That’s just the place where he demonstrates the role-reversal where pitch-control in TZLFM is concerned. I highly recommend watching the entire video, where he demonstrates the different types of FM in depth, including how pitch-control works. It’s the most educational demonstration I’ve ever seen and he explains and demonstrates it far better than I ever could: https://www.youtube.com/watch?v=kkAagua3NN4
Interesting to note as well, that for pure TZLFM you need complete DC-blocking/AC-coupling in place, otherwise it won’t work as expected (just another stab/advert for DC-blocking everywhere at audio rate).
The bottom line for me, where TZLFM is concerned is this:
For all manner of sound effects, percussion, R2D2 sounds and the likes, traditional exponential and linear FM is great.
For musical, tonal playing, with clean pitch, dialing in all those musical timbres, TZLFM is by far the best, and that’s the selling point of Neoni.
I have to check out what is going on there, because this is not how I understood TZ-FM, and how it is implemented elsewhere. In Reaktor Blocks Bento Box VCO for example it works how I would have expected it, and I thought Bogaudio VCO also does TZ-FM how I expected it.
I still have to understand this phenomenon, but this means everything is fine with the module, thought it might be a bug.
I think much of the confusion stems from phase modulation being called FM by the industry. Most TZ-FM (at least in the digital world) is actually phase modulation. I know for sure that the Bogaudio VCO uses phase modulation for linear FM, not true frequency modulation.
I’m with you. I haven’t watched the Instruo video but as I understand it, TZ linear FM is just linear FM that goes TZ… so in the part of the range before it approaches zero (ie relatively small amounts of FM), it should be no different from regular linear FM - I don’t see why the pitch control should swap from carrier to modulator. Might this be being confused with Sync? If you are FMing and you sync the carrier to modulator then pitch control does indeed move to the modulator.
Omri made a good video about how TZFM works.
and this (hardware) video by Nektarios talks about FMing with sync and how the pitch control gets transferred to the modulator.
yes, although I will say in my hacking around I have seen the FM (as opposed to PM) is so sensitive to DC that even a tinny amount will make it go out of tune. So probably the most practical advice it that all FM (vs. PM) VCOs should have a dc block (option?) on their FM input.
I’ve gotten maybe one VCO vendor to get rid of the unintentional DC on the output of their pulse waves, but DC is as you know distressingly common on the outputs of many modules.
OK, so, I watched (big parts of) the video… it took me a bit but I think I figured it out now…
It is just doing TZFM, and it’s doing it the same way as for example Bogaudio VCO, the difference here is that neóni switches to a very low base frequency when switched to TZFM and also has the FM amount coupled to it’s pitch. So it is making sure that the FM always actually goes through zero, where it sounds like a very smooth sync.
So with TZFM the perceived pitch is that of the carrier if it doesn’t actually go through zero (because then it’s just regular FM) and that of the modulator if it actually goes through zero. But neóni is tuned in a way that it is practically impossible to dial in a small enough FM amount at a high enough base frequency to avoid going through zero.
Or not… When I look at it on the scope and try to reach those extremes (with a very high pitch on the V/Oct input and heavy attenuation on the FM input) it seems to be syncing… I’m confused again… does it sync? (with the sync switch in off-position)
Aha, this cleared things up for me. I’ve got a patch going where I set up groups of four TZ/LFM oscillators and sort of stress test them with one of my morphing algorithm modules. I wasn’t having any success taming the Neóni quartet in this patch despite much effort in raising the base frequency in TZ mode and getting the frequencies set to ratios (has anyone figured out the scale/units for the frequency knob?). The result is always unstable pitch in this context.
Now I see that I can replicate Neóni’s behavior in Terrorform: just press Terrorform’s ‘True FM’ button.
Re: those submarine modules in the corner, I’m planning to code up a small module that takes a V/Oct input, has an Octave knob and a Ratio knob, and applies the octave shift and ratio math to a V/Oct output. The Submarine AO-106 does the job for now.
This module would make it easier to use oscillators like WT-VCO, Terrorform, and Neóni in more traditional FM contexts, seeing as they each lack a Ratio knob. With Terrorform and especially WT-VCO in Rack v2 it’s not so much a problem though, since you can just right-click their frequency knobs and type in some math. The scaling on Neóni’s freq knob prevents this from working there.
In the video Jason Lim talks about through zero is actually soft-syncing the VCO-core.
It looks a lot like a windowed sync on the scope.
One thing I cannot recommend enough with Instruo Modules is to set the internal sample rate to at least 96kHz, that completely changes the sound and behaviour of the oscillators and filters.
I think it mostly makes sense for me, without working through the calculus. But I still want to make sure my understanding is correct:
All phase modulation “FM” is inherently through zero, correct? That is because 360 degrees out of phase is equivalent to 0 degrees, and then the cycle repeats through to 720, etc. And the same is obviously true in the negative direction.
Simple linear frequency modulation goes out of tune once the modulation amount exceeds the carrier frequency because there are no negative frequencies. The next part is I think the “aha” moment for me, but I want to verify. Through zero FM reflects the negative frequency around zero, back into the positive realm. The carrier frequency doubled is considered as a Nyquist frequency, and modulated frequencies above this point are also reflected. Thus the modulation remains symmetric, and the pitch stays in tune. Until today, I didn’t realize the positive modulations are also reflected. Do I have this all correct?
I shall refrain from demonstrating where I disagree
Yes, agree, and Neoni has that DC/AC switch on its input, which is really great thinking and I can’t recall seing another hardware module having that. It’s also fascinating to watch Jason demonstrate, that once you get yourself into totally clean TZLFM territory, the carrier basically works as a kind of wavefolder/shaper.
Nope, it doesn’t switch to anything and does not have FM amount and pitch coupled He is simply demonstrating, that if you have DC-blocking on FM input, and you implement true TZLFM, then this is what naturally happens. It’s kind of a physics lesson, no magic or switching involved.
Yes. But not just percieved, the actual pitch. Watch his spectrum analyzer when he does it, there’s a good reason he put it on the screen
I’m having trouble parsing that and I don’t see why that would be the case.
Nope, no (explicit) sync involved But sync is a good crutch on “normal”, non-TZ linear FM, to get into the same territory.
Good tip, I’ll have to try that!
Two things go into stabilizing it: 1) Being actually, truly, through-zero. 2) Putting a DC-blocker (=AC-coupling) on the FM input.
From my understanding of the video traditional FM is modulating the set frequency in both directions (up and down) while TZFM is modulating in a kind of rectified way, where the abs(voltage) is always the maximum and set frequency of the OSC. So the freqency never goes above the set frequency. And whenever the modulating signal crosses 0v, it flips the phase of the OSC and that is creating the sync.
(just another stab/advert for DC-blocking everywhere at audio rate).
When I look at it on the scope and try to reach those extremes (with a very high pitch on the V/Oct input and heavy attenuation on the FM input) it seems to be syncing… I’m confused again… does it sync? (with the sync switch in off-position)