• 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.

Audio Interface Suitable for Measurements

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
6,638
Likes
2,685
Location
Riverview, FL
#1
Ok, my birthday has come and gone.

Nobody bought me anything so I'll have to do it myself.

I bought a little Behringer UMC202HD a couple of years ago.

Here's its ADC/DAC loop-back spectrum. Some of the spikes are the Plasma TV, but still...

1557763171484.png


Meh.

I know some other devices have been recommended in other threads, but please feel free to repeat yourself here.

At least two inputs - line, instrument, phantom power, mic amp, the usual stuff - and USB out should be sufficient. Four channels would be a preference, I think.

Probably prefer a self-powered device (wall wart, whatever).

Post a picture of its performance if you have one handy.
 
Last edited:

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
6,638
Likes
2,685
Location
Riverview, FL
#2
Nobody?


What brought this on:


I can forgive a little noise floor, or work around it, but....

I wanted to see the harmonics coming from the amplifier, if there are any to see, since I hadn't done that before.

The output level is low, a volt or less, and not overdriving the ADC.

So, looking at the amp, I see (top graph) second harmonic at about -25dB to the signal tone.

I think "Whoa... is this why I like my amps? They make a fat 2nd?" but only momentarily, since I'd not seen that in-room.

To confirm, measure the in-room with the microphone, no second harmonic to be seen down to the noise area at -65 or so relative to the signal.

The fifth is added to the test signal, at -50dB, as a cross-check, and is visible at the appropriate level at both the amplifier and in the room.

Therefore, I reason the ADC is adding (at least) the 2nd harmonic, since it isn't in the room, and would be easily visible if it were, based on the measurement of the manually added 5th. and makes my electrical measurement rather useless.

1557786917039.png
 

SIY

Major Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
1,906
Likes
2,878
Location
Phoenix, AZ
#3
I've had decent results with the Focusrite Scarlett. Mine is the 2i2, but they have a four channel version as well.
 

Blumlein 88

Major Contributor
Joined
Feb 23, 2016
Messages
5,452
Likes
4,243
#4
It depends upon what you wish to do with it.

I've pretty good results for the Focusrite Forte which is better than the Scarlett series. Though the Scarlett would let you do pretty well. I have an 18i20 Scarlett. It is spec'd a few db better than something like the 2i2 or 2i4.

You can only get Forte's second hand. Probably can get one for $200. It has its own power supply, but you only need it for phantom power. Just doing line level measures you can power it from the wall or over USB. It is only 2 channels. I could do some REW loops of either for you if you wish. Just say so.

The Clarett series is supposed to be the same mostly as the Forte and they currently sale it and it will get software support better than the Forte.

What kind of budget do you have in mind? MOTU makes some nice interfaces, but a bit more money. If you have a Mac, the Apogee Element series looks pretty good.
https://www.audiosciencereview.com/...nterface-quick-measurements.7522/#post-175521
 

sergeauckland

Addicted to Fun and Learning
Patreon Donor
Joined
Mar 16, 2016
Messages
776
Likes
1,360
Location
Suffolk UK
#5
When my Lexicon measuring interface failed, I bought a UMC 202HD but returned it as being unusable for measurements due to excessive distortion. I thought it might just be a faulty unit, so bought another, the 204 HD this time, but with identical results. I also borrowed a third unit with again identical results, so gave up and bought another Lexicon interface on eBay.

I couldn't get the sort of numbers Amir's measurements showed, even for just the DAC, and on loopback, so including the ADC, I was getting distortion levels of around -50dB if I remember correctly.

For my purposes, I'm happy with distortion figures of -90dB, which the Lexicon does comfortably. I haven't tried other USB interfaces except the little Behringer UCA202 which is quite good as a DAC, but I've never managed to get the ADC part working properly.

S
 

Blumlein 88

Major Contributor
Joined
Feb 23, 2016
Messages
5,452
Likes
4,243
#6
Here is a couple FFTs of the Scarlett 18i20. First is the ADC capturing the 1 khz tone from a Forte. The second is capturing output of the Zen Tour. The ZT has higher distortion. Something worth pointing out. There are very low level tones around -130 db every 1 khz. This is an oddity of the 18i20. They are there even with no input. So the only harmonics above this for the Forte are the 2nd, 3rd and 5th. The rest are these odd low level idle tones. For most of your purposes they are low enough not to matter. Don't know if this is common to the Scarlett line or not. You also see some lines below 1 khz. These are 60 hz hum harmonics. It isn't endemic to the 18i20. I can run the laptop off battery and they disappear just forgot to unplug the power cord on this particular set of measurements.

Forte into 18i20 1 khz.png


Zen Tour feeds 18i20 1 khz.png

Here is a Forte loopback showing a cleaner result.

Forte loopback.png
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
6,638
Likes
2,685
Location
Riverview, FL
#8
It depends upon what you wish to do with it.
Measurements would be one thing - which is the main reason to want a clean ADC.

And I have some musical toys laying around, so both instrument and mic inputs would be desired.


The Clarett series is supposed to be the same mostly as the Forte and they currently sale it and it will get software support better than the Forte.
Looking at Clarett 4Pre...

https://focusrite.com/thunderbolt-audio-interface/clarett/clarett-4pre


What kind of budget do you have in mind?
I tend to regret going cheap more than I worry about spending a bit more...


For my purposes, I'm happy with distortion figures of -90dB
Audibly, that's right around "good enough", but as a measuring tool, maybe not so good.
 

Blumlein 88

Major Contributor
Joined
Feb 23, 2016
Messages
5,452
Likes
4,243
#9
What's with the roll-off above 20 kHz?
The Forte uses minimum phase filters in ADC and DAC. So in a loopback their filters cause a double dose of the little droop. If you switch to 192 khz, you get the same droop at the top of that band just past 90 khz. So it isn't a shortcoming of the analog portion.
 
Last edited:

DonH56

Major Contributor
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
2,662
Likes
2,688
Location
Monument, CO
#10
https://www.sweetwater.com/shop/studio-recording/audio-interfaces/

Apogee, Focusrite, MOTU, RME are all highly-regarded professional interfaces. I personally have had issues with several Behringer components but they did purchase a high-end mixer company some years back. I have a couple of PreSonus preamps/interfaces and they are, as Amir would say, "ok".
 

March Audio

Major Contributor
Manufacturer
Patreon Donor
Joined
Mar 1, 2016
Messages
3,400
Likes
2,768
Location
Perth Western Australia
#11

sergeauckland

Addicted to Fun and Learning
Patreon Donor
Joined
Mar 16, 2016
Messages
776
Likes
1,360
Location
Suffolk UK
#12
Audibly, that's right around "good enough", but as a measuring tool, maybe not so good.
Depends on what the measurements are for. I mostly measure vintage analogue equipment including tape machines, so -90dB is 0.003% in THD and is 20dB better than I need to know. If something measures better than -70dB in terms of distortion, there's not a lot wrong with it. For noise measurements I have a meter that goes down to -100dB dBu so again, quite low enough.

Few bits of vintage analogue equipment have distortion lower than 0.01%, and even those that might have, if it's so low I can't measure it, it's Good Enough.

S
 

JohnYang1997

Addicted to Fun and Learning
Joined
Dec 28, 2018
Messages
546
Likes
162
#13
I think there was a thread here brought this up.
Basically, if thd+n is needed, it's better to measure thd and noise independently.
Forte shows very good ADC performance at around -10dbFS to -30dbFS. DAC is good at -18db. Adding an attenuator in the front, you can measure 0.0001% THD.

A slight higher end options are motu 828es, and apogee element24. They have best adc chip available in the consumer market. And still not expensive at all. DAC chip is 9016 on both not great but can with pre-distortion in rew it's fine to measure amps.

Then it's more refinement than else, which is rme adi2pro and stuff. Not really a cheap option but it's a complete package to measure sub 0.0002% thd+n.

For noise, just build a 20db-30db gain opamp based preamp for noise measurements. And divided by the gain later. A modified o2 should do the job as well(battery is clean power).
 

Blumlein 88

Major Contributor
Joined
Feb 23, 2016
Messages
5,452
Likes
4,243
#16
I think there was a thread here brought this up.
Basically, if thd+n is needed, it's better to measure thd and noise independently.
Forte shows very good ADC performance at around -10dbFS to -30dbFS. DAC is good at -18db. Adding an attenuator in the front, you can measure 0.0001% THD.

A slight higher end options are motu 828es, and apogee element24. They have best adc chip available in the consumer market. And still not expensive at all. DAC chip is 9016 on both not great but can with pre-distortion in rew it's fine to measure amps.

Then it's more refinement than else, which is rme adi2pro and stuff. Not really a cheap option but it's a complete package to measure sub 0.0002% thd+n.

For noise, just build a 20db-30db gain opamp based preamp for noise measurements. And divided by the gain later. A modified o2 should do the job as well(battery is clean power).
On the other end, I use microphone inputs for low level stuff. I'll run a -60 db signal and then whatever I wish to test is that or lower. Shoot it to the mic inputs, apply near 60 db of gain. You'll push the noise floor of the DUT well above the noise floor of the mic inputs. Then digitally knock the levels back to where they belong using that -60 db reference tone. That is how I've been able to confirm excellent linearity right down to the 24th bit in some DACs. So good ADCs with attenuation of careful signal selection can get to some pretty incredible measurements.

Many/most/nearly all affordable interfaces have one gain circuit for the microphones. Switching to line or instrument puts some resistors in the way to drop the signal and uses the same circuit. These usually erode the EIN of the line inputs. For most purposes of the interface meaningless, but they do impact it for measurements of very low levels and low noise.
 

JohnPM

Member
Technical Expert
Joined
Apr 9, 2018
Messages
81
Likes
189
Location
UK
#17
So, looking at the amp, I see (top graph) second harmonic at about -25dB to the signal tone.
That is relative to an extraordinarily low fundamental, the 378 Hz tone is at -128 dBFS. How does it look with a more reasonable test signal level?
 

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
6,638
Likes
2,685
Location
Riverview, FL
#18
That is relative to an extraordinarily low fundamental, the 378 Hz tone is at -128 dBFS. How does it look with a more reasonable test signal level?
I used two instances of REW, one receiving digits from the Behringer, the other using a UMIK-1.

The electrical "test signal" was a couple of volts across the speakers and to the Behringer, in-room level 80dB.

The REW display for the electrical display is uncalibrated. I can try again with calibration, I suppose.

The REW display (including the noise floor) moves up and down as I adjust the input level at the Behringer.

Relative levels - fundamental to displayed 2nd harmonic - remained the same with various adjustments at the Behringer trying to clean up the harmonics, so, it just got left at a reported "extraordinarily" low level.

The level of second reported by the Behringer would be measurable in-room, but wasn't present in-room, verified by adding 5th harmonic at the level of the 2nd, and verifying it was present in-room, indicating the 2nd harmonic reported electrically was fake news.

It served the purpose to show me the Behringer is unsuitable for detailed measurement purposes, because it creates its own harmonics, no matter what you adjust - source level, input level, 10dB pad, whatever, it's ADC aparently is just not very "clean".
 
Last edited:

JohnPM

Member
Technical Expert
Joined
Apr 9, 2018
Messages
81
Likes
189
Location
UK
#19
There's no calibration for dBFS, it reads relative to digital full scale. -128 dBFS is about 21 bits down, more than 30 dB below the lsb of a 16 bit system to add some context, so it is an unusually low level to encounter as a signal. The levels generally in that plot are oddly low. That's not to say there is anything particularly good about the UMC202HD, the few reports I've seen on it weren't particularly complimentary, but nothing remotely as bad as 4% 2nd harmonic. There are some potentially relevant notes here about the UMC204HD, which may have a similar input architecture and hence may have the same issues and benefit from the same workarounds discussed in that post. The UMC202 measurements here aren't great, but not disastrous either.
 
Last edited:

RayDunzl

Major Contributor
Central Scrutinizer
Joined
Mar 9, 2016
Messages
6,638
Likes
2,685
Location
Riverview, FL
#20
The levels generally in that plot are oddly low.
The level controls for the Behringer move the display (both signal and noise floor) up and down the scale. The 2nd harmonic remains.

I'm using a TRS to RCA with short test clips on the end to touch the signal and ground at the amp output, so, there's that... Room measure is UMIK-1. Arbitrary setting for amplifier output, whatever makes in-room 86dB in this case.

No 2nd harmonic in the room, so it isn't really coming from the amp.

1558053861509.png
 
Top Bottom