Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Mark Wilkinson]

Hello fellow cheapskates*:

I thought this would be of interest.

Advertisements for the Focusrite Scarlett 2i2 Gen 2 claim it to be the best selling audio interface of its type in the world. Given its low price, good looks, massive advertising budget and great specs (except the ones they don't mention), I can see why.

Thx Langston, and in keeping with cheapskate....have you measured the UMC404HD ?   Same price...
I know it at least has much better phase performance ...not to mention 4 ch....

[Langston Holland]

Hi Mark:

I'd be glad to measure anything that would have a general interest to folks on PSW, and I think that unit fits on the low end. Anyone that really want this - just PM me and I'll figure out my schedule and you can ship it to me. I'll do it within two business days and ship it back.

There are lots of variables with audio interfaces for measurement that aren't obvious that the Scarlett 2i2 did well with. One is USB jitter and many units, including more expensive ones like the Tascam UH-7000, fail in this area and are unusable with the MLS stimulus that many still use for acoustics.

[Langston Holland]

I have a first gen 2i2 sitting on my desk that I've been using to take measurements, I wonder how they compare?

Hi Nathan:

I just did a little search and found a couple of interesting things about the Gen 1 unit:

1. It was spec'd to 96k max. Thus the (unusable IMO) 192k of Gen 2 is likely driver implemented without hardware upgrades.

2. Gen 1 now uses the Gen 2 driver (4.x series) with the same latency reduction. Maybe the Gen 1 driver (3.x series), which I can't find on the Net, would fix or reduce the phase error in the Gen 2 unit in exchange for increased latency.

It'll be fun if the mfg. fusses at me about these posts, but they won't. It'd put attention on things that would be better ignored from their POV.

[Langston Holland]

Meanwhile, on the other side of the railroad tracks...

Hello fellow buy-once, cry-once heretics!

In case you were looking for an end-game for a two channel audio interface for measurement (and recording), the RME Babyface Pro is a bargain IMO.

Build quality is Apple computer, Lectrosonics, Mercedes, etc. It's just wonderful. The Windows WDM and ASIO drivers are the best and most flexible I've ever seen. The included routing software includes basic, but excellent EQ and effects. This ain't Behringer - everything that comes with this device is something someone is proud of, not something birthed in marketing.

Summary

Sample rates from 32k to 192k show near theoretical noise floors on both input and output. Because USB is an asynchronous animal, SR's can be adjusted in 0.001% increments around the primary frequency for syncing with other stuff that's not playing by the rules (common).

Its channels have identical transfer functions, so dual channel FFT is dead accurate at any SR. Its USB bus powered or externally powered in stand-alone mode (mic preamp, D/A converter, that sort of thing). You can also use it as an interface with an iOS gizmo in class compliant mode.

Unlike the inexpensive Scarlett 2i2 Gen 2, this thing is a solid recording interface and the included EQ and effects in the routing software probably make it the world's best Podcast interface. That's what I'm going to use it for along with D/A conversion at my desk when it's not out making measurements.

Odds and Ends

Measurement gizmos used were the Audio Precision APx515, FIR Capture, Keysight U1282A and U1733C. The APx515 allows you to send or receive to/from a DUT through ASIO, thus allowing closed loop measurement of the analog inputs and outputs independently.

Phantom power is excellent, 47.3V unloaded and 13.4V when both channels have full 10mA loads. The spec for phantom is 48V +/-4V unloaded and 14V +/-1.2V loaded with 10mA. Most phantom powered mics I'm aware of draw 3mA to 5mA. The XL outputs are unaffected by the presence of phantom power. XL input impedance is 2,053Ω from 100Hz to 10kHz, sloping down to 1,850Ω at 100kHz. This impedance is the same with or without phantom power engaged as it should be. Optical digital I/O can do SPDIF or ADAT, MIDI I/O with an adapter cable is included as well. Metering is used for display feedback of operating adjustments as well as level indication. They are nicely visible, and like everything else on this unit, accurate.

Measurements

This unit has analog gain adjustments for both input and output, thus you should adjust both so that you can keep the digital signal as close to full scale as possible. I chose -3dBFS for these measurements.

Flat lines from a dual channel FFT measurement program called FIR Capture at all sample rates:



Phase input and output:





Magnitude input and output:





Noise input and output:





Distortion at -3dBFS:













Finally, my Pin 1 torture test that few devices have ever passed. Passing is when the 60Hz square wave fundamental and harmonics stay below about -100dBFS. The Babyface Pro (why, oh why did they name it that?!), doesn't show any evidence that the square wave is on! The is a best ever result for me. No response whatsoever to a nasty 10V square wave shoved down Pin 1's throat. RME obviously got the memo and connected Pin 1 directly to chassis shared only with the negative power supply rail:

[David Sturzenbecher]

Meanwhile, on the other side of the railroad tracks...


Unlike the inexpensive Scarlett 2i2 Gen 2, this thing is a solid recording interface and the included EQ and effects in the routing software probably make it the world's best Podcast interface. That's what I'm going to use it for along with D/A conversion at my desk when it's not out making measurements.

Tell us more about your podcast...

[Nathan Riddle]

Hi Nathan:

I just did a little search and found a couple of interesting things about the Gen 1 unit:

1. It was spec'd to 96k max. Thus the (unusable IMO) 192k of Gen 2 is likely driver implemented without hardware upgrades.

2. Gen 1 now uses the Gen 2 driver (4.x series) with the same latency reduction. Maybe the Gen 1 driver (3.x series), which I can't find on the Net, would fix or reduce the phase error in the Gen 2 unit in exchange for increased latency.

It'll be fun if the mfg. fusses at me about these posts, but they won't. It'd put attention on things that would be better ignored from their POV.

So you think Gen1 and Gen2 hardware is same just a fresh look & newer driver?

I did have to update to 4.x driver for the G2 to work

I just so happen to have the 2.5.1 driver

https://www.dropbox.com/s/yfsy6tx7grvfb1t/focusrite-usb-2-driver-2.5.1.exe?dl=0


Edit: oh my goodness do NOT run G1 and G2 interfaces at the same time. Instant windows crash

[Helge A Bentsen]

Is there any good alternatives for a 8 preamp interface with internal PSU?

Today I have a Octa-Capture. It works well but I'm tired of lugging around an external PSU.

I found one from Motu that looks good, but it's a bit expensive.
http://motu.com/products/proaudio/8pre-es

 

[Frank Koenig]

http://motu.com/products/proaudio/8pre-es

This one is less than half the price (~US$ 550) and looks like it would do everything I need in a measurement interface. Rack mount, no wall wart, and independent, clearly labeled  buttons for phantom and pad.

--Frank

http://motu.com/products/motuaudio/8pre

[Mark Wilkinson]

Is there any good alternatives for a 8 preamp interface with internal PSU?

Today I have a Octa-Capture. It works well but I'm tired of lugging around an external PSU.

I found one from Motu that looks good, but it's a bit expensive.
http://motu.com/products/proaudio/8pre-es

Probably larger than you're willing to lug....but I like using a x-32 rack for higher mic counts

[Tim McCulloch]

Looking for something to replace my Tascam US-144 mkII, thinking about the Focusrite Scarlett 2i1 3rd gen.  Does anyone know if Focusrite improved the issues found by Langston with the G2 model?  I'm not especially concerned as my only use for this is measurement at 48kHz, but I don't want spend my "stimulus money" on an unsuitable product.

Anyone have insight?

[ProSoundWeb Community]