Tags: posts polarity-music Bitwig Tutorial Mixing

Game-Changing Bitwig Mixing Trick: Pre-Emphasis & Auto Leveling

Tutorial | Jun 27, 2025

In this video, I demonstrate a unique technique in Bitwig using pre-emphasis and de-emphasis with tilt EQs to mimic multi-band limiting and achieve more balanced loudness across frequency ranges, without traditional band splitting. By tilting the analyzing signal fed into auto-levelers or limiters, you can shift how loudness is controlled across your mix, changing the tonal balance solely through volume adjustments. This approach allows for more natural and flexible mixdowns and can be applied not just to limiting but also to compression, reverb, and other effects, potentially streamlining your workflow and achieving more transparent results.

You can watch the Video on Youtube

Short Overview

In this video, I explore a new idea I recently discovered using Bitwig that could be a real game-changer for mixing and mastering. I demonstrate how different analyzers can show different loudness readings based on their tilt settings, and how this affects our perception of sound. Instead of relying solely on multiband compression to control different frequency ranges, I experiment with using pre-emphasis and de-emphasis EQs around a full-range limiter. This technique lets me manipulate which frequencies get limited harder, all without needing to split the band. I then apply a similar concept to auto-leveling by tilting the analyzing signal, allowing me to adjust the overall tonal balance of a mix just by changing volume on each channel, not EQing. It’s an interesting way to quickly rebalance a mix, especially if your tracks are already separated by frequency content. While there are some limitations, like value persistence on bounce, I find this method powerful and surprisingly effective for shaping a mix. I hope this sparks some ideas for your own workflow.

Introduction

In the last few days, I experimented with a new approach in Bitwig, and I believe what I discovered could be a game-changer for mixing and balancing audio. I want to share what I found, as it might be useful or at least spark some inspiration for others. I will start by outlining the problem before walking through my solution, which centers on understanding and leveraging the concepts of pre-emphasis and de-emphasis, commonly found in audio processing but rarely used in the way I demonstrate here.

Understanding the Problem

Frequency Perception and Metering Differences

In Bitwig, I set up a test channel with a tone generator, spectrum analyzer, and peak limiter, using the histogram for level analysis. I noticed that if I play a bass sound at 40 Hz and push the gain so the level reaches -3 dB, both the histogram and the spectrum analyzer reflect this. But if I move the frequency to, say, 4 kHz, the spectrum analyzer shows a big difference in level, while the histogram's reading barely changes.

This is because the spectrum analyzer has a tilt setting. With the tilt at 3 dB per octave, the visual representation of the spectrum slopes downwards, matching how humans perceive audio frequencies. Our ears are more sensitive to mid-to-high frequencies, so the analyzer uses this tilt to display sound in a more perceptually accurate way. The histogram, on the other hand, shows the technical, unweighted loudness.

Why Tilt Matters

If I reset the tilt to 0 dB per octave, white noise displays a flat line, as every frequency has equal energy. This means the peak limiter's histogram represents "white noise" loudness , each frequency has the same energy value, regardless of whether they are more or less perceptible to our ears.

We use tilted spectrum displays to make better EQ decisions, but our limiters and compressors often react to unweighted loudness, which does not always align with our perception. This disparity often pushes producers to use multi-band compression so that different frequency bands can be processed independently, matching the way we perceive their impact on loudness.

My Solution: Pre-emphasis and De-emphasis Processing

The Multiband Workaround

Traditionally, multi-band limiters and compressors are used to process frequency bands differently. We split the signal, process each band with tailored settings, and then sum the output together. My goal was to find a way to get similar results without having to split the audio into multiple bands.

Implementing Pre-emphasis and De-emphasis

The breakthrough idea was to use pre-emphasis and de-emphasis filters. This technique is common in various types of audio and signal processing, especially for improving noise performance, but it is not commonly used for dynamic processing in music production.

Here's how I set it up:

The signal flow is:

  1. Signal passes through pre-emphasis (tilt EQ).
  2. The altered frequency balance dictates how the limiter or compressor reacts.
  3. The original balance is restored after limiting with another tilt EQ.

Practical Uses

This approach allows me to make the limiter react more aggressively to selected frequency ranges. For example, I could emphasize the high frequencies before the limiter so cymbals and hi-hats hit the limiter more prominently, then return the tonal balance after limiting. This mimics a multi-band response with just one full-range device. The same approach can be applied to compressors, using side chain EQing to shape the side-chain signal.

Side-chain Analysis for Volume Automation

I extended this concept to my auto-leveling tool. Typically, it uses a full-range envelope follower to automate gain. By feeding a tilted version of the signal into the side-chain input (while leaving the actual audio untouched), the gain automation is based on a perceptually tilted analysis. This lets me adjust the tonal balance of the mix by changing levels rather than using EQ, providing a powerful and fast mixdown helper.

Macro Controls and Workflow Integration

To make the process user-friendly and flexible, I created macros for common parameters:

I demonstrated how, by copying this macro setup to all main tracks in a project (drums, bass, pads, etc.), I could dramatically change the overall mix balance by simply adjusting the tilt, without ever applying EQ directly. This is particularly useful if each track occupies a specific frequency range.

Limitations and Further Considerations

While this method is powerful for achieving a fast, perceptually balanced mix, it is not a replacement for detailed mixing decisions, like adjusting specific instrument prominence. Also, my auto-leveling tool does not persist its level calculations when exporting audio, causing some issues with bounced tracks, but I am working on a solution. This approach is easy to extend to other effects, such as reverb, where you could remove bass before the verb and restore it afterward.

Conclusion

This exploration into using pre-emphasis and de-emphasis with full-range limiters and compressors opens up a lot of creative potential, offering a way to control tonal balance without complicated routing or multi-band splitting. It’s not a common method in music production but could inspire new tool designs or workflows, especially for quick and transparent mix balancing. I will continue refining the process, and I encourage others to experiment with this idea and possibly integrate it into their own productions or plugin designs.

Full Video Transcription

This is what im talking about in this video. The text is transcribed by Whisper, 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.

Click to expand Transcription

[00:00:00] Hey folks, welcome back to another video. So the last few days I tried something and I found something and it's probably a game-changer
[00:00:07] It's
[00:00:09] Hopefully kind of useful to you. I don't know you can learn something always right so before I show you the solution
[00:00:16] I want to show you the problem. So here in bitwig. I have some kind of test channel
[00:00:22] so I have a test tone here and the spectrum analyzer and we have your peak limiter and we just use here more or less the
[00:00:30] Histogram the level histogram
[00:00:32] So let's say we have here a frequency. I don't know 40 Hertz
[00:00:36] Maybe a bass sound we push up here the gain. So let's say we have the bass at minus
[00:00:42] 3db
[00:00:44] Okay, you can see here the level on the histogram and you can also see the level here on the spectrum analyzer
[00:00:51] so let's change the frequency on that let's say we have instead of 40 Hertz we have maybe
[00:00:58] 4 kilohertz
[00:01:00] Now you can see here the level is way up there
[00:01:07] But here the level stays the same. Okay, so maybe go back
[00:01:11] You can see 40 Hertz
[00:01:15] Slow and level but the level stays the same we go up
[00:01:20] Level is higher here on the spectrum analyzer, but here it stays the same
[00:01:24] So the reason for that is of course pretty obvious the spectrum analyzer here has a tilt setting
[00:01:30] so the tilt is here on 3db per octave which means the
[00:01:34] Slope or the visuals only the visuals go down here or are pushed down
[00:01:40] Because that's how we
[00:01:43] perceive audio or frequencies or ears are just more sensible to
[00:01:49] Upper frequencies around 1k then lower frequencies
[00:01:53] So we just change the visuals on the spectrum analyzer so they match more like how we hear music or how we hear
[00:02:01] audio signals
[00:02:04] but the histogram here doesn't it's always shows you the exactly
[00:02:10] Technical loudness, so when we switch the spectrum analyzer here to let's say 0db per octave, which is by the way a white noise
[00:02:18] So when we switch this here to white noise, you can see it's a flat line
[00:02:22] um
[00:02:24] So with white noise we can change here the frequency to
[00:02:28] 40 Hertz and it stays the same here on this visualizer, of course
[00:02:35] And also stays the same here
[00:02:39] So what this means is the peak limiter actually shows you white noise or white noise tilted
[00:02:44] It's not tilted. It's technically the same loudness for each frequency
[00:02:49] That's how you see it in here, and that's pretty fine. It's pretty normal and it's very useful
[00:02:54] So with the spectrum analyzer we use the tilt setting to actually make more sense or to make more
[00:03:01] Based decisions for how to treat different frequencies with an EQ
[00:03:07] So yeah, let's switch this back to a 3db per octave
[00:03:11] So usually for my drum bass stuff I use
[00:03:18] What's it called?
[00:03:20] multi-band limiters multi-band compressors and so on and that's because because of that, right?
[00:03:27] We use here maybe an FX3 splitter and
[00:03:30] Then we put a
[00:03:34] Limiter in each of these bands and then use different settings for the gain or for the threshold
[00:03:40] To actually counter that we have lower frequencies in the top end than in the low end because that's how we hear music
[00:03:48] Right, so that's why we use basically multi-band compression
[00:03:51] So we can compress different bands differently and act on the different frequencies differently
[00:03:57] That's that's why we have this but my idea was can we actually use one full range?
[00:04:05] Limiter and make something like a multi-band limiter out of this without using
[00:04:13] band splitting and
[00:04:16] the answer is yeah, you kind of do this and
[00:04:20] My idea for this was to use something. That's called pre emphasis and de emphasis and that's used in
[00:04:27] in all kinds of different audio situations, but it's not used in
[00:04:31] At least I haven't seen it in
[00:04:35] Limiting and compression in compression is probably used a lot. So let's do this here. So we use a chain a
[00:04:43] Chain device and we put you at the limiter in there and in front of this we use a tilt a queue, right?
[00:04:49] Also after the limiter we use a tilt a queue and you can probably already guess where this is going. So we use here
[00:04:57] Three emphasis and the emphasis use the same
[00:05:03] Values here plus 24 minus 24 DB
[00:05:07] So now when we change here the
[00:05:11] The gain you can see this here and when we change the frequency
[00:05:15] it stays the same because it's
[00:05:18] Kind of white noise. It's the same technically the same loudness for for all the frequencies
[00:05:24] But what I want to do is I want to emphasize or I want to react to lower sounds maybe
[00:05:32] Faster or slower. So what I can do now is I can say tilt the spectrum here
[00:05:41] get the lows out of the
[00:05:43] limiter and
[00:05:46] Push up the top end bit higher. So now I have a different balance here. Let's change it
[00:05:54] Let's go to this frequency
[00:05:57] gain
[00:05:59] So when we change now here the frequency you can see the level goes down
[00:06:03] Because we go down and frequency and distilled kind of pushes here the frequencies lower in volume down here
[00:06:11] And then after that we count up basically this EQ. So we go with this
[00:06:16] Frequency curve into the peak limiter we limit then the audio signal and then we correct it back to the original balance
[00:06:24] So this is called pre emphasis and the emphasis and it's maybe interesting if you want to just
[00:06:30] Compress the low end. So usually when you have a track you push the limiter to this track and all the bass sounds the kick drum and so on
[00:06:39] You catch basically early on here with then with your threshold because this is what the loudest the loudest signal is
[00:06:46] But sometimes you don't want to have this you want to just you know compress the hi-hats also on so you do this
[00:06:52] You push the hi-hats up right into the limiter and then you push the hi-hats back down where they are before that
[00:06:59] So with this you have kind of a multi-band limiter
[00:07:03] kind of
[00:07:06] Or behaves in kind of the same way
[00:07:08] So the funny thing is you not only can do this with a limiter
[00:07:13] You can also do this with a compressor, right? And here it's
[00:07:16] We can do this here with compressor plus so in compressor plus we can use the side chain FX
[00:07:23] So you can use a tilt EQ in here if you want to and
[00:07:26] I guess some people already do this. You can also use an EQ in here and
[00:07:33] Say get rid of all the bass sounds for the analyzing part here of the compressor
[00:07:39] So the kick drum and the bass sound doesn't you know make the sound
[00:07:44] Pumping in a way so you get the kick drum and the bass out of here
[00:07:48] So this is also something you can do in the compressor plus. I think that the dynamics device also does this
[00:07:55] There's also here. Yeah, there's a side chain FX so you can also put here
[00:08:01] EQs in there tilt EQ
[00:08:03] But there are also some devices where you don't have this so you can say you have your distortion device
[00:08:10] Something like this is a full band distortion years. That's here some kind of band pass in there
[00:08:18] We can also use this with a filter
[00:08:21] Filter plus for instance here with a transfer curve, right? So you can say compress or
[00:08:28] kind of
[00:08:31] Saturate overtones or the top end more than the low end
[00:08:35] It's possible with this, but that's not actually what I want to show you
[00:08:41] I just want to show you what you can do just with this
[00:08:43] pre-emphasis and D emphasis
[00:08:47] What I wanted to do
[00:08:49] Let's delete all of this here
[00:08:52] What I wanted to do is I want to mix down a track
[00:08:56] with some kind of normalization
[00:09:00] but
[00:09:02] The normalization
[00:09:04] Analyzing signal is tilted
[00:09:06] Which gives me in the end. So this was my thought thought process
[00:09:10] Was then better balanced
[00:09:13] So let's say we have your distract and this track is not mixed down at all and
[00:09:17] We put here. Let's say spectrum analyzer on this
[00:09:22] So we have some kind of frequency distribution here, which is probably fine for a track. I don't know
[00:09:36] So my idea was can I actually just twist one button and change all of these tracks in loudness
[00:09:44] Just with volume. No eqing
[00:09:48] Nothing just loudness. Can I change the volume of these tracks?
[00:09:51] So there's the so that the top end is lower or the lower end is higher or the other way around so I can
[00:09:58] Change basically the distribution without using an eq and it turns out you can and this this maybe
[00:10:05] Is just an interesting idea. Maybe
[00:10:08] Someone wants to put this in the plug-in
[00:10:11] So let's go you to pads. Let's put this out here
[00:10:16] And push this yeah
[00:10:18] So let's try this out. So we have here music we have here an arp
[00:10:24] So my idea was let's use actually my auto level
[00:10:31] this one here
[00:10:34] which changes the volume it kind of
[00:10:36] Takes here a threshold and everything that's above the threshold gets pushed up to zero to be you can see
[00:10:43] There's a change in volume. We also have an amount no peers. We can say this is the original loudness and
[00:10:50] This is my
[00:10:53] Changed loudness so we can morph between original volume and my adjusted volume
[00:10:58] So when we go in here we more or less use a envelope follower on on the full range signal
[00:11:12] So what if we actually tilt the signal is this possible and the answer is yes
[00:11:20] This is possible. So my idea for this was to go into the pre FX box here of this auto leveler and I use a tilt eq
[00:11:28] Like this and I put this into an FX layer here and use the second tool device and
[00:11:38] We mute the tilt output here. We just mute the output. So everything that goes
[00:11:43] Into here, maybe let's loop this here for a moment
[00:11:48] We'll be tilted and in here we use a side chain input
[00:11:54] And we grab the signal coming here from the tilted signal
[00:11:59] all music arp
[00:12:02] Auto level chains pre FX
[00:12:07] tilt
[00:12:08] Maybe tilt out. Okay, so now we can
[00:12:12] Use this here in front so we can tilt the signal and the tilted signal
[00:12:19] Goes into this side chain input, but it doesn't go into the auto level thing here with this input, right?
[00:12:27] So we have two inputs now. So this is our side chain input and for the side chain input. We go basically here into the
[00:12:35] Analyzing part so the audio signal is
[00:12:37] Untouched we only change the volume on this
[00:12:41] But the tilt eq changes the analyzing signal and we go into the side chain input here
[00:12:46] And then we analyze the loudness of this and then we change the volume of the original signal that's idea
[00:12:53] Okay, so let's try this out here makes a bit
[00:12:56] Makes a bit louder first for most
[00:13:00] . Mm-hmm. Oh
[00:13:05] , yeah
[00:13:07] , yeah
[00:13:09] You
[00:13:37] So with this tilt setting we can now change how the envelope holder receives audio and then it makes
[00:13:43] Decisions on that and it turns out it works actually really well if you do this on all channels
[00:13:50] So what I want to do now, so I want to use to the project thing and I
[00:13:55] Just create some macros and I say this is maybe the rise time
[00:13:59] This is maybe a fall time
[00:14:05] then
[00:14:07] We want to use
[00:14:10] Maybe the amount so we can see or can hear how it sounds before and after
[00:14:16] Then we of course want to change tilt setting
[00:14:22] What's this 24 DB
[00:14:27] Maybe we want to change also the gain the overall gain of this
[00:14:34] Let's go to go to one here. That's okay
[00:14:41] Yeah, let's gain that's okay something else maybe the threshold
[00:14:54] threshold
[00:14:56] Okay
[00:15:06] So now all we have to do is to just copy this to all other tracks, so let's put this here
[00:15:15] Pets
[00:15:23] Also your precautions kick drum and maybe the bass and maybe put this also on the music bus
[00:15:31] So let's look at the frequency spectrum here and let's change here the
[00:15:43] This is actually this game
[00:15:47] I think this is the tilt setting right for what to change the name here
[00:15:54] Yeah, that's the tilt
[00:15:56] Can you hear the difference now we have lower lows and more highs and
[00:16:11] It's not a cueing. It's just changing the volume on each of these tracks based on the tilted signal
[00:16:19] and
[00:16:21] This works the more or the better you have
[00:16:33] Frequency
[00:16:36] Let's say you have on different channels. You have more like singled out frequencies, of course
[00:16:42] if you have like channels with full distribution of all frequencies so bass and top-end it of
[00:16:49] Course it doesn't work that well, but it's but you know if you have like on each track bass
[00:16:56] Then they have a top bass layer, then you have hi-hats and so on it works pretty well actually
[00:17:03] We can also change at the full time
[00:17:06] Which is more like a compressor actually
[00:17:15] And then here the gain setting is actually not the gain for the master it's changing the gain on each of these tracks individually
[00:17:22] You can see all the volumes are down
[00:17:24] And then all you have to do is you watch this the spectrum analyzer and
[00:17:37] you know change the tilt here and
[00:17:41] Without a cueing you can balance the tonal. Yeah, the tonal balance of the whole mix down
[00:17:47] It's basically a mix down helper kind of of course for a good mix down
[00:17:53] You have to go in here and say all I want to have the melody actually a bit louder
[00:17:58] I want to have this FX sound a bit quieter and so on so you have to make some
[00:18:03] Specific decisions here and there, but it gives you a nice
[00:18:08] Balanced mix down in my opinion just with this tool that the drawback of this always is that my auto level actually doesn't save
[00:18:16] the Calculate calculated values. So if you want to go here to
[00:18:21] Export and bounce the audio you have some volume changes in the beginning because no values are persisted
[00:18:29] And it has to recalculate all the values here
[00:18:32] When the tracks or when the instruments begin to play that this is a bit of a downer
[00:18:37] But I'm working on that. Maybe I find a solution
[00:18:41] In my opinion so far. It's pretty fun to use. So this is what I want to show you it
[00:18:47] I haven't seen anyone doing this so far. Maybe it's a completely stupid idea
[00:18:54] But it gives you so much power here with the project remotes or project modulators
[00:19:01] To actually change the compression on all tracks at the same time
[00:19:07] Tilt the spectrum and change the loudness. Maybe you can also add here some distortion to it
[00:19:13] Right and we can distort on each tracks on each track differently
[00:19:17] Anyway, I'm talking out of my ass here
[00:19:25] That's basically it I want to show you this maybe I put this tilt thing
[00:19:31] On this auto level and put the preset up but you could see it. It's not very
[00:19:36] Complicated you can do this with all kinds of audio effects not only
[00:19:41] The auto level are not only the peak limit. I you can do this also for reverb so you can take out
[00:19:47] The bass in front of the reverb then you go through the reverb and after the reverb you put the bass back in I
[00:19:54] Don't know. Maybe it's useful. I
[00:19:58] Kind of found I found it interesting and I probably want to use the game changer
[00:20:03] Anyway, thanks for watching leave a like leave a subscription to your next video
[00:20:08] Bye.
[00:20:10] [BLANK_AUDIO]