decent one.

Worth mentioning that with 1V or 2V analogue signals the M4 will have a THD+N worse than the one measured by Amir at Full Scale on both DAC-out and ADC-in.

Solo-to-MotuM4 - 1V RMS.png

Solo Gen3 outputting 1V RMS into M4

ARTA 1V RMS Mic1 input Monitor 1 Output.png

M4 outputting 1V RMS into M4

We can easily spot the added noise shape when lowering the DAC volume on the Motu M4. Seems that the analogue volume control in Solo Gen3 handles better this noise. Basically, M4 has this issue only when volume knob is between 10:30 and 1:30 o'clock, otherwise the noise gets lower.

Overall, I find the Motu M4 being a great interface, it has a good amount of gain, a proved low noise when fed with both dynamic and condenser mics and is capable of a clean audio recording and playback.
 
Last edited:

Blumlein 88

Major Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
10,722
Likes
15,256
#2
Amazing the level of performance and usefulness packed into a package this small and for so little money.

Even things like Neve recording consoles and such don't have specs this good. They have many facets and features useful as a a studio recording console, but top quality fidelity isn't part of the package. When you have something as good as the M4 your recordings aren't limited by the device in any meaningful way. It will always be the microphones, the recording venue and the musicians.
 

Χ Ξ Σ

Active Member
Joined
Jul 16, 2020
Messages
146
Likes
414
Location
Bay Area
#3
Wow, this is amazing! As a Motu M2 owner, I am happy to see such a technical evaluation of its bigger sibling. Thanks for sharing your finding of the distortion when the headphones knob passes at 3 o'clock. I have never turned the knob passes 12 so I would have never known. As for the added noise when lowering the DAC volume, the area between 10:30 to 1:30 is unfortunately where I listen to high dynamic classical music. I suppose it does not matter in practice because my active speakers and their built-in amps have higher noise than the DAC.

I am experiencing a high delay when the Motu M2 is changing the sampling rate. It is probably the limitation of being a USB-powered device, but can you shed some light on this issue?
 
Last edited:

q3cpma

Major Contributor
Joined
May 22, 2019
Messages
2,048
Likes
2,492
Location
France
#4
Thanks for this detailed writeup. Just one or two questions:
* Do you know if the volume is reset at each power cycle in UAC mode? This annoys me quite a lot on my id4.
* Guess I'll get the best out of it by setting the minimum input sensitivity on my Genelecs, right?
 
OP
trl

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,317
Likes
1,331
Location
Iasi, RO
Thread Starter #5
* Do you know if the volume is reset at each power cycle in UAC mode? This annoys me quite a lot on my id4.
Not sure I get it, sorry. Is the volume on ID4 rotating 360 degrees, without stopping? If so, then it's a different way of setting the volume than it's on Motu M2/M4. With Motu the volume knob acts like an analogue one, so the volume will always be the one that is shown by the white marking from the knob.

* Guess I'll get the best out of it by setting the minimum input sensitivity on my Genelecs, right?
I did about the same with my Mackie's too, not quite minimum, but pretty close to.
 

q3cpma

Major Contributor
Joined
May 22, 2019
Messages
2,048
Likes
2,492
Location
France
#6
Not sure I get it, sorry. Is the volume on ID4 rotating 360 degrees, without stopping?
Yes. And it's 3 dB per step =(.
If so, then it's a different way of setting the volume than it's on Motu M2/M4. With Motu the volume knob acts like an analogue one, so the volume will always be the one that is shown by the white marking from the knob.
Thanks, I thought this was a downside of all digital coders (as they count steps, not really the absolute position, right?).
I did about the same with my Mackie's too, not quite minimum, but pretty close to.
I see.

Guess I'll get myself a M2 soon enough, seems like a good endgame analogue card if I just want to use it as a balanced DAC + measurement mic input. Only thing I prefer on the Audient is the controls being on top, like a Babyface.
 

bwg

New Member
Joined
Nov 25, 2020
Messages
3
Likes
0
Location
Europe
#7
Great teardown, thanks @trl!

I bought an M4 recently and ran into a ground loop issue when using the MIDI In port. I can measure a dc path from pin 2 of the MIDI In port to gnd, which shouldn't be present according to the MIDI standard. To make my setup work without annoying noises, I had to remove pin 2 of a MIDI cable. Did anyone else run into similar problems with the device?
 
OP
trl

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,317
Likes
1,331
Location
Iasi, RO
Thread Starter #8
Thanks, I thought this was a downside of all digital coders (as they count steps, not really the absolute position, right?).
Motu M2/M4 are reading the absolute position, like the Clarett and others.
 
OP
trl

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,317
Likes
1,331
Location
Iasi, RO
Thread Starter #9
I can measure a dc path from pin 2 of the MIDI In port to gnd, which shouldn't be present according to the MIDI standard.

I see on the above pic that Pin# 2 shouldn't be connected at all, so I wonder if this DC-voltage may come downstream from your MIDI connected device. Or this DC-voltage is there by default, without any cables connected to the M4? What voltage can you measure from Pin# 2 to shield/GND?
 

bwg

New Member
Joined
Nov 25, 2020
Messages
3
Likes
0
Location
Europe
#10
I see on the above pic that Pin# 2 shouldn't be connected at all, so I wonder if this DC-voltage may come downstream from your MIDI connected device. Or this DC-voltage is there by default, without any cables connected to the M4? What voltage can you measure from Pin# 2 to shield/GND?
With the device unpowered and not connected to anything, I measured resistance between Pin 2 of the MIDI In port and GND (I used the outside part of one of the RCA connectors) which is zero. According to
this it should be infinite I think. If I understand this correctly, Pin 2 may be connected to GND at a MIDI Out, but must never be connected at the MIDI In port. If there is a connection, the optical isolation is kind of pointless, I guess...
 
OP
trl

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,317
Likes
1,331
Location
Iasi, RO
Thread Starter #11
the optical isolation is kind of pointless, I guess
I think this protection is for the input balanced signal only.
So, without anything connected to M4, but with M4 powered ON, can you still measure an output voltage between pin-2 and GND on the Midin Input DIN plug?
 
OP
trl

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,317
Likes
1,331
Location
Iasi, RO
Thread Starter #12
Adding below the way the Superlux ECM999 measurements microphone was positioned to record with Motu M4 and Focusrite Solo Gen3 the above four graphs. Speaker used was Mackie MR6mk3 and volume was setup for 90 dB SPL @ 1 kHz

IMG_0606.jpg

A similar setup was used to perform measurements with the SE Electronics X1 S microphone too.

Worth mentioning that I was choosing the SE X1 S for the 50 Hz THD graphs due to the higher distortions provided by this microphone vs. the ECM 999, so easier to visualize any possible differences between the two audio interfaces too.
 

bwg

New Member
Joined
Nov 25, 2020
Messages
3
Likes
0
Location
Europe
#13
I think this protection is for the input balanced signal only.
So, without anything connected to M4, but with M4 powered ON, can you still measure an output voltage between pin-2 and GND on the Midin Input DIN plug?
With the device powered on, the voltage between pin 2 and gnd is zero as expected. The problem is that there is a low resistance connection which should not be there according to this schematic. That's what I find so weird.
 

AnalogSteph

Addicted to Fun and Learning
Joined
Nov 6, 2018
Messages
926
Likes
746
Location
.de
#14
A minor correction:
Despite all the THD measurements done to audio interfaces, home microphone recordings done with different modern audio interfaces will usually measure and sound very similar, because today audio interfaces are very capable, while most microphones are having a THD up to 1% @94 dB-SPL.
Defo not.

Have you seen the max SPL ratings for some mics? Yeah, those are given at 1% or even 0.5% THD, and in general are way higher than 94 dB. The literally cheapest XLR side address condenser you can buy at Thomann for the princely sum of 28€ right now, the t.bone SC300 complete with its luxurious plastic body, is rated for 122 dB SPL @ 1 %, 1 kHz. That's about the lowest rating I could find. (Given the nominal -35 dBV/Pa rating, that's -5 dBu of voltage output.) If you were to get much fancier with e.g. a Sennheiser MKH4, you could get 140 dB max, or 138 dB @ 0.5% with a Neumann TLM 103.

Dynamic mics, being all passive, can reach 150+ dB SPL.

About the only way of falling way short of these regions is using a cheesy electret capsule with common source JFET circuit. Even so, an unmodified Panasonic WM-61 capsule seems to have been able to reach around 115 dB SPL @ 1% distortion in the midrange... with distortion being dominant 2nd order (it's a single-ended JFET circuit, what do you expect), that would be around 0.1% at 94 dB. Add about 10 dB for a Linkwitz modified capsule, and another 10 dB for a B&K measurement mic, apparently.
 
OP
trl

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,317
Likes
1,331
Location
Iasi, RO
Thread Starter #15
Not sure what do you mean, sorry. I was only referring to THD+N and to "home microphone recordings done with different modern audio interfaces" and the two mics tested above at about 90 dB SPL were measuring pretty similar on two different audio interfaces. Not sure what SPL-rating of the mics has to do with these measurements, given that the test was done at around 90 dB SPL.

If you are referring to dynamic, I can't test the final dynamic of the recorded track on mic's highest SPL rating, because I don't have a place nor an audio source to apply 130...150 dB SPL, nor such a quiet room to fully test my SE X1S that has an internal noise of only 9 dB (my room's background has >20dB). So the theoretically dynamic of a performant microphone could probably pass the 120 dB on the paper, but I haven't ever seen anyone being able to fully test this dynamic written on the specs, although I doubt that the final THD will be pleasant when approaching max. SPL (simply from a measurement perspective, given that we are on ASR forum here). Probably in ideal testing conditions the final dynamic of the recorded tracks will differ from one audio interface to another, given that EIN is different too, but the "sound flavour" (harmonic profile + impulse response + freq. response) shouldn't get changed, unless some interface is either faulty either really poor by design.

But hey, this is just my 2 cents, I'm sure others can contribute better on this subject and create a dedicated thread regarding to how microphones are measuring.
 

AnalogSteph

Addicted to Fun and Learning
Joined
Nov 6, 2018
Messages
926
Likes
746
Location
.de
#16
Not sure what do you mean, sorry. I was only referring to THD+N and to "home microphone recordings done with different modern audio interfaces" and the two mics tested above at about 90 dB SPL were measuring pretty similar on two different audio interfaces. Not sure what SPL-rating of the mics has to do with these measurements, given that the test was done at around 90 dB SPL.
I was just taking objection with the "up to 1% at 94 dB SPL" claim, which is clearly not true, even +N. That would be N+D at +54 dB SPL. (What's seen in the measurement is almost all speaker distortion.)
Now even if a mic has self-noise in the 40s A-weighted (which does happen with small electret capsules, as used in some measurement mics), you'll generally be thinking "man, this sucker is noisy". You'll generally be seeing self-noise around 30 dB SPL(A) or better, at times much better as you've noticed.
Mics also tend to have very predictable, well-behaved distortion (well described by intercept points for each harmonic), so with some assumptions you can find a decent approximation for the distortion at any level if the SPL limit and corresponding distortion level are known. The max SPL spec at generally either 1% or 0.5% is one such point. Below this level, distortion is decreasing monotonically and hardly ever slower than 2nd-order - 1 dB per 1 dB, or IOW it's dominant H2. Using this extrapolation, estimated 94 dB SPL distortion levels come out to values ranging from 0.04% (+26 dB SPL) for the lowly SC300 to 0.003% (+4 dB SPL) for the TLM103, with the ECM-999 (max SPL = 132 dB) in the middle at 0.0125% (+16 dB SPL).

So while yes, electronics tend to have lower distortion than mics, it's not quite as cut and dry as you might think. Especially with the M2/M4 inputs and their substantially rising 3rd order HD/IMD towards the high frequencies, while condenser mics at normal volumes tend to be almost all (benign) H2 with very little H3 (even if H3 were to dominate with 1% @ 132 dB SPL, by 94 dB SPL it would be down to -40 dB - 76 dB = -116 dB = 0.00016%). Low-power, complex high-feedback circuits also have a tendency of not being "well-behaved", to the point of virtually constant THD over a substantial level range. Mic circuitry is generally low complexity, low feedback instead, which is why it tends to be as "well-behaved" as indicated.
 
Last edited:
OP
trl

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,317
Likes
1,331
Location
Iasi, RO
Thread Starter #17
I was just taking objection with the "up to 1% at 94 dB SPL" claim, which is clearly not true, even +N.
I see your point now, thank you for clearing this up!

60Hz_ECM999_-1dB_THD_MotuM4_Mic1.png

ECM999 - 60 Hz
60Hz_SE_X1_S_-1dB_THD_MotuM4_Mic1.png

SE X1 S - 60 Hz


10KHz_ECM999_-1dB_THD_MotuM4_Mic1.png.png

ECM999 - 10 kHz

10KHz_SE_X1_S_-1dB_THD_MotuM4_Mic1.png.png

SE X1 S - 10 kHz
In the above pics @60Hz the ECM999 does a much better job than SE X1 S regarding the THD+N, 0.87% vs. 2.03%, same speakers and arm position, same M4.

On 10KHz things are vice-versa, SE X1 S having a lower THD+N than ECM999, 0.14% vs. 0.28%. For 1 kHz SE X1 S is also a bit better than ECM999 too, so not sure how much is the THD of the speakers and how much of the microphone, but positioning was the same for both mics. I will probably try again to place both mics nearby, in front of the same speaker, and redo the above measurements.
 

AnalogSteph

Addicted to Fun and Learning
Joined
Nov 6, 2018
Messages
926
Likes
746
Location
.de
#18
In the above pics @60Hz the ECM999 does a much better job than SE X1 S regarding the THD+N, 0.87% vs. 2.03%, same speakers and arm position, same M4.
Except the X1 S was measured at 50 Hz. :p:facepalm:
On 10KHz things are vice-versa, SE X1 S having a lower THD+N than ECM999, 0.14% vs. 0.28%.
I think this is virtually all owed to +N at this point, I don't even see any 2nd harmonic @ 20 kHz. You would expect the large diaphragm SE mic to perform better than the ECM-999 with its (IIRC) 16 mm capsule in this regard, of course. You'd better switch to twin-tone IMD for frequencies this high, otherwise the otherwise irrelevant mic response >20 kHz as well as potentially imprecise mic axis alignment will give you a lot of grief.
 
OP
trl

trl

Major Contributor
King of Mods
Joined
Feb 28, 2018
Messages
1,317
Likes
1,331
Location
Iasi, RO
Thread Starter #19
Except the X1 S was measured at 50 Hz. :p:facepalm:
Yes, you're right, thank for noticing this. However, there's the same THD+N on both 50 Hz and 60 Hz measurements. There's no way for me to edit that post and change the pic now, so I will leave it as it is.
You'd better switch to twin-tone IMD for frequencies this high
I did the 60 Hz + 8000 Hz, but also 50 Hz + 7000 Hz measurements and ECM999 performs a bit better on the low end too, but I haven't posted here the pics due to a bit off-topic. If you think it worth starting a thread in comparing the mics, then please let me know and I'll start one. Thanks!
 

AnalogSteph

Addicted to Fun and Learning
Joined
Nov 6, 2018
Messages
926
Likes
746
Location
.de
#20
Sure, why not. In terms of twin-tone, I was actually thinking more like 10 + 11 kHz or similar, though if your source is a speaker with multiple ways splitting the test tones accordingly would not bad a bad idea... maybe 1 + 10 kHz or so.
 
Top Bottom