Use key-tracking for Frequency Balance

Tutorial | Mar 12, 2024

In the video, I explained the concept of using a tilt setting of 4.5 dB per octave to achieve a balanced mixdown. I demonstrated how this can be applied within Bitwig's EQ+ and also within the Grid. By using the node signal and modulating the gain, I showed how to achieve the desired frequency balance without relying on external EQ processing.

You can watch the Video on Youtube - support me on Patreon

In my tutorial, I delve into an advanced technique for achieving a well-balanced mixdown in music production by utilizing the EQ+ tilt setting and key tracking modulators in Bitwig Studio and VST plugins. Here's a summary:

1. Tilt EQ Setting for Mixdowns: I introduce a guideline for mixdowns using a slope of 4.5 dB per octave in an EQ, which means every octave above a reference point is reduced in volume by 4.5 dB. This approach helps in achieving a more natural balance in the mix, as our ears perceive frequencies differently than what white noise (equal volume across all frequencies) suggests. The goal is to adjust frequencies so that they align in a straight line when viewed on an EQ, reflecting this slope.

2. Analyzing Frequencies: To simplify EQ adjustments, I suggest using the tilt setting only in the EQ's analyzer part. By doing so, the visual representation of frequencies aligns in a straight line, making it easier to achieve the desired slope without altering the actual EQ curve.

3. Application in Bitwig's Grid: I explore how this concept can be applied directly within Bitwig's Grid environment, bypassing the need for post-process EQ adjustments. By modulating the volume of notes based on their pitch, it's possible to internally balance the frequency spectrum according to the 4.5 dB per octave rule. This involves calculating the volume reduction needed across the pitch range and adjusting accordingly, essentially embedding the mixdown balance within the sound design phase.

4. Practical Example with Key Tracking: I provide a practical example of implementing this technique on a synthesizer patch in Bitwig. Using the key track modulator, I adjust the volume of notes across the keyboard to match the desired frequency slope, ensuring a balanced output directly from the synthesizer.

5. Saving Default Presets: To streamline workflow, I suggest saving these adjustments as default presets for synthesizers in Bitwig, allowing for consistent frequency balance across projects without additional EQ adjustments.

6. Applicability to Other Synthesizers and Sounds: Finally, I mention that this technique isn't limited to Bitwig's Grid or a single instrument but can be applied to any synthesizer or sound source, considering the specific range and capabilities of the key tracking modulator available.

This approach promotes a deeper understanding and application of frequency balance in music production, encouraging producers to think critically about how they achieve a mixdown not just through EQ, but through the initial sound design and synthesis process.

Questions & Answers #

Maybe you dont watch the video, here are some important takeaways:

What is the slope of 4.5 dB per octave and why is it important in mixdowns? #

The slope of 4.5 dB per octave refers to lowering the volume of every octave by 4.5 dB. It is used as a rough guide for achieving a good mixdown by balancing the frequency spectrum. This slope helps create a more natural and pleasing sound by reducing the higher frequencies while maintaining the overall balance.

How is the tilt setting used in EQ+ to achieve the desired slope? #

The tilt setting in EQ+ is used to adjust the analyzing part of the EQ to create a straight line for every frequency in the mix. This straight line represents the desired slope of 4.5 dB per octave. By aligning the frequencies in this way, it becomes easier to achieve a balanced mixdown.

How can the grid in Bitwig Studio be used to achieve frequency balance without using an EQ? #

Inside the grid in Bitwig Studio, the node signals can be used to balance the frequency spectrum. By modulating the gain of the node signal based on the pitch of the nodes, the volume can be lowered by the desired amount per octave. This allows for achieving a perfect frequency balance without the need for an EQ.

How can the frequency balance be applied to synthesizers outside of the grid? #

For synthesizers outside of the grid, such as the Polymer synthesizer, the key track modulator can be used. By setting the modulation amount to match the number of octaves and adjusting the volume accordingly, the frequency balance can be achieved. This can be saved as a default preset for future use.

Transcription #

This is what im talking about in this video. The text is transcribed by AI, so it might not be perfect. If you find any mistakes, please let me know.
You can also click on the timestamps to jump to the right part of the video, which should be helpful.

[00:00:00] When we talk about mixdowns, we all tend to agree that the slope of 4.5 dB per octave
[00:00:05] is actually a nice rough guide for a good mixdown.
[00:00:09] And I talked about this before in some of my videos here.
[00:00:12] I use inside of the EQ+ a tilt setting on the left side here of 4.5 dB per octave, which
[00:00:19] means every octave is lowered by 4.5 dB in volume.
[00:00:26] And this tilt setting only changes the analyzing part, which makes it pretty easy to use the
[00:00:32] EQ because you need to just keep every partial, every frequency in a straight line to have
[00:00:39] a good mixdown or a rough guide for a mixdown.
[00:00:44] So I want to go down a bit deeper here in explaining this because I think not everyone
[00:00:49] gets the idea of this tilt setting here.
[00:00:52] So let's say we have white noise.
[00:00:55] White noise is technically every frequency has the same volume.
[00:01:02] It's a straight line.
[00:01:05] It's mathematically perfect, right?
[00:01:08] White noise, every frequency, same volume.
[00:01:11] And if you ever listen to white noise, it's very bright.
[00:01:15] That's not how our ears perceive sounds.
[00:01:19] So some mixing tutorials go for, let's say, pink noise.
[00:01:25] So pink noise looks more like this.
[00:01:29] It's a slope of 3 dB per octave.
[00:01:32] So every octave is 3 dB lower than the previous octave, which means it results in this kind
[00:01:40] of slope.
[00:01:42] So higher frequencies are much lower in volume than lower frequencies.
[00:01:49] But most of the times, this is still too bright, right?
[00:01:52] So we tend to go for something like this, 4.5 dB per octave.
[00:01:58] So every octave is 4.5 dB lower in volume than the previous octave, which means, of course,
[00:02:06] higher frequencies are much, much lower than lower frequencies.
[00:02:10] So this is the slope we can agree on universally that we want to have every frequency speaking
[00:02:19] around roughly around this line.
[00:02:22] But having this kind of tilted line inside of the EQ is not easy, right?
[00:02:29] So the idea is that we actually just tilt only the analyzer part.
[00:02:34] So what we do inside of the EQ when we change here the setting, the tilt setting to 4.5 dB
[00:02:41] per octave, we actually do this with the analyzer part.
[00:02:44] We put this into a straight line, right?
[00:02:49] But in the background, the EQing part is still the same as before.
[00:02:52] It's still in this angle, right?
[00:02:55] We put only the analyzer part into this straight line.
[00:02:59] So now we can EQ everything here into a straight line and achieve basically the tilt or a slope
[00:03:07] of 4.5 dB per octave, which makes it very easy to EQ stuff inside of the EQ plus.
[00:03:14] And yeah, most EQs tend to agree on this.
[00:03:18] So I have also here a VST called the Clis EQ.
[00:03:23] It's also used by a lot of producers.
[00:03:26] And inside here of the settings, you can see we have the slope here 4.5 dB per octave.
[00:03:34] So it's also a default setting inside of the Clis EQ.
[00:03:37] So everything here in a straight line, perfect.
[00:03:42] OK, so my thinking today was why we use actually on synths inside of Bitwig or here on this
[00:03:51] generative patch, why do we use actually an EQ to balance the frequency spectrum?
[00:03:56] Because we have access inside of the grid to the frequencies or to the nodes and to
[00:04:02] the volume.
[00:04:03] So why not actually level this inside of the grid inside of the patch instead of applying
[00:04:08] here a post process or an EQ process to level out the frequencies.
[00:04:16] And then I came to the conclusion, this is actually why the key track was invented.
[00:04:21] So we have a key track modulator here.
[00:04:25] And with the key track, we can get here all the nodes and depending on which node we are
[00:04:30] playing, we can lower the volume or maybe increase a cutoff frequency or add more reverb
[00:04:40] or whatever we want to do.
[00:04:44] So we want to keep focus here on the grid itself because it's a bit different inside
[00:04:48] of the grid than using here this modulator.
[00:04:51] I will explain this later.
[00:04:54] So inside of the grid here, I have some kind of generative patch running.
[00:04:57] I will play some different sine partials.
[00:05:02] And you can see some of the higher partials here are much louder than the lower ones on
[00:05:07] this graph because we have here the tilt setting of 4.5 dB per octave.
[00:05:11] And as you can remember, everything needs to be in a straight line.
[00:05:15] So having here frequencies or partials much, much higher than lower ones is actually kind
[00:05:21] of wrong.
[00:05:22] So we need to apply an EQ to balance this.
[00:05:24] So that's not what we want.
[00:05:26] So we want to apply this inside of the grid.
[00:05:29] So inside of the grid, we have node signals.
[00:05:32] So let's go here for an oscilloscope.
[00:05:36] And we get the nodes here from this pitch quantizer.
[00:05:38] And the pitch quantizer gets the nodes here from this value.
[00:05:41] And this value is driven by this step modulator.
[00:05:46] So we get some different nodes here.
[00:05:49] And we want to actually see what the highest node is inside of the grid.
[00:05:56] So here you can see we have different nodes.
[00:06:01] And this is the raw value.
[00:06:03] So having a raw value, we can use a constant and use the constant of 1, which is the highest
[00:06:11] possible value for nodes inside of the grid.
[00:06:14] So 1 and 1 equals to the pitch of C13.
[00:06:20] The middle node, actually raw value of 0 is C3.
[00:06:27] So from C3 up to C13, it's exactly 10 octaves.
[00:06:36] So we want to lower the volume by 4.5 dB on every octave.
[00:06:44] So we need to lower the volume on the last octave here, C13 by exactly 45 dB, because
[00:06:53] 4.5 dB octave times 10 equals to 45.
[00:07:00] So what we can do now easily inside the grid is we can just take the modulator here, take
[00:07:08] the node signal, and then use the gain and then lower the volume here by exactly 45 dB.
[00:07:17] So now every partial here is basically balanced to the guide or to the slope or to the frequency
[00:07:25] falloff of 4.5 dB per octave.
[00:07:28] So we don't need to apply an EQ now to this, because higher frequencies or higher pitches,
[00:07:34] higher nodes are much lower in volume than lower frequencies, which is really nice, because
[00:07:43] it makes our life much easier and we can save actually an EQ to balance the frequency spectrum.
[00:07:48] We probably need here still an EQ to cut off some lower frequencies or to remove some
[00:07:53] rumble.
[00:07:54] But the overall balance, it's not needed that we use an EQ for that.
[00:07:59] We can actually put this inside of the grid.
[00:08:03] And we also can apply this not only to generative patches like here in this case, in this example,
[00:08:11] we can also apply this to lead sounds.
[00:08:13] We can apply this to kick drum patches, kick drum presets or sound design stuff.
[00:08:22] And you know, you can balance basically everything just by using the node signal here and then
[00:08:28] changing the volume dependent on the node input.
[00:08:32] Which is really nice actually.
[00:08:36] That's why it's invented and it's super clear that probably a lot of people just use the
[00:08:40] node signal to change the volume.
[00:08:43] And I did this also before, but I never use it in this way that I actually thought about
[00:08:49] frequency balance in this kind of slope way or in this tilt way.
[00:08:54] I want to achieve actually a perfect kind of slow frequency balance from within the
[00:09:01] grid without using an EQ.
[00:09:02] So I'm kind of using here the key tracking in a kind of mathematically high precision
[00:09:10] side chain kind of way.
[00:09:14] So this was my idea earlier.
[00:09:17] And you can just remember this.
[00:09:19] You just use here the node signal and then modulate here the gain by minus 45 dB.
[00:09:27] And then, yeah, the higher the frequencies go, the lower the volume goes and it's exactly
[00:09:33] the perfect tilt setting of 4.5 dB per octave.
[00:09:36] If you want to go for pink noise, you just go for minus 30 dB because we need to lower
[00:09:41] the frequencies by 3 dB per octave.
[00:09:45] We have 10 octaves.
[00:09:46] So 3 times 10 is 13.
[00:09:50] Okay.
[00:09:51] 30.
[00:09:52] Okay.
[00:09:53] And then get the nice balance out of this.
[00:09:56] Okay.
[00:09:57] That's inside of the grid.
[00:09:58] And like I said, it's not only useful for these generative patches here.
[00:10:03] Also nice for lead sounds, bass sounds, kick drum sounds and so on.
[00:10:07] Okay.
[00:10:08] So if we want to use this outside of the grid, let's say on a polymer synthesizer here, I
[00:10:14] play here the note of C3 and I roughly balance this here around this line, this middle line
[00:10:22] here.
[00:10:24] And I use your tool device for this, right?
[00:10:26] Put this down here so everything is perfectly aligned with the straight line.
[00:10:32] The EQ plus is the setting here, tilt setting 4.5 dB per octave.
[00:10:36] And if we now put this here to the highest, to the highest note inside of the piano roll
[00:10:41] is actually not 10, it's actually 5.
[00:10:44] So 1, 2, 3, 4, 5.
[00:10:49] It can't go higher.
[00:10:51] That's the highest octave, C8, right?
[00:10:53] So 5 octaves.
[00:10:55] Also here, the key track plus modulator, if you open up your help itself, set the parameter
[00:11:02] to C3 and then set the modulation amount to +60 semitones representing 5 octaves.
[00:11:11] So instead of 10 octaves like in the grid, we have only 5 octaves.
[00:11:15] So basically we have to put the value in half, the dB value in half.
[00:11:21] But sadly or unfortunately, we can't see here the unit.
[00:11:27] As you can see here, there's some kind of star.
[00:11:29] We have here -0.86 and then a star.
[00:11:33] It's not the unit, it's not dB.
[00:11:35] It's some random value.
[00:11:37] I don't know what it is exactly.
[00:11:39] It's some kind of linear value that you have, at least some kind of value.
[00:11:45] Sadly we don't see the dB setting here or dB unit.
[00:11:49] So we have to roughly match this here.
[00:11:53] You can see here we are way above this line.
[00:11:58] And if you remember when we put this down to C3, we match the line here and then we go
[00:12:05] up.
[00:12:06] 1, 2, 3, 4, 5.
[00:12:08] We are way above this line.
[00:12:10] So all we have to do now is to use this modulator here and pull down the voice volume so we
[00:12:16] hit this line.
[00:12:23] And that's basically it.
[00:12:26] So now we aligned or reference basically the highest octave here with the line of this
[00:12:32] tilt setting of 4.5 dB per octave.
[00:12:34] And we go down here to C3.
[00:12:37] You can see it's still at this line.
[00:12:40] We go lower, C1.
[00:12:42] We roughly hit the line here.
[00:12:46] And everything is perfectly aligned.
[00:12:49] And the best part is you can actually save this here as a default preset.
[00:12:54] So we right click and say save as a default preset.
[00:12:57] And next time you pull up here the Polymer, you don't need to care for the frequency balance.
[00:13:03] It's lowering the volume for the higher notes perfectly fine.
[00:13:07] And it's matching here your EQ curve or tilt setting perfectly fine.
[00:13:12] So you don't need to apply EQ then in the post process to balance the frequencies.
[00:13:18] Of course, if you put in the FX box here some kind of EQs or distortion in different, you
[00:13:25] know, some multi-band FX or something like this, then you get the balance out of balance.
[00:13:32] But out of your synthesizer, you get a nice perfectly balanced frequency spectrum.
[00:13:39] And I thought it's worth mentioning this in a video to explain you this.
[00:13:44] And I think it's probably also a nice way of using here the default setting for the synthesizer.
[00:13:51] And by the way, you can also apply this to other synthesizers, right?
[00:13:54] Not only Polymer.
[00:13:55] You can also do this on Polysynth or whatever you want to use.
[00:14:00] If you use VST devices, maybe you have also a key track in your VST synthesizers.
[00:14:06] But there you also need to look what's the highest octave for the key track and what's
[00:14:11] the lowest octave for the key track.
[00:14:13] And then you have to calculate basically the volume that you need to apply to change for
[00:14:20] the modulation of the modulation amount for the volume knob.
[00:14:24] Okay, that's basically it.
[00:14:26] It was pretty technical, but I wanted to document this because I thought about this while taking
[00:14:33] a walk.
[00:14:35] Okay, that's it.
[00:14:37] Leave a like if you liked the video.
[00:14:39] Subscribe to the channel.
[00:14:40] Ask me questions in the comments down below.
[00:14:42] Thanks for watching and I'll see you in the next video.
[00:14:44] Bye.