• 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!

Measurements of Focusrite Forte recording interface

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,056
I measured the Focusrite Forte recording interface. In this case focusing on the DAC. This was a nice two channel recording interface which has been discontinued. The ADC and DAC chips and much of the circuitry is what is used in the current Focusrite Clarett line of interfaces. It has two microphone inputs with up to 75 db of gain which is higher than usual. Also line level and instrument inputs (for guitars). Headphone, and monitor outputs. It is a 24bit/192 khz ADC/DAC. The Forte has a Display Port dongle for some connections and a touchscreen for many of the controls. It connects via USB using ASIO in windows and Focusrite software in Mac. It sold for most of its life around the $450 to 400 mark.

upload_2018-4-28_19-37-18.jpeg


The measurement FFTs are all 32K FFTs as is the custom on ASR. The spectrograms used a 2k FFT. As is Amir's custom I'll start with a 12 khz -1 db tone. Very tight central spike and not much else. Ignore the little blip at 16.3 khz as that shows up when my ADC is connected to my Macbook Pro. BTW, the ADC used for measurements is the Antelope Audio Zen Tour.
12 khz tone .png


Here is the 0-300 hz spectrum while producing the 12 khz tone. I think you'll agree it is pretty clean.
300 hz.png


A combined chart showing a - 60 db 1 khz tone and the full 12 khz tone to show the noise floor is not modulated by the signal. Most DACs do modulate that just a little and some more than a little. AES17 Dynamic range is 114 db.

12khz noise modulation.png


The ubiquitous 1 khz thd plot. 2nd harmonic is below -120 db and the 3rd is about -106 db. You see some fifth and a little 9th for some reason. SINAD is 104 db.
1khz.png


Here is a twin tone IMD plot with -6db tones at 18 khz and 19 khz. The 1 khz difference signal is -120 db. While the first sidebands around the main tones lie around -100 db. Not too bad I don't think.
18 and 19 imd.png


I like to look at spectrograms because you can spot other spurious signals, aliasing, imaging etc. very easily. I use 2 k FFTs for this as it catches even short lived blips or glitches. In the following you see a single max level tone sweep from 2hz-20khz. Next to it is a two tone sweep at max level. The tones are a constant 1 khz apart. I have it set so at -100 db the background goes to medium gray.

In this case there is very nearly nothing from these tones above the -100 dbFS level. A little IMD creeps in just under that at the upper most frequencies.
spectro sweeps minus 100 db background.png


Next the same thing except I have the background going to medium gray at -120 db. So anything above that level will show in the graph. You can see some harmonic distortion showing up at these levels. Also a touch of imaging in the upper 2 khz. You also see the 16.3 khz idle tone of the ADC in these graphs. If your monitor is good you can see a tiny hint of the 1 khz difference tone in the IMD side of the spectrogram.

spectro sweeps minus 120 db background.png


Though not shown linearity is good to 20 bits. I didn't test lower, it probably is good even lower than this.
 

Attachments

  • spectro sweeps minus 100 db background.png
    spectro sweeps minus 100 db background.png
    86.3 KB · Views: 225
Last edited:

watchnerd

Grand Contributor
Joined
Dec 8, 2016
Messages
12,449
Likes
10,408
Location
Seattle Area, USA
I measured the Focusrite Forte recording interface. In this case focusing on the DAC. This was a nice two channel recording interface which has been discontinued. The ADC and DAC chips and much of the circuitry is what is used in the current Focusrite Clarett line of interfaces. It has two microphone inputs with up to 75 db of gain which is higher than usual. Also line level and instrument inputs (for guitars). Headphone, and monitor outputs. It is a 24bit/192 khz ADC/DAC. The Forte has a Display Port dongle for some connections and a touchscreen for many of the controls. It connects via USB using ASIO in windows and Focusrite software in Mac. It sold for most of its life around the $450 to 400 mark.

View attachment 12392

The measurement FFTs are all 32K FFTs as is the custom on ASR. The spectrograms used a 2k FFT. As is Amir's custom I'll start with a 12 khz -1 db tone. Very tight central spike and not much else. Ignore the little blip at 16.3 khz as that shows up when my ADC is connected to my Macbook Pro. BTW, the ADC used for measurements is the Antelope Audio Zen Tour.
View attachment 12395

Here is the 0-300 hz spectrum while producing the 12 khz tone. I think you'll agree it is pretty clean.
View attachment 12397

A combined chart showing a - 60 db 1 khz tone and the full 12 khz tone to show the noise floor is not modulated by the signal. Most DACs do modulate that just a little and some more than a little.

View attachment 12398

The ubiquitous 1 khz thd plot. 2nd harmonic is below -120 db and the 3rd is about -106 db. You see some fifth and a little 9th for some reason.
View attachment 12399

Here is a twin tone IMD plot with -6db tones at 18 khz and 19 khz. The 1 khz difference signal is -120 db. While the first sidebands around the main tones lie around -100 db. Not too bad I don't think.
View attachment 12401

I like to look at spectrograms because you can spot other spurious signals, aliasing, imaging etc. very easily. I use 2 k FFTs for this as it catches even short lived blips or glitches. In the following you see a single max level tone sweep from 2hz-20khz. Next to it is a two tone sweep at max level. The tones are a constant 1 khz apart. I have it set so at -100 db the background goes to medium gray.

In this case there is very nearly nothing from these tones above the -100 dbFS level. A little IMD creeps in just under that at the upper most frequencies.
View attachment 12403

Next the same thing except I have the background going to medium gray at -120 db. So anything above that level will show in the graph. You can see some harmonic distortion showing up at these levels. Also a touch of imaging in the upper 2 khz. You also see the 16.3 khz idle tone of the ADC in these graphs. If your monitor is good you can see a tiny hint of the 1 khz difference tone in the IMD side of the spectrogram.

View attachment 12404

Do you know what chipset it uses?

It seems a lot of the mid-tier and up interfaces are using generic DSP chips (like SHARC); my Apollo Twin Duo used SHARC chips.
 
OP
Blumlein 88

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,056
Do you know what chipset it uses?

It seems a lot of the mid-tier and up interfaces are using generic DSP chips (like SHARC); my Apollo Twin Duo used SHARC chips.
Not certain this is correct, but AK4413 DAC and AK5388 ADC.
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
Nicely done.
I believe this is the same unit that Archimago uses for the measurements on his blog.
 
  • Like
Reactions: trl
OP
Blumlein 88

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,056
Yes
Nicely done.
I believe this is the same unit that Archimago uses for the measurements on his blog.
Yes is the same. I've had mine longer. I'll put up more measures when I get time. Linearity and maybe some other sample rates.
 

mshenay

Active Member
Joined
Feb 9, 2018
Messages
177
Likes
206
Good to see some interface measurements and thanks @amirm for putting it on FP! I'd love to see more
 

ManglerFi

Member
Joined
Feb 7, 2018
Messages
59
Likes
16
Nice work Blumlein88! Could you point me to an explanation of how to read the last two sets of graphs you posted?
 
OP
Blumlein 88

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,056
On reading the spectrogram:

The left hand vertical axis is frequency from 0 to 24 khz. The horizontal axis is time.

The position indicates frequency at a given time. The color represents signal level. Dark gray is no signal, brown is low signal level, orange is a middle signal level, green is a high signal level, and black is max signal.

In the above I swept a max signal from 2 to 20,000 hz at 500 hz per second. In the first case the lower threshold is minus 100 dbfs. Anything at minus 101 dbfs and less is ignored.

Now in the second graph I lowered the threshold to minus 120 dbfs. You see some additional low level lines show up. Those sloping upward at a stepper angle than the main tone are harmonics. Actually harmonics distortion.

Look closely and there is just a tiny bit of 2nd harmonic. The steeper and more visible lines are the 3rd and 5th harmonics. So they are somewhere between minus 100 and 120 dbfs.

The slight vertical hash around the main signal is an artifact of fft windowing. The low level horizontal line at 16 khz is a very low idle tone in the ADC so ignore it.

Notice how the 3rd harmonic line hits 24khz and then goes downward? That is imaging in the DAC output because the filter isn't steep enough. That is an imaging reflection of the third harmonic reaching 24 through 26 khz and reflecting downward in frequency. The output filtering does keep anything higher from leaking in at lower frequencies.
 
OP
Blumlein 88

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,056
Spectrogram with two tones:

I swept a pair of tones low to high frequency separated by 1000 hz. You see some low level tones parallel to the main tones. They are the sum and difference sidebands caused by a small amount of odd order intermodulation distortion. You also see the steeper lines from harmonic distortion of the main tones. Almost visible is the 1 khz line caused by 2ñd order imd which in this case lies at a very low minus 120 dB.

If you have other questions, or something wasn't clear in my explanation, just ask.
 
OP
Blumlein 88

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,056
Had a chance to look at low level linearity of the Forte.

I listed rounded values to the nearest .1 db as more decimal places is false precision at low levels I feel. The number after the measured value is how far it differs from the correct amount to the nearest tenth db.

−60.2 ::::::0.0
−66.2 :::::0.0
−72.2 :::::0.0
−78.3::::::0.0
−84.3::::::0.0
−90.3::::::0.0
−96.3::::::0.0
−102.3:::::0.0
−108.4::::-0.1
−114.3::::::0.1
−120.5::::-0.1
−126.1::::::0.4
−132.2::::::0.3​



I didn't feel I could reliably get a handle on the 23rd and 24th bit. According to Amir's .1 db limit this is a 20 bit DAC. To the left is the measured value and next to it is how much it differs from the theoretical number. To give you an idea here is a spectrogram of the low level signals. I use 12 khz tones for my linearity testing. I have the depth of the spectrogram set to -160 db and using a 32k FFT so the background is slightly lit by the basic noise of the unit. It starts with - 60 db and goes down one bit at a time. Just to the right of the vertical cursor you can see the 22nd bit in the noise floor faintly as the straight horizontal line at 12 khz. The other solid horizontal line is an idle tone at 16.3 khz when the ADC is connected to my Macbook Pro. The very faint horizontal line at maybe 14.7 khz near -130 db is something I haven't noticed before.

In any case this good linearity along with the signal not modulating the noise floor is an example of good DAC performance at low levels.

Forte low level linearity.png


Here is a 128k FFT of the 20th bit.

Forte 20th bit FFT.png
 
Last edited:

Stump909

Member
Joined
Apr 9, 2018
Messages
80
Likes
53
Wow, so overall this is a fairly solid product, despite being discontinued. How's the headphone amp stack up? Any idea what the output impedance is? Also, can you control the line-out with the knob?

Edit: Found some specs:

Output Impedance: 8ohms

30mW @ 150Ohms

Dang, I thought this would a contender but that' OI is a little high for my IEMs and the power a bit too weak for my HD650s.
 
Last edited:
OP
Blumlein 88

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,056
Wow, so overall this is a fairly solid product, despite being discontinued. How's the headphone amp stack up? Any idea what the output impedance is? Also, can you control the line-out with the knob?

https://global.focusrite.com/usb-audio-interfaces/forte/specifications

Yes, you can control the headphone out, and the monitor/line outs separately from the knob or via software. The touchscreen lets you select which if you use the knob.

Headphone spec is listed as less than 8 ohms, but I've not confirmed that. Their other specs seem reliable. They suggest headphones of 24 ohms or more. It puts out 30 mW so not the most robust headphone amp. It does sound good with low noise and distortion.
 
OP
Blumlein 88

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,056
Took a look at what upsampling does for the output of the Forte DAC. I recorded the Forte output at 48 khz native sample rate for the twin tone IMD, and also for the same signal upsampled in software to 96 khz, and played back at 96 khz. The recordings of both were done at 192 khz in the ADC.

What you see as the only real difference is imaging artifacts. At the 48 khz playback rate the imaging from the DAC is lower in frequency and higher in signal level. In this case the highest levels of imaging were still -101 db below the signal level. With upsampling to 96 khz and playback at 96 khz the imaging artifacts are twice as high in frequency and at most -130 db below signal level. You see the same features with a single max level tone as well.

Though not shown I looked at THD, IMD and other parameters which are all essentially the same either way.

BTW, the color code on this spectrogram is different. Background goes to light gray at -160 db. Light blue at low levels, then purple for midlevels, red at high levels and white at max level. Upper pane is 48 khz playback. Lower pane is upsampled 96 khz playback.

The horizontal line at 31 khz is a low level idle tone in the DAC.

Forte upsampling spectrogram.png
 
Last edited:

Jimster480

Major Contributor
Joined
Jan 26, 2018
Messages
2,880
Likes
2,032
Location
Tampa Bay
Wow really awesome and detailed review, loving these ADC graphs. Giving you an even better idea of what a product might sound like including artifacts!
The performance is rather impressive considering the ADC included @ $400?
 
OP
Blumlein 88

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,056
Wow really awesome and detailed review, loving these ADC graphs. Giving you an even better idea of what a product might sound like including artifacts!
The performance is rather impressive considering the ADC included @ $400?

Yes it also has an ADC, and some really nice high gain low noise microphone preamps. I've seen them second hand now for about $200. So the headphone out isn't robust, but is very high quality. If you don't need powerful headphone amps it is good for those. Plenty of output for active speakers or power amps via monitor out. I may eventually show the ADC performance it is similarly good with a couple little gotchas though nothing that will intrude audibly.
 

Jimster480

Major Contributor
Joined
Jan 26, 2018
Messages
2,880
Likes
2,032
Location
Tampa Bay
Yes it also has an ADC, and some really nice high gain low noise microphone preamps. I've seen them second hand now for about $200. So the headphone out isn't robust, but is very high quality. If you don't need powerful headphone amps it is good for those. Plenty of output for active speakers or power amps via monitor out. I may eventually show the ADC performance it is similarly good with a couple little gotchas though nothing that will intrude audibly.
Do you recommend it over anything else in the $400 range? Or even in the $200 range? I need an ADC setup for my wife and her violin, I bought her a cork mic that embeds into the bridge of the violin and it requires a powered ADC to work.
 
Top Bottom