Phonec3 issues (resolved!)

[cstooch]

Edit: ISSUES RESOLVED!!

I'm wondering if anyone else is experiencing these problems.. I've sent this list off to the developer's email:

(Using the fully updated Reaper and Windows 11 25H2)
- I press a note down on my MIDI keyboard, Phonec plays that note.. if I release then press a note (same or other) within a second or two, that second note does not play. I press again right away, and a note plays.
- I press and HOLD a note... it plays... continue to hold, hold down second note, it does not play, hold 3rd note, it plays... hold more notes and Reaper crashes (this seems repeatable). Windows Event Viewer shows phonec3 as the faulty module that crashed.
- Unrelated problem, the arrows for changing presets on the main plugin view do nothing.. but I can change click on the plugin name and browse through presets that way.

I used it just 2 or 3 minutes and found 2 bugs that are pretty significant (although probably related, I'd imagine) and one that should have been noticed... so I'm wondering if it's just my luck here or what. I have hundreds of plugins that are all working great, just this one here is a problem.

[Vortifex]

Tried in Reaper on Windows and am not experiencing any of these issues. However, loading it in FL Studio causes the DAW to crash every time.

[cstooch]

hmm, so this might just be isolated to something on that computer. I'll try a few different things and see if I can get it working, or anything.

I installed it on a different PC of mine.. it's an older Reaper version, but also Win 11 25H2, and I didn't experience the problems I spoke of on that PC.

[MCMorrise]

Running Phonec 3 as AU under macOS Sonoma / Logic Pro 10.7.7, I have noticed the following issues:

If sustain pedal (CC64) is active when notes are released, the notes will continue to sound indefinitely after sustain pedal release. This occurs regardless of whether Phonec 3 receives the note-on message before or after sustain pedal activation.

In the modulation section under the CTRL tab, the boxes for LFO1 depth under both MODWHEEL and AFTERTOUCH have the same problem: they control both LFO1 and LFO3. So, the LFO3 depth boxes currently appear to do nothing.

When using portamento, the filter behaves unexpectedly: the filter cutoff does a chromatic glide rather than a standard portamento from note to note. This is noticeable when key-tracking is enabled and resonance is turned up to self-oscillation. Kind of sounds like a Yamaha CS-80, except it's time-constant rather than rate-constant as the CS-80 is. Still, it's not documented in the manual, so I'm calling it a bug.

(In Phonec 2, when portamento was active, only the oscillators moved. That is, when a note was instantiated, the filter would already be sounding at the target pitch and the oscillators would glide to it. I'm not sure which of these behaviors is preferable; neither are better than nominally normal portamento behavior, which is not occurring. I'd love if we could have either time-constant portamento or rate-constant chromatic glide, but let it occur with osc and filter simultaneously, and let it be well-documented.)

Phonec 3 patch translation from v2 is not fantastic. I made a Phonec tutorial for Groove3 in 2020, so I spent a lot of time rolling my own patches, most of which I kept for personal use. Between mine and the libraries I've purchased from The Unfinished and blortblort, most sounds are very different if not unrecognizable from v2 to v3.

Translated cutoff values are fairly different. Not sure why a filter that's fairly open in v2 suddenly sounds so dark in v3. (Maybe the modifier amounts have changed? I see we now have key scaling up to 200%, so that could be it.) Resonance behavior is more drastically altered. The v3 SVF may be technically better than the v2 biquads, but I'm struggling to replicate the tonality of the v2 filter in v3.

I wrote a few more notes that are less about overt functionality issues and more nit-picking about what's been changed in v3 (a lot of things I was using in my patches got removed!). But I will hold off on those until the primary issues are addressed. [Edit: no I won't. lol]

[MCMorrise]

I was planning to wait until bug fixes came to make a video review, but I think everything I've written is worth posting now, just so Phonec users can know about the changes between the versions. So, for users of Phonec 2 and 3, here is a lengthy, yammering (the kids call it yapping) review of most of the changes in v3.

Before all else: I deeply appreciate Jack and Chris for their commitment to updating a great synth to work on newer systems. This is never an easy task, and Jack has been through the wringer with regards to the health of beloved family pets, etc. We all remember developers who completely ghosted us, but Jack has persisted against all obstacles. That commitment is very honorable.

The v3 aesthetics are great. The new color scheme is enjoyable. I love to see vector GUIs for future-proofing. Nice touch adding a third ENV to the modulation section. Note sequencer addition is good. Interval map is useful, though it is unintuitive as to how the intervals set in the note grid map to the intervals specified by the map. (Does that make sense? I've revised that sentence at least 4x and I remain unsatisfied.) Pitch wheel quantize feature is a pleasant novelty. There are many good things to be said about Phonec 3.

I have a few issues based on the loss or major alteration of features that I used pretty extensively in my personal Phonec 2 programming. They probably don't apply to all or even most Phonec 2 users who have transitioned to v3, but I consider them noteworthy. (Please also see my previous post re: LFO CTRL behavior, patch translation, and filter behavior. Also, note that Phonec 2 banks cannot yet be read by Phonec 3.)

Adelante:

In Phonec 2, LFOs are bipolar. In Phonec 3, LFOs are unipolar. There doesn't appear to be a way to change that. I suspect that, like the filter, LFO polarity is a key player for many sounds made in v2 that do not successfully translate to v3. To Jack's credit, in v3 all LFOs now have keytracking from 0 to 200%. (In Phonec 2, only LFO3 had it, it was only available when not in BPM sync mode, it was a toggle, and it was some arbitrary percentage that did not scale in a sensible way and could even be inverted so LFO3 was faster on lower notes. I kind of miss that jankiness. It would take some mod matrix remapping to recreate that, and Phonec 3 does not have this capability.)

Looks like LFO3 and HFO rates are no longer available as mod wheel / aftertouch destinations. This is a little sad. I did aftertouch > LFO3 rate and depth on a lot of string and pad patches and sometimes reached for aftertouch > HFO rate for FX patches.

In Phonec 2, the HFO had destinations for both OSC pitch and OSC FM and they behaved quite differently. Based on how they sounded, my supposition was analogue-like (let's call it pleasantly chaotic) FM vs. DX-like (phase modulation: harmonic, "musical") FM, respectively. But the v2 manual did not explain the difference between them, so I don't know. In Phonec 3, the HFO has only the more musical but more sterile OSC FM. (Same deal with Main Pitch and Sub Pitch HFO destinations, which were also removed. Their behavior was distinct enough that it was cool to have them.)

On the flipside, I'd say the removal of the more esoteric LFO destinations was a great choice for simplicity's sake. You could make the case that modulating attack and/or decay of amp and/or filter could help simulate uncalibrated voice cards in an analogue polysynth. But I never reached for them. An uncrowded modulation destination list is probably good. (Anyhow, it would be preferable if amp/filter attack and decay were trim pots like on DIVA, but I assume that would require major changes to Phonec's polyphony allocation scheme, so I will not ask for it.)

In Phonec 2, the sequencer had a pitch toggle on each step linked to its own little glide amount, so you could do little 303-like slides on only the notes with the toggle active. In v3, this is gone, and you only have global portamento (as on v2). In practice, this means that when the arp is active and portamento is opened, every arp step has portamento. You can get around this by automating portamento amount, but this quickly became a hassle for me, and I think I'm as patient as the average user.

Curiously, portamento in Phonec 3 does not occur when the NOTE section of the sequencer is active, only when Phonec receives a new MIDI note. This behavior is not very intuitive, and it's not documented in the manual, so I assume it's a bug.

In Phonec 2, we also had separate time bases for the arp and sequencer. This made it possible to do, among other things, polyrhythmic effects by setting the time bases differently on each. I miss that in v3, but ultimately this is a small quibble.

Another technically small quibble, but one that tweaks my knobs: I see OSC drift is gone now. Phonec 3 has no parameter(s) to replicate the subtle phasing / beating interactions attained via drift, as MELT in both Phonec versions changes master pitch only. Drift could be replicated if (1) LFO rate were a modulation target (2) capable of being modulated by a different LFO. Setting LFO1 and 2 to random and gently modulating LFO1 > LFO2 rate and LFO2 > OSC1 pitch would get similar effects to drift. But both LFOs would need slew control to smooth out their modulations, as the SUBTLE toggle isn't enough. (Also, both OSC1 and OSC2 drift.) Either way, it'd be a shame to use 2 of the 3 LFOs to do a task which was previously managed on the backend and adjusted with just one knob.

The late great Echoshifter has been truncated to Echo. This is a major loss, in my opinion. It was fun to have a sort of mini Eventide H910 inside Phonec 2. I used LFO > Echo tune to get a patch vaguely reminiscent of David Bowie's flanger-addled "Ashes to Ashes" piano; it was quite good for a synth trying to sound like a piano trying to sound like a synth.

I concede the loss of separate MELT A/B modes as yet another win for simple design. While they were distinct (MELT B was truly special for getting weird sounds out of the HFO, and it appears to not exist in v3), the new MELT is a sort of distillation of the most useful parameters from both.

A couple new issues I have found with v3 workflow:

When adjusting an OCT or SEMI control via scroll wheel, I expect scrolling upward to raise the value (as when scrolling up on ARP RANGE), but the value goes negative instead. Maybe this is a macOS exclusive issue, and/or maybe I'm just picky.

When using fine control on a knob (with CMD held on macOS, or CTRL on Windows), clicking and dragging with the mouse either produces either no value change or a sudden move. It's a little frustrating.

My suggestions:

If I had my druthers, I would say Phonec 3 needs a mod matrix. The current per-source tabbed setup is hard to work with. Serum 2 is the prime example of modulation done well, with ENVs and LFOs in separate tabs, but all destinations and scaling together on one page. (Also, it would be nice to use OSC1/2/SUB as sources since they're already there, but that might be a lot more work for Jack so I'm not asking for it.) I don't think it's a good idea to offer 64 slots; I'd be content with as few as 4. (After all, one of the nice things about Phonec is its inherent simplicity.) But a consolidated matrix would beat the current workflow.

Also, the sequencer could stand to see some changes. It doesn't need to have as many bells and whistles as the sequencer in Pigments (or Serum 2 or SynthMaster or …), but I think it would be improved if pitch, gate time, and mod amount (and glide, preferably) were each their own lane on the same 16-step track. Again, consolidation is a win for users.

Again, making Phonec compatible with Apple Silicon was a huge workload, it's great that Phonec 3 happened at all, and props to Jack and Chris for all of your work. I am (I hope understandably) concerned at how many of the small details from Phonec 2 — details that made it one of my favorite synths — are now gone, and I hope my critiques prove useful for the improvement of one of my favorite synths.

[cstooch]

V 3.0.1 is out now, just got the email.

I've downloaded and installed and played for only a couple mins.. so far so good (*knock on wood*). I came here though just to post that it sounds AMAZZZZING. The sound design on the factory presets is really quite good and very inspiring.

Well done, I hope the major bugs are fixed now. Will report back if I find anything.

[jack [psychicmodulation]]

I know this is an older thread based on the state of Phonec3 prior to the recent update, but I should address this. And sorry for not being as attentive on this forum, I kind of forget to check it.

Interval map is useful, though it is unintuitive as to how the intervals set in the note grid map to the intervals specified by the map. (Does that make sense? I've revised that sentence at least 4x and I remain unsatisfied.)

I'll have to re-read that a few times to process, but I think maybe what you're saying is that the way the intervals translate into the note grid are hard to follow. I spent quite a bit of time trying to figure out the best method to implement this feature within the Note Sequencer. I think it works well, though may be confusing as to which notes are being corresponded to. I will continue to contemplate this feature and see if there is room for improvement. And I am always open to suggestions.

Pitch wheel quantize feature is a pleasant novelty. There are many good things to be said about Phonec 3.

I think that the pitch wheel in quantize mode utilizing the Interval Map is a functionality that will get completely overlooked but that some will discover to be quite interesting.

In Phonec 2, LFOs are bipolar. In Phonec 3, LFOs are unipolar. There doesn't appear to be a way to change that. I suspect that, like the filter, LFO polarity is a key player for many sounds made in v2 that do not successfully translate to v3.

This has been fixed. It was something I overlooked initially, but it has been resolved. Please note: As was pointed out recently, the Melt modulation is also modulating unipolar and will be fixed in the next update.

Looks like LFO3 and HFO rates are no longer available as mod wheel / aftertouch destinations. This is a little sad. I did aftertouch > LFO3 rate and depth on a lot of string and pad patches and sometimes reached for aftertouch > HFO rate for FX patches.

I'll keep this in mind and see if I can reimplement them in the GUI space available.

In Phonec 2, the HFO had destinations for both OSC pitch and OSC FM and they behaved quite differently. Based on how they sounded, my supposition was analogue-like (let's call it pleasantly chaotic) FM vs. DX-like (phase modulation: harmonic, "musical") FM, respectively. But the v2 manual did not explain the difference between them, so I don't know. In Phonec 3, the HFO has only the more musical but more sterile OSC FM. (Same deal with Main Pitch and Sub Pitch HFO destinations, which were also removed. Their behavior was distinct enough that it was cool to have them.)

This has been remedied in the 3.0.1 update. FM is now working properly as it was in v2, and PM (pitch modulation) has been added to the mod target list.

On the flipside, I'd say the removal of the more esoteric LFO destinations was a great choice for simplicity's sake. You could make the case that modulating attack and/or decay of amp and/or filter could help simulate uncalibrated voice cards in an analogue polysynth. But I never reached for them. An uncrowded modulation destination list is probably good.

Yeah some of those I found to be unnecessary so I left them out. Useful in unconventional situations, but not necessary and possibly confusing.

In Phonec 2, the sequencer had a pitch toggle on each step linked to its own little glide amount, so you could do little 303-like slides on only the notes with the toggle active. In v3, this is gone, and you only have global portamento (as on v2). In practice, this means that when the arp is active and portamento is opened, every arp step has portamento. You can get around this by automating portamento amount, but this quickly became a hassle for me, and I think I'm as patient as the average user.

Not quite. The Glide amount in v2 was universal for Pitch, Filter and Amp in the Sequencer. I will be incorporating Glide into the v3 sequencer, but needed to give it more thought. Not sure if I'm doing a single universal glide again, or two glides (one for Pitch, one for Mod).

Curiously, portamento in Phonec 3 does not occur when the NOTE section of the sequencer is active, only when Phonec receives a new MIDI note. This behavior is not very intuitive, and it's not documented in the manual, so I assume it's a bug.

This is because the Note sequencer operates the same as the Pitch Sequencer in v2. These are note-based pitch offsets which complement the arp.

In Phonec 2, we also had separate time bases for the arp and sequencer. This made it possible to do, among other things, polyrhythmic effects by setting the time bases differently on each. I miss that in v3, but ultimately this is a small quibble.

It's never too late to add a useful feature. I'll look into it.

Another technically small quibble, but one that tweaks my knobs: I see OSC drift is gone now. Phonec 3 has no parameter(s) to replicate the subtle phasing / beating interactions attained via drift, as MELT in both Phonec versions changes master pitch only. Drift could be replicated if (1) LFO rate were a modulation target (2) capable of being modulated by a different LFO. Setting LFO1 and 2 to random and gently modulating LFO1 > LFO2 rate and LFO2 > OSC1 pitch would get similar effects to drift. But both LFOs would need slew control to smooth out their modulations, as the SUBTLE toggle isn't enough. (Also, both OSC1 and OSC2 drift.) Either way, it'd be a shame to use 2 of the 3 LFOs to do a task which was previously managed on the backend and adjusted with just one knob.

This has been mentioned by another user, and yes Drift will make a comeback. I didn't really intend to leave this out, I just kind of forgot about it. We'll get it back soon.

The late great Echoshifter has been truncated to Echo. This is a major loss, in my opinion. It was fun to have a sort of mini Eventide H910 inside Phonec 2. I used LFO > Echo tune to get a patch vaguely reminiscent of David Bowie's flanger-addled "Ashes to Ashes" piano; it was quite good for a synth trying to sound like a piano trying to sound like a synth.

Yes the Echo is quite different and is not the same effect. While I do intend to expand on the Echo in v3, I'll admit it is pared down quite a bit. I ultimately decided to leave out the pitch shifter and venture more in the direction of EchoMelt3. However, time constraints led me to move along and I never really did much with it other than incorporating the basics. I definitely plan to revisit this and expand on it over time. We'll see where it goes.

I concede the loss of separate MELT A/B modes as yet another win for simple design. While they were distinct (MELT B was truly special for getting weird sounds out of the HFO, and it appears to not exist in v3), the new MELT is a sort of distillation of the most useful parameters from both.

Both Filter and HFO can utilize Melt. The option to do so is both on the hidden Melt panel to control the levels, and on the HFO/Filter panels themselves as 'Melt' buttons to turn this on or off. The functionality is there, it's just laid out differently.

When using fine control on a knob (with CMD held on macOS, or CTRL on Windows), clicking and dragging with the mouse either produces either no value change or a sudden move. It's a little frustrating.

I'll look into it. Should be an easy fix.

If I had my druthers, I would say Phonec 3 needs a mod matrix. The current per-source tabbed setup is hard to work with. Serum 2 is the prime example of modulation done well, with ENVs and LFOs in separate tabs, but all destinations and scaling together on one page. (Also, it would be nice to use OSC1/2/SUB as sources since they're already there, but that might be a lot more work for Jack so I'm not asking for it.) I don't think it's a good idea to offer 64 slots; I'd be content with as few as 4. (After all, one of the nice things about Phonec is its inherent simplicity.) But a consolidated matrix would beat the current workflow.

We probably won't see a massive overhaul of the modulation section, but I can consider ways to improve on it.

Also, the sequencer could stand to see some changes. It doesn't need to have as many bells and whistles as the sequencer in Pigments (or Serum 2 or SynthMaster or …), but I think it would be improved if pitch, gate time, and mod amount (and glide, preferably) were each their own lane on the same 16-step track. Again, consolidation is a win for users.

I was mainly separating the two to be used separately without confusing the placement of the Interval Map. My idea was keeping Pitch in it's own lane with the Mods underneath made the most sense.

Again, making Phonec compatible with Apple Silicon was a huge workload, it's great that Phonec 3 happened at all, and props to Jack and Chris for all of your work. I am (I hope understandably) concerned at how many of the small details from Phonec 2 — details that made it one of my favorite synths — are now gone, and I hope my critiques prove useful for the improvement of one of my favorite synths.

All of my plugins typically evolve over time. This happened with Phonec2. Look how many features I packed into it from v2.0 to v2.4. This is my plan for Phonec3, but I want to put more thought into some aspects before I jump into anything. Some things I planned to take into a different direction, but as I learn how much they are missed, I may reconsider. I assume you have tried the v3.0.1 update by now, and I hope it meets your expectations for the most part. Phonec3 is in the very early stages, there is much more to come.

[pekbro]

I am seriously enjoying Phonec3 myself. Also everything else! Great stuff Jack!