Inflators, exciters, saturators and all that

[Guy Rowland]

There was a thread recently that mentioned Oxford Inflater as something unique and wonderful. I've had it for many years and I've more or less stopped using it - I find that for hyping things other products get me there better.

I understand that under an oscilloscope it's basically a wavefolder. It has a dramatic effect on a sine wave, but the more complex the waveform the less noticeable it is.

I thought it would interesting to spend a hour comparing some tools on a full commercial mix, crucially ALWAYS gain compensating.

Fist was Inflator, and I turned the effect up to 100%. The original master was quiet, so I added gain beforehand. Then by ear I compensated the output so that it sounded the same loudness-wise. Once I'd done that, the honest truth of it was - I couldn't tell the slightest difference between bypassed and in-circuit, and I'd have definitely failed a double blind test. Also there was no difference in Loudness reading either according to iZotope Insight - it didn't magically increase perceived loudness without being detected.

I found much the same thing on Zynaptiq's Intensity when I demoed it. It sounds amazing until you gain compensate, then it sounds like nothing at all that you cant do with other basic tools.

I then went to Waves' BB Tubes, which offers odd (Beauty) and even (Beast) harmonics. Up until I hit the red I couldn't hear any difference here either when gain compensated. However there was very clear difference when pushed, with the nice and nasty distortion was clearly audible.

But distortion - even nice - isn't really what I was after. So next up was iZotope Neutron 3's Exciter. What does this do? Distorts, more obviously and with different flavours, and also compresses. I can clearly hear it working when pushed, but I didn't like it as a general excitement thing because when I perceived a change I just perceived a distortion with compression.

At the end of the little test, I'm pretty cynical about all of it. Compression is available in a billion nice and nasty variations, obviously. So is distortion. But I don't think there's a whole lot of magic beyond that that will take a master and make it sound louder for the same measured loudness or more exciting without distortion. "Warmth" is just odd harmonic distortion I think.

Am I wrong? What am I missing? Be honest with me, is it just old ears?

[Lawrence]

Try the TDR Limiter GE, a Piet recommendation. I’m definitely able to push my loudness with it before things distort.

The Inflator does nothing for me personally and never has.

BB Tubes are ok for a little warmth. I like Satin better.

[Guy Rowland]

Try the TDR Limiter GE, a Piet recommendation. I’m definitely able to push my loudness with it before things distort.

But it's just a comp/limiter, right? I have several that can push very well without distortion, but of course they proportionally increase technical loudness. The magic trick would be - can it make things sound louder without making it technically louder, or more exciting in a way that isn't just distortion?

[Lawrence]

Yeah, doubtful.

[wst3]

it depends on what you are trying to accomplish.

If you just want to sound louder the easy path is a compressor pushed till it flatlines the waveform. Energy density. Sadly it will sound like the remnants of a horses breakfast.

For other things there are other tools.

Exciters are an interesting category. I've used them since the days of hardware, like the Aphex Aural Exciter. Used judiciously it could add a little sparkle, or even a little clarity, and yeah, you got the impression of being a little louder - again energy density, but this time that density occurs in the upper octaves where the harmonics live. My favorite trick was to apply a LOT of aural excitement and then send it through a dynamic low pass filter. Instant noise reduction, and pretty effective at that, with minimal artifacts. Of course we ca do a LOT better with more modern toys.

My experience with Zynaptiq Intensity is mixed - it can do a lot of the same things the old Aural Exciter did, but better, and less obvious (which I guess is better?). I do use it from time to time, mostly on tracks that I want to bring out a little bit (and only a little bit).

My experience with the Oxford Inflator was "meh". I think my expectations were set a wee bit high, but I was disappointed.

For me I think it comes down to tools looking for a problem to solve. But then I am not competing with a lot of over-compressed mixes. (ask me some day about my time in broadcasting!)

[Piet De Ridder]

(...) or more exciting in a way that isn't just distortion?

I’ll often expand audio before adding other processing plug-ins. And drastically so. Because in my experience, it’s not those processing plug-ins that are the problem, it’s the source audio most of us work with — samples — which is. Samples, by way of being, are the main reason why so many plug-ins disappoint, in my opinion. (The same plug-ins which, when used on live audio, will produce excellent results.) There’s no life in samples. None whatsoever. So there’s also nothing for processors to work with. With the result that saturators, exciters, tape simulators and especially most dynamic processing tends to sound very flat and lifeless as well. (And that’s a sound I really, really hate.)

Apply a compressor/saturator/pre-amp/simulator/exciter/whateverator to, say, a sampled drumkit (or any other kind of sampled audio, for that matter) and then apply that same plugin to a recording of a real drumkit, and it’ll sound like you used an entirely different processor. It's a completely different world of sound. Amplifier & cabinet plug-ins also respond totally differently to real guitars then they do to virtual ones. There’s simply no comparison. And it’s all because of that huuuuge difference between dead and/or sterile audio (samples and/or modelling) and rich, living audio (recordings of real instruments, or live input).

I’ll do anything to increase the dynamic contours of sample-based audio and inject some semblance of life and organic, non-linear energy into sample tracks. And as soon as I’m making noticeable progress with that, I find that subsequent processing — even standard EQ’ing — sounds much more satisfying too. To my ears anyway.

Next to expansion (which really is on just about everything I record and mix), another tool I frequently use is a tremolo. Tremolo, yes. Don’t laugh. The Goodhertz Tremolo, to be more specific. You can shape the envelope of its tremolo effect in great detail and you can also choose which part of the audio it will affect. With those two options it’s perfectly possible to make audio behave a bit more sprightly in sync with the tempo of the track. Works really well. (A similar result, and much more of course, can be obtained with Cableguys’ ShaperBox, a fantastic piece of software and, in my world, an essential plug-in.)

Another good trick is to use an audiofile of speech as the side chain input for any type of processor that accepts side chain input. I do this a lot. Any kind of spoken word will do. The nice thing about spoken word is that it’s completely irregular (with random silences and random bursts of energy) which results in completely irregular and unpredictable behaviour of the processor that listens to it. You have to experiment with this a little bit before you get the hang of it, but trust me, it works. Also: key is not to overdo it. Subtle random changes, in some cases almost subliminal ones, in whatever parameter that is controlled by the side chain, is what we want.

A coating of saturation just for the sake of saturation — or for the sake of the preposterous idea that saturated audio somehow sounds more glued, more warm and therefore more appealing — is something I *never* do. Really don’t like it. The last thing I look for when mixing, is ‘glueing’ a sample-based production because samples, again by way of being, are more than sticky enough already all by themselves (giving you that typical smeary, porridgy summed sound). In fact, a big part of my mixing routine is to find ways to UNGLUE the stuff so that the mix has separation and the sounds can breathe.

But again: in irregular, random doses, tastefully applied, saturation can indeed be quite nice. If I were forced to keep just two saturators, it would be FF’s Saturn 2 and Goodhertz Tupe. Both of these are not just very fine saturation tools (capable of many different flavours) but surprisingly capable dynamic tone shapers as well. Saturn 2 is unquestionably the more powerful of the two, but Tupe (or ‘TupeWow’, as it since has become) has a certain tonal stamp and adds a certain flavour and texture which I happen to really, really like.

__

[Tanuj Tiku]

This is a very interesting topic and as we know there are many options these days. I will try to make some audio examples and see how we all feel about things.

I agree with Piet in that samples have an inherent disadvantage when working with plugins. One of the main reasons in my opinion is that they are already pre-processed. Tape saturation, hardware pass through and many other processes have already been applied. If anything, I find myself finding ways to keep the integrity of the sound intact by making cuts rather than boosts. You sort of have to live with what the sound is. Any signal that has been fairly processed carries that risk of not being very favourable to further tweaking.

As an example, let us take the case of CSS. The library has a dark and dull sound to begin with. You can remedy this in a few ways but only within reason. Exciters and EQ's can help bring out things or change the balance but what it cannot do (tastefully) is really add something that isn't there. So, if you keep pushing the HF in CSS with exciters, you will start encountering a little unpleasant noise. Whereas a library that is already bright (Cine Strings for example) will give you a favourable result because you either need to boost just a little bit or in many cases cut the HF.

I also like to remind myself that exciters and other such tools only contribute about 5% to the overall result. It isn't something that will completely change the sound and if you are needing to push it excessively then it is either an 'effect' or you will run into clarity issues.

What is fun however is that after many years of doing this, you can lean into that 5% and passionately discuss tools and their effect within that 5% resolution. But it is useful to not get carried away too much!

[Guy Rowland]

Piet and Tanuj - some video examples would be fantastic here if you have the time. Particularly the notion of expansion first, that's new to me, as is the wild sidechaining idea.

I am so cynical / sceptical about it all. Your last sentence chimes with me, Tanuj, though I usually put the figure at 1%. But I'm always interested in being proved wrong, I'm definitely curious about hearing samples->saturation vs samples->other->saturation.

[Piet De Ridder]

Would love to do some videos, Guy, and I would have done so already if it weren't for the fact that, ever since I upgraded to Sonoma, I can't get Screenflow to work decently. Recording video is no problem, but there's always something wrong with the sound. Either there is none or, when there is, it's severely distorted. Spent already many an hour trying to get it to work, but my efforts have so far all been fruitless. Or bootless.

I'll have another go at it one of these days, hopefully with better results. The moment I'm able to make a video, I certainly will do so.

__

[wst3]

An interesting observation Piet!

I can't remember who, but several years ago a friend and I were discussing his extreme disappointment with various plugins. I don't remember the details, but I remember being really quite happy with the same plugins. Eventually we dug deep enough to discover that I was processing 'live" recordings and he was processing samples.

Out of sheer laziness I never went any further with it. Perhaps I should, and take a look at the crest factors of the various sources.

[Tanuj Tiku]

I meant to post some examples here long ago but forgot about it.

Anyway, I have created some examples here:

https://www.dropbox.com/scl/fo/lmac8urc ... 90i0n&dl=0

There are two folders. One is a quick electronic example with a groove, a pad and some glitchy sounds. The other is an excerpt from my demo mock-up to Symphonic Dances for a new template I built a few months ago.

In each folder the title will tell you which is the original source file which will playback as is. The others are labelled with famous processors. Then there is a file called 'Multi'. This means that I have used a combination of plugins to achieve a more specific sound, either using bands and/or other methods. In the case of the electronic example, I did some custom processing on each of the three elements. For Symphonic Dances, I only had the master.

Please keep in mind that Symphonic Dances already has processing from the template.

My opinion is the same as before, these tools add less than what musicians typically think they do. What is more important is how, when and in which context you use them along with other steps. The more musical decisions, including filter movement and performance will have a greater effect. You may find that the multi examples seem to have been affected more than others. Again, a more specific and intention-ed use, leaning on the sweet spot of the said plugins will yield a more obvious result.

When I buy plugins today, I largely look for the following things:

1. It is from a reputable company and the design is robust. It will not further degrade my sound unless that is what it is supposed to do.

2. It does something I cannot otherwise achieve on a purely technical level and/or has new functionalities. For example, the inflator is now very outdated if you want to do multi-band processing and have multiple algorithms to choose from. Saturn and other such processors are now way more advanced and nuanced.

3. A breakthrough has been achieved or it offers genuinely new colours which I don't have access to.

I would rather today spend money on synths and other sound creating sources than plugins unless there is something that catches my eye.

Plugins used for all of the examples:

* Oxford Inflator (I rarely use it now but it was mentioned in this thread and I have it!)

* FF Saturn 2

* U-he Satin