UAD SSL 4000 E Channel Strip Plug-in Analysis

The SSL 4000 is famous for being the console employed on more Platinum-selling records than any other. Its expansive, punchy, deep and assertive character have made it an industry standard of modern recording.

I/O Stages
The preamp module offers a true reproduction of the famed SSL 001 preamp circuit and its full-of-character Jensen input transformer
Accurate emulation of the dbx output VCA circuit, including clipping.
There are two distinct EQ circuit topologies available in the SSL E Channel Strip. The classic Bbrown EQ is less exacting and musical, while the more modern Black circuit is more surgical and clean.
The Brown EQ bands can be boosted/attenuated by ±15 dB whereas the Black has a range of ±18 dB.
In Brown mode, the slope of the high pass filter is 6 dB per octave. In Black mode, the slope of the high pass filter is 18 dB per octave.
The slope of the low pass filter is 12 dB per octave in both Brown and Black modes.
A neat trick for toggling your filter settings in/out is to click the ‘Out’ text.
The compressor allows for a wide variety of dynamics response, from transparent to aggressive.
At heavy compression settings with quick release times, the SSL design has a similar room-expanding quality as the UREI 1176 on signals such as drum overheads and room mics.
Release time is continuously variable between 0.1 seconds and 4 seconds.
Compression attack time is program dependent and automatically varies between 3 ms – 30 ms. When Fast is engaged, the attack time is fixed at 3 milliseconds for 20 dB of gain reduction.
Exp/gate attack time is fixed at 1.5 milliseconds per 40 dB. When Fast is engaged, the attack time is fixed at 100 microseconds per 40 dB of gain reduction.

To analyse the non-linear behaviours of the SSL 4000 E Channel Strip, I ran a 100 Hz sine wave through the plug-in at various settings.

100 Hz sine wave at 0 dB through Line input at Unity

100 Hz sine wave at - 12 dB through Line input at +12 dB

100 Hz sine wave at -12 dB through Mic input no transformer (XFMR) at +2.4 dB

100 Hz sine wave at -12 dB through Mic input XFMR engaged at 3.3 dB

As you can see, the SSL I/O generates both even and odd lower order harmonics, with a prominent second at -60 dB. Hence the smooth and musically sympathetic sound of this channel strip. What I did find very interesting is that the harmonic generation is almost identical for all of the above setting.

Lets see how the dbx output VCA circuit modifies the sound…

100 Hz sine wave at -12 dB through Line input at +2 dB and VCA output at +10 dB

The second harmonic is approximately 10 dB lower than previous settings, with the subsequent odd ordered harmonics increasing by approximately the same. Odd ordered harmonics are often associated with ‘richness’ and ‘depth’, so if you want a smooth, constant sound drive input for a harder, richer sound drive the output VCA.

For this test I ran a drum loop through the SSL at various settings. I have compensated for level so that the drum loop peaks at -1 dBFS for maximum resolution of the waveform in Pro-L.

Dry Drum Loop -12 dB

Drum loop at -2.2 dB ran through SSL Line input at Unity

Drum loop at -12 dB ran through SSL Line input at +9.8 dB

Drum loop at -12 dB ran through Line input at Unity and boosted 9.7 dB post SSL

Surprisingly they all look the same, but they don’t sound the same! More of that to come later. The SSL subtly softens transients with some high frequency saturation, just look at how the hats have been attenuated. There is also an overshoot of about a dB during a small drum fill.

Lets overload the preamp section and attenuate VCA output fader.

Certainly different saturation characteristics, there is more transient rounding and less saturation/attenuation of high frequencies.

Drum loop at -12 dB ran through Mic input no XFMR

Drum loop at -12 dB ran through Mic input XFMR engaged

Very subtle differences in waveform, the most noticeable difference was the 1 dB attenuation when the XFMR was engaged.

I bet you are thinking shut up Paul and let me have a listen for myself…

Dry Drums (RMS matched)
SSL Drums Dry (RMS match).aif (678.3 KB)

Line Input
SSL Drums Line.aif (678.3 KB)

Mic Input (no XFMX)
SSL Drums Mic No XFMR.aif (678.3 KB)

Mic Input with XFMX
SSL Drums Mic XFMR.aif (678.3 KB)

So if the harmonic generation and saturation characteristics are very similar, you may be asking why the unit sounds different at different settings. Well…
What I am looking for is the frequency and phase response of the plug-in at unity and how particular settings change this relationship.
Spectrum Colours
Yellow - noise generated by the plug-in analysis software to analyse the plug-in
White - frequency response of the plug-in at that particular setting
Red - phase shift of the plug-in at that particular setting

Line input at Unity

Line input at Maximum

Mic Input at Unity (no XFMR)

Mic input at Unity (with XFMR)

Mic input at the midway point (with XFMR)

Mic input at maximum (with XFMR)

Wow! Just look at the difference in frequency and phase responses. These properties were continuously variable when adjusting I/O parameters.

It is worth noting that Dyn In and EQ In (at unity) had no effect on frequency or phase response.

I hope you enjoyed this article, I certainly enjoyed creating it and will do many more when time allows. If you have any suggestions for what plug-in to analyse next, please let me know.


This is by far the best analog modelling plug in analysis i have ever seen anywhere. Amazing job, Paul!.
Ok, so within the UAD collection we have the following channel strips that are modelled with the pre-amp coloration: api vision, neve 1073, neve 88, manley VOXBOX, and the new SSL E…am i missing any?

Thanks Danny!

Yes, for the channel strips. There are the UA 610s as well, which are very colourful preamps indeed.

that’s awesome what a difference, would or could you sidechain to master buss?

it’s funny but almost none of these channelstrips or analog compressors for that matter have an external sidechain input…i am not sure why actually…

I believe it is a technical limitation, as the side-chain input is not delay compensated. For example, when you solo the side-chain in Pro-C2 there is a delay between the source and the side-chain and Pro-C2 does not induce much latency, whereas UAD plug-ins are rather latent, thus rendering SC unusable.

That makes perfect sense.

I hear this argument a lot, you can use the same logic for many different situations e.g. what is the point on paying 2k+ for a preamp when it only records one source at a time, why pay 3k for a Culture Vulture when it can only be used on one stereo track. Well, you do because they sound great! For me, UA sounds better than the competition and therefore, I am happy to work around its quirks. For others these quirks might be too much, I know some people hate Ilok and refuse to purchase any plug-ins utilising this security. I don’t mind it at all. The UAD plug-in code is obviously optimised for the DSP chips, it would be interesting to see how many instances you would be able to load natively.
On the practical side, you can turn ‘Load Lock’ off and any disabled sections of the plug-in will not use up any DSP. You can always commit, I much prefer this method of working now, just like you had to back in the day with a hardware based studio, if you didn’t have multiple units.

1 Like

To echo @Paul,

I’m 6 months in to being a uad satellite octo owner… once you hear the analog modeling of UAD plugins and start using them to make your mixes come alive… there is no going back.

I initially thought they were expensive, and they are, BUT they are a fraction of the cost of their hardware counterparts. It’s not a light investment, but it IS an investment in your craft and sound. Throw an ATR-102 on your mixbus with @Danny’s magic settings and try and not like UAD after that