PreSonus Blog

Category Archives: Uncategorized


Make a Splitter for Studio One Artist

One of my favorite Studio One Professional features is the Splitter, and quite a few of my FX Chains use it. If you own Studio One Artist, which doesn’t have a Splitter, you may look longingly at these FX Chains and think “If only I could do that…”

Well, you can implement most splitter functions in Studio One Artist, by using buses. All the following split options are based on having a track that provides the audio to be split, along with pre-fader sends to additional buses. Note that the track’s fader should be turned all the way down.

 

Normal Split

 

The Splitter’s Normal mode sends the input to two parallel paths, which is ideal for parallel processing. For Artist, we’ll duplicate this mode with two buses, called Split 1 and Split 2 (fig. 1).

 

Figure 1: How to create a Normal split in Artist.

 

The sends to the buses are pre-fader, and panned to center. One send goes to Split 1, and the other to Split 2. Now you can insert different effects in Splits 1 and 2 to do parallel processing.

 

Channel Split

 

The Channel Split mode also splits the input into two parallel paths. One path is for the left channel, while the other path is for the right channel.

 

Figure 2: How to create a Channel Split in Artist.

 

The setup is the same as for the Normal Split (fig. 2), except that each bus has a Dual Pan inserted. The Dual Pan for the left channel has the Input Balance set to <L>, while the Dual Pan for the right channel has the Input Balance set to < R>. I recommend the -6dB Linear Pan law so that if you pan either of the buses, the level remains constant as you pan from left to right.

 

Frequency Split

 

This is tough to duplicate, because the Splitter can split incoming audio into five frequency bands. If other DAWs don’t do it, we can’t expect Artist to do it. But, we can do a three-way, tri-amped split into low, mid, and high frequencies (fig. 3).

 

Figure 3: Tri-Amp Frequency Split.

 

This split is like the Normal Split, except that there are three buses and pre-fader sends instead of two, and each bus has a Pro EQ2 inserted. Each EQ covers its own part of the frequency spectrum—low, mid, and high (fig. 4). Using 6 dB/octave slopes doesn’t provide as much separation between frequency ranges as steeper slopes, but the gentler slopes are necessary to make sure the frequency response is flat when you mix the three channels together.

 

 

Figure 4: (Top to bottom) low, mid, and high curves.

 

The only filter sections we need to use are High Cut and Low Cut—you can ignore everything else. Fig. 5 shows the settings. All bands have 6 dB/octave slopes.

 

Enable the Low band’s Pro EQ2 HC (High Cut) filter, and choose 200 Hz for frequency. Enable the Mid band’s Pro EQ2 LC (Low Cut) filter, and set it to 200 Hz; also enable the HC filter, and set it to 4.00 kHz. Finally, enable the High band’s Pro EQ2 LC filter, and set it to 4.00 kHz. These frequencies are a good starting point, but you may want to modify the split frequencies for different types of audio sources. Just make sure that the low band HC frequency is the same as the mid band’s LC frequency, and the Mid band’s HC frequency is the same as Hi band’s LC frequency.

 

 

Figure 5: Filter control settings.

Granted, setting up these splits takes more effort than dragging a Splitter plug-in into a channel, but the result is the same: cool parallel processing options.

The Dynamic Brightener—Reloaded

In April 2019, I did a Friday Tip called The Dynamic Brightener for Guitar. It’s kind of a cross between dynamic EQ and a transient shaper, and has been a useful FX Chain for me. In fact, it’s been so useful that I’ve used it a lot—and in the process, wanted to enhance it further. This “reloaded” version makes it suitable for more types of audio sources (try it with drums, bass, ukulele, piano, or anything percussive), as well as less critical to adjust. It also lessens potential high-frequency “smearing” issues—the original version applied large amounts of boost and cut, with a non-linear-phase EQ.

Although the original version could have been built using a Splitter, I did a bus-based implementation so that it would work with Studio One Artist. This new version needs to use the Splitter (sorry, Artist users), but that’s what allows for the improvements.

Another interesting aspect is that by using the effects’ expanded view in the channel inserts, you don’t even need to open the effect or Splitter interfaces, to do all the necessary tweaking. This makes the reloaded version much easier to edit for different types of tracks.

How It Works

Fig. 1 shows the FX Chain’s block diagram.

Figure 1: The Reloaded Dynamic Brightener’s block diagram.

 

Splitter 1 is a normal split. The left split provides the track’s dry sound, while the right split goes to Splitter 2, which is set up as a Frequency Split. The Frequency Split determines the cutoff for the high frequencies going into the right split. Splitter 2’s left split, which contains only the split’s lower frequencies, is attenuated completely. Basically, Splitter 2 exists solely to isolate the audio source’s very highest frequencies.

These high frequencies go to an Expander, which emphasizes the peaks. This is what gives both the transient shaping and dynamic EQ-type effects. Because the high frequencies aren’t very loud, the Mixtool allows boosting them to hit the desired level.

Fig. 2 shows the initial Expander and Mixtool settings. But, you won’t be opening the interfaces very much, if at all…you don’t even need Macro Controls.

 

Figure 2: Initial parameter settings for the Expander and Mixtool.

 

Using the Reloaded Dynamic Brightener

In the short console view, open up the “sidecar” that shows the effects. Expand the effects, and set the mixer channel high enough to see the ones shown in fig. 3.

Figure 3: The Reloaded Dynamic Brightener controls.

Here’s how to optimize the settings for your particular application:

  1. Turn off Splitter 1’s output 1 power button. This mutes the dry signal, so we can concentrate on the brightener’s settings.
  2. Adjust Splitter 2’s Frequency Split to isolate the optimum high-frequency range for brightening. This can be as low as 1 kHz or less for guitars with humbucker pickups, on up to 6 kHz (or more) to emphasize drum transients.
  3. Set the Expander’s Ratio and Threshold parameters for the desired amount of brightening and transient shaping. Higher Threshold settings pick off only the top of the boosted high-frequency peaks; the Ratio parameter controls the transient shape. The higher the ratio, the “peakier” the transient.
  4. After editing the high frequencies, re-enable the dry signal by turning on Splitter 1’s output 1 button.
  5. Mix in the desired amount of brightening with the Mixtool Gain parameter. In extreme cases you may want to increase the level control at the end of the Splitter 2 branch, or the output level from Splitter 2 output 2, but this will be needed rarely, if at all.
  6. As a reality check to determine what the brightener contributes to the sound, turn off either Splitter 1 or Splitter 2’s output 2 power button to mute the brightened signal path.

 

This is a tidier, easier-to-adjust, and better-sounding setup than the original dynamic brightener. Download the FX Chain here—the default settings are for dry guitar, and assume a normalized overall track level. With lower track levels, you’ll need to lower the Expander Threshold, or boost output 2 from Splitter 1. But feel free to tweak away, and make the Reloaded Dynamic Brightener do your bidding, for a wide variety of different audio signals.

The Virtual “Back of the Tape Box.”

In the 20th century, tape reels came in boxes. Engineers wrote information about tracks, running order, timing, credits, and such on the back of the box. And because it was a box, you could fold up some sheets of paper and include lyrics, notes, and other information

These days, when you open a project, it’s just like you left it. But what mics did you use? How was the tone control set on the bass? And you got those loops from…which sample library? If you ever need to re-visit a track, fix a glitch, do an overdub, or weeks pass before you can finish a project, you’ll need to know these details. Let’s talk about taking notes, and while we’re keeping track of things, let’s also create a lyrics track.

 

Taking Notes

To access Studio One’s virtual “back of the tape box,” choose Song > Song Information.

  • The Info tab shows selected meta data, info from the Song Notes tab, and any image you uploaded to Song Setup. The image can show an album cover, but also a miking setup, an analog processor’s control settings, or a cool picture that inspires you.
  • For developing song lyrics, I keep the Song Notes tab open, and do my writing/editing there. Saving lyrics with a song is convenient. However, it’s also useful for other notes—session personnel, web site URLs with reference info, and so on.
  • Track Notes (fig. 1) is where you can include all track-related information. It’s ideal for info that’s not included in a preset, like the vocal mic of choice, analog processor control settings, and which pickup you used on a guitar.

 

Figure 1: Track notes are key to documenting a session.

 

 

 

Track Notes Access Shortcuts

In addition to accessing track notes in the Song Information menu, you can scroll down in the Inspector to the field below automation, and click on it to open the corresponding Track Notes (fig. 2). Or, right-click anywhere within a Console channel, and select Edit Note from the context menu. This is the fastest way to view or edit Track Notes, compared to going to Song Information, selecting a tab, and then clicking on the track label.

Figure 2: Open track notes from within the Inspector by clicking, or in a mixer Channel by right-clicking.

 

 

Furthermore, you can supplement Track Notes by entering information in the track name itself. Hovering over the name in the Arrange window or a mixer Channel (fig. 3) shows whatever you entered, which can be quite long if needed. This is useful for temporary notes, like if you recorded several similar parts, and need to differentiate among them.

Figure 3: You can see long track names by hovering over them.

Pipeline Docs

Analog processors…you love ‘em, right? But when they don’t have presets, and you’re using them with Pipeline, you’ll want to know how the controls were set. Pipeline’s image icon (lower left) allows uploading an image of control settings, while the pencil icon allows adding a note (fig. 4).

 

Figure 4: The rack of effects from my book “Electronic Projects for Musicians,” being used with Pipeline.

 

A photo larger than 1200 x 1200 will be scaled to fit the image space, but click on the image, and it expands to the original size. You may even be able to see all the controls on a rack-mount piece of gear.

If you have an analog synthesizer with so many controls they won’t fit in a picture, no problem. Pipeline doesn’t have to be used for its intended purpose. You can take a picture of the synth’s oscillators, another of the filters, another of the envelopes, Then, stack multiple Pipelines within a track, to be used solely as a picture gallery. However, each instance does add latency—when you don’t need to see the pictures, disable the track, and hide it to reduce clutter.

Of course, you could always put the images in a folder, and include that folder in your song folder. But it’s kinda cool that everything you need to know about a song can be stored within that song.

 

The Lyrics Track

A lyrics track is helpful, because you always know where you are in the song—not just in the “chorus” or “verse.” It can be located right above a vocal, so it’s easy to find and select a particular part that needs editing, overdubbing, deleting, etc. Although Studio One doesn’t have a lyric track per se, you can put one together in two ways (fig. 5). There are pros and cons to each approach.

 

Figure 5: Lyric track made with markers (top) and with a dedicated “lyrics track.”

 

 

Marker-based lyric track. Lyric markers are quite readable, and have useful songwriting functions:

  • Select the Marker track, open the Inspector, and jump to a particular phrase by clicking on it. It’s a quick way to get where you want to go.
  • Right-click on any marker, select “Create Arranger Sections from Markers,” and now you have the start of an Arranger Track based on the lyrics.
  • Right-click on a marker, and choose “Stop at Marker” to make sure that playback won’t continue past that point.

Event-based lyric track. This might be best if you already have a lot of markers inserted, and don’t want to add more. Create a dummy track, and populate it with events whose lengths correspond to phrases. One advantage is that you can color-code the events to help guide you through a vocal by emphasizing certain phrases. Another advantage is that if you zoom in or out, the Event will continue to span the length of the chosen phrase. A Marker is always anchored to the beginning.

The Ampire Sweetener

 

I work a lot with amp sims, and I love ’em. Well, except for one thing: Almost all of the ones that involve distortion exhibit what I call “the annoying frequency.” It’s hard to describe, but when it’s removed, you can definitely tell what’s missing—kind of a whistling sound, but without a sense of pitch. I have no idea why this particular type of artifact happens. It doesn’t go away if I increase the sample rate, choose a different pickup, switch guitars, or change my socks. And it’s worse with some amp sims than others; when reviewing a [particular amp sim by a PreSonus competitor] and I made the product manager aware of the annoying frequency, a subsequent expansion pack included a parametric equalizer so users could notch it out.

Granted, the 3rd gen Ampire is light years ahead of the 1st gen, as well as a lot of other amp sims out there. But we can still make it better, because the goal of the Friday Tip is to make things better, right?!? Besides, I’m an unreasonably picky guitar player.

Adding the EQ

Download the preset Ampire Sweetener.preset , and load it into the Pro EQ2 (Just open the .zip and double click the .preset file to install).  It will now have the curve shown in Fig. 1. Insert the Pro EQ2 after your Ampire amp and cab of choice, and the sound will magically lose its artifact.

Figure 1: Insert these notch filters after Ampire and its cab.

You have every right to skeptical—after all, you are reading this on the internet—so let’s listen to an audio example. The first half is with the EQ following the MCM800 amp and 4×12 MFB cab. The second half is with the EQ bypassed, but everything else the same. Both examples in the audio file are normalized to the same level. I’m pretty sure you’ll hear the artifact in the second half. Another way to hear the difference is play some power chords, and bypass the EQ stages to hear what they contribute to the sound.

 

The EQ’s curve isn’t only about the dual notches. There’s no need for super-high or -low frequencies, so those are reduced as well. Also, because the notches are in the high frequencies, adding a slight treble shelf compensates for the reduced amount of highs.

Figure 2: Reduce the high-shelf level and the two notches to sound more like the original amp sim/cab tone.

Now, this doesn’t mean you’ll like the end result better. You might prefer the sound with the artifact, and that’s fine. However, the artifact persists through the various amps and cabs. Inserting the Ampire Sweetener EQ removes that common element, which emphasizes the unique character, and tonal quality, of the individual amps and cabs. However, you can also “split the difference” by dialing back the parameters outlined in white (Fig. 2).

Finally, if you use other amp sims, many (if not most) will also benefit from one or two steep notches at the output. They probably won’t be the same frequencies, but they’ll likely be pretty close. The bad news is quite a few of them have far more prominent artifacts than Ampire, but the good news is the higher level makes it easier to hear them, so you can dial in their frequencies more quickly to notch them out.

Mid-Side Meets Reverb

The post on using mid-side processing with the CTC-1 garnered a good response, so let’s follow up with one of my favorite mid-side techniques: M-S reverb.

 

To recap, mid-side processing separates sounds in the center of a stereo file from sounds panned to the sides, processes them individually, then puts them back together again into stereo. It isn’t a perfect separation, because the mid is the sum of the left and right channels. Although this boosts the center somewhat, the mid still includes the sides. However, the side channel is quite precise, because it’s derived from putting the right and left channels out of phase—so the center cancels.

 

Applying Mid-Side Reverb

 

Before getting into how to make M-S reverb, here’s why it’s useful. Some productions have an overall reverb to provide ambiance, and a second reverb (often plate) dedicated to the vocal. The vocal is usually mixed to center, so it’s competing for space with the bass, snare, and kick. If they’re contributing to the overall reverb, and the vocal is creating its own reverb, that’s a lot of reverb in the center.

 

One popular fix is adding a highpass filter prior to the overall reverb, set to around 300 Hz. This keeps the bass and kick from muddying the reverb. However, it doesn’t take care of midrange or high-frequency sounds that are panned to center, like snare. These can compete even more with the vocal if they’re in the same frequency range.

 

While some reverbs let you tailor high- and low-frequency reverb times with a crossover, this doesn’t cover all the processing you might want to do, nor does Studio One’s Room Reverb include these parameters. Mid-side reverb, with different reverbs on the mid and sides, is a more flexible solution for customizing an overall reverb ambiance.

 

 

Assembling the Mid-Side Reverb

 

Download the FX Chain, or if you want to roll your own, start by dragging the MS-Transform FX Chain into a bus (of course, this also works for individual channels). Then drag a Room Reverb into each split (Fig. 1). The default reverb preset is a good place to start, but if the FX Chain is in a bus, remember to set the Mix controls for 100% wet sound. I also like to insert a Binaural Pan after the second MixTool to widen the overall stereo image.

 

Figure 1: Mid-Side Reverb FX Chain, which adds two Room Reverbs and a Binaural Pan to the MS-Transform FX Chain.

 

The reverb on the left handles the center, while the reverb on the right processes the sides. Lower the fader after the left reverb; Fig. 1 shows -6 dB, but adjust to taste. This alone will open up some space in the center for your vocal and its reverb. However, where this effect really comes into its own is when you tweak the reverb parameters for each reverb. For example…

 

  • If you still want reverb on the kick and low end, vary the mid reverb’s Length parameter. Shorter lengths tighten the kick more, while longer lengths give that Kick of Doom reverb sound.
  • Increase Length on the sides for a more atmospheric reverb sound.
  • Increase pre-delay on the sides, to make space for attacks on the vocal track. Consonants benefit from the extra clarity.
  • For this application, Eco mode sounds fine but try HQ as well.
  • Turn up the Binaural Pan after the second Mixtool. I often turn it up all the way, because it sounds great in stereo, and there aren’t any phasey issues of the output collapses to mono.

 

By adjusting the two reverbs, you can sculpt them to give the desired overall reverb sound. If you then place a vocal in the center with a sweet plate, I think you’ll find that the vocal and overall reverb create a smooth, differentiated, and conflict-free reverb effect.

 

 

 

 

 

 

 

Recording ReWired Programs

I had a bunch of legacy Acid projects from my pre-Studio One days, as well as some Ableton Live projects that were part of my live performances. With live performance a non-starter for the past year, I wanted to turn them into songs, and mix them in Studio One’s environment.

 

Gregor’s clever video, Ableton Live and Studio One Side-by-Side, shows how to drag-and-drop files between Live and Studio One. But I didn’t want individual files, I needed entire tracks…including ones I could improvise in real time with Live. The obvious answer is ReWire, since both Acid and Live can ReWire into Studio One. However, you can’t record what comes into the Instrument tracks used by ReWire. Nor can you bounce the ReWired audio, because there’s nothing physically in Studio One to bounce.

 

It turned out the answer is temporarily messy—but totally simple. First, let’s refresh our memory about ReWire. 

 

Setting Up ReWire

 

Start by telling Studio One to recognize ReWire devices. Under Options > Advanced > Services, make sure ReWire Support is enabled. In Studio One’s browser, under the Instruments tab, open the ReWire folder. Drag in the program you want to ReWire, the same way you’d drag in an instrument. (Incidentally, although you’re limited to dragging in one instance of the same ReWire client, you can ReWire two or more different clients into Studio One. Suitable clients includes Live, Acid Pro, FL Studio, Renoise, Reason before version 11, and others.) 

 

After dragging in Ableton Live, open it. ReWired clients are supposed to open automatically, but that’s not always the case.

 

Now we need to patch Live and Studio One together. In Ableton Live, for the Audio To fields, choose ReWire Out, and a separate output bus for each track. In my project, there were 9 stereo tracks (Fig. 1).

Figure 1: Assign Ableton Live’s ReWire outputs to buses. These connect to Studio One as track inputs.

 

Then, expand the Instrument panel in Studio One, and check all the buses that were assigned in Ableton Live. This automatically opens up mixer channels to play back the audio (Fig. 2). However, the mixer channels can’t record anything, so we need to go further.

Figure 2: Ableton Live loaded into Studio One, which treats Ableton Live like a virtual instrument with multiple outputs. 

 

Recording the ReWired Program

 

As mentioned, the following is temporarily messy. But once you’re recorded your tracks, you can tidy everything up, and your Live project will be a Studio One project. (Note that I renamed the tracks in Studio One as 1-9, so I didn’t have to refer to the stereo bus numbers in the following steps.) To do recording:

 

  1. In each Studio One track, go to its Send section and choose Add Bus Channel. Now we have Buses 1-9—one for each track.
  2. Our original instrument tracks have served their purpose, so we can hide them to avoid screen clutter. Now Studio One shows 9 buses (Fig. 3).

Figure 3: The buses are carrying the audio from Ableton Live’s outputs.

 

  1. Create 9 more tracks in Studio One (for my project, these were stereo). Assign each track input to an associated bus, so that each of the 9 buses terminates in a unique track. Now we can hide the bus tracks, and record-enable the new tracks to record the audio (fig. 4).

Figure 4: Studio One is set up to record the audio from Ableton Live.

 

  1. Now you’re ready to record whatever is in Ableton Live over to Studio One, in real time. 
  2. Fig. 5 shows the results of unhiding everything, narrowing the channels, and hitting play. At this point, if everything transferred over correctly, you can delete the ReWired tracks, remove the buses they fed, close Ableton Live, and you’re left with all the Live audio in Studio One tracks. Mission accomplished!

Figure 5: The Ableton Live audio has completed its move into Studio One. Now you can delete the instrument and bus channels you don’t need any more, close Ableton Live, return the U-Haul, and start doing your favorite Studio One stuff to supplement what you did in Live. Harmonic Editing, anyone?

 

Bonus tip: This is also the way to play Ableton Live instruments in real time, especially through Live’s various tempo-synched effects, while recording them in Studio One. And don’t forget about Gregor’s trick of moving Studio One files over to Live—this opens up using Live’s effects on Studio One tracks, which you can then record back into Studio One, along with other tracks, using the above technique.

 

Granted, I use Studio One for most of my multitrack projects. But there’s a lot to be gained by becoming fluent in multiple programs.   

 

Mid-Side Meets the CTC-1

I’ve often said it’s more fun to ask “what if I…?” than “how do I?” “What-if” is about trying something new, while “how do I” is about re-creating something that already exists. Well, I couldn’t help but wonder “what if” you combined the CTC-1 with mid-side processing, and sprinkled on a little of that CTC-1 magic? Let’s find out. (For more information on mid-side processing, check out my blog post Mid-Side Processing Made Easy. Also, note that only Studio One Professional allows using Mix Engine FX.)

 

One stumbling block is that the CTC-1 is designed to be inserted in a bus,  and the Mid-Side Transform FX chain won’t allow inserting Mix Engine FX. Fortunately, there’s a simple workaround (see Fig. 1).

 

  1. Copy the stereo track you want to process, so you have two tracks with the same stereo audio. One will provide the Mid audio, and the other, the Sides audio.
  2. Insert an MS-Transform FX Chain into each track (you’ll find this FX Chain in the Browser’s Mixing folder, under FX Chains)
  3. Create a bus for each track.
  4. Assign each track output to its own bus (not the main out). However, the bus outputs should go to the Main out.
  5. Add a CTC-1 Mix Engine FX in each bus.

Figure 1: Setup for adding mid-side processing with the CTC-1 to a mixed stereo file.

 

  1. To dedicate one bus to the mid audio, and the other to the sides, open up the Splitters in the MS-Transform FX Chains.
  2. Mute the sides output for the Mid track (top of Fig. 2, outlined in orange). Then, mute the mid output for the Sides track (bottom of Fig. 2, also outlined in orange).

Figure 2: One bus is Mid only, the other is Sides only.

 

Now you can add the desired amount of CTC-1 goodness to the mids and sides. And of course, you can vary the bus levels to choose the desired proportion of mid and sides audio.

 

Audition Time!

 

The following example is an excerpt from the original file, without the CTC-1.

 

 

 

Next up, CTC-1 with the Custom option on the Mid, and the Tube option on the Sides. Fig. 3 shows their settings—a fair amount of Character, and a little bit of Drive.

Figure 3: CTC-1 settings for the audio example.

 

 

 

If you didn’t hear much difference, trying playing Audio Example 1 again after playing Audio Example 2. Sometimes it’s easier to tell when something’s missing, compared to when something’s been added.

 

The more you know about the CTC-1, the more effectively you can use it. The bottom line is I now know the answer to my “what if” question: get some buses into the picture, and the CTC-1 can be hella good for processing mid and sides!

 

The Multiband X-Trem

Finally! People are becoming aware of the Splitter. Although the Splitter can act like a Y-cord or split based on channel, the coolest Splitter feature for me is being able to split based on frequency. This is what makes creating multiband FX Chains in Studio One sooo easy. 

 

Check out the audio example to hear a taste of what this can do with a pad and drum part. The first four measures are unprocessed, while the second four measures use the same Multiband X-Trem settings on the pad and the drums.

 

 

 

The block diagram (Fig. 1) is pretty simple—the Splitter creates three bands, Lo, Mid, and Hi, with crossovers at 332 and 854 Hz. (There’s nothing magical about those particular frequencies, choose what works best for the audio you’re putting through it.)

Figure 1: Block diagram for the Multiband X-Trem.

 

The real magic in this FX is the way the crucial parameters are brought out to the control panel that’s available in Studio One Pro (Fig. 2). However, Studio One Artist users can still load the FX Chain, and edit individual parameters. Although it’s more time-consuming, you can end up with the same sonic results.

Figure 2: Multiband X-Trem control panel.

 

How to Use It

 

This FX Chain assumes you’re going to sync it to tempo. Each of the three bands has a control to choose the Beat (tremolo rhythm) and Waveform, along with buttons to choose each band’s mode (Pan or Tremolo) and waveform Phase Flip. So far, that’s pretty simple.

 

The Mix section toward the right, with two knobs and their associated switches, is a little more complex. There aren’t enough control panel knobs to have a Depth control for each band, however in use, I’ve found that I usually adjust the depth for the Mid and Hi bands together, and the Lo band by itself. So, the Lo band has its own Depth control, while the Mid and Hi bands share a Depth control. There are also buttons to bypass the X-Trem for the Mid and/or Hi band. This is almost as good as having individual Depth controls, because you can remove depth for either band as needed.

 

We’ll close out with some additional tips…

 

  • The sawtooth wave defaults to positive-going (i.e., the level ramps up from nothing to full). Flipping the phase makes a more percussive effect.
  • A slow rhythm for the Lo band gives a sort of “rolling” effect. Faster speeds seem to work best for the Mid and Hi bands.
  • Feel free to jump in and do tweaks—like change the Gate or Step waveform levels, vary the levels of the bands within the Splitter module, or change the crossover frequencies.

Happy download! Grab the Multiband X-Trem FX Chain preset here.

A Slick Trick for Thick Kicks

Imagine if you had a mold for sound, the same way you can have a mold for Jell-O—and whatever you poured into your “sonic mold” took on those particular characteristics. Well, that’s pretty much what convolution processors do. When they load their “mold,” which is called an impulse response, it shapes whatever sound they’re processing. 

 

Studio One has two convolution processors. Ampire uses one to load speaker cabinet impulse responses. For example, when Ampire wants to sound like it’s going through a 2 x 12 speaker cabinet, it loads a 2 x 12 cabinet impulse response. The other convolution processor, Open Air, is optimized for creating acoustic spaces. So if the impulse is of a concert hall, sound processed through Open Air sounds like it’s in a concert hall. If the impulse is a blues club, the the sound takes on the characteristics of being in a blues club.

 

What’s perhaps not as well known is that you can load pretty much any WAV file into Open Air and use that as your sonic mold. So, this month’s tip is for the  EDM and hip-hop crowd, because we’re going to load big-sounding kick drums into Open Air. Then, we’ll use them as molds to turn wimpy kicks into giant, thick kicks that smash through a mix, while leaving a trail of sophisticated destruction in their wake. But don’t take my word for it—check out the audio example, which has no EQ or compression. 

 

 

 

There are five two-measure examples. The first example is from a kick track. The second, third, and fourth examples process the kick using this technique. The fifth example repeats the first example, as a reminder of how the sound started. 

 

The secret is processing the kick track through the Open Air reverb, using a kick drum sample as the impulse (Fig. 1). Just like how a cabinet impulse response imparts the sound of a cabinet onto a guitar amp, these kick drum impulses shape the kick track to have an entirely different character.

Figure 1: The Open Air reverb has a kick impulse loaded, and imparts that sound to the kick track.

 

Studio One’s Sound Sets have lots of kick drum samples. Here are the ones I used for the second, third, and fourth two-measure examples. The Open Air Mix control hovered around 30% for these.

 

Acoustic Drum Kits and Loops > Samples > TM Pop Rock Kit > DW 20.24 Pop Rock Kick  > DW 20.24 Pop Rock Kick 1.wav

 

Acoustic Drum Kits and Loops > Samples > TM Thuddy 70’s Kit > Gretsch 14×22 Vintage Thuddy Kick > Gretsch 14×22 Vintage Thuddy Kick 1.wav

 

909 Day Studio One Kits > Samples > F9 909 Detroit Kick.wav

 

So What’s the Catch?

 

Kick drum impulses can overload the Open Air pretty easily. The first two examples used the softest-velocity kick, but the F9 909 Detroit Kick was way too loud (well, unless you like horrific distortion). Most convolution reverbs are happiest with impulses that peak at around -12 dB.

 

So, the solution is simple. Drag the kick drum impulse into a Studio One track, use the gain envelope to cut the gain to about -12 dB peak, hit ctrl+B to make the change permanent, and then you can drag this impulse from the Studio One track right into the Open Air reverb.

 

Of course, you don’t have to limit yourself to kick, but it does seem kicks are where this technique shines the brightest. I also fooled around with using a floor tom as an impulse, and open hi-hat impulses on closed hi-hat tracks. The results aren’t always predictable…but that’s what makes it fun, right?

The Presence 12-String Electric Guitar

 

It’s difficult to sample a 12-string. The core Presence content includes a 12-string acoustic guitar, but there are no 12-string electrics—so let’s construct one. 

 

One of my favorite guitars ever is the Rickenbacker 360 12-string. Back in my touring days, it travelled tens of thousands of miles with me (Fig. 1). 

 

Figure 1: The mighty Rickenbacker 360 12-string guitar. Nothing else sounds like it.

 

I thought it would be a challenge to try and emulate that iconic sound with Presence. Listen to the audio example, and hear the results.

 

 

How It Works

 

The sound starts with one Presence instance, which uses a 6-string electric guitar preset. Then, we create a second, multi-instrument track with two Presence instances that use the same electric guitar preset. Transposing one of the instances up creates the octave above sound; however, a real 12-string guitar doesn’t have octaves on the 1st and 2nd strings. So, we use the final Presence for a unison sound, and edit the ranges in the multi instrument so they don’t overlap.

 

Step-by-Step Guitar Construction

 

  1. Create an Instrument track with Presence, and load the Guitar > Telecaster > Telecaster Open preset. This guitar sound is closest to a Rickenbacker, but we’ll do some EQ tricks later to get it closer.
  2. Create a new Instrument track with Presence, and load the same preset. Drag a second instance of Presence into the same track. When asked whether you want to “load the instrument or combine the instrument Presence,” choose Combine. This opens the Multi Instrument window. Load the same preset into the new Presence instance as well.
  3. In the multi instrument, drag the upper end of one Presence key range down to A#2 (we’ll call this the “Octave Presence”). Drag the lower end of the other Presence key range up to B2 (we’ll call this the “Unison Presence”). Fig. 2 shows the multi instrument window.

Figure 2: The multi instrument window has two instances of Presence—one for the octave above strings, and the other for the unison strings.

 

  1. Open the Octave Presence preset. Set Transpose to +12, and Pitch Fine Tune to +5 cents. Then open the Unison Presence preset, and change Pitch Fine Tune to -2 cents.
  2. Insert an Analog Delay in the multi Instrument channel , with the settings shown in Fig. 3. The reason for the 20 ms delay is because the higher string in a pair of strings gets hit just a little bit late. (We can’t use the Delay in Presence itself, because the mix needs to be 100% delay—no dry sound.) Without this delay, the emulated 12-string doesn’t sound right.

Figure 3: The Analog Delay emulates the delay caused by hitting the octave strings just a little bit later.

 

Note the High Cut setting—this reduces some of the brightness caused by transposition. The Width settings give a big stereo image, but for a more “normal” sound, turn ping-pong mode to Off.

 

Your mixer should look like Fig. 4, with two channels (basic guitar, and multi preset).

Figure 4: Mixer channels for the 12-string guitar.

 

Additional Tweaks

 

The Pitch Fine Tune settings in the multi instrument instances emulate the reality that a 12-string is seemingly never in tune, which accounts for that beautiful shimmering effect. Feel free to adjust your virtual 12-string so that it’s more or less in tune.

 

Another important tweak is to set the multi instrument channel’s fader about -6 dB below the main guitar sound. The octave strings on a 12-string are thinner than the strings with standard pitch, so they generate less output. This isn’t true of the 1st and 2nd strings, but that’s fine. With the octave strings a little lower, there’s a better balance.

 

Bring on the EQ

 

And finally…the coup de grâce to get us closer to the iconic Ric sound. On the main Presence instance, use the EQ in the Bass range. Boost 3 dB 3200 Hz, and pull the lowest slider down all the way. On both multi instrument instances, pull down the highest and lowest sliders (Fig. 4). Then, insert a Pro EQ in each mixer channel.

Figure 5: These EQ settings help get “the” sound. Clockwise from top: EQ on main Presence, EQ on the two multi instrument Presence instances, and Pro EQ placed on both mixer channels.

 

The narrow cut in the Pro EQ at 3.27 kHz helps reduce what sounds like some bridge “ping” in the original Telecaster samples. But all the EQ settings shown are suggestions. Between the broad EQ in Presence and the surgical nature of the Pro EQ, you can shape the sound however you want.