One of the most innovative plugins around bass mixing and mastering is without a doubt Basslane Pro from Tone Projects. I linked up with co-creator and renowned mastering engineer, Ian Stewart from Flotown Mastering, who was kind enough to provide detailed answers to my interrogation-like barrage of technical questions. What a champ. Thanks for indulging us Ian, and respect to Rune at Tone Projects for handling the code and DSP of this bass-beast, as well as many of the design and feature ideas. Here we go!
What inspired you to collaborate on Basslane Pro with Tone Projects?
Ian Stewart [IS]: About six years ago I was doing some investigation into existing tools that claimed to mono low frequencies, to see how they performed on different types of stereo signals, and found that pretty much all of them had non-ideal behaviors to greater or lesser extents. That got me thinking about what the ideal behavior for such a tool should be, and how it could be accomplished. In the end, I built an Ableton effect rack that could more or less do what I was after, albeit with some limitations. My original blog post detailing my investigations is still up here.
Eventually, during the summer of 2021, I connected with Rune from Tone Projects via a mutual friend. We chatted a bit about it and realized we had kind of similar ideas on how to approach low-frequency width management and started work on Basslane Pro, an update (really, a complete rebuild from the ground up) to the very first Tone Projects plugin from back in the early 2000s. In the end, it incorporated not only my original ideas to preserve level and antiphase content when narrowing the low end, but also an auto mode to manage channel priority when doing so, stereo harmonics generation for mono-compatible low-end perceived width, multiple dynamics modes, and some other cool goodies.
What are the core problems that it addresses?
[IS] If you’re just trying to center a bass signal that’s panned off to one side, the standard solution of using an elliptical filter (e.g. a high-pass filter on the side, or difference channel) can accomplish that, but you lose some level. Depending on how far it’s panned, that could be up to 6 dB. You can try to compensate for this with a boost in the mid, or sum channel, but that means you’ll end up boosting elements that were originally center-panned as well. In other words, the balance between bass and kick drum can get changed when using an elliptical filter and there’s not much you can do about it.
If a bass signal is panned beyond hard-left or hard-right, meaning it has an antiphase component to it, the level drop will be even greater. In the extreme case where a bass signal is entirely in antiphase — meaning only in the difference channel, opposite polarity, etc. — it will disappear completely when mono-ed.
When do you sum the low end fully to mono vs. other approaches?
[IS] Ultimately, I think this has to be informed by genre and what the predominant playback scenario(s) are likely to be, coupled with what you’re hearing (caveat, you need to be listening on a system that’s capable of true, extended low-frequency stereo playback without excessive room interference — much easier said than done). Listening in mono can be a helpful way to find the right balance though as it helps highlight any mono incompatibilities down low. Often, the low-frequency content in a mix comes from just a few elements, so if one loses a lot of level due to mono incompatibility, you could end up with a big hole in the frequency spectrum down there. Comparatively, many more elements tend to contribute to spectral density as you pass 100, 200, 300 Hz, so mono incompatibilities can be a bit more forgiving up there.
What are the other (less heavy-handed) approaches you may use to process low end width vs. a HP side filter?
[IS] A reductive low shelf on the side channel can be a gentler way to tame low end width than an elliptical filter, but it’s still subject to the level loss that can happen with an elliptical, albeit to a lesser extent. The width control on Basslane Pro essentially blends between a reductive low shelf and high pass though, so you don’t have to pick and choose, and mono recovery works for both.
Do you ever leave the low end untouched in terms of side information?
[IS] Absolutely! It’s not something I go hunting for, I only address it if I hear something that sounds off.
When you sum to mono, do you often see a noticeable level loss?
[IS] It’s definitely very context dependent but yes, I’ve seen level loss anywhere from 6 dB to full on cancellation.
What type of content are you finding is in the very low end in terms of side information?
[IS] There are definitely some bass synths that create substantial antiphase content. It’s also not unusual for LF content to get into reverbs which can just end up kind of muddying things up (although that can often be better filtered out than recovered to mono). Low toms and other percussion can also end up out near the edges in the low end, which can cause problems on vinyl.
What side content is worth getting rid of and what is worth keeping? How can producers or engineers make better decisions here?
[IS] This may sound simplistic, but from a mastering perspective, if it’s intentional it’s worth keeping (although that doesn’t mean it couldn’t benefit from some repositioning), while if it’s unintentional — think about the reverb example above — it might be worth filtering out entirely.
The Basslane Pro manual refers to “Lost Dimension” when cutting the side fully. Could you elaborate on what you mean by that? What information is in there that you’d miss with stripping out the side completely?
[IS] While humans aren’t great at perceiving the directionality of very low frequencies, the transition frequency where imaging becomes apparent can vary by individual. Coupled with the fact that there’s probably some intention behind low-frequency stereo information, we found it was useful to have a way to create harmonics that psychoacoustically reinforce the intended stereo information, while not compromising mono compatibility.
You mentioned that you have a lot of drum and bass pre-masters coming in with up to a 400 Hz side cut and that you’re using the stereo harmonics feature to fix that. Could you elaborate?
[IS] Definitely. Maybe in part because of some of the online hyperbole around monoing the low end, some mastering engineers have noticed that some mixes are coming through with no stereo content below 200, 300 or even 400 Hz. In these cases, the stereo harmonics section can allow an ME to add some mono-compatible, musically-related width to the frequency range of their choosing.
How are the upper bass harmonics generated? What harmonics specifically?
[IS] The code specifics are above my pay grade, but the harmonics generation is fed by the filtered low-end signal dictated by the main frequency and slope parameters, along with the sub filter. There are a few different harmonics algorithms, 2nd/3rd, even/odd, and crunchy, that have different balances of harmonic content.
How, specifically, are they made “wide”? Is there a panning process? A haas process? Any “under the hood” insights would be great.
[IS] Alternating harmonics are panned in opposing directions and then an M/S balance is applied to control the overall panning and width.
How does the bass harmonics generation work differently in side mode and mid mode?
[IS] The mid/side switch determines which portion of the filtered signal is fed into the harmonics generator. By using the side signal you can recreate or reinforce the sense of width created by low-frequency side channel content. Conversely, by using the mid signal you can create mono-compatible width for mid-channel content.
The Stereo mode for harmonics generation is actually dual mono A is it not? Could you elaborate on that and why that’s a neat aspect to the design? (EX: harmonics following a tom fill with the drums all panned differently)
[IS] Stereo mode is in fact dual mono, fed by the left and right channels of the filtered signal. This means that the harmonics follow the original panning of the low-end, so anything that moves around in the low end can have its original position psychoacoustically reinforced by the harmonics, even if they’ve been narrowed or monoed entirely.
You mentioned Ableton’s Operator in your interview with Nicholas from Panorama Mixing & Mastering, with respect to wide bass harmonics – the additive engine vs. FM. Neat idea! For us Ableton Live users, could you elaborate on what you were doing there?
That video interview can be seen here.
[IS] Sure. My proof of concept for stereo harmonics was to set up 3 instances of Operator in an instrument rack. One was panned center and just playing the fundamental, one was panned right with only even harmonics, and the last was left with only odd harmonics (or maybe the directions were flipped, I don’t exactly recall).
In terms of the Dynamics section, I noticed there’s no attack parameter, but that “detection speed” and transient response are changed by another parameter. Given the more minimal set of controls here, compared to a regular compressor, could you explain the setup and any tips you have on how to compress the low end?
[IS] Sure. The dynamics section is basically running a custom tuned version of Unisum (the Tone Projects mastering compressor) in the background, and Rune (with the help of our beta testers) spent a lot of time setting the parameters to work well for low-end. When set to stereo mode it behaves like the low band of a multiband compressor, which as often as not is how I use multiband compression (low band only). This means you can basically just pull down the threshold till you start getting some compression, and then reduce the Amount control if you want to back off the ratio. There are lots of interdependencies though, like cutoff frequency to attack time, etc. We recognize that the control set might feel pretty limited for tweakers and control freaks (I count myself in those categories), but for that certain type of punchy, gluey low end compression, it’s set up to work really well with just that limited control set.
The crossover filters are linear phase, so do you ever have any issues with seeing pre-ringing on things like kick drums?
[IS] In almost 2 years of use (counting development alpha versions) I have yet to encounter a scenario where I noticed audible pre-ringing. There are some good technical reasons for this too, so it doesn’t just have to be anecdotal. First, when the low band and high band are at unity, the polarity of the pre-ringing is opposite in each band. So, when you sum them back together the pre- (and post) ringing cancels. Second, the crossover is on the side channel so anything in the mid channel will pass through unaffected. Third, even as you start to process the side channel and it strays from unity, the pre-ringing will only be that contributed by the delta signal, which is generally substantially lower in level. It just might be the closest thing to a free lunch I’ve ever had!
Many DAWs and simplistic plugins have a “bass mono” function that usually involves a minimum phase side HP. Could you discuss the drawbacks of this and considerations in use? And why Basslane’s approach is better?
[IS] Sure. First, the level loss thing. Panned signals will inherently lose level when monoed in this way, to the extent of disappearing entirely if they’re in antiphase. Second, stereo positioning in mid/side is very sensitive to the phase relationship between the mid and side channels. So, when you put a minimum phase HP that shifts the phase of the side signal by 90, 180, 270, etc. degrees around the cutoff frequency, you end up with a smeared image well above the range you were trying to mono. With higher order filters like 12, 18, 24 dB/oct the left/right image can actually flip entirely at the cutoff. Basslane Pro’s linear phase filters avoid this, and also allow for mono recovery.
