• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Focusrite Scarlett 2i2 Audio Interface Gen 3 Review

Roland68

Major Contributor
Joined
Jan 31, 2020
Messages
1,364
Likes
1,129
Location
Cologne, Germany

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,338
Likes
3,278
Location
.de
Steinbergs have rather high output impedance of ~80 ohms so are not ideal for anything impedance-critical. Other than that the output is not bad though, and running it into a headphone amplifier should give great results. While the Focusrites would give substantially more accurate sound with impedance-critical IEMs (though possibly a bit of hiss), their arch nemesis are inefficient low-impedance headphones. @Julian Krause has looked into this in some detail in his latest 18i20 gen3 review, from 08:54:
18i20gen3-hpdist-krause.png

The circuitry appears to be virtually the exact same across the entire range from 2i2 to 18i20. At 32 ohms, distortion becomes excessive beyond a few hundred microwatts. If we say that's 200 µW, this would be 80 mV / 2.5 mA rms. I would actually be fine with that given regular old dynamic driver cans, but it's not exactly what you would call luxurious.
 
Last edited:

Music1969

Major Contributor
Joined
Feb 19, 2018
Messages
4,645
Likes
2,813
I would actually be fine with that given regular old dynamic driver cans, but it's not exactly what you would call luxurious.
I think regular old dynamic driver cans is exactly what this headamp section was designed for.

I don't think luxury cans were in focus here during design.

It's not high on the list of these designers (so far).

Maybe future Gens get more focus on headamp, who knows.
 

shoto

Member
Joined
Jul 4, 2021
Messages
59
Likes
12
Would Solo be as good as 2i2 for making measurements?
Is there ever really need for 2 inputs?

BTW Am I right thinking you could use any ADC as analysier to make accurate measurements of another device up to the limits of it's own performance or is more complicated than this? Im interested in the scarlet for this.
 
Last edited:

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,588
Likes
38,291
Location
Gold Coast, Queensland, Australia
Would Solo be as good as 2i2 for making measurements?
Is there ever really need for 2 inputs?

Yes, for simultaneous left and right, channel separation and of you want measure impedance, you would need two identical inputs.

Otherwise solo is perfectly good.
 

shoto

Member
Joined
Jul 4, 2021
Messages
59
Likes
12
Thanks, I made an edit just now.

hmm why is 2 channel need for channel seperation (aka crosstalk?)? If you look at one output you would see the crosstalk from other channel on it, or no?
 

jhaider

Major Contributor
Forum Donor
Joined
Jun 5, 2016
Messages
2,823
Likes
4,522
What is loopback correction exactly?
You connect the second output to the second input, to keep timing accurate.

That said, REW’s timing reference does a pretty good job, too. I find it useful for, e.g. measuring a room correction system when you inject the test signal over HDMI. If I’m measuring a loudspeaker I’ll want a loopback though.
 

Leiker535

Active Member
Joined
Jul 7, 2020
Messages
259
Likes
317
Hello all. I'm considering getting this interface for some different use-case scenarios. Can anyone confirm that the TRS line out is balanced? I intend to connect it to an a30 pro bal in through a trs to xlr cable, would that be possible?

I'll be using it for instrument work, but also for digitizing my LPs and also as a general purpose DAC.
 

LTig

Master Contributor
Forum Donor
Joined
Feb 27, 2019
Messages
5,760
Likes
9,442
Location
Europe
Hello all. I'm considering getting this interface for some different use-case scenarios. Can anyone confirm that the TRS line out is balanced? I intend to connect it to an a30 pro bal in through a trs to xlr cable, would that be possible?

I'll be using it for instrument work, but also for digitizing my LPs and also as a general purpose DAC.
What does the user manual say?
 

Leiker535

Active Member
Joined
Jul 7, 2020
Messages
259
Likes
317
What does the user manual say?
It does that that it is balanced, but with the maximum output of +10 dBu. My question is whether the Scarlett will supply the 4v to the a30 pro, like amirm could get on his dashboard, or whether it'll limit to 2V.
 

staticV3

Master Contributor
Joined
Aug 29, 2019
Messages
7,540
Likes
12,014
@Leiker535 the 6.35mm differential line outputs on the back of the 2i2 3rd gen can output 4.6Vrms with 0dBFS input and volume set to 100%.
You can get the full voltage into the A30Pro using TRS to XLR cables.
You can reduce the voltage to 4.0Vrms in a number of ways, for example with a -1.2dB software preamp.
 

vitalyx

Member
Joined
Dec 25, 2022
Messages
6
Likes
2
So is it still better to plug your heaphones into this Scarlett rather than a macbook?
Apple claims high impedence heaphones support in the recent models: https://support.apple.com/en-gb/HT212856

EDIT: also, would it matter if used with something like Airpods Max with the cable? https://www.apple.com/uk/shop/product/MXK22ZM/A/lightning-to-35mm-audio-cable-12m-white

There is an ADC in the cable. Does this mean that there is no impedance to speak of? At least I wasn't able to find any information anywhere.
 
Last edited:

staticV3

Master Contributor
Joined
Aug 29, 2019
Messages
7,540
Likes
12,014
So is it still better to plug your heaphones into this Scarlett rather than a macbook?
Apple claims high impedence heaphones support in the recent models: https://support.apple.com/en-gb/HT212856
The DAC and headphone amp that's built into the new MacBook Pro models is better than the 2i2 G3.

EDIT: also, would it matter if used with something like Airpods Max with the cable?
No.

There is an ADC in the cable. Does this mean that there is no impedance to speak of?
Only the impedance of the ADC's input, which is most likely 20kΩ or more.
 

vitalyx

Member
Joined
Dec 25, 2022
Messages
6
Likes
2
The DAC and headphone amp that's built into the new MacBook Pro models is better than the 2i2 G3.


No.


Only the impedance of the ADC's input, which is most likely 20kΩ or more.
Thank you! So basically I can continue using AirPods Max via cable with 2i2, but with another pair of headphones I'd be better off switching to the MBP's headphone jack?
Is the reason for it not mattering in case of AirPods Max the very high 20kΩ ADC impedance you mentioned (compared to a ~32-600Ω in a pair of heaphones)?

EDIT: oh, I guess, it's because the preamp is in the headphones themselves?
 
Last edited:

teashea

Addicted to Fun and Learning
Joined
Dec 23, 2022
Messages
695
Likes
763
Location
Nebraska
This is a review and detailed measurements of the Focusrite Scarlett 2i2 Audio Interface (DAC/ADC). I was talked into buying it recently. The 2i2 costs US $140 including free Prime shipping so very reasonable for amount of functionality.

For a budge device, the Scarlett 2i2 looks pretty good:


Yes, I have left the plastic cover on the front and back. :) Leaving it there in case I decide to sell the unit.

I found the controls more logical than Behringer budget products. Like the separate volume control (small one) for headphone and the larger one for the line out for example.

The inputs are kind of strange in that depending on what type of cable you use, the gain changes. If you use XLR, it assumes microphone input and hence higher gain. If you use TRS/1/4 inch plug, then it assumes "line" input. This works but you better have the right cables to mate with it this way.

The back panel doesn't have much other than a USB-C jack that I like and TRS line outs:
Note: I spelled Scarlett wrong in all the graphs below. Too lazy to edit them all now.

DAC Audio Measurements
Let's test each subsystem independently starting with DAC:
View attachment 41079

This is pretty good! We actually beat the company spec by good bit. Max output is I think 4.6 volts so this is not a powerhouse like higher end audio interfaces are.

SINAD of 102 is enough to place the Scarlett 2i2 in our competent bucket:
View attachment 41081

Dynamic range is good enough for 17 bits or so:

View attachment 41080

Noise level is not competitive with better desktop audiophile DACs though:
View attachment 41082

Jitter test shows the same elevated noise level but otherwise clean:
View attachment 41083

Linearity is near excellent:
View attachment 41084

Multitone shows increasing distortion with frequency:

View attachment 41085

Filter response is more or less one expects from default DAC chip settings:
View attachment 41086

We hit what is a seemingly major snag with we test at much higher bandwidth of 90 kHz:
View attachment 41087

Measured distortion+noise goes through the roof, rising to as much as 0.3%! Is the 2i2 this bad? Yes, and no. Let's first perform a wideband spectrum analysis of a 1 kHz tone:
View attachment 41088

Noise level sharply rises above 28 kHz. This is a feature of many DACs where noise in audible band is pushed into ultrasonics. The levels though seem low rising to "only" -85 dB. But that is misleading because FFT analysis artificially lowers the noise output of the device under test. Let's use a simple noise meter and see what we get:
View attachment 41089

At default 22.4 kHz I use for the dashboard measurement, the noise level is 25 microvolts (millionth of a volt). Increase the bandwidth to 90 kHz to capture the noise shaping and that value jumps 650 times higher to 16.3 millivolts (thousands of a volt).

Fortunately we don't hear ultrasonics and even if we did, this is pretty low level so it is more of a bother for our eyes in measurements than ears.

Headphone Amplifier Measurements
Jumping right into the heart of the matter with THD+N versus power into 300 ohm we get:
View attachment 41090

This is no good. We are getting power levels that we can get out of a smartphone or a $9 dongle attached to it.

The situation doesn't improve with 33 ohm load:
View attachment 41093

Now we have tons of distortion in addition to low output level.

Thankfully output impedance is low:
View attachment 41095

ADC Audio Measurements
Digitizing analog inputs is the main purpose of these interfaces so let's see how we do with a 1 kHz tone at 2 volts using XLR connection (i.e. "mic"):
View attachment 41096

This is quite good actually, once again landing the 2i2 into the competent bucket:
View attachment 41097

The above required gain settings at minimum so I thought I test the line input using TRS jacks:
View attachment 41098

Gain was lowered but distortion actually increased! Not sure why.

Either way, you can't use professional levels of 4 volts or more as you can with higher end products. 2 Volts is it.

Dynamic range is good with 2 volts in and near 0 dBFS capture:

View attachment 41099

Frequency response is very flat and nice, albeit stopping early on 192 kHz:
View attachment 41100

Linearity shows the better noise performance of XLR/mic input:
View attachment 41101

THD+N versus level shows the superiority of the Scarlett 2i2 compared to consumer interfaces:

View attachment 41102

Test using "jitter" signal (j-test" shows clean but elevated noise level:
View attachment 41103

1 kHz FFT spectrum shows the same:
View attachment 41104

A bit of oddness in THD+N versus frequency:

View attachment 41105

Conclusions
I can see why the Focusrite Scarlett 2i2 is such a popular audio interface. Its DAC and ADC both deliver competent performance with no dark corners. The only item performing well below expectation is the headphone amplifier. If you care about headphone monitoring, best get a proper amplifier to pair with it. With a single output though this may be a bit challenging.

Overall, I had low expectations going into this review and came out pleased. I am happy to recommend the Focusrite Scarlett 2i2 a very good budget audio interface.

--------
As always, questions, comments, corrections, etc. are welcome.

I understand getting Christmas presents for your pets is "a thing" now. With two dogs and half a dozen panthers, this bill adds up big time. So I appreciate generous donation using : https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
The very poor headphone amplifier makes it useless. The cheap plastic construction does not help.
 

harrisonjr98

Member
Joined
May 20, 2021
Messages
17
Likes
7
The headphone output seems to be capable of at least 1.4 V RMS (~6.5 mW RMS into 300 Ohm) output into high impedance loads. That should be enough to connect it to a standalone amp. With the volume set correctly, that would still give a SINAD of about 85 dB or better.
I am trying to use the 2i2 as a dac to go out to a JDS Atom, and attempting to understand why one would go the route this commenter suggested instead of just running out of the monitor outs with the knob on the front dimed (presumably that would be line level.)
 

daftcombo

Major Contributor
Forum Donor
Joined
Feb 5, 2019
Messages
3,687
Likes
4,068
I am trying to use the 2i2 as a dac to go out to a JDS Atom, and attempting to understand why one would go the route this commenter suggested instead of just running out of the monitor outs with the knob on the front dimed (presumably that would be line level.)
With the 2i4 2nd Gen you would have a fixed level output, which would be even better for that purpose.
 

HarmonicTHD

Major Contributor
Forum Donor
Joined
Mar 18, 2022
Messages
3,326
Likes
4,830
I am trying to use the 2i2 as a dac to go out to a JDS Atom, and attempting to understand why one would go the route this commenter suggested instead of just running out of the monitor outs with the knob on the front dimed (presumably that would be line level.)
The 2i2 has line outs. I would use those with an adapter. No need to use the headphone out.

This makes only sense if you have difficulties driving your headphones from the 2i2 because I see no other reason why not leave out the Atom pre-AMP altogether.
 
Last edited:
Top Bottom