GGWP's DBX 286s guide for spoken word/voice/podcasting/streaming -- Setup help thread!

The DBX 286s has a learning curve, so I've written a guide that everyone can contribute to and ask questions in! I'd love to hear your advice on this great tool!

TLDR: TBD

Here are my setup tips:

Plan to iteratively change your settings many times to achieve your final goal. Each setting affects the others, and it is not realistic to expect perfection on the first try.

Plan to record a spoken sample for each processor at its lowest and highest settings, then listen to them to see what extremes the processor can go to. You'll find that the 286s has smart and reasonable limits, so a lot of these processors can be driven hard and still not sound weird. Listen to what they do, then choose what you'll do with them.

Less is more. It's better to record the source audio in a natural form because there's no undoing what you record there. You can always process the recording afterwards in a different tool, or tweak the 286s settings later. For now, just make it sound like a better version of reality, not a different version of reality.

Take a picture of the front panel with your phone when you find a good configuration, and also right before you CHANGE a good configuration. Anything worth having is worth backing up!

Stage 1, Mic Preamp input gain: Set this so your loudest moments hit the yellow LED. DBX staff can confirm or deny me on this one, but I find that it's better to turn the mic preamp gain as high as possible at the start of setting up your 286s, even if it means turning the output gain way down. I find this for two reasons: 1) it gives the various processors in the middle (compressor, etc) more juice to work with which improves their results, and 2) changing the input gain in the future affects the processors, which could mean re-tweaking the whole setup. So pump up that input gain as high as you can now, and as each new processor you enable lowers the final gain, turn that input gain up even more to compensate. Then, at the end, use the output gain knob to finalize the output to the right level. Just don't let the red clipping LED turn on.

Stage 2, Compressor: Put simply, this lowers the volume when you get too loud. The DRIVE is how low it will go (more drive is lower volume). DENSITY is a weird name but it's how fast it goes back to normal (higher density means more speed returning to default volume). Having a slow/low DENSITY sounds good but will interfere with testing the other processing features of the 286s right now, so set your DENSITY to 10 at first, and only change it down to 0-6 when all other tweaking of the 286s is complete. Test your DRIVE at OFF and 10, both while making your quietest (and furthest from the mic) and loudest (and closest to the mic) sounds. If the volume of those two extremes is too far apart for you, raise your DRIVE. Lower the output gain knob (stage 6, not 1) this whole time to make sure you aren't blowing your ears off or clipping at your analog-to-digital device/computer. The DRIVE at 10 is perfect for someone who varies how far they are from the mic, how loud they are, which direction they speak, and so on. A low DRIVE (no compressor at all) is not what most of us need (that'd be for an orchestra or a voice actor in a booth).

Stage 3, De-Esser: Most of us don't need this one. Only turn it on if you have noticed or been told that your speech has too much "sssss" sound in it. It can also be used to tone down a lisp. The 286s will detect the FREQUENCY you choose and lower the recorded volume a little bit, according to the THRESHOLD you set. You can see the peak frequency of your "s" sounds by recording "sssss" in a tool (like the free Audacity) and analyzing the spectrum which will show that it's around 6-8 kHz. Think of this as a second Compressor which examines only the area around one high frequency.

Stage 4, Enhancer: This is like Bass and Treble controls. And like the De-Esser, it's purely a taste thing. Since microphones fail to catch more low frequency the further away you are, people who are further from their mic can use the LF DETAIL knob to add that missing LF (bass) back in. But once you play with the LF DETAIL knob and hear yourself as a radio announcer who swallows his mic, you'll start to notice other people online who add too much bass to their voice, and you might agree with me that "less is more" on this one.

Stage 5, Expander/Gate: This is another relative of the Compressor. Put simply, this lowers the volume when you get too quiet. Also known as a noise gate, this is for muting background noise. It also works in tandem with the Compressor to prevent "noise rushups", which are when the Compressor's muting effect wears off and the background noise becomes audible. We use this to make the background noise no more noticeable to the listener when you're not speaking as when you are. The more you need this, the harder it is to set correctly, because those people have background noise that is close in volume to their voice, which makes it easy to accidentally cut off the voice or have other weird behavior. The THRESHOLD value determines what decibel level the audio needs to go down to before the Expander/Gate drops your volume off a cliff. The RATIO determines how fast the volume is dropped, and how hard it hits the ground. The proprietary mathematical formulas that go into this would boggle the mind. The RATIO actually controls so many different variables that the 286s manual refers to it as two separate functions in one: a "downward expander" for voice when RATIO is set to under 2:1 (the bottom half), and a "switch gate" for instruments when RATIO is set above 2:1 (the top half). This means that even if you have the noisiest background ever, you still keep RATIO under 2:1 for voice and never go above that... and the 286s manual actually recommends a RATIO of 1.4:1 to 1.8:1 for voice. Try setting the RATIO to 1.6, then adjust the THRESHOLD so the background noise when you're not talking is perceived as equally quiet as when you are talking (and thus distracted from it). I'm not sure how useful the LED is, but I set THRESHOLD so that only my loudest sounds turn the +LED green. Ultimately this processor results in the Expander and Compressor constantly fighting each other, but that's a good thing, so I prefer to think of it as massaging. Your quietest moments should sound quiet but normal, and your voice shouldn't ever cut off. Then any amount of background noise quieting when you're not talking is gravy. Less is again more on this one.

Stage 6, Output gain: I know of two free and easy ways that anyone can set their output gain to industry-standard levels. Audacity and OBS Studio are both free and open source software. Both show a db meter for the mic input. Clicking the meter in Audacity will begin monitoring without recording. A small mark will show the highest recent volume level. You want that max input mark to be around -9db. In OBS Studio this is shown as the top of a yellow bar between -20 and -9, denoting that they want your normal performance to be between -20 and -9. The -9 is a good ballpark figure that leaves room for error so you don't clip your or other peoples' equipment.

Once you've done all this, go back and do it again. My results keep getting better the more I experiment and learn!

GL HF out there!