PreSonus Blog

Category Archives: Friday Tip of the Week


Friday Tips: The Center Stage Reverb

If you’ve ever played a large venue like a sports arena, you know that reverb is a completely different animal than what the audience hears. You hear your instrument primarily, and in the spaces between your playing, you hear the reverb coming back at you from the reflections. It might seem that reverb pre-delay would produce the same kind of effect, but it doesn’t “bloom” the way reverb does when you’re center stage in a big acoustical space.

This week’s tip is inspired by the center stage sound, but taken further. The heart of the effect is the Expander, but unlike last week’s Expander-based Dynamic Brightener tip, the Expander is in Duck mode, and fed by a sidechain. Here’s the Console setup.

 

 

In the audio example, the source is a funk guitar loop from the PreSonus loop collection; but any audio with spaces in between the notes or chords works well, especially drums (if the cymbals aren’t happening a lot), vocals that aren’t overly sustained, percussion, and the like. I deliberately exaggerated the effect to get the point across, so you might want to be a little more tasteful when you apply this to your own music. Or maybe not…

The guitar’s channel has two sends. One goes to the FX Channel, which has a Room Reverb followed by an Expander. The second send goes to the Expander’s sidechain input. Both are set pre-fader so that you can turn down the main guitar sound by bringing down its fader, and that way, you can hear only the processed sound. This makes it easier to edit the following Room Reverb and Expander settings, which are a suggested point of departure. Remember to enable the Expander’s Sidechain button in the header, and click the Duck button.

The reverb time is long—almost six seconds. This is because we want it going constantly in the background, so that after the Expander finishes ducking the reverb sound, there’s plenty of reverb available to fill in the spaces.

 

 

To tweak the settings, turn down the main guitar channel so you can monitor only the processed sound. The Expander’s Threshold knob determines how much you want the reverb to go away when the instrument audio is happening. But really, there are no “wrong” settings—start with the parameters above, play around, and listen to what happens.

This is a pretty fertile field for experimentation…as the following audio example illustrates. The first part is the guitar and the reverb effect; the reverb tail shows just how long the reverb time setting is. The second part is the reverb effect in isolation, processed sound only, and without the reverb tail.

 

 


This is a whole different type of reverb effect—have fun discovering what it can do for you!

Friday Tips: The Dynamic Brightener for Guitar

When you play an acoustic guitar harder, it not only gets louder, but brighter. Dry, electric guitar doesn’t have that quality…by comparison, the electrified sound by itself is somewhat lifeless. But I’m not here to be negative! Let’s look at a solution that can give your dry electric guitar some more acoustic-like qualities.

How It Works

Create an FX Channel, and add a pre-fader Send to it from your electric guitar track. The FX Channel has an Expander followed by the Pro EQ. The process works by editing the Expander settings so that it passes only the peaks of your playing. Those peaks then pass through a Pro EQ, set for a bass rolloff and a high frequency boost. Therefore, only the peaks become brighter. Here’s the Console setup.

 

The reason for creating a pre-fader send from the guitar track is so that you can bring the guitar fader down, and monitor only the FX Channel as you adjust the settings for the Expander and Pro EQ. The Expander parameter values are rather critical, because you want to grab only the peaks, and expand the rest of the guitar signal downward. The following settings are a good point of departure, assuming the guitar track’s peaks hit close to 0.

 

The most important edit you’ll need to make is to the Expander’s Threshold. After it grabs only the peaks, then experiment with the Range and Ratio controls to obtain the sound you want. Finally, choose a balance of the guitar track and the brightener effect from the FX Channel.

The audio example gets the point across. It consists of guitar and drums, because having the drums in the mix underscores how the dynamically brightened guitar can “speak” better in a track. The first five measures are the guitar with the brightener, the next five measures are the guitar without the brightener, and the final five measures are the brightener channel sound only. You may be surprised at how little of the brightener is needed to make a big difference to the overall guitar sound.

Also, try this on acoustic guitar when you want the guitar to really shine through a mix. Hey, there’s nothing wrong with shedding a little brightness on the situation!

Friday Tips: Crazee Snare Enhancements!

You never know where you’ll find inspiration. As I was trying not to listen to the background music in my local supermarket, “She Drives Me Crazy” by Fine Young Cannibals—a song from over 30 years ago!—earwormed its way into my brain. Check it out at https://youtu.be/UtvmTu4zAMg.

My first thought was “they sure don’t make snare drum sounds like those any more.” But hey, we have Studio One! Surely there’s a way to do that—and there is. The basic idea is to extract a trigger from a snare, use it to drive the Mai Tai synth, then layer it to enhance the snare.

Skeptical? Check out the audio example.

 

ISOLATING THE SNARE

If you’re dealing with a drum loop or submix, you first need to extract the snare sound.

  1. Create an FX Channel, and insert a Pro EQ followed by a Gate.
  2. Add a pre-fader send from your mixed drums to the FX Channel. Aside from providing a more consistent level for triggering, a pre-fader send lets you turn down the main drum track. This way you hear only the FX Channel, which makes it easier to tweak the EQ and isolate the snare.
  3. With the Gate bypassed, tune the Pro EQ to the snare frequency. Use the LC and HC with a 48 dB/octave slope to provide the preliminary isolation, then use a super-sharp bandpass setting to zero in on the snare frequency (Figure 1). The EQ’s background spectrum analyzer can help by making sure the bar in the range you’re boosting goes as high as possible. In stubborn cases, you may need to double up the bandpass filter with a second sharp bandpass filter.

 

Figure 1: Use the Pro EQ and Gate to extract a snare drum trigger.

 

  1. Enable the gate, and click on Active to enable the trigger output. Set the Note and Velocity as desired (however when using noise with Mai Tai, the specific Note isn’t that critical). Adjust the Gate’s settings so that it triggers on the snare hits. Like the Pro EQ, the controls here are critical as well.
  • A short attack is usually best.
  • Keep release relatively short (e.g., 10 ms), unless you want to mix in some of the peaked/gated sound from this channel along with the Mai Tai and drums.
  • Hold times around 50 ms can help prevent false triggering. But you can also get creative with this control. If you don’t want to trigger on hits that are close together (e.g., fills), a long Hold time will trigger on the first snare of the series, but ignore subsequent ones that fall within the hold time.
  1. Insert the Mai Tai. Set its input to Gate, and enable the mixer’s Monitor button. Figure 2 shows the finished track layout.

Figure 2: Track layout for snare drum extraction.

 

TWEAKING THE MAI TAI

Now the fun begins! Figure 3 shows a typical starting point for a snare-enhancing sound.

 

Figure 3: Starting point for a cool snare drum sound with Mai Tai.

The reason for choosing Mai Tai as the sound source is because of its “Character” options that, along with the filter controls, noise Color control, and FX (particularly Reverb, EQ, and Distortion), produce a huge variety of electronic snare sounds. The Character module’s Sound and Amount controls are particularly helpful. The more you play with the controls, the more you’ll start to understand just how many sounds are possible.

BUT WAIT…THERE’S MORE!

If the snare is on a separate track, then you don’t need the Pro EQ or FX Channel. Just insert a Gate in the snare track, enable the Gate’s trigger output, and adjust the Gate Threshold controls to trigger on each snare drum hit. The comments above regarding the Attack, Release, and Hold controls apply here as well.

Nor are you limited to snare. You can isolate the kick drum, and trigger a massive, low-frequency sine wave from the Mai Tai to give those car door-vibrating kick drums. Toms can sometimes be easy to isolate, depending on how they’re tuned. And don’t be afraid to venture outside of the “drum enhancement” comfort zone—sometimes the wrong Gate threshold settings, driving the wrong sound, can produce an effect that’s deliciously “right.”

 

Friday Tips: The Limiter—Demystified

Conventional wisdom says that compared to compression, limiting is a less sophisticated type of dynamics control whose main use is to restrict dynamic range to prevent issues like overloading of subsequent stages. However, I sometimes prefer limiting with particular signal sources. For example:

  • For mixed drum loops, limiting can bring up the room sound without having an overly negative effect on the drum attacks.
  • With vocals, I often use a limiter prior to compression. By doing the “heavy lifting” of limiting peaks, the subsequent compressor doesn’t have to work so hard, and can do what it does best.
  • When used with slightly detuned synth patches, limiting preserves the characteristic flanging/chorusing-like sound, while keeping the occasional peaks under control.
  • Limiting is useful when following synth sounds with resonant filters, or with instruments going through wah or autofilter effects

THE E-Z LIMITER

Some limiters (especially some vintage types) are easy to use, almost by definition: One control sets the amount of limiting, and another sets the output level. But Studio One’s limiter has four main controls—Input, Ceiling, Threshold, and Release—and the first three interact.

If the Studio One Limiter looked like Fig. 1, it would still take care of most of your needs. In fact, many vintage limiters don’t go much beyond this in terms of functionality.

Figure 1: If Studio One’s Limiter had an “Easy Mode” button, the result would look something like this.

To do basic limiting:

  1. Load the Limiter’s default preset.
  2. Turn up the input for the desired limiting effect. The Reduction meter shows the amount of gain reduction needed to keep the output at the level set by the Threshold control (in this case, -1.00 dB). For example, if the input signal peaks at 0 dB and you turn up the Input control to 6 dB, the Limiter will apply 7 dB of gain reduction to keep the Limiter output at -1.00.

Note that in this particular limiting application, the Threshold also determines the maximum output level.

THE SOFT CLIP BUTTON

When you set Threshold to a specific value, like 0.00 dB, then no matter how much you turn up the Limiter’s Input control, the output level won’t exceed 0.00 dB. However, you have two options of how to do this.

  • With Soft Clip off, gain reduction alone prevents the waveform from exceeding the ceiling.
  • With Soft Clip on, clipping the peaks supplements the gain reduction process to keep the waveform from exceeding the ceiling (Fig. 2).

Figure 2: The left screen shot shows the waveform with the input 6 dB above the Threshold, and Soft Clip off. The right screen shows the same waveform and levels, but with Soft Clip turned on. Note how the waveform peak is flattened somewhat due to the mild saturation.

 

While it may sound crazy to want to introduce distortion, in many cases you’ll find you won’t hear the effects of saturation, and you’ll have a hotter output signal.

ENTER THE CEILING

There are two main ways to set the maximum output level:

  • With the Threshold set to 0.00, set the maximum output level with the Ceiling control (from 0 to -12 dB).
  • With the Ceiling set to 0.00, set the maximum output level with the Threshold control (from 0 to -12 Db).

It’s also possible to set maximum output levels below -12.00 dB. Turn either the Ceiling or Threshold control all the way counter-clockwise to -12.00 dB, then turn down the other control to lower the maximum output level. With both controls fully counter-clockwise, the maximum output level can be as low as -24 dB.

SMOOTHING THE TRANSITION INTO LIMITING

Setting the Ceiling lower than the Threshold is a special case, which allows smoothing the transition into limiting somewhat. Under this condition, the Limiter applies soft-knee compression as the input transitions from below the threshold level to above it.

For example, suppose the Ceiling is 0.00 dB and the Threshold is -6.00 dB. As you turn up the input, you would expect that the output would be the same as the input until the input reaches around -6 dB, at which point the output would be clamped to that level. However in this case, soft-knee compression starts occurring a few dB below -6.00 dB, and the actual limiting to -6.00 dB doesn’t occur until the input is a few dB above -6.00 dB.

The tradeoff for smoothing this transition somewhat is that the Threshold needs to be set below 0.00. In this example, the maximum output is -6.00 dB. If you want to bring it up to 0.00 dB, then you’ll need to add makeup gain using Mixtool module.

Studio One’s Limiter is a highly versatile signal processor, so don’t automatically ignore it in favor of the Compressor or Multiband Dynamics—with some audio material, it could be exactly what you need.

Friday Tips: Frequency-Selective Guitar Compression

Some instruments, when compressed, lack “sparkle” if the stronger, lower frequencies compress high frequencies as well as lower ones. This is a common problem with guitar, but there’s a solution: the Compressor’s internal sidechain can apply compression to only the guitar’s lower frequencies, while leaving the higher frequencies uncompressed so they “ring out” above the compressed sound. (Multiband compression works for this too, but sidechaining can be a faster and easier way to accomplish the same results.) Frequency-selective compression can also be effective with drums, dance mixes, and other applications—like the “pumping drums” effect covered in the Friday Tip for October 5, 2018. Here’s how to do frequency-selective compression with guitar.

 

 

  1. Insert the Compressor in the guitar track.
  2. Enable the internal sidechain’s Filter button. Do not enable the Sidechain button in the effect’s header.
  3. Enable the Listen Filter button.
  4. Turn Lowcut fully counterclockwise (minimum), and set the Highcut control to around 250 – 300 Hz. You want to hear only the guitar’s low frequencies.
  5. You can’t hear the effects of adjusting the main compression controls (like Ratio and Threshold) while the Listen Filter is enabled, so disable Listen Filter, and start adjusting the controls for the desired amount of low-frequency compression.
  6. For a reality check, use the Mix control to compare the compressed and uncompressed sounds. The high frequencies should be equally prominent regardless of the Mix control setting (unless you’re hitting the high strings really hard), while the lower strings should sound compressed.

 

The compression controls are fairly critical in this application, so you’ll probably need to tweak them a bit to obtain the desired results.

If you need more flexibility than the internal filter can provide, there’s a simple workaround.

 

 

Copy the guitar track. You won’t be listening to this track, but using it solely as a control track to drive the Compressor sidechain. Insert a Pro EQ in the copied track, adjust the EQ’s range to cover the frequencies you want to compress, and assign the copied track’s output to the Compressor sidechain. Because we’re not using the internal sidechain, click the Sidechain button in the Compressor’s header to enable the external sidechain.

The bottom line is that “compressed” and “lively-sounding” don’t have to be mutually exclusive—try frequency-selective compression, and find out for yourself.

Friday Tips: The Sidechained Spectrum

You’re probably aware that several Studio One audio processors offer sidechaining—Compressor, Autofilter, Gate, Expander, and Channel Strip. However, both the Spectrum Meter and the Pro EQ spectrum meter also have sidechain inputs, which can be very handy. Let’s look at Pro EQ sidechaining first.

When you enable sidechaining, you can feed another track’s output into the Pro EQ’s spectrum analyzer, while still allowing the Pro EQ to modify the track into which it’s inserted. When sidechained, the Spectrum mode switches to FFT curve (the Third Octave and Waterfall options aren’t available). The blue line indicates the level of the signal going through the Pro EQ, while the violet line represents the sidechain signal.

As a practical example of why this is useful, the screen shot shows two drum loops from different drum loop libraries that are used in the same song. The loop feeding the sidechain loop has the desired tonal qualities, so the loop going through the EQ is being matched as closely as possible to the sidechained loop (as shown by a curve that applies more high end, and a slight midrange bump).

Another example would be when overdubbing a vocal at a later session than the original vocal. The vocalist might be off-axis or further away from the mic, which would cause a slight frequency response change. Again, the Pro EQ’s spectrum meter can help point out any differences by comparing the frequency response of the original vocal to the overdub’s response.

The Spectrum Meter

Sidechaining with the Spectrum Meter provides somewhat different capabilities compared to the Pro EQ’s spectrum analyzer.

 

With sidechain enabled, the top view shows the spectrum of the track into which you’ve inserted the Spectrum Meter. The lower view shows the spectrum of the track feeding the sidechain. When sidechained, all the Spectrum Meter analysis modes are available except for Waterfall and Sonogram.

While useful for comparing individual tracks (as with the Pro EQ spectrum meter), another application is to help identify frequency ranges in a mix that sound overly prominent. Insert the Spectrum Meter in the master bus, and you’ll be able to see if a specific frequency range that sounds more prominent actually is more prominent (in the screen shot, the upper spectrum shows a bump around 600 Hz in the master bus). Now you can send individual tracks that may be causing an anomaly into the Spectrum Metre’s sidechain input to determine which one(s) are contributing the most energy in this region. In the lower part of the screen shot, the culprit turned out to be a guitar part with a wah that emphasized a particular frequency. Cutting the guitar EQ just a little bit around 600 Hz helped even out the mix’s overall sound.

Of course, the primary way to do EQ matching is by ear. However, taking advantage of Studio One’s analysis tools can help speed up the process by identifying specific areas that may need work, after which you can then do any needed tweaking based on what you hear. Although “mixing with your eyes” isn’t the best way to mix, supplementing what you hear with what you see can expedite the mixing process, and help you learn to correlate specific frequencies with what you hear—and there’s nothing wrong with that.

Friday Tips: Synth + Sample Layering

One of my favorite techniques for larger-than-life sounds is layering a synthesizer waveform behind a sampled sound. For example, layering a sine wave along with piano or acoustic guitar, then mixing the sine wave subtly in the background, reinforces the fundamental. With either instrument, this can give a powerful low end. Layering a triangle wave with harp imparts more presence to sampled harps, and layering a triangle wave an octave lower with a female choir sounds like you’ve added a bunch of guys singing along.

Another favorite, which we’ll cover in detail with this week’s tip, is layering a sawtooth or pulse wave with strings. I like those syrupy, synthesized string sounds that were so popular back in the 70s, although I don’t like the lack of realism. On the other hand, sampled strings are realistic, but aren’t lush enough for my tastes. Combine the two, though, and you get lush realism. Here’s how.

  1. Create an instrument track with Presence, and call up the Violin Full preset.
  2. Drag Mai Tai into the same track. You’ll be asked if you want to Replace, Keep, or Combine. Choose Combine.

  1. After choosing Combine, both instruments will be layered within the Instrument Editor (see above).
  2. Program Mai Tai for a very basic sound, because it’s there solely to provide reinforcement—a slight detuning of the oscillators, no filter modulation, very basic LFO settings to add a little vibrato and prevent too static a waveform, amplitude envelope and velocity that tracks the Presence sound as closely as possible, some reverb to create a more “concert hall” sound, etc. The screen shot shows the parameters used for this example. The only semi-fancy programming tricks were making one of the oscillators a pulse wave instead of a sawtooth, and panning the two oscillators very slightly off-center.

 

  1. Adjust the Mai Tai’s volume for the right mix—enough to supplement Presence, but not overwhelm it.

 

That’s all there is to it. Listen to the audio example—first you’ll hear only the Presence sound, then the two layers for a lusher, more synthetic vibe that also incorporates some of the realism of sampling. Happy orchestrating!

Friday Tips: Studio One’s Amazing Robot Bassist

When Harmonic Editing was announced, I was interested. When I used it for the first time, I was intrigued. When I discovered what it could do for songwriting…I became addicted.

Everyone creates songs differently, but for me, speed is the priority—I record scratch tracks as fast as possible to capture a song’s essence while it’s hot. But if the tracks aren’t any good, they don’t inspire the songwriting process. Sure, they’ll get replaced with final versions later, but you don’t want boring tracks while writing.

For scratch drums on rock projects, I have a good collection of loops. Guitar is my primary instrument, so the rhythm and lead parts will be at least okay. I also drag the rhythm guitar part up to the Chord Track to create the song’s “chord chart.”

Then things slow down…or at least they did before Harmonic Editing came along. Although I double on keyboards, I’m not as proficient as on guitar but also, prefer keyboard bass over electric bass—because I’ve sampled a ton of basses, I can find the sound I want instantly. And that’s where Harmonic Editing comes in.

The following is going to sound ridiculously easy…because it is. Here’s how to put Studio One’s Robot Bassist to work. This assumes you’ve set the key (use the Key button in the transport, or select an Instrument part and choose Event > Detect Key Signature), and have a Chord Track that defines the song’s chord progression.

 

  1. Play the bass part by playing the note on a MIDI keyboard that corresponds to the song’s key. Yes, the note—not notes. For example, if the song is in the key of A, hit an A wherever you want a bass note.
  2. Quantize what you played. It’s important to quantize because presumably, the chord changes are quantized, and the note attack needs to fall squarely at the beginning of, or within, the chord change. You can always humanize later.
  3. Open the Inspector, unfold the Follow Chords options, and then choose Bass (Fig. 1).

 

Figure 1: Choose the Bass option to create a bass part when following chords.

  1. Now you have a bass part! If the bass part works, choose the Edit tab, select all the notes, and choose Action > Freeze Pitch. This is important, because the underlying endless-string-of-notes remains the actual MIDI data. So if you copy the Event and paste it, unless you then ask the pasted clip to follow chords, you have the original boring part instead of the robotized one.
  2. After freezing, turn off Follow Chords, because you’ve already followed the chords. Now is the time to make any edits. (Asking the followed chords to follow chords can confuse matters, and may modify your edits.)

The bottom line: with one take, a few clicks, and (maybe) a couple quick edits—instant bass part (Fig. 2).

Figure 2: The top image is the original part, and yes, it sounds as bad as it looks. The lower image is what happened after it got robotized via Harmonic Editing, and amazingly, it sounds pretty good.

Don’t believe me? Well, listen to the following.

 

You’ll hear the bass part shown in Fig. 2, which was generated in the early stages of writing my latest music video (I mixed the bass up a little on the demo so you can hear it easily). Note how the part works equally well for the sustained notes toward the beginning, and well as the staccato parts at the end. To hear the final bass part, click the link for Puzzle of Love [https://youtu.be/HgMF-HBMrks]. You’ll hear I didn’t need to do much to tweak what Harmonic Editing did.

But Wait! There’s More!

Not only that, but most of the backing keyboard parts for Puzzle of Love (yes, including the piano intro) were generated in essentially the same way. That requires a somewhat different skill set than robotizing the bass, and a bit more editing. If you want to know more (use the Comments section), we’ll cover Studio One’s Robot Keyboardist in a future Friday Tip.

Friday Tips: Demystifying the Limiter’s Meter Options

Limiters are common mastering tools, so they’re the last processor in a signal chain. Because of this, it’s important to know as much as possible about its output signal, and Studio One’s Limiter offers several metering options.

PkRMS Metering

The four buttons under the meter’s lower left choose the type of meter scale. PkRMS, the traditional metering option, shows the peak level as a horizontal blue bar, with the average (RMS) level as a white line superimposed on the blue bar (Fig. 1). The average level corresponds more closely to how we perceive musical loudness, while the bar indicates peaks, which is helpful when we want to avoid clipping.

The TP Button

Enabling the True Peak button takes the possibility of intersample distortion into account. This type of distortion can occur on playback if some peaks use up the maximum available headroom in a digital recording, and then these same peaks pass through the digital-to-analog converter’s output smoothing filter to reconstruct the original waveform. This reconstructed waveform might have a higher amplitude than the peak level of the samples, which means the waveform now exceeds the maximum available headroom (Fig. 2).

 

Figure 2: How intersample distortion occurs.

For example, you might think your audio isn’t clipping because without TP enabled, the output peak meter shows -0.1 dB. However, enabling True Peak metering may reveal that the output is as much as +3 dB over 0 when reconstructed. The difference between standard peak metering and true peak metering depends on the program material.

K-System Metering

The other metering options—K-12, K-14, and K-20 metering—are based on a metering system developed by Bob Katz, a well-respected mastering engineer. One of the issues any mix or mastering engineer has to resolve is how loud to make the output level. This has been complicated by the “loudness wars,” where mixes are intended to be as “hot” as possible, with minimal dynamic range. Mastering engineers have started to push back against this not just to retain musical dynamics, but because hot recordings cause listener fatigue. Among other things, the K-System provides a way to judge a mix’s perceived loudness.

A key K-System feature is an emphasis on average (not just peak) levels, because they correlate more closely to how we perceive loudness. A difference compared to conventional meters is that K-System meters use a linear scale, where each dB occupies the same width (Fig. 3). A logarithmic scale increases the width of each dB as the level gets louder, which although it corresponds more closely to human hearing, is a more ambiguous way to show dynamic range.

Figure 3: The K-14 scale has been selected for the Limiter’s output meter.

 

Some people question whether the K-System, which was introduced two decades ago, is still relevant. This is because there’s now an international standard (based on a recommendation by the International Telecommunications Union) that defines perceived average levels, based on reference levels expressed in LUFS (Loudness Units referenced to digital Full Scale). As an example of a practical application, when listening to a streaming service, you don’t want massive level changes from one song to the next. The streaming service can regulate the level of the music it receives so that all the songs conform to the same level of perceived loudness. Because of this, there’s no real point in creating a hot master—it will just be turned down to bring it in line with songs that retain dynamic range; and the latter will be turned up if needed to give the same perceived volume.

 

Nonetheless, the K-System remains valid, particularly when mixing. When you mix, it’s best to have a standardized, consistent monitoring level because human hearing has a different frequency response at different levels (Fig. 4).

 

Figure 4: The Fletcher-Munson curve shows that different parts of the audio spectrum need to be at different levels to be perceived as having the same volume. Low frequencies have to be substantially louder at lower levels to be perceived as having equal volume.

 

The K-System links monitoring levels with meter readings, so you can be assured that music reaching the same levels will sound like they’re at the same levels. This requires calibrating your monitor levels to the meter readings with a sound level meter. If you don’t have a sound level meter, many smartphones can run sound level meter apps that are accurate enough.

 

Note that in the K-System, 0 dB does not represent the maximum possible level. Instead, the 0 dB point is shifted “down” from the top of the scale to either -12, -14, or -20 dB, depending on the scale. These numbers represent the amount of headroom above 0, and therefore, the available dynamic range. You choose a scale based on the music you’re mixing or mastering—like -12 for music with less dynamic range (e.g., dance music), -14 for typical pop music, and -20 dB for acoustic ensembles and classical music. You then aim for having the average level hover around the 0 dB point. Peaks that go above this point will take advantage of the available headroom, while quieter passages will go below this point. Like conventional meters, the K-Systems meters have green, yellow, and red color-coding to indicate levels. Levels above 0 dB trigger the red, but this doesn’t mean there’s clipping—observe the peak meter for that.

 

Calibrating Your Monitors

 

The K-System borrows film industry best practices. At 0 dB, your monitors should be putting out 85 dBSPL for stereo material. Therefore, you’ll need a separate calibration for the three scales to make sure that 0 dB on any scale has the same perceived loudness. The simplest way to calibrate is to send pink noise through your system until the chosen K-System meter reads 0 dB (you can download pink noise samples from the web, or use the noise generator in the Mai Tai virtual instrument). Then, using the sound level meter set to C weighting and a slow response, adjust the monitor level for an 85 dB reading. You can put labels next to the level control on the back of your speaker to show the settings that produce the desired output for each K-Scale.

 

But Wait! There’s More

 

We’ve discussed the K-System in the context of the Limiter, but if you’re instead using the Compressor or some other dynamics processor that doesn’t have K-System metering, you’re still covered. There’s a separate metering plug-in that shows the K-System scale (Fig. 5).

 

Figure 5: The Level meter plug-in shows K-System as well as the R128 spec that reads out the levels in LUFS. Enabling TP converts the meter to PkRMS, and shows the True Peak in the two numeric fields.

Finally, the Project Page also includes K-System Metering along with a Spectrum Analyzer, Correlation Meter, and LUFS metering with True Peak (Fig. 6).

Figure 6: The Project Page metering tells you pretty much all you need to need to know what’s going on with your output signal when mastering.

 

 

Friday Tips—Blues Harmonica FX Chain

If you’ve heard blues harmonica greats like Junior Wells, James Cotton, Jimmy Reed, and Paul Butterfield, you know there’s nothing quite like that big, brash sound. They all manage to transform the harmonica’s reedy timbre into something that seems more like a member of the horn family.

To find out more about the techniques of blues harmonica, check out the article Rediscovering Blues Harmonica. It covers why you don’t play blues harp in its default key (e.g., you typically use a harmonica in the key of A for songs in E), how to mic a harmonica, and more. However, the secret to that big sound is playing through the distortion provided by an amp, or in our software-based world, an amp sim. I don’t really find the Ampire amps suitable for this application, but we can put together an FX Chain that does the job.

Check out the demo to hear the desired goal. The first 12 bars are unprocessed harmonica (other than limiting). The second 12 bars use the FX Chain described in this week’s tip, and which you can download for your own use.


The chain starts with a Limiter to provide a more sustained, consistent sound.

 

Next up: A Pro EQ to take out all the lows and highs, which tightens up the sound and reduces intermodulation distortion. (When using an amp sim, blues harmonica is also a good candidate for multiband processing, as described in the February 1 Friday Tip.)

 

Now it’s time for the Redlight Dist to provide the distortion. For the cabinet, this FX Chain uses the Ampire solely for its 4 x 10 American cabinet—no amp or stomps.

 

After the distortion/cabinet combo, a little midrange “honk” makes the harmonica stand out more in the mix.

 

For a final touch, blues harp often plays through an amp with reverb—so a good spring reverb effect adds a vintage vibe.

You can download the Blues Harp.multipreset and use it as it, but I encourage playing around with it—try different types of distortion and amps, mess with the EQ a bit, and so on. For an example of a finished song with amp sim blues harmonica in context, check out I’ll Take You Higher on YouTube.

 

Click here to download the multipreset!