How to measure EIN (Equivalent Input Noise) of an interface's Mic preamp

[nanook]

So EIN and max. undistorted input level plotted vs. gain would give the dynamic range vs. gain and represent the interesting parameter in the real world?

 

[Rja4000]

So EIN and max. undistorted input level plotted vs. gain would give the dynamic range vs. gain and represent the interesting parameter in the real world?

Both EIN and DR vs gain are interesting values, IMO.

With EIN vs gain, you learn where the sweet spot is between safe overload margin and noise benefit (when plot is flat, there is nothing to save in terms of noise by pushing the gain any more), in function of the level you expect out of your microphone.

With DR, you'll see the performance for high levels, like condenser mikes recording a high dynamic range Orchestra.

Typically, high end mic preamp will allow higher dynamic range at high levels.
Much higher, sometimes.

 

[Blumlein 88]

Actually, he's explaining the same I'm trying to explain.
And his bargraphs are similar to my sensitivity graphs, except he's just looking at the max gain value, and doesn't say anything about overall max value, which I also include.

This

View attachment 388991
Is the same than this
View attachment 388993
but without the left value (which I think is also interesting)

Yes I agree he is explaining the same thing. I prefer your chart to his. The point about this in regard to your post and AnalogSteph's posting is that you don't quite get the info you need with a simple number like only EIN or by testing with only a single input level or two. With the sensitivity chart and EIN vs gain you have what you need. If really splitting hairs you would also want SNR at 0 db gain.

So we are somewhat back to what is the simplest and fewest useful tests that Amir could perform when he tests interfaces or that we can do if we want to contribute useful information for others? I think the resulting info we need would be the sensitivity chart above and the chart of EIN vs gain setting.

You'll need to measure max input at 0 db gain. You might as well get the SNR or noise floor in dbFS while you are doing this.
You'll need to measure EIN at max gain and I think at least every 20 db below that.

So any criticisms or additions you think needed? Or something I'm missing?

Of course it looks like other than a few uncommon poor designs the makers of this gear have it pretty well covered. So perhaps to weed out only bad examples while giving less info Max gain at 0 db, EIN at max gain, and EIN at 30 db less than max gain will do about as well.

 

[Rja4000]

That Yamaha AD8HR must have a very funky gain structure for its curve to look like that. It almost looks like input amplifier gain is switched in ~12 dB steps (or perhaps it's actually an input attenuator? Almost makes more sense), followed by another stage with much finer ones. Quirky. The theory of it using an input attenuator would be supported by the fact that the (mic) input can accept levels up to +30 dBu when mic input amplifiers tend to have a minimum gain of 0 to +10 dB (I very much doubt that the unit would be running internal levels of +30 to +40 dBu). The noise penalty of this approach is quite obvious.

Found this on the maintenance manual
RY101 and RY102 are relays.

 

[Rja4000]

Yes I agree he is explaining the same thing. I prefer your chart to his. The point about this in regard to your post and AnalogSteph's posting is that you don't quite get the info you need with a simple number like only EIN or by testing with only a single input level or two. With the sensitivity chart and EIN vs gain you have what you need. If really splitting hairs you would also want SNR at 0 db gain.

So we are somewhat back to what is the simplest and fewest useful tests that Amir could perform when he tests interfaces or that we can do if we want to contribute useful information for others? I think the resulting info we need would be the sensitivity chart above and the chart of EIN vs gain setting.

You'll need to measure max input at 0 db gain. You might as well get the SNR or noise floor in dbFS while you are doing this.
You'll need to measure EIN at max gain and I think at least every 20 db below that.

So any criticisms or additions you think needed? Or something I'm missing?

Of course it looks like other than a few uncommon poor designs the makers of this gear have it pretty well covered. So perhaps to weed out only bad examples while giving less info Max gain at 0 db, EIN at max gain, and EIN at 30 db less than max gain will do about as well.

What I suggest is that we measure FS level (in dBu) and noise (in dBFS) for 4 gains:
Max gain and Min gain
And 2 fixed level gains, like the gain giving you (the closest to) 10mV and 100mV for FS.
(I now think adding 1V would also make sense, though)
I suggest we do it at least shorted and unweighted, since we can still deduct the 150 ohm load from theory.
CCIR-2k and A-weighted being optional.

Doing it shorted will also directly give you the preamp's own noise and will probably be more discriminant.

With the above, you also have your dynamic range for those Gains (that's your Noise measurement in dBFS)

Note that preamp having a coarse gain setting may be penalized a little (or the opposite) vs preamp with 1dB gain steps.

And if we want a single value for ranking (?), we may still use the 10mV one, with or without weighting.

 

[nanook]

What I suggest is that we measure FS level (in dBu) and noise (in dBDS) for 4 gains:

This will allow to choose a gain that is a good compromise between risk of overload and noise.

I suggest we do it at least shorted and unweighted, since we can still deduct the 150 ohm load from theory.

I agree, shorted will separate the really quiet ones from the good ones and the current noise contribution at 150 Ohm should be negligible for most input stages.

I suggest we do it at least shorted and unweighted,

In case the noise is mostly flat above a few kHz (that's my expectation), the A-weighting will make an almost fix difference in dB.
1/f noise will anyways not be visible in the EIN unless there's something really wrong.

Going for noise densities (as promoted by me earlier) is less important compared to covering noise (EIN) and max. level vs. a couple of gain settings.

 

[Blumlein 88]

I think the noise of these are going to be close to white. Here is one for the Babyface. This is shorted with max gain of 65 db. I think the worst of those I had on hand had a very slight tilt up at 400 hz and lower. You can see including below 20 hz only increased the noise level by .1 dbFS.

This is the same thing showing 0-20 khz in log scale where you can see noise increases a bit at maybe 40 hz and lower.

This is the same info with 150 ohm resistor input. 20-20,000 hz.

0-20,000 hz

If you are wondering about the straight line at low frequencies it is from using 64k FFT and 5 averages. Here is a 4 meg FFT and 8 averages. The ADC is AC coupled of course. Looks like a corner frequency of maybe 2 hz. This is with 150 ohm resistor.

 

[nanook]

Thanks, this measurement confirms my expectation that there may be an increase of the noise density towards lower frequencies, but that it's white in the audio band otherwise.
So we should even be ok when we measure un-weighted and give a calculated A-weighted value.
(or is my understanding wrong, that for a white noise distribution noise in dB vs. dBA differs by a fix number?)

 

[Blumlein 88]

Thanks, this measurement confirms my expectation that there may be an increase of the noise density towards lower frequencies, but that it's white in the audio band otherwise.
So we should even be ok when we measure un-weighted and give a calculated A-weighted value.
(or is my understanding wrong, that for a white noise distribution noise in dB vs. dBA differs by a fix number?)

Well I can report what was measured. If you look at the screenshots the dba is to the right of the db level. In the case of this interface A-wtd is 2 db better. The other interfaces were in that range. 2.0 to 2.5 db. The difference didn't vary with gain level.

 

[Blumlein 88]

So I'm still puzzling over the least number of things we can test to tell us what we want to know about microphone preamps practically speaking. Here is my current thinking.

A Shure SM7b is about the least sensitive microphone you run across. Most modern ribbons are more sensitive. It puts out 1.12 mV at 94 dbSPL at 1 khz. Shure suggests you need 60 db of gain minimum to make it work.

The more sensitive LDC microphones are 28 or 29 mV at 94 dbSPL at 1 khz. Let us call that a 30 to 1 ratio roughly. Which is also roughly 30 db vs an SM7b. So if we know how things work at max gain and 30 db less than max gain we have most microphones covered.

Test #1
We need to test for max input at minimum gain. You can do that with a wide range of voltages. I suggest .775 volts as that is 0 dbu. You can read it straight off the dbFS graph to see what is the max dbu input level. That and the gain range is enough for this part.

Test #2
EIN with 150 ohm resistor at max gain and 30 db less than max gain.

Test #3
EIN at max gain with a short.

(Do we need noise level at minimum gain with a short?)

Would anything be missing to tell us most of what we want to know for practical purposes?

I'm hoping to make this simple enough Amir will do this when he tests audio interfaces. Generally speaking the majority of audio interfaces use the same microphone circuit padded down for line level. Amir already tests those for distortion, frequency response and dynamic range.

So what do you guys think? Will these three tests do the job?

@Rja4000 @AnalogSteph @nanook @KSTR @restorer-john @SIY

 

[Rja4000]

A Shure SM7b is about the least sensitive microphone you run across. Most modern ribbons are more sensitive. It puts out 1.12 mV at 94 dbSPL at 1 khz. Shure suggests you need 60 db of gain minimum to make it work.

There are less sensitive ones (like the Audix OM7 - an excellent mic - with 0.8mV/Pa), but what really makes the SM7b a challenge is that it's often used for spoken voice, and, even, from some distance (10-15cm).
The cumul of those conditions explains why it's considered such a critical mic.

 

[Blumlein 88]

There are less sensitive ones (like the Audix OM7 - an excellent mic - with 0.8mV/Pa), but what really makes the SM7b a challenge is that it's often used for spoken voice, and, even, from some a distance (10-15cm).
The cumul of those conditions explains why it's considered such a critical mic.

Yes, but it is only about 3 db less sensitive than the SM7b. The ratio between it and 29 mV is only 32 db so not greatly different than 30 db between them. The fact OM 7s are typically held with hands wrapped around them very close to the mouth more than makes up for that 3 db.

 

[Rja4000]

A Shure SM7b is about the least sensitive microphone you run across. Most modern ribbons are more sensitive. It puts out 1.12 mV at 94 dbSPL at 1 khz. Shure suggests you need 60 db of gain minimum to make it work.

The more sensitive LDC microphones are 28 or 29 mV at 94 dbSPL at 1 khz. Let us call that a 30 to 1 ratio roughly. Which is also roughly 30 db vs an SM7b. So if we know how things work at max gain and 30 db less than max gain we have most microphones covered.

You're right about sensitivity, but I'm pretty sure the range we need to cover is wider than that:
You also have to take into account the SPL of the signal reaching the mic membrane.
Between a SM7b used by a vlogger and a LDC used to record a nervous drum overheads (think RHCP), there is, in my opinion, more than 30dB.

Test #1
We need to test for max input at minimum gain. You can do that with a wide range of voltages. I suggest .775 volts as that is 0 dbu. You can read it straight off the dbFS graph to see what is the max dbu input level. That and the gain range is enough for this part.

I suggested 1V above.
But 0dBu or 4dBu also work.

Test #2
EIN with 150 ohm resistor at max gain and 30 db less than max gain.

Sure.
But then I'd go for 40dB rather than 30dB, and I'd rather chose a fixed FS voltage, like 100mV.

Test #3
EIN at max gain with a short.

Sure

(Do we need noise level at minimum gain with a short?)

Yes, or AES-17 DR, which is probably better.
Than, load would be the AP's output impedance (which should be specified)

Would anything be missing to tell us most of what we want to know for practical purposes?

I still think EIN for 4 or 5 gains makes more sense.

I'm hoping to make this simple enough Amir will do this when he tests audio interfaces.

That's the goal, indeed

Generally speaking the majority of audio interfaces use the same microphone circuit padded down for line level. Amir already tests those for distortion, frequency response and dynamic range.

If only one input is tested, it has to be the Mic input, without pad, I agree. That's clear

So what do you guys think? Will these three tests do the job?

@Rja4000 @AnalogSteph @nanook @KSTR @restorer-john @SIY

 

[SIY]

Test #1
We need to test for max input at minimum gain. You can do that with a wide range of voltages. I suggest .775 volts as that is 0 dbu. You can read it straight off the dbFS graph to see what is the max dbu input level. That and the gain range is enough for this part.

Test #2
EIN with 150 ohm resistor at max gain and 30 db less than max gain.



So what do you guys think? Will these three tests do the job?

@Rja4000 @AnalogSteph @nanook @KSTR @restorer-john @SIY

They will as long as you have a source impedance not much bigger than 150R. If you do, current noise that may be lost at 150R might be significant at, say, 50k.

Re: mic dynamic ranges, my 1cm diameter lab condenser mikes run as high as 50mV/Pa. My ribbon mikes are more like 2mV/Pa, as are my high-SPL 1/8" condenser mics, so there's a good 28dB spread right there.

 

[Blumlein 88]

They will as long as you have a source impedance not much bigger than 150R. If you do, current noise that may be lost at 150R might be significant at, say, 50k.

Re: mic dynamic ranges, my 1cm diameter lab condenser mikes run as high as 50mV/Pa. My ribbon mikes are more like 2mV/Pa, as are my high-SPL 1/8" condenser mics, so there's a good 28dB spread right there.

Are there mics with 50 k output impedance? Most common is 50 ohm to 200 ohm. I've seen a few at 300. I think I recall one at 600 ohms, and an old ribbon that was 1000 ohms.

I do note some expensive stand alone analog mic preamps specifically for ribbons have input impedances of 47k ohm to 100k ohm.

 

[SIY]

Are there mics with 50 k output impedance? Most common is 50 ohm to 200 ohm.

I don't know, there may or may not be, but there's certainly guitar pickups like that, so there well could be some dynamic...

If everything really is 150R or below, the short circuit input noise should be enough since the current noise would have to be brutal to be significant with impedances that low.

 

[Blumlein 88]

You're right about sensitivity, but I'm pretty sure the range we need to cover is wider than that:
You also have to take into account the SPL of the signal reaching the mic membrane.
Between a SM7b used by a vlogger and a LDC used to record a nervous drum overheads (think RHCP), there is, in my opinion, more than 30dB.

I suggested 1V above.
But 0dBu or 4dBu also work.

Sure.
But then I'd go for 40dB rather than 30dB, and I'd rather chose a fixed FS voltage, like 100mV.


Sure

Yes, or AES-17 DR, which is probably better.
Than, load would be the AP's output impedance (which should be specified)

I still think EIN for 4 or 5 gains makes more sense.

That's the goal, indeed

If only one input is tested, it has to be the Mic input, without pad, I agree. That's clear

There is wider range than 30 db between podcasting and drum overheads, but even with an SM7b or OM7 you won't be using maximum gain with drum overheads. Another common use is miking a guitar cabinet. Also you are not using max gain for those.

Regarding using -30 db or something else or a fixed voltage like 100 mV, which I can see the attraction for, I looked at what most interfaces are. On average you see 60 db gain range, 8 dbu max input. Using 100 mV (-17.8 dbu) you would end up with gain set at -25.8 db so -30 db gain I suggested is a bit better for testing. With some interfaces with 75 db gain range you usually have higher input max levels. The RMEs are max 18 dbu so 100 mV would be - 35.8 from max gain. I had one interface that was max input 12 dbu and 75 db gain. That one would be -29.8 db from max gain for 100 mV. The Arrow interface has 1.5 dbu max input so with 100 mV you would be at -19.3 db from max gain so again -30 db might be a better test. A Metric Halo has 90 db gain range with 20 dbu max input. 100 mV would be at -37.8 db. If one preferred using a fixed level for each interface, I would suggest -20 dbu (77.5 mV) as it probably is in a slightly better direction for most interfaces.

While I would agree using 4 or 5 different test levels would be better I don't think Amir will test that many. I would prefer EIN done every 5 db myself. But I'm trying to get a consensus on a minimal number of tests that are useful in say 90% of the cases. I think the three or four tests or something similar to them would allow someone wanting to know if a given interface works for them to figure that out.

My apologies to Rja4000, maybe I've hijacked your thread for a different purpose. Didn't intend that. Maybe this belongs in the thread Amir had asking for what to test for. That turned into the Tower of Babel for mic pre testing. As a result Amir hasn't indicated any additional testing of interfaces on the preamp side.

 

[AnalogSteph]

Are there mics with 50 k output impedance? Most common is 50 ohm to 200 ohm. I've seen a few at 300. I think I recall one at 600 ohms, and an old ribbon that was 1000 ohms.

Lots of cheapie dynamics use 600 ohm capsules. That's about the highest you're commonly going to encounter, however, explaining why current noise tends to be a secondary concern unless you're leaving the input open. I have seen preceding pad circuits using resistors in the low kOhms though.

On the opposite end, there's some dynamic mics around 50 ohms... most of them very old, but I think you can still buy at least one model like that.

The choice of 600 ohms for cheapie dynamics seems to be for achieving the required sensitivity in spite of lower inherent driver sensitivity. They generally aren't expected to be driving super long cables either (mostly being for personal use, home studios or karaoke), unlike what an SM58 on a festival stage might be seeing.

 

[Sokel]

Ok,I did E-MU per your guide but can't find gain in specs:


Shorted


150 Ohm


Specs

 

[Blumlein 88]

Ok,I did E-MU per your guide but can't find gain in specs:

View attachment 389544
Shorted

View attachment 389545
150 Ohm

View attachment 389546
Specs

If the EIN spec is accurate gain must be about 55 db. You could loopback a signal you know and reduce it digitally by say -70 db and figure out the gain by running it thru at 0 and max gain.