After more silence on my end, I do have one major project involving my work I can finally announce:

Burnout has still persisted insofar as music making, sadly. Will do my best to keep everyone in the loop there.

That said, however, I'm happy to report that I've successfully dumped all of the SL-JD80 cards and, even bigger, successfully merged into one SR-JV80 ROM, complete with the PN-JD80 presets for each, formatted for the JD-990. The time saving approach between tests also revealed that removing a few duplicate samples across the first three volumes proved unnecessary for storage capacity: there's still some space left in the card after squeezing in all of the waveform data, and all 512 JD-formatted presets as well (which busts the myth that the SR-JV80 ROM's were limited to 256 per supported model)!

The biggest caveat with this approach, other than the method JD-990 owners are used to of no support for bulk dumping patches from these boards, is no rhythm sets. Unfortunately, while I admit my knowledge of the rhythm set format is limited, all attempts to get my 990 to read the four across these eight cards directly from the board failed, my suspicion being it's due to the limitations of the 990 OS (only patches and multisamples are clearly available when installed, and even for the former, only individually).

This in mind, to save people some effort if they do manage to get it installed, the bottom of this post includes a small archive of SysEx dumps, including for the rhythm sets. These will only sound the way they should with my ROM because of the specified multisample numbers, though by all means, feel free to do what some 990 owners did with the SL multisamples versus presets in the past and mixmatch them with other SR-JV80 cards to see what happens, such as SR-JV80-04, and if you're saavy enough with a hex editor, you can also try tweaking the dumps for SL-JD80-02 sounds onwards to get the right SL-JD format number, provided you have the right waveform card already. As for my ROM itself, out of respect for Roland's copyright for the underlying samples, I have no plans to host that on this website, though if I do somehow get clearance from them in the distant future, that may change for the better later.

That's all for now.

While music had taken a major downturn for various reasons, even if I think I am recovering from the burnout aspects now, it's not to say I've nothing at all related to music.

The latest example: several months ago I finally acquired the budget to purchase an overpriced Roland JD-990 (which I might review and dispel myths over later), and while importing the presets verbatim into a RolandCloud VST would be impossible due to the different architecture, I did get the idea to do the next best thing for the sound archivists out there: archiving what I can of expansion cards meant for it other than the famous SR-JV80 ROM cards.

Rather, I'm taking a light interest in archiving the SL-JD80 library meant for the JD-800 and JD-990.

The SL-JD80 series came in sets of two cards. One contained patches and, in some cases, drum/rhythm sets, and looks like a very mildly conventional Linear Flash card from the time (early 90s). The other card, however, which contained the waveform data, is a type of card so obscure that I could only find two bits of relevant documentation: the service manual for the JD-990 for its pinout, and the data for a very similar but not identical card type (the card is apparently housed in a format meant for JAE JC20-C40 connectors, yet the closest thing I could find was for the 45-pin variant, JC20-C45). Especially tragic is that the connector and the shape gives the impression that this was CompactFlash before it was officially available for the masses (CompactFlash came out officially in 1995, while these cards were created a few years before).

Technical research about the cards itself was one hassle to try to archive. The other was how to extract the data. Even the desoldering approach proved fruitless due to the one card I tried having a thick potting covering what I presume to have been its main memory. Pictured above was the alternative approach, which, while requiring a lot of thin enough wires, has the benefit of, if done right, keeping the card still useable due to no soldering skills required.

It took me many hours, and one allnighter, to wire up three out of the eight cards decently (the other five in the series I simply haven't bought yet), and even after that, the process wasn't perfect. The pictured programmer, a T48, is not only intended for IC's, the closest thing its 40pin internal ZIF socket supports without adapters is a 27C080 EPROM which uses 32 pins instead of 40 (37 if you consider 3 are unusued). The 27C080 also only holds up to 1 MB whereas the cards were capable of holding twice as much, but the good news there is that Roland never took advantage of that capacity for the SO-JD80 cards, though they did for two SO-PCM1 cards for their early JV synthesizers.

Long story short, I managed to finally make working dumps of three of the eight cards in the series, one of which I don't know if anyone else has pulled off (SO-JD80-06 "Grand Piano"). Much to my surprise, despite being intended for a different synthesizer with a different architecture, having examined the data of these dumps, the only major difference from if they were stored on an SR-JV80 card is the size (The SR-JV80 cards generally held 8 MB of data), so, provided the sample data is little enough, it might be feasible to squeeze the entire series onto a single SR-JV80 ROM, though will confirm that when I do find all of the cards in the series, which, if I succeed, could make a nice unofficial ROM for others to use who can't find the SL-JD80 library for their JD-990's (which is a very real possibility down the line knowing data preservation issues in general...)

That's it for now, though hoping things will pick up in the new year. Happy Holidays to all!

I realize it's once again been months since I touched this blog, and my lack of ideas for music lately haven't helped the problem, even if I think I might be past the worst of that funk for now. Still, I suppose it is time I admit one of my more technical projects I've been up to.

I'm fully aware that many people who've run into this blog came here solely for my discoveries with the RolandCloud VST's, and because of that, I've been hesitant to disclose it, but I suppose it is time: that's not the only project I've been caught up with insofar as Roland gear.

A year or two ago I was informed of this third-party product that makes it possible to read and write SR-JV80 ROM images. While I did ultimately buy both the programmer and a spare blank expansion card by them, it was clear what was currently known about the ROM's format was limited from what the product pages originally said, so I took it upon myself to do some disassembly and reassembly work.

Having finished up that aspect, and acquired a JD-990 for additional testing, I can verify that it is possible to create your own SR-JV80 ROM images if desired via a GitHub repo. I make no claims that the process is perfect, especially given the nature of delta PCM audio, but as crude as it is, it works, which I know for a fact having tested with E-mu Proteus samples loaded up in the video at the bottom of this post.

That's one of the projects I admit I've been caught up with lately. I promise I will try to update this blog again within the next month or so with at least one of the others.

I fully admit I'm a huge fanatic for the 1980s powerhouses of music production, and it's a pleasant surprise the past few years to see some of those developers trying to preserve their legacies.

Among the most active of these legacy-prolonging brands is the remaining staff from New England Digital, best known as the developers of the Synclavier systems from the late 70s to the early 90s, with virtually unprecedented digital sound, and the first to support 16-bit sampling (even Fairlight fans had to wait until 1985 for the CMI 3).

Needless to say, when I was first informed a few months ago that they were revisiting hardware synthesizers with some one-ups from the Arturia VST recreation, I was immediately tempted. Fortunately, my family was supportive and generous enough that they were willing to pre-order it as a late Christmas gift (the production has apparently been slow for customers, meaning I only got my own unit a few days before this was published).

Enter the Regen.

At first glance, I admit I was a little deterred: desktop synthesizers are not normally my thing over the old rack-mountables, but despite the form factor, it seems to definitely be capable. Powering up took a much longer time than I'm used to for a hardware synthesizer (barely less than a minute without external storage), but then, this is the first "new" synthesizer I've had in a long time.

Like the classic Synclavier, the Regen is a multitimbral synthesizer (albeit 12 parts instead of the usual MIDI standard of 16), capable of sampling as well as true synthesis (including the innovative resynthesis of audio recordings they innovated in the mid 80s), and unlike the Arturia recreation, it seems to support literally every sound-design-related parameter of the classics, particularly multisample support (up to 48 samples per partial, just like the old systems!). It doesn't stop there, however: it also adds some features unseen in NED products before. For the first time, Synclavier samples can have controllable filters applied (effectively making them passable for their old Fairlight competition), and in addition, analog modeling a la many 90s synthesizer classics are also a sound generation option, as well as built-in reverb! This is all in addition to supporting 12 partials a la the Arturia VST as well (the original hardware supported only up to four per timbre).

How does it handle all of this power at one's fingertips? The manual suggests it does so by computing a variable maximum of resources depending on the complexity of the session. The most obvious case stated: polyphony has a probable max of over 90 voices, but it does state the polyphony limit will be lowered with enough effects applied, the two clear cases being the new filters and reverb algorithms. Haven't tried to test these limits through a multitimbral production yet (a matter I hope to fix soon), but upon pressing the Volume button in the Master section on the unit, you can also find a meter that is generous enough to inform you of the CPU load, so it's not overly difficult to figure out if you're doing something too CPU-heavy or not.

Many of the classic presets and sounds are included (from the famous Michael Jackson "Beat It" FM/additive gong to the pizzicato violin multisample in the "Princess Bride" soundtrack), as well as some new ones by some new sound designers, but let's face it: many people will want to create their own, especially given each bank still only has up to 64 timbres. With an SD card (not included with my own order), it is possible to add more timbres and samples, and best of all, I have confirmed via the Synclavier3 software that it is easily feasible, albeit a bit time-consuming for samples, to convert all of the classic timbres into a Regen-compatible folder of text files! I was a bit surprised that the format of the timbres is that simple, especially in contrast to the binary code of the older format, but I'm certainly not one to complain about that, especially if I feel it best to make a small edit on my computer for whatever reason.

My only ongoing gripes with it: for user sound storage (timbres and samples alike), the Regen relies exclusively on SD card storage. It would be nice if there were alternative storage options, or better yet, the ability to edit sounds in realtime through a computer connected to the Host USB port (which does transfer MIDI, at least). In addition, the Regen only supports two pairs of outputs out of the box: one balanced, and one unbalanced. The manual suggests USB accessories that can work with the Regen include a Bluetooth audio transmitter and a USB audio interface, which could help mitigate this issue, but it still could've made a better impression to me if it supported more audio outs, especially given we're talking a multitimbral module. Still, this criticism I can let slide given the relatively small form factor: multitracking is an option for the desperate, anyway, just like in the old days.

Overall, well done to Cameron Jones and the rest of the Synclavier team for keeping their legacy alive, and with some pleasant surprises as a bonus!

The Synclavier Regen can be ordered from their website. As of this writing, production is paused atm, but will resume next month.

While inspiration is still coming back to me insofar as music creation, may as well update this blog with another update relating to music tech.

Roughly a year ago I was lucky enough to be able to acquire an E-mu Proteus/1XR with the Protologic expansion made by InVision Interactive, as reviewed in a prior post. True to my word, as of yesterday, I finally was able to successfully extract and read the sound chips from it, as well as the standard Proteus 1 and 2 sound ROM's.

This process was more gruelling than one might expect: while the sound ROM's were all thru-hole DIP's a la EPROM's and thus easier to extract than the more widespread SMD's (which also was also changed to for the Proteus MPS and the Proteus 3), the brutal truth is that I never desoldered before, so I admit I spent a few months trying to manage the courage to start despite paranoia about damaging anything.

And then it happened: toward the end of that procrastination period, someone was generous enough to list a Proteus 1 motherboard for sale on Reverb. I was absolutely certain I'd damage the board and possibly the components with the skillset I had, so I figured I'd purchase that as well for the sake of practice.

Lots of slight burns on the board and some chips with clipped pins later, I finally felt confident enough to try on the reason I took on the project in the first place: the Protologic expansion. Perversely, despite later extracting from the Proteus 2 as well, this was my most successful attempt at desoldering: all pins were intact, and I failed to notice major damage to the board for the most part, though thankfully, resoldering them back in proved unnecessary due to someone else also selling another Protologic board separate from their unit. Might consider that anyway just to learn how to do it, though...

Anyway, most of the remaining time on the project was spent trying to salvage the Proteus 1 and 2 chips that were damaged during extraction, a process that only concluded last night.

Pictured was the solution for the worst case: a test clip and lots of cables due to nearly a quarter of all the 32 pins being broken, even if the leg bases were all still workable. Needless to say, these particular chips should probably never be used again, though thankfully, I was still able to read from them with enough effort.

Alas, out of respect for copyright of E-mu Systems, Creative, and Digital Sound Factory, I have no plans to host the dumps for the standard Proteus 1 and 2 sounds here: those sounds can still be aquired legally through the last entity, anyway. The consolation, however: InVision Interactive went out of business long before even E-mu did, so I think it's safe to assume that those rare sounds are true abandonware at this point.

This zip file includes dumps of the ROM's as well a few other goodies.

Thanks everyone for bearing with the ramble, and before I wrap up: know that I do have another probable game music project in the works, though I can't publicize the details about it at this time.

Now that it's been almost exactly a month since my big music project came out, I'll start this post by reflecting.

Brok: The InvestiGator is doing well. In fact, while CowCat Games has repeatedly expressed concerns about it not being viral, I honestly feel it's doing far better than I expected for something I was involved in. Granted, part of the publicity was the result of discovering scammers posing as Steam curators, but even beyond that, I'm honored to be involved with a project that's doing as well as it is.

All this in mind, I received an interesting offer from CowCat the other day that ties into my hobby-level interest in technology in general. As of that day, I'm also doing a chunk of the porting of Brok to the macOS platform. To clarify, however, I'm not touching the code for the most part: we just reached an agreement to use my 2019 Mac Pro, which is apparently newer than the Mac Mini he has, for compiling and deploying it all.

The process has been interesting so far, to say the least. First, until this point, I was used to compiling Mac programs for personal use and thus didn't need to worry too much about Apple's steep validation requirements: all I needed beforehand was a valid Developer account in order to compile, and codesign locally. Now, however, I have multiple certificates on behalf of CowCat Games, plus, the apps need to be submitted to Apple's servers for notarization, an aspect that, unfortunately, the engine used to make Brok (GameMaker Studio) cannot handle automatically.

Second, it was only as of late last night that we finally managed to get the Steamworks API cooperating with the builds. Granted, this phase isn't required for DRM-free stores such as GOG and itch.io, but even so, initial builds saw failure to load the API properly: XCode debugging revealed that it would look for the library in the wrong place. For any other gamedevs hoping to build Steam games on Macs via GameMaker Studio, I only barely found the solution: while a Hardened Runtime is required for notarization, there need to be two runtime exceptions specified:

-"Allow DYLD Environment Variables"
-"Disable Library Validation"

In my experience, only the latter option is required for testing builds, but the other is required for when it's submitted for notarization.

I realize this is a long read over matters that are technical but not musical, and I apologize for that, but in the end, this is related to a project I did compose for, so I feel it is worth a mention. Needless to say, an official Mac build of Brok is also guaranteed in the works.

There also is more on my plate, but I'll save that information for a future entry. In the meantime, Brok is currently available for Windows via Steam, GOG, and itch.

After another period of silence on this blog, I'm happy to say that one reason for my relative inactivity the past few years is now finally available to the general public.

Last week, on August 26th, Brok: The InvestiGator was released for PC on both Steam and GOG. I had the pleasure of being able to make all of the music (and a small few other contributions) to this project. I'd say more about my experience with it, but I intended my Bandcamp release, out as of now, to handle that via its liner note bonus item and the descriptions for each track.

Note: this is not literally the complete-score: CowCat and I agreed that that would be a Steam DLC exclusive. Still, enjoy, and below is the link to the Steam page for the game itself!

https://store.steampowered.com/app/949480/BROK_the_InvestiGator/

Admittedly things have been very slow musically, lately, and in order to try to keep this blog alive, one reason for that, other than my dayjob of just over two years being a priority, is that I've spent some time experimenting with the modding of older games. This post will be covering the modding of one game in particular, one that I fully admit is a very unusual choice for a mod: Myst.

No, I'm not talking the various remasters and remakes: even the ScummVM recreation is pretty lacking due in part to it only supporting the Windows versions. I'm talking about the very original Mac version of the game, possibly one of the most complex successful HyperCard stacks ever made.

The idea for my mod was fairly simple: improve the sound quality by making everything 16-bit (the original game was 8-bit), roughly doubling the typical samplerate (an unusual 11127 Hz to 22050), and stereo when possible over mono.

Compiling all of the higher quality sounds from a sound library that it turns out made most of the sound effects (The Works, currently distributed by Sound Ideas), and the iPhone port of the game for most of the rest, was the relatively easy part. The trickier part was reimporting them via ResEdit on the Mac OS 9 emulation side, and the open-source ResForge project on my host Mac's side.

This process resulted in three obstacles:

1. Some sounds occassionally failed to play. After some diagnosing, I discovered the problem was that the sounds of concern were too large for the current memory settings to handle, but fortunately, doubling the memory requirements for the game by editing the SIZE resource via ResEdit seems to have all but fixed that.

2. The game originally stored most of the non-looping sounds in a single resource file, which turned out to be a problem because said resource files seemingly cannot exceed a filesize of 16 MB. Therefore, the only solution for this was to split it up into multiple files.

3. The movie files that made use of sound changed their timestamps after reimporting the sound, and consequently, scripts that made use of video timestamps had shorter durations, roughly 3/5 to be exact.

Obstacles 2 and 3 could only be fully overcome by loading up HyperCard (2.4, in my case) and changing the scripting marginally, which, in retrospect, is risky territory, hence, why I hesitate to share the edited stack files in particular.

Still, I am satisfied with the results, and if there is enough interest, I am willing to share videos of what I've done, at the very least.

For anyone else with an interest in modding other old Mac games, ResForge is definitely a great way to ease into it!

After celebrating my 30th birthday a few days ago, I'm fully ready for another blog post.

Anyone familiar with early 90s electronic music, especially video game music, has probably heard of the E-mu Proteus series. What isn't as readily known is that there was also a 4 MB sound expansion for the Proteus 1, and by a third party at that: ProtoLogic, by InVision Interactive.

Due to the additional components, and the consequential heat generated, many expanded Proteus 1 models also came with a 2U faceplate extension with the InVision branding on it. How I managed to find one expanded Proteus 1XR without this is a good question, that said, but regardless, the sounds are there.

In a nutshell, most of the Protologic ROM is dedicated to picking up where the sound ROM of the famous Proteus 1 left off, meaning additional pop-oriented sounds, most notably some synth pads, distorted guitars, some TR-808 samples, and a few ethnic percussion sounds. There are lots of mixed thoughts about how the Proteus 1 sounds without it, so in my opinion, this made it much more worth collecting, even if it is still constrained to the programming limits of the Proteus series (i.e. no easily-programmed filters). It comes with an additional 128 ROM presets to get started with as well, which makes having it for a Proteus 1XR all the more recommended since the XR series inherently has more RAM presets (a total of 512 presets combined!).

Alas, one reason why I'm making a big deal out of this: there is no modern software alternative to these particular sounds, not even from Digital Sound Factory. Aside from InVision being a third party sound developer, they faded into obscurity shortly after this came out, and even their sample CD's seem to have been better remembered despite many being based on other soundware brands of the time such as Sound Genesis and Northstar. The consolation: I've put in the effort of sampling the ROM as best as I could, and while I hesitate to publicly share my results in the event that the copyright owners are still around and take notice, know that some archiving of these rare sounds have been done, complete with an attempt of recreating the presets in SFZ format. I'm also considering dumping the firmware EPROM's, which, rest assured, I am more willing to share if I do for the sake of the longetivity of these rare units.

Thanks for reading!