Focusrite Liquid 4Pre
I own a pair of quite special 4 channels microphone preamplifiers from Focusrite
The model is discontinued for some time now...
What's interesting with this preamp is that it may emulate 40 different preamps, using convolution.
If you're interested, you may find some official information on Focusrite's page or download the descriptive PDF.
It may also add harmonics to taste.
So I thought I'd give it a try.
First, I use the Focusrite software - allowing to connect to the preamp through LAN.
This is pretty much a copy of the front panel.
Then, using APx 500 Flex, I measured Multitone 32 for various emulations.
I mentioned the Preamp model Focusrite measured for each emulation.
Why Multitone ?
Because it shows you immediately various types of distortions at all frequencies.
I ran them at high level, to maximize differences.
In real life, the level -and the effect- would be lower.
All measurements are done at minimum gain on Mic input.
Measurement setup: RME ADI-2/4 Pro SE > Analog 1/2 > Focusrite Liquid 4 Pre Mic 1/2 > AES cable > RME ADI-2/4
Note that I also sync the clock of the 4Pre using its AES input, with a signal coming from the ADI-2/4
First using the "FLAT" preamp - no emulation.
API 3124+ emulation
(I also keep the "FLAT" curve for reference)
API 3124+ "HOT"
Same but with convolution measured with a very high input level
Neve 1073
Neve 1073 High-Z
SSL 4000
Millennia SST-1 (tube + transformer mode)
Focusrite ISA 110
Telefunken V72
Telefunken V76
Here, we should note the level drop at low frequencies,
more visible here:
Universal Audio M610
AVALON VT-737SP
Note that all the above are with "Harmonics" setting at 0
"FLAT" preamp, with Harmonics at full (15)
Fun, isn't it ?
Used as a digital effect
Interestingly, the convolution engine also works in full digital:
You may send a signal to the 4Pre through ADAT or AES (2 input channels only for AES) and receive it back as a digital signal.
This may be interesting to emulate some preamp and/or add distortion without requiring to insert a PC with a VST plugin, as an example.
It may be usefull in a live environment, as a digital mixer insert to up to 4 channels.
Neve 1073 emulation (full digital)
I tried to measure added latency and found 40 samples (0.8ms) for the above emulation compared to the ADI-2/4 digital loopback.
Total as seen from the PC, including ADI-2/4, will vary with the RME driver buffer size, with a minimum at 3.125ms @48kHz, which looks very reasonable.
RME ADI-2/4 > SPDIF (L) > RME ADI-2/4
compared to
RME ADI-2/4 > AES (L) > 4Pre > AES (L) > RME ADI-2/4
Other measurements
Another way to look at it is to compare dashboard for the
"FLAT" preamp , compared to the UA M610
And how do they behave with level ?
THD+N @1kHz vs Level for various emulations
Dashed line is "FLAT" + added Harmonics
We see the raise in distortion starts sooner.
For other emulations, it shows at much higher level, with various profiles.
Now for different level of Harmonics
Level of Harmonics 1-7 is adding even harmonics.
Level of Harmonics 8 and above is adding even and odd harmonics.
We see the difference is mostly at the extreme level for the highest Harmonics settings.
Conclusion ?
A common feedback -that I share- is that the different Preamp emulations are adding little audible difference, if any.
But Harmonics are more audible.
Looking at above plots, we may understand better why:
Emulations mostly change the sound at very high level, closer to the digital clipping - that you want to avoid at all cost, since it REALLY sounds bad when it occurs.
Typical wisdom says you should record at -20dBFS to -18dBFS on average.
With this kind of level, the effects of emulations remain quite limited, while the Harmonics addition may become audible.
If you want to maximize the effect of those emulations, you'd better feed the 4Pre with a pretty hot signal.
Beware of digital clipping though.
If anyone has some free-of-rights music or tracks I could process through those emulations and publish for AB-X testing, I'm interested.
I measured this one since I'm considering saling one of the two I own.
If anyone in EU is interested, drop me a message.
I own a pair of quite special 4 channels microphone preamplifiers from Focusrite
The model is discontinued for some time now...
What's interesting with this preamp is that it may emulate 40 different preamps, using convolution.
If you're interested, you may find some official information on Focusrite's page or download the descriptive PDF.
It may also add harmonics to taste.
So I thought I'd give it a try.
First, I use the Focusrite software - allowing to connect to the preamp through LAN.
This is pretty much a copy of the front panel.
Then, using APx 500 Flex, I measured Multitone 32 for various emulations.
I mentioned the Preamp model Focusrite measured for each emulation.
Why Multitone ?
Because it shows you immediately various types of distortions at all frequencies.
I ran them at high level, to maximize differences.
In real life, the level -and the effect- would be lower.
All measurements are done at minimum gain on Mic input.
Measurement setup: RME ADI-2/4 Pro SE > Analog 1/2 > Focusrite Liquid 4 Pre Mic 1/2 > AES cable > RME ADI-2/4
Note that I also sync the clock of the 4Pre using its AES input, with a signal coming from the ADI-2/4
First using the "FLAT" preamp - no emulation.
API 3124+ emulation
(I also keep the "FLAT" curve for reference)
API 3124+ "HOT"
Same but with convolution measured with a very high input level
Neve 1073
Neve 1073 High-Z
SSL 4000
Millennia SST-1 (tube + transformer mode)
Focusrite ISA 110
Telefunken V72
Telefunken V76
Here, we should note the level drop at low frequencies,
more visible here:
Universal Audio M610
AVALON VT-737SP
Note that all the above are with "Harmonics" setting at 0
"FLAT" preamp, with Harmonics at full (15)
Fun, isn't it ?
Used as a digital effect
Interestingly, the convolution engine also works in full digital:
You may send a signal to the 4Pre through ADAT or AES (2 input channels only for AES) and receive it back as a digital signal.
This may be interesting to emulate some preamp and/or add distortion without requiring to insert a PC with a VST plugin, as an example.
It may be usefull in a live environment, as a digital mixer insert to up to 4 channels.
Neve 1073 emulation (full digital)
I tried to measure added latency and found 40 samples (0.8ms) for the above emulation compared to the ADI-2/4 digital loopback.
Total as seen from the PC, including ADI-2/4, will vary with the RME driver buffer size, with a minimum at 3.125ms @48kHz, which looks very reasonable.
RME ADI-2/4 > SPDIF (L) > RME ADI-2/4
compared to
RME ADI-2/4 > AES (L) > 4Pre > AES (L) > RME ADI-2/4
Other measurements
Another way to look at it is to compare dashboard for the
"FLAT" preamp , compared to the UA M610
And how do they behave with level ?
THD+N @1kHz vs Level for various emulations
Dashed line is "FLAT" + added Harmonics
We see the raise in distortion starts sooner.
For other emulations, it shows at much higher level, with various profiles.
Now for different level of Harmonics
Level of Harmonics 1-7 is adding even harmonics.
Level of Harmonics 8 and above is adding even and odd harmonics.
We see the difference is mostly at the extreme level for the highest Harmonics settings.
Conclusion ?
A common feedback -that I share- is that the different Preamp emulations are adding little audible difference, if any.
But Harmonics are more audible.
Looking at above plots, we may understand better why:
Emulations mostly change the sound at very high level, closer to the digital clipping - that you want to avoid at all cost, since it REALLY sounds bad when it occurs.
Typical wisdom says you should record at -20dBFS to -18dBFS on average.
With this kind of level, the effects of emulations remain quite limited, while the Harmonics addition may become audible.
If you want to maximize the effect of those emulations, you'd better feed the 4Pre with a pretty hot signal.
Beware of digital clipping though.
If anyone has some free-of-rights music or tracks I could process through those emulations and publish for AB-X testing, I'm interested.
I measured this one since I'm considering saling one of the two I own.
If anyone in EU is interested, drop me a message.
Last edited:
