Solid State Logic SSL 12 first impressions

[CuriousAli]

Given there is relatively less info on SSL 12 here, I'm posting my first impressions:

1. Line-in Inputs:
I have used its line-in analog inputs to connect both my turntable and LCD display's audio out (for consoles and Apple TV) and both sound very good, with no noise, no coloration, of course these are my subjective impressions.

2. Headphone Outputs:
It has two headphone outs, while I couldn't find independent measurements, its specs says <1 Ω and it has four settings in its SSL 360 software:

• Normal
• High Sensitivity
• High Impedance
• Line Out

I saw these settings are similar to SSL 18 and if their measurements are close to that, that's a pretty good headphone amp for an audio interface.
I have the Sony MDR-MV1 headphones with 24Ω impedance, I originally tried the normal mode and it was fine but it got loud too soon and I wanted more granular control, so I changed to the high sensitivity mode and now even at that I don't have to crank up the volume even to half, it is loud enough before that, and I don't know if its placebo effect or what, but I think it is a bit more clear with less overwhelming bass, while not sounding thin at all, compared to its normal mode, of course that's just subjective test and not measured.
Now this is a big deal for me because my iD24 had too much output impedance (22Ω?) which I felt it affected my headphone's sound, now I can listen to music from Mac, turntable and other devices with my headphones and get the best sound out of them.
I also have the HD 560S (120Ω) but have yet to test them with it.
The interesting part is it can have these settings separately for each headphone out, and also you can change the headphone outputs to become line outs, not my use case though.

3. ADAT input
I wish it directly accepted SPDIF too, but it is still better than nothing. I can connect my SPDIF devices to Audient iD24 and use its ADAT output to give them to SSL 12's ADAT input, also it can be used as an extra analog input source, maybe connecting my cassette tape deck too, so I'm glad this ADAT option is there, but it wouldn't hurt them to make SSL 12 SPDIF compatible, it seems they have done it for SSL 18.

4. 32-bit
I know the 32-bit claim is mostly marketing, and for 32-bit integer the difference isn't that dramatic and it still doesn't reach the limits of 24-bit in practice, but I read somewhere where they guessed SSL 12 is using a cheap old AKM Codec ADC/DAC chip which is 24-bit in ADC and 32-bit in DAC, but on Audio MIDI setup of macOS it shows as 32-bit on both:

So that rules it out? Or it may show as 32-bit still then?

5. No sleep/delay/pops
Something that scared me of the new DACs intended for home use is their practice of using some kind of power save/sleep, so after inactivity for a while if you play something you either get delayed start of the sound or sometimes a pop sound too, I got it with my Fiio KA17 and heard some other devices do it too. I have heard there is some driver option on Windows to disable that feature on some DACs, but no such setting exists on Mac, so I'm glad SSL 12 doesn't do it, at least not encountered it yet.
But please beware that SSL12 can make a loud sound in cases of firmware update/restart of the device, so be careful.
Thankfully SSL12 has an on/off button too so you can prevent it most of the times I guess.

6. USB Bus-powered
It just uses the USB port for power (USB 3.0 900mA for power, 2.0 compatible for audio) which I know people hate for low headphone output power, but for my headphones it is more than enough and I love not having to plug another device to my power outlet.

A Question:
If you have been patient and read all this, I have a question too:
I have my CD rips, Spotify Lossless and iTunes purchased music at 44.1 kHz, so I have set the output to it for bit-perfect playback.
But doesn't it negatively impact my line-in input performance for Turntable and other sources?
Any sample rate suggestion for my use case?

Feel free to ask me try stuff with it if you are curious.
And if you want measurements, please tell me what exactly to do in REW or other Mac compatible software and I will try my best.

 

[AnalogSteph]

3. ADAT input
I wish it directly accepted SPDIF too, but it is still better than nothing.

Having ADAT but not SPDIF in is a bit unusual and a minor bummer, but oh well.

4. 32-bit
I know the 32-bit claim is mostly marketing, and for 32-bit integer the difference isn't that dramatic and it still doesn't reach the limits of 24-bit in practice, but I read somewhere where they guessed SSL 12 is using a cheap old AKM Codec ADC/DAC chip which is 24-bit in ADC and 32-bit in DAC, but on Audio MIDI setup of macOS it shows as 32-bit on both:

As you wrote, it was someone's guess. They may have been wrong. You'll only know for sure when somebody takes a peek inside.

Have you considered taking some loopback measurements (line out to line in)?

A Question:
If you have been patient and read all this, I have a question too:
I have my CD rips, Spotify Lossless and iTunes purchased music at 44.1 kHz, so I have set the output to it for bit-perfect playback.
But doesn't it negatively impact my line-in input performance for Turntable and other sources?
Any sample rate suggestion for my use case?

Apparently CoreAudio resampling has been perfectly fine since at least OS X Leopard more than a decade ago, and you may still be able to find an upsampler with less CPU overheard for your player of choice, so I don't think going to 48 or 96 kHz (depending on ADC digital filter performance) should pose any issues at all.

 

[CuriousAli]

Having ADAT but not SPDIF in is a bit unusual and a minor bummer, but oh well.

I know, I even tried of testing it in practice but it only accepts iD24's output in its ADAT in if I set iD24 to output ADAT.

As you wrote, it was someone's guess. They may have been wrong. You'll only know for sure when somebody takes a peek inside.

You are right.

Have you considered taking some loopback measurements (line out to line in)?

I don't know if I have done it right? I tried my best by reading guides, and I have disabled the monitor out for this test to not cause a loop.
The best I got was THD+N 0.0006% at -0.03 dBFS (to prevent clipping).
Please let me know if I'm measuring it right:
Left output/Input1:

Right output/Input1:

Also Right output/Input2 (at 0.0 dBFS):

Apparently CoreAudio resampling has been perfectly fine since at least OS X Leopard more than a decade ago, and you may still be able to find an upsampler with less CPU overheard for your player of choice, so I don't think going to 48 or 96 kHz (depending on ADC digital filter performance) should pose any issues at all.

That's a relief, thank you!

 

[jkim]

Your loopback measurements look right. If you had measured the performance of either ADC or DAC separately using higher-quality measurement rigs, you would've obtained a few dBs better results. Other than that, your results look accurate.

As @AnalogSteph said, you do not have to stick with bit-perfect (or more precisely bitrate-perfect) playback. There is no difference in digital signal quality between upsampling by the host (either Mac OS or Windows) and by the DAC's internal algorithm. Any signal at a lower bitrate will be upsampled, anyway, by the DAC's reconstruction filtering.

 

[CuriousAli]

Your loopback measurements look right. If you had measured the performance of either ADC or DAC separately using higher-quality measurement rigs, you would've obtained a few dBs better results. Other than that, your results look accurate.

Thank you, glad to hear it!
I tried the same measurements with iD24 too and it got worse results, so I think so far SSL 12 is my best measurement device.

As @AnalogSteph said, you do not have to stick with bit-perfect (or more precisely bitrate-perfect) playback. There is no difference in digital signal quality between upsampling by the host (either Mac OS or Windows) and by the DAC's internal algorithm. Any signal at a lower bitrate will be upsampled, anyway, by the DAC's reconstruction filtering.

I see, thank you.

 

[CuriousAli]

Sorry @jkim I'm new to doing such measurements, I tried my best to follow your example on the Cirrus Logic hump thread.
Can you please see if I have done it right?

Here is the result with iD24 which I think the unusual raise of TD+N shows the infamous hump right?

And doing the same test with SSL 12 thankfully didn't result in a hump:

 

[jkim]

Yes, the iD24 shows the hump. Someone else already measured the iD24's hump in my review thread. It shows up in your measurements, too. You may want to use smaller steps to see it more clearly, though. And use the dBr scale on REW's main window to make the results look similar to my measurements.

The SSL is not supposed to produce a hump since it is not based on a Cirrus Logic chip.

 

[CuriousAli]

Thank you, yes I knew from your thread that I should get the hump on iD24, so it was a way for me to make sure I'm testing it right.
I will try smaller steps too.

I wanted to make sure SSL 12 didn't have the Cirrus Logic chip since there is no teardown yet, so I'm glad the results prove it isn't having the issue and most likely not using the chip.

 

[AnalogSteph]

I am a bit surprised that you can't seem to be getting more than about 105 dB and change worth of dynamic range in loopback. This suggests to me that there is a potential level mismatch between output and input. You did use TRS patch cables, right? Maybe the output was not turned up all the way? According to specs, the line outputs can dish out +24 dBu, and the line inputs can accept as much, and let's say I would expect a bit more action in the distortion department at these levels.

If the unit sticks to its specs, maximum attainable loopback SNR should be about 110.9 dB(A).

If I had to hazard a guess, I'd say there's two AK4621 or CS4272 codecs for inputs plus headphone outs plus some extra DACs for the line outs that could be a lot of things (AK4493, ES9039Q2M, CS4398 perhaps). Could be totally wrong, of course. A loopback measurement of the headphone output at 192 kHz and closely inspecting loopback frequency response for periodic ripple at 44.1/48 kHz may provide some clues.

 

[jkim]

I am a bit surprised that you can't seem to be getting more than about 105 dB and change worth of dynamic range in loopback. This suggests to me that there is a potential level mismatch between output and input. You did use TRS patch cables, right? Maybe the output was not turned up all the way? According to specs, the line outputs can dish out +24 dBu, and the line inputs can accept as much, and let's say I would expect a bit more action in the distortion department at these levels.

If the unit sticks to its specs, maximum attainable loopback SNR should be about 110.9 dB(A).

If I had to hazard a guess, I'd say there's two AK4621 or CS4272 codecs for inputs plus headphone outs plus some extra DACs for the line outs that could be a lot of things (AK4493, ES9039Q2M, CS4398 perhaps). Could be totally wrong, of course. A loopback measurement of the headphone output at 192 kHz and closely inspecting loopback frequency response for periodic ripple at 44.1/48 kHz may provide some clues.

His measurements look alright for loopback results. See my measurements of SSL 2 MKII. In hindsight, I believe there was some cabling issue, but it would not have made night-and-day differences. Also have a look at @Rja4000's measurements of Topping E2x2's ADC. His measurements were done using a higher quality signal from Topping D50 III. Still, line-in SINAD does not reach 105 dB.

I believe what @CuriousAli obtained are convincing lookback measurement results.

 

[CuriousAli]

You did use TRS patch cables, right? Maybe the output was not turned up all the way? According to specs, the line outputs can dish out +24 dBu, and the line inputs can accept as much, and let's say I would expect a bit more action in the distortion department at these levels.

I used a long 3 meters TRS cable with reasonable quality but not the best, I don't have a shorter higher quality cable (patch cables are shorter?) to test, will it make a noticeable difference?
I have used its direct routing mode so it is at the maximum level and it is the only device I have that shows the same input and output levels at REW's level check in its Soundcard settings, others have about 7 to 12 db lower output compared to SSL 12's max output.

If I had to hazard a guess, I'd say there's two AK4621 or CS4272 codecs for inputs plus headphone outs plus some extra DACs for the line outs that could be a lot of things (AK4493, ES9039Q2M, CS4398 perhaps). Could be totally wrong, of course. A loopback measurement of the headphone output at 192 kHz and closely inspecting loopback frequency response for periodic ripple at 44.1/48 kHz may provide some clues.

I think the Line out dual tone test rules out the CS4398 as it doesn't show the Cirrus Logic hump, will test the headphone out too.

I believe what @CuriousAli obtained are convincing lookback measurement results.

Good to hear!

 

[Sokel]

I used a long 3 meters TRS cable with reasonable quality but not the best, I don't have a shorter higher quality cable (patch cables are shorter?) to test, will it make a noticeable difference?
I have used its direct routing mode so it is at the maximum level and it is the only device I have that shows the same input and output levels at REW's level check in its Soundcard settings, others have about 7 to 12 db lower output compared to SSL 12's max output.


I think the Line out dual tone test rules out the CS4398 as it doesn't show the Cirrus Logic hump, will test the headphone out too.


Good to hear!

You can use this silly, little guide for the basics:

Horrible Crackling from Topping E1x2

Hi! I lurked on this forum a lot when deciding on which interface to buy, and I ended up deciding on "upgrading" from my rudimentary Behringer UM-2 to a Topping E1x2. The first day, it worked great! The next day however, I noticed a horrible crackling noise when I was playing a game and...

www.audiosciencereview.com

-make sure you use Exclusive I/O as shown
-try a little lower level like -1dBFS or -3dBfs, some ADC's don't like more than a Volt or so.
-minimize all ADC gain,etc,play with ground.
-also go to distortion settings and change to dB units, it's easier to read and make sure that your selected range is 20Hz-20kHz.

Great first attempts so far though!

 

[CuriousAli]

A loopback measurement of the headphone output at 192 kHz and closely inspecting loopback frequency response for periodic ripple at 44.1/48 kHz may provide some clues.

Here is the headphone out at High Impedance mode and 192 kHz (with input gain used too):

But I got better results with 48kHz with High Impedance mode (I had to use gain on input too, although a different gain, so maybe that caused better results):

Also if I set the headphone out as Line Out (gain used just before clipping):

And I tried the Dual Tone measurement with both Headphone out set as High Impedance and Line Out and there was no Cirrus Logic Hump:

(Please note that I'm using the same TRS cable for the headphone tests too, not using special cables or setting ohm with a tool in the way).

You can use this silly, little guide for the basics:

Thank you, if only I had seen your guide sooner! I used mostly the same settings, now I changed averages to 8 and the overlap percentage and setting HP limit to 0Hz to match your guide.

-make sure you use Exclusive I/O as shown
-try a little lower level like -1dBFS or -3dBfs, some ADC's don't like more than a Volt or so.
-minimize all ADC gain,etc,play with ground.
-also go to distortion settings and change to dB units, it's easier to read and make sure that your selected range is 20Hz-20kHz.

Great first attempts so far though!

Thank you for the kind words!
- REW on Mac doesn't seem to have the exclusive/hog mode, but I had set the macOS sounds to go through MacBook Pro's own DAC to not interfere.
- Gain was set to 0 during my line out tests, had to raise it about 5 levels for headphone out tests though.
- I have tried -1dBFS and -3dBFS for Line out and it wasn't as good as reaching near 0, here are the results:

 

[CuriousAli]

Please correct me if I'm doing it wrong, I don't have special tools for ohm change and I'm using the same TRS cable for headphone tests too, but I think my results for High Impedance and Headphone Line out modes are similar to Julian Krause's measurements for SSL 18, right?

 

[CuriousAli]

They have released a firmware update and SSL 360° update:

https://support.solidstatelogic.com/hc/en-gb/articles/4408123894417-SSL-360-Downloads-and-Release-Notes#ssl-360-downloads-and-release-notes-0-0

These parts are related to SSL 12:

SSL 12 and SSL 18: Custom Monitor Control Assignments​

Ensure you update your SSL 12 / SSL 18 firmware to access this new feature

• Allows you to customise the analogue outputs that are affected by the front panel monitor control or disable it entirely.
• Useful for surround monitoring with multiple analogue outputs.
• Can also be used to affect the S/PDIF output (SSL 18 only).

I think it is this new option:

I guess by enabling it for outputs 3 & 4 they can also be controlled at the same time by the volume knob on the device.

There is also this update which is a mouse scroll behavior setting on SSL 360:

SSL 12 and SSL 18 Mixers​

• Mouse Scroll can now affect mixer controls and faders.

There is this firmware bugfix for Windows too:

SSL 12 Firmware​

• [FIXED] SSL 12 - Some reported cases of SSL 12 not being discovered as a USB audio device on Windows and needing to be power cycled before it gets discovered