SSL Channel

Excellent Suite that I often use for my mixes.

the SSL Channel Plugin from Waves simulates the channel strip of the SSL 4000G console.

I am attaching a photo and will comment on every Pot.

At the top left, under "FILTERS" which also includes the EQ, you will find the high-pass filter on the left and the low-pass filter on the right with the white pots. The one on the left cuts all frequencies below the one indicated by the Pot. If you set it to 70 Hz, for example, you go to remove all frequencies below 70 Hz. The low-pass filter works exactly the opposite way. By selecting a frequency with the Pot, the entire band below that frequency will pass undisturbed and the part above will be cut. Obviously, this occurs within a transition band that we do not know. However, know that the filter is higher quality the smaller its transition band is. It would be interesting to run some tests to see the width of the transition band. I should note that since we are in digital, it wouldn't be difficult to make a filter with a transition band practically equal to zero; however, since these plugins must simulate the behavior of the filters present on the analog console, it is very likely that there is a transition band different from 0.

By selecting "OUT" with the Pots, the filters are disabled.

Moving down we find (red pots) on the left the pot for gain and on the right the pot for center band selection. We are obviously within the high-frequency (HF) equalization section. The bell button serves to determine the band selectivity (the Q factor). By activating it, the filter becomes more selective; by deactivating it, wider. Moving down further we find HMF and LMF (High-medium frequencies and Low-Medium frequencies), i.e., the mid-high and mid-low section. Here you can also adjust the Q factor.

Finally, with the black Pot, you have the low frequencies, for which the same applies as for the highs.

Under EQ TO you have two buttons. The one on the left is used to turn the EQ on and off; specifically, by selecting ByPass it turns off, and by deactivating it, it turns on. Nearby is the DYN S-C button, which would be the Dynamic Side-Chain in case you wanted to use the SSL plugin in sidechain with some other plugin.

At the top right we have the DYNAMICS section.

The three white pots are for compressor adjustment. In the first one, you can choose the compression ratio. The numbers you see are obviously the value of the denominator: 4 stands for 4:1. To the right you find the pot for adjusting the compression threshold; further down on the left is the release time. You will then see that just below there is a small F-ATK switch, which stands for Fast Attack. On the real SSL 4000 console, this button does not exist and fast attack is selected by turning up the pot. Since in the compression stage of the SSL 4000 Channel Strip there is no POT for adjusting attack times, this switch allows you to select a fast or slow attack time.

Further down (green pots) you find the expander, which works the opposite of the compressor; that is, above a certain threshold that you can set, instead of compressing as the compressor did, it expands the waveform. By pressing the GATE key, you can activate the GATE by selecting the operating threshold with the Threshold pot.

Then you have two meters. One gives you information on the gate threshold and the other on the compression.

The Bypass button, as in the case of the filters, serves to disable the module; the other, CH OUT, instead serves to move the Equalization stage post-EQ. In essence, even if you see the compressor on the left and the EQ on the right, the signal is normally compressed first and then equalized, following the signal routing of the analog SSL console. By pressing that button, the compression/expansion stage is moved after the equalization stage.

On the right, the INPUT and OUTPUT buttons are used to select whether we want to consider the signal before it enters the filters or after, and this is essential for the Gain. Increasing the signal level before it is filtered is one thing; increasing it after having filtered it is another.

On the right you find the track volume, on the left the gain pot. The red button is the phase inverter which flips the signal by 180°.

Finally, the analog switch serves to tell the plugin whether to behave as a simulator of the real console (with all the relative problems of analog) or as a digital filter, without ripple phenomena in the stopband.

TheSimon wrote:

Excellent Suite that I often use for my mixes.

il SSL Channel Plugin from Waves simulates the channel strip of the SSL 4000G console.

I am attaching a photo and will comment on every Pot.

At the top left, under "FILTERS" which also includes the EQ, you will find the high-pass filter on the left and the low-pass filter on the right with the white pots. The one on the left cuts all frequencies below the one indicated by the Pot. If you set it to 70 Hz, for example, you go to remove all frequencies below 70 Hz. The low-pass filter works exactly the opposite way. By selecting a frequency with the Pot, the entire band below that frequency will pass undisturbed and the part above will be cut. Obviously, this occurs within a transition band that we do not know. However, know that the filter is higher quality the smaller its transition band is. It would be interesting to run some tests to see the width of the transition band. I should note that since we are in digital, it wouldn't be difficult to make a filter with a transition band practically equal to zero; however, since these plugins must simulate the behavior of the filters present on the analog console, it is very likely that there is a transition band different from 0.

By selecting "OUT" with the Pots, the filters are disabled.

Moving down we find (red pots) the gain pot on the left and the center band selection pot on the right. We are obviously within the high-frequency (HF) equalization section. The bell button serves to determine the band selectivity (the Q factor). Activating it makes the filter more selective, deactivating it makes it wider. Moving down further we find HMF and LMF (High-medium frequencies and Low-Medium frequencies), i.e., the mid-high and mid-low sections. Here you can also adjust the Q factor.

Finally, with the black Pot, you have the low frequencies, for which the same applies as for the highs.

Below EQ TO you have two buttons. The one on the left is used to turn the EQ on and off; specifically, by selecting ByPass it turns off, and by deactivating it, it turns on. Nearby is the DYN S-C button, which would be the Dynamic Side-Chain in case you wanted to use the SSL plugin in sidechain with some other plugin.

At the top right we have the DYNAMICS section.

The three white pots are for compressor adjustment. In the first one, you can choose the compression ratio. The numbers you see are obviously the value of the denominator: 4 stands for 4:1. To the right you find the pot for adjusting the compression threshold; further down on the left is the release time. You will then see that just below there is a small F-ATK switch, which stands for Fast Attack. On a real SSL 4000 console, this button does not exist and fast attack is selected by turning up the pot. Since in the compression stage of the SSL 4000 Channel Strip there is no POT for adjusting attack times, this switch allows you to select fast or slow attack time.

Further down (green pots) you find the expander, which works the opposite of the compressor; that is, above a certain threshold that you can set, instead of compressing as the compressor did, it expands the waveform. By pressing the GATE key, you can activate the GATE by selecting the operating threshold with the Threshold pot.

Then you have two meters. One gives you information on the gate threshold and the other on the compression.

The Bypass button, as in the case of the filters, serves to disable the module; the other, CH OUT, instead serves to move the Equalization stage post-EQ. In essence, even if you see the compressor on the left and the EQ on the right, the signal is normally compressed first and then equalized, following the signal routing of the analog SSL console. By pressing that button, the compression/expansion stage is moved after the equalization stage.

On the right, the INPUT and OUTPUT buttons are used to select whether we want to consider the signal before it enters the filters or after, and this is essential for the Gain. Raising the signal level before it is filtered is one thing; raising it after having filtered it is another.

On the right you find the track volume, on the left the gain pot. The red button is the phase inverter that flips the signal by 180°.

Finally, the analog switch serves to tell the plugin whether to behave as a simulator of the real console (with all the relative issues of analog) or as a digital filter, without ripple phenomena in the dark band.

Excellent!!! Congratulations!! You are giving me a huge hand!!!

Great, detailed, precise and of great help everything you have written!!!

Thank you!!!!"}

Compliments to TheSimon for the deep knowledge of the Waves SSL plugin, and also for having demonstrated that he surely knows the original one as well.

I just wanted to add a piece of advice for jenn: I believe it is not worth starting to learn how an equalizer works in general from this plugin; first of all, one must understand how it works on a convenient digital equalizer, in this case Logic's, which is also very well made; in this case, learning to build and shape curves, feeling the effect, and also seeing the result on the integrated spectrum analyzer.

Then all the points excellently explained by TheSimon will make sense; in fact, the plugin in question is used not so much for the result that can be obtained in terms of pure equalization, but because it simulates a famous circuitry that had a particular sound; for some reason, trained ears might already notice differences in sound just by passing the signal through it without doing anything, precisely because it goes through an emulator of a specific circuit...

Sorry for the typos... I'm writing on iPad...

Liukkes wrote:

Compliments to TheSimon for the deep knowledge of the Waves SSL plugin, and also for proving that he surely knows the original one as well.

I just wanted to add a piece of advice for jenn: I believe it is not worth starting to learn how an equalizer works in general from this plugin; first of all, one must understand how it works on a convenient digital equalizer, in this case Logic's, which is also very well made, learning in this case how to build and shape curves, feel their effect, and also see the result on the integrated spectrum analyzer.

Then all the excellent points explained by TheSimon will make sense; in fact, the plugin in question is used not so much for the result that can be obtained in terms of pure equalization, but because it simulates a famous circuitry that had a particular sound; to an absurd degree, trained ears might already notice differences in sound just by passing the signal through it without doing anything, precisely because it goes through an emulator of a specific circuit...

Thank you Liukkes (I know I actually asked too much starting with SSL).

Now I am studying Logic x's parametric EQ as I was missing some basics that I am learning bit by bit; I refer you to a very fresh discussion on the subject -> http://www.pianoconcerto.it/forum/index.php?/topic/5870-equalizzatore-parametrico/

Liukkes wrote:

absurdly, prepared ears might already notice differences in sound just by passing the signal through it without doing anything, precisely because this travels through an emulator of a specific circuit...

This reasoning is not absolutely true... What you are saying is only valid in the analog domain. Plugins like the Waves SSL simulate (they do not emulate) the behavior of analog circuitry digitally only "at operating levels". If on the plugin you leave everything at 0, the signal isn't even touched; what you have as input is exactly what you get as output. Obviously, in digital, loading a plugin that does nothing only results in the use of system resources, and on a two-track session it causes no problem, but it is good to be careful when working on sessions with a large number of tracks, especially software instruments. At the level of the signal, what you have as input and output from the plugin is exactly the same.

So... I might be wrong, but I've noticed that some studio engineers insert these plugins, even if they don't apply any adjustments; I ventured to ask why, and the answer was:

These Plugins simulate the signal path of the original SSL console, so just the fact of passing the signal through this simulated circuitry, on which, among other things, Waves has done painstaking work, "should" contribute its part.

If we think about it, this reasoning holds up; I don't understand why you maintain that zero is equal to zero in a context like this.

In fact, I would ask you for clarification on the matter, because the math doesn't add up for me: if we did the same with a digital equalizer, then yes, because being digital and treating sound without any simulations of discrete circuitry, zero is indeed zero.

But the SSL, in addition to providing its classic adjustments in terms of both equalization and dynamics, should (according to some engineers) also simulate the circuitry even at zero, with all its merits.

Now, I admit I don't have your level of expertise in the matter, but I can't help but think of Arturia's Mini Moog; in fact, the synth offers oscillators, filters, and other adjustments that are essentially very simple and are present in a sea of plugins—DCO, filters, resonance, cutoff, ADSR, DCA, etc... except that the sound is different from any pseudo-analog synth; that sound, besides making the various sections of the synth work like all synths do, does so by simulating the original Moog circuitry, and thus that particular full timbre characteristic of dear old Moog emerges.

Excuse me, but I am particularly interested in this topic and would be interested in an in-depth discussion... thank you.

So I recently spoke with the sound engineer in question, and the error is mine, in the sense that he told me he performed this maneuver with UAD plugins, and that Waves work differently, so he essentially thinks like you.

However, I still don't understand how Waves simulates the SSL circuitry then.

Let's use the informal 'tu', I'm still young, don't make me feel old...

We can discuss choices made in a mix as much as we want (within certain limits, anyway), but in this case, the matter is mathematical and therefore unequivocal; it is not subject to opinions, and if you want, I can prove it to you in 10 seconds. In any case, to answer your curiosity, normally when developing a plugin that simulates the behavior of an analog circuit, one can choose to operate in two ways.

The first is to create a plugin that tries to best simulate the corresponding analog circuit, thus performing painstaking analysis on the analog circuit.

The second is to do the same while trying to improve it, therefore starting directly from a digital analysis and seeing where it can be improved.

Taking the SSL EQ as an example, an EQ is nothing more than a set of filters having specific circuit characteristics. Anyone who has built even a simple RLC filter knows very well that reactive components have, among other things, a phase shift effect on the signal. No problem if you are doing mastering, i.e., working on a single track; the problem might arise when you are mixing because introducing a phase shift to the original signal could cause cancellations.

There is also another problem, very visible even in low-pass or high-pass filters. Creating a very selective filter implies increasing the number of poles, but increasing the number of poles increases ripple issues in the stopband. That is, in the stopband, by definition, no signal should pass, but if you have ripple, it means that small amounts of signal will pass anyway, and that is why good equalizers cost so much money—precisely because to limit, rather than eliminate, these problems, the designs are often very complex.

Now coming to the analog SSL, you understand well that even though the signal is not being processed (i.e., if all filters are set to 0), it still passes through active and passive electronic circuits that somehow modify it.

If I were in the position of someone designing a plugin, would this be an added value for me or not? Obviously, the answer is that this is one of the problems with analog, and if I can avoid it in digital, I will gladly do so, also because the CPU state would change based on unrequested processing, thus unnecessarily increasing latency just to process a signal that I actually (since I leave everything at 0) do not want to process.

The simulation of the SSL circuitry therefore occurs "at operating temperature" [a regime], meaning when parameters other than 0 exist, and there are two approaches to doing this...

The first one I mentioned is to analyze the SSL circuitry in analog, i.e., by starting to send different types of signals (sine, triangle, and square waves which have different Fourier transforms) into the input using a signal generator, and seeing how these are returned at the output as frequency varies. Since we will then have to convert this analog signal into digital, we will need to extrapolate an algorithm from the analysis that simulates the behavior of the analog circuitry.

The second approach is faster and cleaner, even if it does not return exactly the behavior of the analog circuit, precisely because it is not an Analog-to-Digital approach but a Digital-to-Digital one. In essence, with circuit diagrams in hand, the circuit is reconstructed in Spice or in EveryCircuit, and from there, the tests from the previous approach are performed using computer simulations. In this case, many problems are eliminated, but obviously the filter's fidelity will be lower because if we were to measure the value of a resistor on an analog circuit, that resistor would probably not be 100Ω but 101Ω. You understand well that with a single resistor the problem doesn't arise, but with the entire circuitry made up of hundreds of resistors, inductors, capacitors, transistors, etc., even a 1% error begins to be felt. It must also be said that the values of electronic components vary over time, and this could also skew the analysis.

Regarding Universal Audio plugins, sincerely, I have never used them so I cannot express an opinion, and even if it seems at least strange to me, anything is possible. In light of what has been said, you will have understood that it is simply a matter of making choices during the plugin design phase. But bringing analog problems into digital, between us, does not seem like a smart choice to me, and that is exactly why it seems strange to me...

Ok to the "tu"... and thank you for the extremely exhaustive response; in fact, I'll take this opportunity to compliment you on your expertise. I haven't been on this forum for very long, but I think you will see me often.

Regarding what you wrote, I think I understood that the choice of Waves—which was also confirmed by my sound engineer friend—was to have any potential circuit simulations intervene only at the moment when one begins to make various types of adjustments to the plugin itself, in order to avoid unnecessarily overloading the system and thus avoiding generating phase errors. On the other hand, if we are talking about UAD, then the discussion is different because, by relying on external hardware, the circuits have been meticulously reconstructed rather than simulated, so they provide their own sound without burdening the system.

In this particular situation, could we therefore assert that perhaps UAD has simulated the famous Solid State circuits better than Waves?

Typos due to the iPad autocorrect which is more of a hindrance than a help...

Don't worry about the errors.

Regarding UAD, I don't feel comfortable expressing an opinion because I have never tried them and on these matters I am like Saint Thomas: if I don't see it, I don't believe it, and if I don't personally test something, I prefer to stay quiet. I know friends who have them and they told me that Waves works better, so out of my own laziness, I trusted them and never bothered to take measurements. However, this is not a given, and I reserve the right to perform tests if I get the chance. If Universal Audio has done things the way they should be done, they should have used FPGA architecture for runtime filters. If that is the case, then the "cost" in terms of latency is zero, so they could have implemented this possibility even though, I repeat, I don't see the ultimate utility in it.

Livio Argentini, a very dear friend and pioneer in the field of analog outboard gear—whom you should know if you are in the industry—was very helpful in assisting me with my thesis work (which was specifically focused on plugin development).

He, like all audio hardware design engineers, believes that the fewer signals passing through, the better, precisely because they aren't "dirtied" by analog components. In fact, for these reasons, he is a pioneer of modular systems. He has his own philosophies in the audio field, whether or not they are shared, but I believe everyone agrees on this point.

If this can be avoided in the digital realm, I don't see why not to do so. In my opinion, it is therefore not a matter of "better" or "worse," which implies objective technical reasons; I think the point of interest should be shifted to "more faithful or less faithful" to analog, and in this case, the issue is purely subjective.

I read with attention (and admiration), noting the high level of preparation on the subject!

Without lingering, I immediately return to my extended parametric EQ (kindly offered by Logic X) with the hope (after many hours of hard work and lessons) of learning the art of mixing, in search of that infamous and annoying frequency to emphasize or cut!!

Thank you!!!

Hi jenn, but what do you mean by annoying frequency, to be emphasized or cut? On which instrument?

Liukkes wrote:

Hi jenn, but what do you mean by annoying frequency, to be emphasized or cut? On which instrument?

Great question!!

I am trying to find the right balance in terms of frequencies on a rather loose bass drum (60s Beatles style)

Mmmm... So, for the kick drum, I think you should first treat it with a compressor, in order to emphasize but keep the dynamics under control; then, if there is ambience in the sample you are using, the compressor could bring it out while simultaneously leveling it, perhaps by acting on the release.

For equalization, first I would analyze the Logic spectrum analyzer, eliminating all irrelevant zones; you will see that you will use the low-pass filter heavily, but not too much, otherwise you might darken the sound—absurdly enough for a kick drum, despite the lows being clearly present, you also have peaks higher up that provide, precisely, the mechanical noises of the kick itself.

Knowing how to use it properly, just like the SSL in the thread, one could then do both things: both the EQ and the dynamics, since it has a compressor.

I haven't reached studying the compressor yet, I am stuck at the equalizer

Does the spectrum analyzer help me understand which frequencies to emphasize or attenuate?

Gianluca's indications are absolutely correct. Not by chance, in some situations, a pinch of dereverb can be added after the compression stage.

Thanks TheSimon, to answer jenn's question, the spectrum analyzer specifically indicates the frequency spectrum passing through the expected range, 20 - 20,000 Hz, and their intensity; it doesn't tell you what to add or what to remove, it is just a precious indication to have a precise idea of which frequencies are used by that specific sound timbre.

It is usually recommended to remove those not used by the timbre, but the most important thing for sculpting sound is listening to it; in fact, as mentioned before, if you press the first low note on the piano, you will have a peak in a very low spectrum, but you will also have other peaks at higher frequencies—these are the harmonics of the sound itself, those that characterize one timbre rather than another. So, cutting frequencies yes, but always listening to what comes out of it, because you could unintentionally change the very timbre of the instrument itself.