SSL 2+ USB Interface Measurements

[tochnia]

I decide to do line REW loopback measurement of SSL 2+ USB Interface not expecting miracles given its 210GBP /280$/ price.

Between 100Hz and 20Khz it is almost straight line which is excellent result. But below 100Hz things are getting funky, with +/-5db starting 43Hz at if levels are on middle for line level.

Four Measurements are taken with with pots for line level on position 2/3/4/5 out of 10.
2 was with lowest distortion, 5 with largest. On position 6 clipping occur.

Position 2 Flatness -> +1.4dB/-0.2dB for 20Hz to 20kHz
Position 3 Flatness -> +2.0dB/-0.2dB for 20Hz to 20kHz
Position 4 Flatness -> +3.9dB/-0.1dB for 20Hz to 20kHz
Position 5 Flatness -> +10.8dB/-0.1dB for 20Hz to 20kHz

Distortion at position 2:

Distorsion at position 5:

Pity that I cannot measure Mic preamplifier as loopback is clipping when I switch to Mic setting.
If anyone have any idea how I can measure Mic preamp loopback please let me know.

 

[AnalogSteph]

These frequency response wiggles scream setup issue to me. Check whether you don't have input monitoring turned on by any chance. (Monitor mix should be turned all the way to USB.) For obvious reasons you don't want to have the input played back immediately. I also vaguely remember that there were some default settings in REW you want to change... maybe window size / window function?

Besides, sample rate settings between REW and hardware seem to be mismatched. REW's default Java output on Windows uses shared mode I/O, so either switch to ASIO output or manually make sure the settings for REW, playback device, recording device and possibly SSL control panel are in sync.

If the mic in still clips with input gain at minimum and you don't have a balanced line-level attenuator at hand (which by any chance you probably don't), reduce generator level in REW until you're out of clipping. Isn't the default like -20 dBFS anyway? According to the SOS review the interface can output +12.5 dBu balanced and will take up to +5.5 dBu in mic mode (vs. +18.5 dBu padded down as a line-in), so you won't need any massive attenuation. You won't be able to measure EIN like that, of course, as input noise would be swamped by output noise.

For reference, Julian Krause has reviewed the SSL 2 sans + and found good performance... input dynamic range about 111 dB(A), output similar. This is about what one would expect from something based on an AK4621 in real life... good midrange performance.

 

[H-713]

Still wondering about that whole "4K" button. I know it's supposed to "emulate" the sound of a 4000 series (not sure which), but I'm struggling to figure out why that would be desirable.

 

[tochnia]

Still wondering about that whole "4K" button. I know it's supposed to "emulate" the sound of a 4000 series (not sure which), but I'm struggling to figure out why that would be desirable.

It raise curve in highs and have nice audible difference. Will put measurement with 4K button too when I have time.

 

[H-713]

In that case, it really isn't emulating the 4000.

Actually, my original post was more of a joke than anything... there's a bit of a fierce divide about the sonic characteristics of SSL desks compared to other consoles. Reality is that most of the "SSL Sound" can probably be attributed to having dynamics processing on every channel. It may come as no surprise, then, that SSL desks became popular right around 1980.

 

[tochnia]

I got new NTI Analyzer and while I'm still learning it, here are some examples of 4K button results on THD.
So it added 50-100 times more noise

Without 4k Engaged THD at 159Hz:

4k Engaged THD at 159Hz:

Without 4k Engaged THD at 5kHz:

4k Engaged THD at 5kHz:

 

[tochnia]

I did new measurements of SSL 2+.
Monitor mix all the way to USB, line setting on input,

Here it measurements are taken with with pots for line level on position 3/5 out of 10.

Position 3 Flatness -> +1.1dB/-1.5dB for 20Hz to 20kHz:

Position 3 with 4K Button Engaged Flatness -> +3.1dB/-1.3dB for 20Hz to 20kHz:

Position 5 Flatness -> +0.4dB/-1.5dB for 20Hz to 20kHz:

Position 5 with 4K Button Engaged Flatness -> +3.1dB/-1.3dB for 20Hz to 20kHz:

And all 4 measurements together:

 

[Raqs]

For reference, Julian Krause has reviewed the SSL 2 sans + and found good performance...

Link, please?...

 

[tochnia]

6 Month later I have Prism dScope audio analyzer so we can compare to REW measurements:

Microphone pre-amp measurement on position 2 out of 10 on knobs:

It looks very good but there is around 1dB difference even both knobs were visually aligned to same position

Microphone pre-amp measurement on position 3 out of 10 on knobs:

On more than 3 it start clipping out.

Here is measurement on position 2 out of 10 on knobs but with button 4K enabled:

This curved graphs quite a lot in high frequencies with almost +3dB.

Line pre-amp measurement on position 2 out of 10 on knobs:

Minimal difference between channels and very flat line, close to -0.1dB only.

Line pre-amp measurement on position 3 out of 10 on knobs:

Again very flat lines, but channel difference to 1dB, it looks that there cannot be any precision with small knobs.

Line pre-amp measurement on position 5 out of 10 on knobs:

Almost same curve as above tests with minimal dip below 30Hz

Line pre-amp measurement on position 5 out of 10 on knobs and 4K button enabled:

 

[tochnia]

Now THD Figures again measured with Prism dScope III:
Surprisingly for me THD barely pass dScope testing only at position 7 of 10 on knobs /line setting on SSL2+/.
Above this it start clipping so it looks best THD+N which can be achieved is -82dB:

Here it is with position 3 out of 10 on knobs it fail test:

 

[Lambda]

Have you used ALSA or WASAPI and the native sampling rate?
And whats IR window setting?
Are you using timing reference signal?

I had the same problems with REW trying to measure a DAC without timing reference

 

[AnalogSteph]

Link, please?...

It looks very good but there is around 1dB difference even both knobs were visually aligned to same position

That's pots for ya. Not uncommon.

Now THD Figures again measured with Prism dScope III:
Surprisingly for me THD barely pass dScope testing only at position 7 of 10 on knobs /line setting on SSL2+/.
Above this it start clipping so it looks best THD+N which can be achieved is -82dB:

Here it is with position 3 out of 10 on knobs it fail test:

Something is not adding up here. I think you should be adjusting the generator level on the dScope each time you are changing input gain on the SSL, aiming for a constant output level on the digital side (say, -3 dBFS). Make sure the dScope's output attenuator is being used properly. Accuracy at very low levels is going to require external passive attenuation, as residual output noise is spec'd at -115 dBu and as such substantially above any half-decent mic input. (If it's any consolation, much the same applies to AP systems.) If you want to capture the input noise floor with any kind of accuracy, I would suggest at least 25 dB worth of attenuation, with moderate resistor values so that source impedance seen by the input remains low (not above the typical 150 ohm spec - e.g. 1k5 / 75R x2 would work).

Recording input sensitivity (dBu in for a certain dBFS) for gain settings from 0 to 11 may prove a decent exercise in dScope operation. (It's not too much of a challenge with a constant generator level and starting out at min gain though.)

I would suggest some more RTFM and asking the support if necessary - if memory serves @Julian Krause was rating them quite highly.

I have generally found mic inputs to have lowest distortion in the 20-26 dB gain range or so... at very low gain performance may be compromised by common-mode distortion, and at high gain amplifier GBW runs out. Results would vary depending on circuitry though, e.g. a Mackie mixer would be essentially unimpressed even at maximum gain while a more basic Behringer counterpart would show substantially rising distortion up there.

I would look at crosstalk vs. frequency to see whether it really is much of an issue. Even -76 dB is by no means a terrible value (ever seen the specs for FM tuners or phono cartridges?)... in a typical recording situation acoustic bleed between mics is likely to be a much more severe issue.

 

[bambadoo]

Hi, just bought a SSL 2+ for about 150usd for testing.
Just a quick measurement.
Focusrite solo gen3 into SSL 2+ 1khz

Seems as if XLR in is somewhat worse that trs.. Doing loopback (trs from output to XLR input does nok look good)Will get some ok trs cables for the purpose later.

 

[bambadoo]

Just another quick.
Loopback (trs - trs)

Loopback with trs-xlr. Seems like when using xlr in it detects as a mic and with trs line input.. Guess I have to read the manual

 

[bambadoo]

Here is a good teardown of the SSL 2 (not +)
Looks decent enough.

Solid State Logic SSL 2 Teardown

SSL2 is the first audio interface from Solid State Logic, which is famous for its large analogue mixing consoles. What components did the legendary...

www.ifixit.com

 

[bambadoo]

Regarding the first post .
Here are mye linearity measurements (loopback mode)
44.1, 48, 88.2, 96, 176.4 and 192khz

And distortion vs frequency for 44.1khz

 

[bambadoo]

Here is another great review with also "teardown" of the SSL2+

Solid State Logic SSL 2+ - interfejs audio USB 24-bit/192 kHz do domowego studia - e-muzyk.EU

Solid State Logic SSL 2+ - interfejs audio USB 24-bit/192 kHz do domowego studia

e-muzyk.eu

He also likes the headphone out even with 30ohm headphones.
I must say, I really like the sound using my Philips Fidelio X2 headphones also.

 

[itayoron]

For reference, Julian Krause has reviewed the SSL 2 sans + and found good performance... input dynamic range about 111 dB(A), output similar. This is about what one would expect from something based on an AK4621 in real life... good midrange performance.

Any idea if SSL 2/SSL 2+ changed their chips after the fire at AKM as Audient did with iD4/iD14 mkii?