Master Your Mix: The Ultimate Tonal Balancer in Bitwig!

Tutorial | Mar 26, 2025

In this video, I introduce a tool I created in Bitwig called Tonal Balancer, which automatically balances the tonal distribution of a track by using noise sources and EQ to match levels across frequency bands. I demonstrate how to use features like the target knob, custom curves, and rise/fall settings to adjust how the EQ adapts to changes in the audio signal. The preset, available for free on my GitHub, is not yet finished, as I plan to implement a threshold feature to prevent unwanted amplification of quiet parts.

You can watch the Video on Youtube

• support me on Patreon
• download the preset
• inspired by Balance by zynaptiq
• based on Volume Level Match

Summary #

Maybe you don't watch the video, here are some important takeaways:

In this video, I introduce a tool I created in Bitwig called Tonal Balancer, which is designed to automatically balance the tonal distribution of your track. The interface is quite basic, consisting mainly of some remote controls. The main goal of this tool is to adjust the tonal balance of the track towards a predefined target curve, like pink noise or a 'smiley' curve.

I demonstrate how the Tonal Balancer sounds by playing a track without any alterations and then by applying the auto EQ using the target knob to show its effect. The tool allows you to switch between different target curves, such as pink noise or a smiley curve, with the option to create and select custom curves. The rise and fall knobs adjust how quickly the EQ responds to changes in the audio, potentially transforming it into a compressor if set too fast, but the primary intention is for it to function as a slow leveler.

The concept is based on a target level preset I shared previously, which extracts volume from a sound source to apply it to another. In this case, instead of a standard sound source, I'm using a noise source—specifically a test tone in pink noise mode. An EQ is applied to this pink noise, which is then split into seven separate frequency bands using convolution reverb to maintain linear phase and avoid phase distortion.

Once split, I apply the same band-splitting technique to the input track. Each frequency band of the track is adjusted to match the levels of the corresponding band of the pink noise, effectively using the noise signal as the source level. There's also the functionality to manage the stereo unlink, allowing you to match the left and right channels separately or bring them to mono, and to toggle RMS on or off for more experimental sound processing.

The Tonal Balancer includes a noise level control, impacting the amplitude across all frequency bands. You can visualize these changes as the match level plugin adjusts the volume up or down. The target knob also influences the volume modifications—fully down means the input and output signal levels remain unchanged, whereas increasing it leverages the envelope follower of the noise source.

I provide a live demonstration of these features, highlighting how the plugin acts as an indirect EQ by using the EQ'ed pink noise to influence the tonal profile of the audio signal. There's still progress to be made, particularly adding a threshold feature to prevent the amplification of the noise floor during quieter segments of the track.

Although the Tonal Balancer is still a work in progress, it functions effectively at this stage, and I'm excited to share it on my GitHub for free. I encourage viewers to try it out, create their custom curves, and share their feedback. To conclude, I appreciate everyone's support and invite them to subscribe or leave a thumbs up if they enjoyed the video.

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] Hey, folks, welcome back to another video.
[00:00:03] So I want to show you something here I made in Bitwig.
[00:00:05] It's called Tonal Balancer.
[00:00:07] This is how it looks like, very basic,
[00:00:10] just some remote controls here.
[00:00:12] And it kind of balances the tonal distribution
[00:00:16] of your track automatically.
[00:00:18] So this is the idea behind it.
[00:00:20] So I show you first how it sounds here in the background.
[00:00:22] You can see I have some kind of track.
[00:00:24] And this is how it sounds without.
[00:00:28] (upbeat music)
[00:00:31] And then I can bring in the auto EQ here with the target knob.
[00:00:41] (upbeat music)
[00:00:43] (upbeat music)
[00:00:46] I can also change here the target curve at the moment.
[00:01:07] It's, I think, pink noise.
[00:01:09] We can also switch this to smiley.
[00:01:11] (upbeat music)
[00:01:14] And these rise and fall buttons or knobs here
[00:01:17] change how drastically or how fast the EQ adapts
[00:01:22] to changes in the audio signal.
[00:01:27] (upbeat music)
[00:01:41] So at some point, if you make it too fast,
[00:01:42] it becomes a compressor, right?
[00:01:44] But you want to have a leveler.
[00:01:46] So it's not really a compressor, it's kind of both.
[00:01:50] So it's a very slow leveler, in my opinion,
[00:01:54] if you have the rise and fall here at least 50%
[00:01:57] and use RMS.
[00:01:59] So this thing is based around my target level preset.
[00:02:04] I showed you last week where I just extract the volume
[00:02:09] from another track or from another sound source
[00:02:13] and then apply it to your current sound source
[00:02:15] to match the level.
[00:02:17] But here, I'm using basically a noise source.
[00:02:22] So let's go to noise here.
[00:02:25] Well, let's open this up.
[00:02:28] So there's a noise source, which is the test tone.
[00:02:32] And test tone here is in pink noise mode.
[00:02:34] So it puts out a pink noise sound, okay?
[00:02:39] So on that pink noise sound, I put an EQ.
[00:02:42] You can see here's an FX selector.
[00:02:44] This is basically what you select here,
[00:02:46] pink and smiley, right?
[00:02:47] So in smiley, there is an EQ on this noise.
[00:02:52] So we change it to a slight smiley curve.
[00:02:55] So we amplify here the tops and the lows
[00:02:58] and reduce a bit in the mids here around five and thirds.
[00:03:02] And pink noise is just pure pink noise.
[00:03:06] So you can, of course, create your own curves here.
[00:03:08] You can duplicate this and say a custom curve
[00:03:12] and say, I want to reduce here this even more.
[00:03:15] Maybe I don't want to have that much top end.
[00:03:19] So you can create your own curves multiple ones here.
[00:03:23] And then you can select it, of course, here
[00:03:25] in the selector or in the dropdown.
[00:03:28] So this is the source.
[00:03:31] This is basically your target curve.
[00:03:34] That's where you want to have your frequencies
[00:03:38] in your audio signal.
[00:03:41] So this one, so this changed pink noise curve then here,
[00:03:44] or pink noise audio signal,
[00:03:46] is then split it into multiple bands here.
[00:03:49] You can see 22, 88/3, 88 to 200/3, and so on.
[00:03:54] So it's one, two, three, four, five, six, seven bands.
[00:03:59] I split this into seven bands.
[00:04:00] I think it's enough for the master.
[00:04:02] You don't want to have more bands.
[00:04:04] You can do infinite bands if you want to.
[00:04:07] And for the splitting here,
[00:04:10] I also use my idea with the convolution reverb
[00:04:15] to use, yeah, impulse responses to split the audio signal,
[00:04:20] which works pretty neat in my opinion.
[00:04:23] So there's also no phase, yeah, it's linear phase basically.
[00:04:29] So there's no phase distortion happening there.
[00:04:33] So this is then split it up into multiple seconds here,
[00:04:37] and that's basically it.
[00:04:38] So this is the noise thing.
[00:04:41] And then here on the sound,
[00:04:43] I'm using the input sound, which is our track here.
[00:04:46] And I also split this into the same kind of frequency bands,
[00:04:51] of course.
[00:04:52] And then on each of these bands,
[00:04:55] there is here one of my match level preset
[00:04:59] to match basically the level of the sound,
[00:05:03] of the pink noise or the, yeah,
[00:05:06] the noise sound from the first channel here,
[00:05:08] from this one, right?
[00:05:09] So I match basically the level on each of these bands
[00:05:14] and corrected upwards or downwards.
[00:05:16] So it's a leveler.
[00:05:18] And the source or the source frequency level
[00:05:23] is basically on the first track.
[00:05:25] And then you match the level here on this sound thing.
[00:05:29] You have also here a dry signal.
[00:05:31] So you can blend in the dry signal,
[00:05:32] also a major with the same,
[00:05:35] with the same technique with the band splitters here,
[00:05:40] but this is the full signal.
[00:05:41] But the compensation is in there, right?
[00:05:44] It's also 87 milliseconds.
[00:05:46] So it's, yeah, it's not finished.
[00:05:50] That's why I want to tell you.
[00:05:51] So you don't want to use dry
[00:05:53] or you don't want to use here the mix knob at all.
[00:05:57] So that's basically it.
[00:05:59] So there's nothing, nothing more to it, right?
[00:06:02] Just a band splitter and then matching each of these bands
[00:06:06] to a noise source.
[00:06:09] Then you have something you're like a stereo unlink
[00:06:13] where you can say, I want to match the left channel
[00:06:17] and the right channel in separately
[00:06:20] or I want to bring it back to mono, right?
[00:06:23] So unlink is basically this.
[00:06:26] Then we can switch on and off your RMS from the leveler
[00:06:31] and you usually want to have this on all the time,
[00:06:35] but if you want to make some sound experiments
[00:06:37] or you want to react faster to the audio signal,
[00:06:41] then you can turn this off.
[00:06:43] This is the noise level here.
[00:06:45] So we can also change the level of the noise.
[00:06:48] So if you pull this down,
[00:06:51] all the volume of all the frequency bands
[00:06:54] are going down or up and let's play this here for a moment.
[00:06:59] So you can see on each of these bands
[00:07:11] what this much level plug in is doing.
[00:07:13] You can see you're on change.
[00:07:15] I'm increasing here the volume, right?
[00:07:17] And here I'm slightly doing nothing
[00:07:23] and sometimes it's going down.
[00:07:25] So you can see the volume change
[00:07:27] if it's going up or down, right?
[00:07:29] Also interesting is this target knob.
[00:07:39] If you have this all the way down,
[00:07:41] then inside of these levelers, maybe decrease here,
[00:07:48] inside of these levelers, I'm using the same envelope follower
[00:07:53] for the multiplication instead of using
[00:07:56] the audio side chain signal for the multiplication.
[00:08:00] So what this means is if you have this all the way down,
[00:08:02] you make no volume change to the input signal.
[00:08:07] So input signal is output signal more or less.
[00:08:10] And if you then pull this up,
[00:08:13] then I'm using the envelope follower here
[00:08:16] of the noise source.
[00:08:18] All right, you can see this here.
[00:08:21] So let's go here to target 100%,
[00:08:31] which is probably not what you want to do.
[00:08:34] And then go here into the pink noise curve
[00:08:37] and just play around with this.
[00:08:38] Maybe make here the changes a bit faster
[00:08:41] so you can hear what I do.
[00:08:43] (upbeat music)
[00:08:46] So right, I'm not a cueing here the audio signal.
[00:09:00] I'm a cueing the pink noise sound
[00:09:04] from the test tone device here.
[00:09:07] And then splitting the signal and analyzing the volume,
[00:09:10] and then I'm forcing it onto my audio signal.
[00:09:13] So this is like an indirect EQ, all right?
[00:09:18] So this is, you can see here,
[00:09:20] I'm having noise all the time.
[00:09:22] So you can EQ, but basically your audio signal
[00:09:27] with pink noise.
[00:09:28] (upbeat music)
[00:09:39] Or maybe use here the custom curve.
[00:09:41] More bass.
[00:09:47] (upbeat music)
[00:09:50] So like I said in the beginning,
[00:10:06] something is missing and this is threshold.
[00:10:09] I want to implement your threshold.
[00:10:10] So it's not leveling up the noise floor
[00:10:15] if you play this here on a quiet part.
[00:10:20] It tries to raise here basically the noise floor
[00:10:28] and brings up all the quiet noise,
[00:10:31] which is not what you want.
[00:10:33] So I want to implement your threshold
[00:10:35] that only certain parts are getting amplified.
[00:10:39] So this is my idea behind it.
[00:10:42] But it works quite well so far.
[00:10:44] So you can try it out.
[00:10:45] I put this on my GitHub.
[00:10:48] So this is completely free.
[00:10:49] You can just download it, try it out for yourself,
[00:10:52] create some curves here if you want to,
[00:10:54] and give me some feedback, okay?
[00:10:57] So that's the tonal balancer preset for Bitwig.
[00:11:01] Let me know what you think.
[00:11:02] Leave a thumbs up, leave a subscription.
[00:11:04] Thanks for watching. Bye.