DrSynth

Jalene, Check your PM Manny

Next, it appears that the Oscillators are all phase shifted relative to one another. Oscillators 1 & 2 are a quarter cycle (90 degrees) out of phase from each other. Oscillator 3 appears to be an eighth cycle (45 degrees) out of phase from 1 & 2. This based on FMing Square wave Osc 1 with Saw wave Osc 2 resulting in standard pulse width modulation depth with Index. FMing either Osc 1 or Osc 2 Square waves with Osc 3 Saw wave results in a variant of pulse width modulation that gives identical results for both Osc 1 & 2, hence the phase differential must be the same betwen Osc 3 & 1and for Osc 3 & 2. Thus Osc 3 would be at 45 degrees as that's the same differential to Osc 1 (0 degrees) and Osc 2 (90 degrees). This was likely done to 'analogue up' the sound of the VA engine, as a phase 'base point' in concert with the "Analog Feel" parameter which by my ear introduces random amounts of detuning, phase shift and key tracking acoss the oscillators. So, no way the Hydra will ever sound like a Yamaha DX except for the simple dual 2-Op pair 'algorithm' That's a good thing... though for the challenge I did finally manage to wrestle some classic DX Brass timbres out of it creating some "intra Mutant" feedback loop structures Manny Upon further review, the aforementioned phase description is likely inaccurate. The actual Osc's may not be phase shifted relative to one another. I assumed it was an offset amongst the Osc's, as offset phase, detuning and keytracking is a common technique used in VA oscillators to create 'analog feel'. The FM Saw->Square creating PWM behavior may be a result of a phase shift introduced via the Mutant process. I'm trying to confirm with ASM devs... regardless of precisely 'how' it's a fun thing to play with ! Manny

Exactly. If serious FM is what you're after, you'd probably feel somewhat hamstrung by the lack of configurable carriers/modulators anyway, even with the variable waveforms... dB There were two reasons I got the Hydrasynth - Poly aftertouch, and the ability to use wavetables as FM Operators (Actually, price didn't hurt...). Using all the available Mod sources, esp. the tempo syncable LFO's and Envelopes to modulate those wavetables within the FM paradigm is very unique and very, very cool. There's a LOT to be found in there... Manny

Thanks for that post, Manny. You're absolutely right -- it is its own thing. I like this "own thing" a lot, actually. Most fun I've had on a synth since the Wavestation first came out. Of course, I haven't been buying synths much over the past 20 years. Shameful, I know, but I still love my older gear and ... well yeah. Good to see that someone else had already considered graphing out the stuff and then realized it was too different as well. I thought it was just me who couldn't ken it. Took me years to get my head around Yamaha's FM synthesis... and that was on a simple 4-op Yamaha V50, which wasn't even mine. LOL. Turned out, it was a lot easier when someone finally explained it in a way that made sense to me -- thank you, Internet! I'm pretty good with manuals; I can figure out most things on my own, but FM threw me for a massive wobbly because like most, I started with subtractive synths. As for the Hydras mutants and how they're processed, that's well out of my depth of experience and knowledge. I'm curious, though, so if you do find out, please share it. I'm not even going to try to assume how it works. So, some deep diving has revealed a few things to explain the Hydrasynth's FM Mutant behavior. First thing was to confirm whether it uses phase modulation or actual frequency modulation. Based on FMing Osc 1 with a sub audio rate Osc 2, since Sine waveform for Osc 2 gives vibrato, but Square wave give a 'pulse train' of clicks and not a trill, it is a phase modulation implementation. I assumed this as all the other Mutants are phase modulation tricks. Next, it appears that the Oscillators are all phase shifted relative to one another. Oscillators 1 & 2 are a quarter cycle (90 degrees) out of phase from each other. Oscillator 3 appears to be an eighth cycle (45 degrees) out of phase from 1 & 2. This based on FMing Square wave Osc 1 with Saw wave Osc 2 resulting in standard pulse width modulation depth with Index. FMing either Osc 1 or Osc 2 Square waves with Osc 3 Saw wave results in a variant of pulse width modulation that gives identical results for both Osc 1 & 2, hence the phase differential must be the same betwen Osc 3 & 1and for Osc 3 & 2. Thus Osc 3 would be at 45 degrees as that's the same differential to Osc 1 (0 degrees) and Osc 2 (90 degrees). This was likely done to 'analogue up' the sound of the VA engine, as a phase 'base point' in concert with the "Analog Feel" parameter which by my ear introduces random amounts of detuning, phase shift and key tracking acoss the oscillators. So, no way the Hydra will ever sound like a Yamaha DX except for the simple dual 2-Op pair 'algorithm' That's a good thing... though for the challenge I did finally manage to wrestle some classic DX Brass timbres out of it creating some "intra Mutant" feedback loop structures Manny

I spent a lot of time exploring how FM is implemented, and I started to map out the DX style 'algorithm' configuration possibilities, but realized the FM implementation of Mutant cross modulations is far too different from what Yamaha does with their Operator calculations within Algorithms -- and you need the Mutants to start building 3 & 4 Operator Stacks. In the end, if the goal is translate DX timbres, you're essentially limited to two basic 2 Operator Pairs, and can add in a second branched Modulator to either or both pairs, though it's not clear when you do that if the additional Modulator is interacting with the prior Mutant Oscillators' "Carrier" only or the total calculated product of that prior Mutant Oscillator. The feedback implementation is completely different as well, and when added to the lack of detailed keyboard output scaling, along with antialiasing filters that are IMHO far to broad (strong) in their implementation killing the high harmonics in the upper note ranges far too quickly, the Hydrasynth's strength is not in the DX's timbral direction. The Hydra is it's own thing, with an interesting unique & cool twist on FM. Manny

So, I've had this little beauty for about a month, _really_ digging it. The user interface works really well for me navigation wise, quick to get where you need to go and set up the mod matrix. I wish there was some 'ultrafine' resolution of the encoders as sometimes it is fussy going too high or too low in trying to get a specific value which on some of the Mutants a difference of 0.6 in a value can be significant. Soundwise my inital impression playing the presets was that it was a little more 'digital' sounding than other VA's with which I'm familiar (the u-he stuff mainly). But after diving in found it capable of being more lush/rich without the 'edge'. That said, there is an 'edgy' timbre space that it can fall into very quickly due to the nature of the FM, Sync & various PW Mod Mutants. And IMHO too many of the presets are in that techno-y timbre space of edgy/harsh analog/digital hybrid sounds. But using things in moderation, and it become quite the chameleon across a wide timbre palatte. Poly AT is the bomb, and I'm finding I use the Ribbon as a mod source in 'hold' mode quite often. The Mod Matrix is very complete, and all that adds on to the very capable realtime control from the Macro Knobs. It can do some pretty impressive sound morphing, and I've experimented with controlling the Hydrasynth from a Yamaha VP1 and the Hydra does extremely well keeping pace with the VP (and is a TON easier to program!) It's really cool using a Wavetable as an FM Modulator and using the Poly AT to scan the table as it varies the Index The worst feature - the aforemention short cable on the wall wart. The (subtle) best feature - TWO headphone jack sizes _on the front_!! A few things for the wish list -- 'chaining' the envelopes together; ability to modulate the dedicated Vibrato LFO (and have have the same value settings as the other LFO's), negative keyboard pitch tracking for the oscillators, and a user configurable modulation source or two to set up things like your own keyscaling 'curves' or non-linear scalars for other parameters. Hooking this puppy up in MPE mode when I get my Osmose (hopefully by the end of the year... ) should be quite the experience. Manny P.S. -- Mike, great job on the Videos, and at some point I hope to post some sound examples

Found this nice suprise on the porch when I got home from work ... it may be a late night Manny

Great job on the videos, Dr. Mike! They finally got back in stock last wek ago at Sweetwater, I can't wait for mine to arrive shortly... It surely will keep me busy until the Osmose arrives (hopefully by the holiday season). Manny

As long as you've revived the zombie, this has been a fave and the Velveeta level 80's cheese video only makes it better... [video:youtube] Manny

Eric, The VP1 is quite the quirky animal. It's been very interesting going back in and programming it (it's been like 27 years since I last voiced a proto in Hamamatsu). It never got the 'love' the VL1 and it's offspring did, and after its (extremely) limited release there was minimal to zero further explorations of it's capabilities. I've only been able to verify 3 of whatever # units sold had the interface to allow programming. Dave Polich, who did the US contributions for the final voicing summed it up well in a recent conversaion we had -- when given the unit to do the presets, the engineers said to him "... this is the VP1, we do not know what it can do" I had a chance to look at a lot of the preset data, and from what I see there seems to be a single paradigm for voice design. That wouldn't be unusal given the constraints of getting X number of sounds in a short time frame for the factory presets. Having a chance to dive back in again, I see at least 2 other paradigms for the voice design that are basically unexplored. There's a lot of untapped potential here... The mystery with generalized physical modeling it just what that potential actually sounds like, or in what contexts it would be useful ! Manny

UPDATE - Took longer than expected given the current situation, but a little over a week ago I got the full editing rig (Mac OS9) up and operational with Reinhold's VP1. Was able to do a full Factory Data Restore, and took the opportunity to record all 10 of the original on-board demo songs. If you're interested, you can check out the playlist on YouTube: Manny

Watched this the other day, well done, informative and very enjoyable! First synth was a blackface Rev 2 Odyssey. Always felt it was a more capable synthesizer than the Minimoog. Had the opportunity to babysit a Chroma for a number of years, regret not buying it when I had the chance. Looking forward to getting my "Ballad of Don Lewis" DVD in a few weeks Manny

Instrument wise, picked up an AN1X a little over a month ago. Instrument related wise, lots of bits and bobs to rehab older gear with new diplays, USB drive replacements and that sort of stuff given all this stay of stay at home time. The significant addition that recently arrived was this 90's vintage Mac serial to PC parallel converter (NIB on eBay, go figure...) to complete a synth patch programming rig: Manny

The 32 presets it ships with are no doubt geared to the retro 80's DX vibe. And for many the mini keys are a "tolerate them / hate them" kind of thing, but Yamaha's are by far the most musically playable. Contrary to others' comments here, it's unique 4-Op FM implementation is actually stealthily versatile. It only sounds as "dated" as you want it to be. If you program yourself you'll be pleasantly suprised at its vast timbral range -- it is no DX9! If you don't program, Yamaha's SoundMondo internet ecosystem has tons of patches available. There's like 400+ patches I've uploaded there, of which the vast majority do not have anything in common with the sterotypical "DX sound", plus there's a lot of others' 'contemporary' patches as well. FM isn't just basses and bells and THAT 'electric piano' sound anymore. But, then again I'm definitely the outlier here re: FM synthesis... Ultimately, what need are you filling with a new piece of gear? For about $320 these days it's actually a very versatile, good "bang for the buck" hardware synth. Manny

Absolutely the right decision, even at your age do not downplay your risk factors, esp the lung/fungal history. In the course of calling our patients to cancel future visits, we've already learned that two older gentlemen w/ health issues have passed form COVID19 and it's heartbreaking. And half expecting there may be more... Meeting digitally is the way to go.

So, not sure this is the right place, but it is a continuation of the topic in a way... The last month I've spent considerable time to track down the resources etc. to get Reinhold Heil's VP1 here in Los Angeles back up and fully functional. Completed the requisite battery and floppy drive replacements, and we're currently getting by on the very faded display (replacement _extremely_ tedious if you're familiar with an SY77 or SY99) I've got it almost there -- it's impressive to watch Reinhold play it as he has considerable experience with the instrument. But we're stymied by the inability to get the model editing software functional. Waiting for some vintage serial-> parallel converters to arrive from eBay Wondering if anyone here has any leads or contacts to track down who actually currently owns or has access to a VP1 to aid in this Quixotic quest? We got the Owner's Manual and Performance Guide via Daniel Forro who was a Yamaha Product specialist for Yamaha Europe. Through Facebook I found a gentleman in Japan - Takahashi Yasunari - that had a fully function system including a copy of the Voice Editing Guide for the operation of the editing software, but he's apparently sold it sometime in the last 2-3 years, but he's no longer active on Facebook apparently We've tracked down 4 VP1's -- Reinhold's here in LA, the one at Innovation Road at Yamaha Japan, another at Yamaha Europe (Both with the parallel editing interface) and the fourth with the interface owned by Kurt Ader of KAPro in Germany. My Yamaha contacts are working on it as well, but they don't know of any other VP owners out there. I'm hoping since members here get around someone may know someone who knows someone... Manny

The discussions here mirror those those we had on the VL1 dev team back in the early 90"s. THE most important aspect of a 'modeled instrument illusion" is idiomatic playing. Timbre is actually second, though if you"re not a natural player of the instrument being modeled you will focus on timbre first. Thus, since Yamaha abandoned the VL technology, the shift has been to sample based engines with performance modeling. The prior VST"s that do this with sophisticated sample switching has gotten quite capable with the current state of computer horsepower, even without any actual acoustic modeling going on per se. And since the users of these are not typically the natural instrument players, the timbral accuracy makes you overlook the still missing articulative acousical nuances. This is where the SWAM stuff is interesting, combining acoustic and performance modeling. I have the Strings bundle, their pure modeled engine. It"s advantage over the VL1 is the implementation of performance modeling in processing the note input â whatever algorithms/AI they have going on to enhance the keyboard input articulations and phrasing to the idiom of the 'instrument" is pretty cool. Plus the timbres are noticeably improved as well, likely due to better resonator modeling. Going back and reprogramming some VL1 violins I"ve got a few that are very close to the SWAM violins in timbre, but if you can"t nail the violin phrasing and articulations from the keyboard, SWAM plays better. The SWAM reed and brass are a little different and use samples for the driver portion of the model, so to what degree they interact with the resonator part is unknown as you can"t get under the hood of the model. I assume its a hybrid of typical performance based sample switching with realtime resonator modeling which would give advantages in timbral performance nuances compared to sample performance switching alone. IMHO the difference in market success of the VL1 vs the VL70M was less the $$ but that the VL70M was developed focused on wind controller players. We did new voicing with L.A. orchestra/session guys playing their instruments side by side with them playing the VL70 emulations on WX controllers to improve the behavior of the sounds, and it made a big difference. Thus a lot of the 'illusion" as it were comes down to controllers - a keyboard is a lousy controller for reed, brass and string modeled instruments, even with fancy AI between the input and the model. As good as your chops may be, playing a reed model on a wind controller will sound more accurate than from a keyboard. The Yamaha VL Big Band at the 1997 NAMM show is a prime example of natural brass and reed players in action. Claudio is no slouch as a player, but controlling the SWAM reeds with a WX could take it into the uncanny valley of instrument simulation*. Ultimately, while this is all fun to explore & deconstruct, my preferred application of modeling is creating acoustic instrument expressivity and nuance in new and different timbral spaces. Manny *. except, as those noted above in the thread that it"s 'never good enough' for the natural instrument _you_ play⦠though I have seen that opinion get more grey when we"ve tweaked and personalized the VL models for a particular WX players style.

Well that did it -- deposit paid. Looking forward not to just the EaganMatrix engine but interfacing it with Alchemy, the u-he stuff and Falcon. Manny

Exactly ! Though the specific methods are different one can recreate the WSA-1 type of timbral shaping and control in VAST. BTW - The "Smith" in question re: Yamaha's physical modeling is Julius O. Smith at CCRMA, not Dave Smith of Sequential Manny

The VP1 is quite a strange beast. What it did, it did extremely well and it sounded incredible and unique once you learned how to use all the controllers to interact with the sounds. But that is also what made it probably the most 'niche' synth ever made. If you didn't put the time into learning how to play it, and it's timbral and behaviorial space didn't hit your hot spots, it would come across as a big 'meh.' The 'failure' of the VP1, if you want to use that term, was that there was so few people who understood & needed what it did and could afford it. The rarity adds to it's legend. There's no exact known number produced & sold, it's been said up to 20 were made. There is at least 3 surviving 'in the wild' - one privately owned here in LA; the one in the video at Innovation Road in Japan; and another apparently recently tracked one down by Yamaha Germany for their version of the SynthSpace room. The WSA-1 is a total apples and oranges comparison. It's been a long time, and Reezekeys may remember a lot more, IIRC it was samples processed by waveguide 'resonators' but the implementation of the resonators were very basic and were essentially just fancy waveshapers -- basic modeling a few types of resonator bodies and allowing for some general timbral motion but no real tweaking of the physics. Thus they didn't effect anything about the behaviors of the sound, which is what VL and VP were all about. A reasonable sounding ROMpler with a twist, but a K2000 could do similar things within VAST. Manny

Actually 4-6 protos were built of the V80 IIRC, as we had at least 2 here in the US and 2 in Hamamatsu. We demoed 2 in a private room at the Hilton for 1989 Anaheim NAMM showing to press and dealers. I was really fond of it, as is pushed the limits of what 6 OP sine wave FM was capable. Right call to cancel it as it was a somewhat kludgey stopgap to wait for the SY77 to come online. I'd posted on Soundcloud a few months ago the original demos I did with Scott Plunkett. Best aspect of it was the 'modular' expansion feature that evolved into the later PLG cards, though that was never operational before the plug was pulled. I can confirm they were all crushed. Knowing its fate I conveniently managed to 'forget' to take mine back to Yamaha for quite a while, and it was the last one to go... Manny Hi Manny, Thanks very much for this information! I have a dear friend and colleague in LA who, at the time of the 1989 NAMM Show, had managed to wangle an invite to the private room and absolutely fell in love with the V80FD. He spent years and years desperately trying to find out if one existed so he could buy it. I think eventually he got an FS1R; he thought the SY77 was cheating, as it had samples in it. mike I can understand his 'crush' on it -- I really dug the aesthetic of the V80FD as it was quite unique & retro modern in a kinda IBM Selectric sort of way. The solid 'clunk' of the large lighted buttons made programming it feel like you were programming a VAX or PDP mainframe. I hoped Yamaha would keep the design look for the SY77, but they decided to follow the more familiar V50 design style. I've been trying to find polaroids I took of it before it got crushed but no luck so far... Manny

Actually 4-6 protos were built of the V80 IIRC, as we had at least 2 here in the US and 2 in Hamamatsu. We demoed 2 in a private room at the Hilton for 1989 Anaheim NAMM showing to press and dealers. I was really fond of it, as is pushed the limits of what 6 OP sine wave FM was capable. Right call to cancel it as it was a somewhat kludgey stopgap to wait for the SY77 to come online. I'd posted on Soundcloud a few months ago the original demos I did with Scott Plunkett. Best aspect of it was the 'modular' expansion feature that evolved into the later PLG cards, though that was never operational before the plug was pulled. I can confirm they were all crushed. Knowing its fate I conveniently managed to 'forget' to take mine back to Yamaha for quite a while, and it was the last one to go... Manny

I would die to put my hands on one of those. I used to have a VL1m Version2 which I had to sell (aaaaaaaaaarrrrrrrrgggggggghhhhh) and I heard every demo of the VP1... seems to sound gorgeous. The VP1 was quite quirky compared to the VL1, due the the different model implemented. The VL was a 'continuous' energized system, the VP was an 'impulse' energized system. It inherently created non-sustaining timbres. The manipulation of the model parameters allowed unique behaviors such as realtime control of damping/excitation resonances and inharmonicities within the waveform harmonics. It really had to be played with multiple controller input - just playing it as a regular poly synth it was actually quite boring. Which IMHO was the reason they didn't actually put it into larger production - far too 'niche' for something so expensive. A few years later when processing power dropped significantly in price, Yamaha made an 8 note poly version of the VP1 (the original was 16), packaged in an AN1x / CS1x 5 octave keybaord / standard controller format to make a $1995 pricepoint -- kinda like the VL70m was a trickle down of the VL1. We showed it to various artists / studio players down at Buena Park, and to call the reception 'tepid' would be an overstatement, all because without the controllers of the original, the sound is flat with just velocity and a single mod wheel input. The VL1/VL70m model in contrast needs only three significant inputs - Breath Controller, velocity and note articulation (i.e. legato) to get 90%+ of it's expressivity, which is why the VL70m was so successful at IIRC $895 compared to the $4995 VL1 & $2995 VL7. Plus, they sold a lot of VL PLG boards... Manny

In my experience, features & specs we request are very much put to the number crunching filter of net unit profitability. The actualy math of that number crunching is very dependent on the resources of the specific company, especially as it comes to the CAPEX, as you can start/stop/start production runs based on demand and control the OPEX A large company like Yamaha had enormous resources at their disposal, not just $$, but also ability to design and fab their own components within their various divisions. I don't believe any other company had the ability to bring something like the DX7 to market with what was needed to develop Chowning's research to a mass (musical) market product. During my hardcore time with them from DX7II through the SY's, VL's, CS/AN/EX & FS1r, I would say most of their CAPEX was 'internal' amongst their various divisions and subsidaries, which IMHO allowed them to take risks others couldn't in developing & bringing unique synth technologies to market. When you think about something like the VP1, even at $30K that price point, the hardware inside was only possible as they did their own VLSI's -- think of the power of a 2008 era 2.5 GHz Xeon CPU in 1993 dollars, and it had the equivalent of 8 of them. The EX5, the cool yet glitchy Kronos of the time -- is actually a good example where they missed the mark in specs vs costs as they invested the CAPEX & kept the features but didn't commit enough OPEX resources to the hardware as by then it had to cost-compete in the market ruled by straightforward ROMplers. OPEX will limit things like number of switches, knobs, buttons, display size, etc but not necessarily the overall features or capabilites of modern "synth on a chip" gear, which as you note is dependent on the willingness to invest in the CAPEX. Fortunately, because of the modern powerful / cheaper off the shelf options of chips like general purpose DSP's & FPGA's the CAPEX is in the coders salaries not the chip dev as in my past days with Yamaha. Which is related to why there's so many softsynths out there compared to hardware... Just my 2 cents... Manny

Can't comment on the sounds since I'm at work and can't play the video at the moment. BUT, the original DX7 can certainly have a hard split. Can't remember if it was Gary Luenberger or Bo Tomlin that did the programming, but there were a series of sounds that consisted of real, hard splits, not keyboard scaling, but real splits. How they did this I don't know, these two guys were the masters of the DX7 when it first was introduced. I will comment on the lack of filters and the inability to create filter sweeps. It is possible to create a patch that mimics a filter sweep. Will it be as good as a true analog synth,um...no. But it will get the point across. The DX7 was an incredible machine. It opened up an entire new set of sonic capabilities that didn't exist prior to it's introduction. Did it sound like an analog synth, no. Then again, an analog synth can't possibly create the timbres created through FM. It comes down to this. If authentic analog filter sweeps float your boat, then get an analog. Frankly, I enjoy a good filter sweep every now an again, but it gets very tiring when I check out some new analog synth video and all they seem to show is the opening and closing of a filter while an arpeggiator is playing some seemingly random group of notes. "Sorry, didn't mean to jump ship on the original topic". Enjoy your new DX-7. And if you need help programming, I'm sure there are one or three folks around here that see the value in the instrument and love programming the beast. I did the "hard split" programming hack, and voices using it were sold as part of one of Bo's libraries. From the front panel, the Operator Level range is 0-99. Via MIDI Sysex you could send values from 100-127, and certain values in that range would randomly glitch the OS to bypass the "normal" Output Level and +/- Lin or Exp level keyscaling behavior and access the underlying "output level set for each 3-note key range" tables as in the GS1/GS2. So you could do hard splits by allocating the Operators to the 'high' or 'low' sound. You could store the sound, and change the Split point by the Level scaling "Break Point" value, but if you accidentally even just viewed that Operator's Level on the front panel, the OS would correct the 100-127 value to 99 and the hard split would be lost. So, Operator "level" had to be changed by altering all of it's Envelope Levels. Re: Filter sweep emulation -- can be very faithfully done in pure FM, but very dependent on combination of Cutoff & Resonance and the modulation rate you're trying to copy. A DX7 (traditional 6-Op) can do the basics. The 4-Op engines of the DX11/TX81Z and especially the Reface DX are better at it in having the additional non-sine waveforms. AFM in the SY77/99 with the 3 patchable feedback loops and phase control can be setup to be a true virtual analog even without using the filters, and FM-X in the Montage nearly so. Even after all these years the ignorance about the capabilities and timbral palette of FM is still widespread, though programmers like the guy who did the aforementioned video are helping to inform the masses. FM really is the most usably versatile synthesis method out there. Manny