Author: Starthief

Yesterday morning I woke early, annoyed by the cat (this happens a lot). I was sitting in the big easy chair in the living room, not really dozing but not ready to be officially awake either, messing with SynprezFM on my phone. That’s a surprisingly okay Yamaha DX7 emulator which a whole bunch of presets (but limited editing). I happened to find a pad sound I liked in that moment, and wanted to record it on the computer.

So I grabbed a stereo-to-dual-mono splitter cable, which didn’t fit with the case on. Taking that off, I noticed the back of the phone was peeling away… uhoh. I’m not 100% sure that battery swelling was what did it, because it wasn’t the super dramatic “spicy pillow” look, but it seems like the most likely cause. So I found a place that does same day battery replacements (got lost on the way, then got quasi-lost again on the way back to pick it up…)

After reading up on it, it turns out a little swelling isn’t unsafe, but charging a battery that’s a little swollen can lead to a battery that’s a lot swollen, and at that point it’s at risk of a “thermal event.” And nobody wants a thermal event.

So I got to experience, for just part of a day, what it’s like in 2023 to suddenly not have one’s phone available. My spouse still had her phone, and I could still turn mine on briefly when absolutely necessary… so I wasn’t completely cut off from civilization. But still, reaching for the phone is a reflex.

Want to know what time it is? What the temperature is officially? Want to use GPS? Play music in the car? Check if you have a message from the repair shop? Kill a couple of minutes with a puzzle or web browsing? Look up an unfamiliar Korean dessert? Log in to work (we use two-factor authentication)?

I had thought about replacing my phone. My Samsung S10e is now 4 years old, which is about 80 in phone years. But this cost a lot less, and aside from needing the battery I don’t really feel like my phone is insufficient in any way. When it’s slow, it’s not the processor but the network; when photos don’t come out well it’s not the megapixels but the teensy lens. (Granted, lenses are more typically dime-sized now rather than notebook paper hole sized… but still far from the bazooka-sized lens for decent astrophotography.)

Anyway, I got my recording done, and the next album just passed the 25 minute mark so far. The “synth choir” thing is pretty abstract in places, but that’s fine; it’s serving more of a sort of ritual purpose than anything else.

I’ve been doing another round of Guild Wars 2, with my usual pattern of deleting an old character, starting a new one, leveling to 80 and quitting. I thought I was going to run a Ranger next… but the current Necromancer absolutely breezed through 1-80 with zero deaths and soloed several open-world Champions, and did it in style as a gender nonconforming Sylvari. So I switched to Reaper and that’s a new experience — new skills and rotation to learn. I’ll probably stick with them for a bit.

Superbooth is over, and the wave of slightly delayed videos from it has crested. (It’s hard to be at a huge awesome event like that, shoot video and also edit and publish it immediately if you’re not relatively big like Sonic State.) The wave of videos about Spectraphon is continuing. Oh and yes, I put in my pre-order.

I’m pretty confident I understand how the module works:

• Where most additive synthesis uses large numbers of sine waves from separate oscillators, this is using the method from the rare Buchla Touche: Chebyshev polynomial waveshaping. When fed a sine wave with a particular amplitude, these not-too-complex mathematical functions will result in a sine wave at a multiple of that frequency. You can put those formulae in a lookup table and it’s super efficient. With several of these in parallel, you can create a whole harmonic series from one oscillator, then mix them together at the desired levels.
• That means that unlike some additive synthesis, you don’t have any inharmonic partials. (The equivalent on Odessa would be no Tension knob.) Note though, just because all partials are an integer ratio of the fundamental, they’re not all integer ratios of each other
• Conjecture: this means you don’t need to do an FFT analysis to get the spectral data, just a series of bandpass filters tuned to harmonic ratios, with envelope followers. Much like the Buchla 296, one of the other major inspirations for the module. The advantage here is lower latency and “clean” data that fits exactly the needs of the oscillator.
• In SAM mode, Slide tunes the filters, and Focus slews the amplitudes. Pretty simple.
• In SAO mode, instead both knobs move through the stored array of amplitudes (as coarse and fine controls). I’m curious why they didn’t keep Focus as a slew for changes between array locations, but perhaps it does interpolation, and most likely the design choice will become clear when I start using it in practice.
• The FM seems most likely to be implemented with phase modulation. This is super efficient in digital modules, and allows the sine output to go unchanged (which is important for how it feeds the FM bus, and also useful in patching).

That said, a solid theoretical understanding is different from the actual sonic consequences and the feel of using it. The module is very chamelonic, and aside from the obvious vocoder stuff, the Sarah Belle Reid video sounded like SBR, the James Cigler video sounded like Cigler, the Cinematic Laboratory video sounded like Cine Lab. I’ve little doubt that, while it will influence and shift my music making, it’ll still sound like Starthief.

---

I feel like the last couple of years have been big for delay effects, at least personally.

• Inertia Sound Systems Hinder from September ’22. A BBD-ish plugin, it features a filter, compressor and overdrive in the feedback path
• The plugin version of Imitor is from October ’22. It’s greatly expanded in features from the Versio firmware, with much more control over the taps and can make a nice psuedo-reverb, wobbly chorus etc. among other things.
• The Moogerfooger plugin bundle was also from October, and includes a very nice delay.
• In January ’23, the Humble Bundle let me pick up Objeq Delay super cheap. A modal resonator inside of a delay… neato.
• Retronaut was from February ’23. It’s not marketed as a delay, but a “lo-fi nostalgia machine” with vibrato and chorus. But it’s implemented with delays, and you can increase the “lag” time to reveal that, and get warbly reverby echoes; it does the same sorts of thing one often wants a “retro” delay for anyway.
• I picked up Jroo Loop in February as well. It’s designed as a sound-on-sound looper, but keep it recording and feeding back and it’s effectively a delay. (a nicely lo-fi one too if you slow down its sample rate).
• Dedalus Delay is from March. It’s got optional granular features for modulating delay time, plus filters and overdrive and just generally nice mojo.
• Yester Versio dropped in April. A lovely 3-tap delay with chorus and folding.
• So did Phonolyth Cascade. It lives somewhere between delay and reverb, as it has extensive control over diffusion.

There were a couple of other promising delays that I passed over because it just gets to be too much. Keeping in mind that this list joined the power team of Mimeophon, Beads, Stega, Echo Cat, PlexiTape, RatshackReverb, Rift, Stardust 201, Sandman Pro, Valhalla Delay, Wires, and Bitwig’s flexible Delay+.

I hope plugin developers delay development of delays for a while!

I’ll point out though, the plugins were cheap (in some cases with early bird discounts), Yester Versio was a free firmware for a module I originally got as a beta tester (but I bought the front panel), and Jroo Loop was pretty reasonable. Up until I ordered Spectraphon, I had sold a bit more gear than I bought this year, including software. I’m not sure what I might wind up selling as a result of getting Spectraphon, but it’s likely to balance out pretty well.

Superbooth 2023 began yesterday, and with it a flood of new product announcements, demos, and so on. This year it seems the synth industry is bouncing back from its relative slump with a vengeance.

There’s a lot and I’m not interested in all of it, but:

• WMD which went out of business last year, is kind of coming back in the form of AMMT, a partnership of four designers including William Matheson. They’ll make some of the more popular WMD modules and possibly create new versions of some of the good ideas that could use an update, as well as creating new things. So that’s pretty promising.
• Noise Engineering announced the Roucha firmware for their Legio platform. It’s a filter and wavefolder, which I thought might be redundant since I like Lacrima Versio so much already. But no, it’s a great combination and my new favorite Legio firmware. Legio and Versio just keep on giving!
• Korg showed the Acoustic Synthesis Phase 5, a very cool little box that’s like a miniature Rhodes electric piano with an Ebow for every tine, for haunting sustained tones. It’s just an R&D prototype, but maybe they’re showing it to gauge the public interest. This particular public is very interested.
• TipTop has a couple of big Buchla collaborations, as expected… but they also announced a new thing called ART that carries polyphonic note messages on a cable within Eurorack. It is kind of crazy, apparently proprietary and incompatible with anything else. Personally, I don’t much like the trend of polyphonic MIDI stuff masquerading as modular, and this is worse because they’ve invented a whole new format for it that can’t talk to anything else.
• Out of Xaoc’s new stuff, the one item of interest to me is Rostock — a shift register or delay line for the Liebniz system. It has a max of 64 stages, which is decent-sized for sequencing but miniscule for an audio delay. But the potential here is high. You can loop the data, and optionally XOR it with incoming data. There are independent clock inputs for each bit, which would be great for mutating patterns and potentially wild for audio. The stage length is constructed from 32, 16, 8, 4, 2, and 1-stage sections, so as you change the length you switch which buffers are active and introduce glitches. It’s the kind of thing I would like to experiment with, but I’m not sure what I would drop out to make room. Not Synchrodyne because it’s also a useful clock for Liebniz.

But the biggest news to me by far is the Make Noise Spectraphon.

This is a DPO-style complex oscillator, but replace the two analog VCOs with digital spectral additive resynthesis…! In SAM mode it works as a sort of vocoder, analyzing input audio and translating it in real time to additive partials, with control over the range of analysis and a sort of spectral blurring or even a resonator kind of feel, with even and odd partial outputs. In SAO mode it uses stored arrays of additive data to synthesize, letting you scrub or step through the snapshots and offset the frequency of the odd and even partials. And in either mode you can FM it (apparently phase modulation since it leaves the fundamental sine output unaffected, and cross-FM is stable).

This sort of module is a serious chameleon, so some of the demos have sounded fantastic and very relevant to my interests, and some have been less to my liking. It all depends on the source material and what you do with it. The potential seems very high to me here. I can certainly see integrating it more with the software end of things, playing sample loops into it or using software drones as the source for its spectral data.

The FM-with-chords thing sounds very Akemie’s Castle to me. But being a complex oscillator with lots of timbral control puts it in Shapeshifter’s zone… and as an additive oscillator it’s also trading on Odessa’s ground. Obviously it’s not a direct equivalent of any of those, and it does some things none of those do either. It’s 34 HP, bigger than everything else I have except Akemie’s Castle. But aside from its size, it has the potential to disrupt more stuff; I feel like the best choice is to get one, try it out, and see what shakes loose.

The new keyboard is here, and the 2.4G dongle doesn’t work with Teletype anyway, so I could have kept the cheaper open-box unit and saved some hassle.

But as it turns out, I prefer Akko’s Jelly Pink switches in this one over the Jelly Purple. The feel is a little smoother and the keys are a little bit quieter. With o-rings it should be fine.

Like I said, the wired connection will work with Teletype just fine, and all connection types work great with the PC. As long as I occasionally plug the USB cable into the PC to keep the keyboard charged, it should be fine.

Typing on it does take a little getting used to, after years with the old one. The keys are the same standard size, but the old one had the Fn key to the left of the spacebar while this one has it on the right — and somehow that makes me want to line my hands up differently on the home row, and makes it feel like Enter and Backspace are a bigger reach for my pinky (they aren’t). I expect I’ll get used to it quickly though.

I read, and will heartily recommend, Light From Uncommon Stars. I don’t want to spoil the plot, but… it reminded me of Steven Universe for a couple of reasons, though the cover blurb calls it “Good Omens meets A Long Way To A Small Angry Planet.” I’m not sure about that description. It does somewhat fit in “cozy SF” territory, while at the same time getting uncomfortably real with the traumatic stuff the main character deals with.

It’s beautiful though, in its take on musical performance and on food. I cried, I sighed, I grinned. I had borrowed the ebook from the library, but I’m going to buy it because I want to read it again and I want to support the author.

---

I’ve been using a Logitech K780 keyboard at home for several years. It’s compact, wireless and can switch between two Bluetooth devices and one using their (proprietary but ironically named) Unifying wireless dongle. That way, a single keyboard can control my PC and my Monome Teletype, saving a lot of desk space.

But I’ve never loved the layout or the feel of that keyboard. With the letters starting to wear off, and increasing frustration with accidental switching between devices, and several other options available, it was time to look elsewhere.

I settled on an Akko mechanical keyboard. Even narrowing it down to those that support multiple devices, there were many options. I chose the 3084B Plus for its relatively complete but compact layout, and because someone was selling an open box unit. Turns out its 2.4G wireless dongle (needed for Teletype) was missing, so I returned it and ordered a new one.

The feel was pretty great. The sound of keys bottoming out, and also springing back, was fairly loud though — so on the new one I decided to go for linear “Jelly Pink” switches rather than the tactile “Jelly Purple” switches. I’ve also ordered some o-rings to soften the bottom-out impact, like I’ve done to the G710+ at work that has Kailh Brown switches. Hopefully that’ll make a difference. I want a nice feel, but I’m not looking for “clack” or “thock” like a lot of mechanical keyboard (aka “keeb”, ugh) enthusiasts. There’s a lot of lubing, taping, modding etc. one can get into, as well as replacing switches and keycaps and such, and there are people who spend thousands building and modding multiple mechanical keyboards. It makes me feel better about my main hobby!

---

I had the odd notion to start every track on this next album with a “choir” sort of sound. In the first track it’s Waldorf Streichfett, in the second there are multiple instances of Plogue Chipspeech, but also Akemie’s Castle sounding very choir-ish due to its treatment. I was planning on Mellotron and some Kontakt samples as well. (For the odd chain of history behind an iconic Mellotron choir sound, watch this. BRB, listening to Kraftwerk now.)

The second track had a sort of psuedo-arpeggio with 0-Ctrl that just didn’t sit right with the rest of things. To rescue it I wound up sandblasting the whole mix in that section with super heavy reverb and adding a couple of Akemie’s Castle parts over it, and I’m pleased at how well that worked.

That thing about the next album being comforting? I recorded a first track last night and it turned out spooky and tense. But nice.

My spouse was listening to a Dracula Daily podcast. The story begins on May 3, but on May 4 (*) Jonathan Harker is warned about it being St. George’s Eve and spooky stuff goes down at midnight. And, well, this recording sounds like spooky stuff is about to go down.

(There’s some conflicts about Harker’s Gregorian vs. local Julian calendar having an 11 day difference, which Stoker seems to have simply ignored, and the year is non-specific in the text. There are good arguments for it being 1893, other arguments for it being 1897 when the book was published, and another for the late 1880s. This only matters to me because I was going to use the date for a title.)

---

Superbooth, by far the world’s biggest and most important modular synth convention for the past few years, is next week and the product teasers and pre-announcements have been flying. Of interest to me:

• Make Noise is up to something and have been teasing it heavily. There’s a lot of speculation about what it could be, and I’ve personally picked several likely possibilities — nowhere have they specified it’s just one new product they’re announcing. Some intriguing sounds for sure, and they have my attention.
• Xaoc Devices has previously said they’ve got 6 modules to be released this year, and teased the blank backs of panels. They make some cool stuff and are one of my favorite brands, so the chances that something will be of interest are pretty decent.
• Noise Engineering, another favorite brand, vaguely teased something new in their “meet us at Superbooth” post. But before that, they showed a photo of a circuit board about the size of a small desktop unit, or else it’s a quite large module. Color me curious…
• A few other scattered things, and no doubt a couple of items that haven’t been hinted at yet.

Obviously, the modular case is full, and there’s not really a good place for another desktop synth; we’re back on the “anything new has to replace something old” thing. So far, though I hear nice stuff in the Make Noise videos, a lot of the best of it sounds like a cleaner Akemie’s Castle. I just need to remind myself that however cool the new stuff is, my “old” stuff is absolutely great. Maybe I can find inspiration in the new stuff without actually buying anything else. Or maybe it’ll be time to make changes. Either way… that’s fine.

I signed up for Knobcon this coming September. I haven’t been since 2019 and I figure it’s about time.

Registration is still the same price, which is kind of amazing to me. But the banquet/keynote is a whopping $75 this time. Last time I was unimpressed with the (buffet-style) food, the wait for the food, the relatively good stuff running out quickly, got much more out of Dr. John Chowning’s separate presentation on FM much more than his keynote speech, and missed an invite to hang out elsewhere with people who’d probably have been more fun. So, I’ll skip that part.

There was no signup for DIY woskhops. I was kind of hoping to get in on one this time, something with basic soldering and assembly. But either they’re just not doing those anymore, or that registration will be first-come-first-served at the con itself.

There’s still the Friday night performances (and Saturday too if I’m not too tired, which usually I am), ton of exhibitors, and some interesting talks.

---

My second Stephen Baxter novel has been Manifold: Time, and I have mixed feelings about it. At the time it was written, Termination Shock had not been, Elon Skum was not yet a household name, Virgin Galactic wasn’t a thing yet, and the Space Shuttle program was still active. So the idea of a reckless libertarian billionaire space cadet who wants to save the world and get all the profit/credit from it, was probably a novel thing and not so tiresome as it is now.

The science is weird, to say the least — really pretty neat for a while, but then it gets to a point where it requires suspending a lot of disbelief, and then it sprints right past the breaking point into utterly absurd “oh, come on” territory. And the way our sympathies are bounced around is pretty fast and loose too.

There were some moments where I thought, whoa, this is a fun and neat book, and I’m afraid those moments have passed. There is still a clever twist that I thought of and am waiting for, but I’m afraid the book just didn’t live up to my hopes. I’ll probably skip the rest of this particular series and try a different one.

I have been poking at Akemie’s Castle a bit, patching it up and mostly just chilling and enjoying the sounds, but I don’t feel like I have a lot to write about.

As far as I’m concerned there are no “tricks” to using Akemie’s Castle, except:

• to me, it sounds best when osc A and osc B are working together as mid/side channels, for big thick unison, chords or clusters. (But it’s not bad to pretend it’s two separate VCOs, either — nor to compromise in the middle with different rhythmic pulses, call/response, etc.)
• Izotope RX Declick works wonders on the inevitably steppy transitions in operator level when you modulate them.
• Modulate a lot of stuff. LFOs, envelopes, sequences, whatever. It sounds best in motion IMHO.

I was patching it last night and found I had to go ahead and record it to layer in with something else later. Which I guess breaks me out of the “study” thing and puts me back into music-making mode.

And I think I need it. I’ve noticed some depression creeping in — not super crushing stuff but I’ve been feeling unmotivated and generally down. Feeling a bit overwhelmed by stuff but plodding through it. Wanting to eat more, not from actual hunger. Being disappointed that caffeine and energy drinks and lots of hydration aren’t perking me up more. Wanting to try new, or re-enjoy old video games and not quite finding what I want from it. Not sleeping all that well, and wanting to blame the cat’s antics but knowing I could sleep through / not be bothered by some of his usual bullshit if it weren’t some other factor. Weird-ass dreams, maybe related? It’s all quite possibly seasonal, with longer days and more sun and sometimes warmer weather (temperature has been all over the place). But I know music can bring me to a happier place.

This morning I’ve already listened to two Cocteau Twins albums and have started in on my own Float. I may have to put Vivaldi’s Four Seasons concerto in the rotation too. It’s time for comforting the disturbed, not disturbing the comfortable. I expect that’s the direction I will be taking my next album.

Maybe I’ll sneak some mini-studies in there as I go? We’ll see. My thoughts about gear right now are that I have a lot of it, it’s awesome and a little overwhelming. I could very easily make an album with just Shapeshifter and a few support modules and FX, likewise with Castle, or Kult, or Aalto, or the BrutePest setup. But I feel oddly like that would be betraying my other gear, which probably isn’t a great way to think of things.

I no longer care about Serge GTO or Klavis Grainity or Rossum Morpheus, or any oscillators that I don’t already have. But Blukac Endless Processor is intriguing again. Too bad the “modules in reserve” thing I tried before didn’t work well for me. I still do have some space to spare in the Pod 60 if I can work out some kind of arrangement that makes sense. The cycle never ends…

After Synchrodyne and Shapeshifter, I had planned to dig into Akemie’s Castle next. My issue with it has always been:

• It’s big. Not too big for what it does, but big enough that I could fit plenty of other options in there. (Or put another way: FOMO.)
• It’s very much an FM specialist, and I have lots of other hardware and software that can do FM extremely well.

Every time I patch it, I love what I hear, and I keep it around. I even made a whole album to prove it to myself… but it seems I have to keep proving it.

If I just compare 2-op FM on it vs. the FM in other modules or in software, it’s a little quirky but not impressive:

• Most notably, the amplitude for each operator is quantized to a few levels so modulating it is steppy rather than smoothed — something I can imitate in other modules if I choose to.
• There’s a certain fizzy/dusty quality to it which is probably mostly due to the resolution of the phase accumulator, but possibly also the sample rate, resolution of the sine lookup table, etc. To a lesser extent that’s also audible in other modules, except Shapeshifter since it’s such a high-specced beast in that department. I can hear it in the SynPrezFM app for Android too, and it’s more obvious at lower pitches.
• The maximum depth of modulation is pretty darned strong.
• The frequency ratios are locked to multiples of a common frequency, so you can’t do the slight detuning that I like, nor super-clangorous stuff, or separate sequencing of carrier and modulator. Which is fine, because it allows easy sequencing/modulation of the multiplier, and I have plenty of other opportunities elsewhere for freely tuned ratios.

But Akemie’s Castle has 4 operators, not 2. That’s much less common in Eurorack, though I suppose I could patch it with Just Friends and Shapeshifter in a pinch. It’s considerably easier in software, and I feel like maybe its closest competitor in my setup is Bitwig FM4, or a Grid patch. (Plogue OPS7 is a lovely DX7 emulation, but you can’t just hold a droning note and modulate the operator amplitudes, which is a shame IMHO.) So I guess that’s what I need to compare it to.

[UPDATE]: no, FM4 sounds nothing at all like it.

AC’s operator levels are never exactly 0, and that’s more obvious when using the 4-op serial algorithm. How you set the multipliers and waves matters, and that not-quite-subtle influence coloring the output honestly just adds to the module’s charm.

AC is also duophonic — the A and B outputs can have separate pitches but share a common set of operator controls. B also has a chord mode. There are all kinds of ways to combine the two outputs, especially with the more “parallel” algorithms and different multiplier settings, for really thick chords and clusters, super nice stereo or mid-side sound, or using them with different envelopes/filters/etc.

I’m not sure if I’m going to write up Akemie’s Castle any beyond what’s in this post. Its complexity is more of an intuitive thing than Synchrodyne’s eccentricities or Shapeshifter’s many forms. But I guess I’ll do a quick AC vs. FM4 test, even though I expect I’m going to want to keep AC.

---

This phase of studying one module at a time in depth is about a month old now, and honestly it feels kind of weird. I’ll think about another module or plugin or the BrutePest setup and catch myself wondering why I haven’t used it recently.

I also feel a little overwhelmed, like I have too much stuff and it’s too powerful and I’m staring into the abyss. I think that’s anxiety talking. To compare synth gear is to compare different infinite sets, and how they overlap both in a general sense and in the specifics of their possible sounds. In the end you just have to pick something that appeals to you and run with it.

And I also wonder how this kind of study is going to apply to actual music-making. I went through the process with Beads, wrote up a bunch of patch ideas and then kept to my old, comfortable, limited habits with it anyway. Going back over those notes and revisiting those patches to see if that shakes something loose is on my list.

Honestly though, what matters is that the gear serves my musical needs, not that I’m taking full advantage of all the possibilities each piece of gear can offer me. Beads’ usefulness to me was never in doubt. Nor was Shapeshifter’s, but I thought its indispensability might be. Synchrodyne’s usefulness was in doubt, but now that I understand its usage and character better, it’s now more a question of will I than can I.

I kind of want to go back to music making again, but I feel there’s plenty more I could learn about my hardware, and also several plugins. I don’t know if I want to plow forward with the studying or return to it at another time.

To be honest I haven’t gotten in a lot of bass practice, either. I kind of feel like, not recording means not feeling the limits of my skill so much, therefore I’m not as motivated to practice. Maybe I need to dedicate some time specifically to bass practice, not try to split it between module studies and bass practice.

---

My parents got me a Rokr 3D wooden puzzle for Christmas — a steam locomotive model, with a wind-up spring and lots of wooden gears. It’s all laser-cut plywood (plus the metal spring coil and a few rods) and IKEA-like assembly instructions. Everything fits together without glue or additional tools. I finally started working on it over the weekend and am most of the way done. Building it has been enjoyable minus a couple of slightly frustrating moments — the difficulty is rated 5/6 after all. I don’t know what I’m going to do with a somewhat delicate wind-up model train when it’s done, though.

I see they also make marble runs, music boxes, an orrery, clocks, a gramophone, etc. I may have to keep those in mind when considering what I can put on my wish list.

That cool SSI 2130 oscillator chip that many different synth builders have been using lately lists thru-zero linear FM and phase modulation among its features.

Linear FM changes the core frequency up and down equally by some number hertz per volt. So if the modulation signal has a +/-1V range a 200 Hz carrier might swing from 100 to 300 Hz, which averages out to 200. At +/-2V, it will swing from 0 to 400 Hz, which is still in tune, no problem! But at +/-3V, it’ll try to go from -100Hz to 500Hz.

If the carrier oscillator doesn’t support thru-zero, it’s going to have a minimum rate of 0Hz. So the average will be 250Hz, which means it goes sharp. The FM sidebands are “wrong” too.

If it does support thru-zero, “-100 Hz” is just 100 Hz with a reversed phase — so it starts going backwards at 100Hz, and the average frequency works out to 400Hz and it sounds right. The SSI2130 does this using a simple external circuit that feeds a “Time Reverse” input.

But their literature also claims it does phase modulation… and then shows simply using the TZFM with a DC-blocking capacitor (aka a highpass filter) in front. That’s cheating.

*If* you have a sine wave for the modulation signal, and it’s above the cutoff frequency, you’ll get the same results with TZFM and PM. (Except that PM is less strong than TZFM given an equivalent amplitude.)

But there are uses for phase modulation at less than audio rates. For instance, controlling the relative phase of two different oscillators, because you’re going to mix them in some way. Or simply using a slow LFO for gentle modulation.

And if your modulation signal isn’t a sine, you get serious differences.

This is Shapeshifter, with the carrier as a sine wave and the modulator at a 2:1 ratio. First we hear a sine wave as the modulator, using TZFM and then PM. Very similar (aside from the PM having to go through conversion at a lower rate and resolution than the internal workings of the module).

Then at :09 I switch the modulator to a triangle and again, demonstrate TZFM and then PM. Very different sounds there, which is the math at work.

At :18 I switch the modulator to a square and repeat. Notice TZFM with a square sounds a lot like PM with a triangle? A steady linear increase in phase is the same as a static increase in freqeuncy, and a square wave is just flipping up and down between two levels, so this totally makes sense.

And then at :26 the modulator is a sawtooth. It’s pretty awesome sounding with TZFM and nasty with PM.

I realize that doesn’t argue that Sound Semiconductor is cheating people out of something great — the TZFM just sounds better anyway. But I don’t think they should have called it phase modulation.