Lurking within this humble arranger keyboard is a completely new form of synthesis...
Arranger keyboards have come a long way since the days of Raw Sex and John Shuttleworth. Nevertheless, they don’t often find their way into the pages of SOS. The ability to press a button and have a pre‑cooked backing track emerge is impressive, but I’d imagine most readers of this magazine prefer to create music in a more granular way.
Casio’s CT‑S1000V is, by any yardstick, an arranger or home keyboard. It has features such as built‑in speakers and battery operation that aren’t usually found on pro instruments; and it lacks quite a few other things that you’d expect to find on a pro‑level workstation. And it forms part of the huge Casiotone product family, which extends right back to the early ’80s and includes numerous Shuttleworth favourites. So why are you reading about it here?
Well, the CT‑S1000V is the first product to incorporate a technology Casio are calling Vocal Synthesis. And, although its capabilities as an arranger keyboard might not be all that interesting to SOS readers, I think that Vocal Synthesis could be.
Synthesizing the human voice is an area that has seen rapid development in recent years, thanks mainly to the explosion in machine learning. Dreamtonics’ Synthesizer V, reviewed in the March issue, is capable of stunningly realistic results, while Respeecher allows anyone’s demo vocals to be replaced with a very believable AI‑generated alternative. But what most of these products have in common is that they can’t easily be used in real time. The unique feature of the CT‑S1000V, as far as I’m aware, is that it allows vocal lines to be played from the keyboard.
Take a listen to this sound demo lifted off our NAMM 2023 video.
Or check out this short video from Jordan Rudess playing this Vocal Synthesizer keyboard.
Words & Music
The home keyboard tendencies of the CT‑S1000V are apparent when you first set it up. It’s relatively lightweight and made largely of plastic; it’s powered by an external DC supply when not running on batteries; and it has built‑in speakers that remain tenaciously active until you do some pretty deep menu‑diving. Thankfully, it’s possible to store up to four User Setups for the whole instrument, which include settings like speakers on/off and which can be set to load on startup.
A pair of buttons labelled Tone Mode Select to the left of the main LCD screen offers the choice of Lyrics or Instrument. The second of these accesses your choice of 999 instrumental sounds, a complement that includes numerous drum and percussion banks, a full General MIDI set, ethnic instruments from around the world and classic Casio CZ and VZ patches as well as a wide‑ranging selection of keyboard, synth and bass sounds. Instruments can be split, layered and arpeggiated; they cannot be edited, as such, but key parameters such as filter cutoff and resonance can be assigned to the two dials directly above the Tone Mode Select buttons, and to the modulation knob above the sprung pitch‑bend wheel. There’s also a basic form of user sampling.
The CT‑S1000V measures 930 x 258 x 91mm and weighs in at 4.7kg.
The instrument sounds are for the most part pretty decent, but I’d imagine most SOS readers will have even more variety and certainly more controllability available in software. The real interest, and most of the CT‑S1000V’s unique features, lie behind the Lyrics button.
Shadow Play
There’s a total of 150 preset slots in the Lyrics domain. As the title suggests, each of these stores a lyrical phrase, plus associated information. The first 100 come pre‑populated, mostly with very short phrases of just a few words, while the last 50 slots are blank for your own creations.
After selecting a Lyric, the two most important decisions are the choice of a Vocalist and whether to operate in Play or Note mode. There are 23 Vocalist slots, of which all but one are populated by default. The factory Vocalists begin with four Choir variants; then there are two Vocoder and a Talkbox preset followed by a mixed bag that includes Female, Boy, Opera, Child, Husky and Cute. The last few fall squarely into novelty territory, including the hilarious Death Voice and the frankly silly Animals. All are polyphonic.
Once you’ve selected a Vocalist, there are various things that you can do to modify the sound. You can adjust Gender, Age and Octave Shift, as well as all of the same filter, attack/release, portamento and modulation parameters that are available for instrument sounds. These don’t change the fundamental character of the sound but do permit a fair range of variation, and there’s a lot of fun to be had with the vibrato in particular. However, the user interface can be obscure. The menu structure, for example, has a second ‘shadow’ level that is accessible only through long presses on certain buttons. There’s no on‑screen indication of how to reach these pages, so unless you read the manual, you won’t find them.
Even within the main menu structure, not all values can be edited directly using the main data entry controls. Some can be adjusted only by selecting Knob from the main menu and then assigning them to one of the three assignable controls. Annoyingly, this includes some of the most important editing controls, such as the filter settings and the send levels for the built‑in effects. These settings are stored as part of a Registration, which is effectively the top level of preset, but not as part of a Vocalist or Lyric.
You Hum It, I’ll Play It
In Play mode, which is the default, holding down a note or chord or playing a legato phrase causes the CT‑S1000V to step through the selected lyric at the current tempo. The syllables within a lyrical phrase each have a duration specified in musical values so, for example, if you select the preset “I wanna be with you” and hold down a key, you’ll hear “I”, “be” and “with” as quavers, “wanna” as two semiquavers and “you” sustained indefinitely until you let go of the key. If you let go of the key before the lyric has completed, it’ll jump back to the start next time you press a key.
In Note mode, by contrast, the internal rhythm of the lyric is ignored. Instead, each syllable sustains for as long as you hold down a note, and the lyric only advances to the next syllable with a new key press. If you’re playing lead lines, this progression happens regardless of whether you play legato or not; if you’re playing chords, a setting called Syllable Stay Note tells the CT‑S1000V not to advance to the next syllable if you’re holding down more than a certain number of notes. This is rather buried in a shadow menu that also includes some interesting options such as triggering syllables in a random order.
For greater control over how your Lyrics are played back, you can also enable the Lyric Tone Controller Keyboard. This, in essence, repurposes the bottom octave and a half of the CT‑S1000V’s keyboard as a series of keyswitches. The white notes allow direct selection from the first 11 syllables in the phrase: so, for example, hitting A2 will bring up the message “Syllable Position:6” on the LCD, and that will be the syllable you hear next time you play a note. The black notes perform a variety of interesting control functions, such as retriggering them at your choice of musical division.
In theory, Note mode and the Lyric Tone Controller Keyboard increase the real‑time flexibility of the vocal synthesis, making Lyrics more adaptable to different backings and tempos. In practice, this is certainly true, but there’s a down side, which is that the intelligibility of the lyrics suffers considerably. Unless you choose one of the wackiest Vocalists, nearly all the preset Lyrics are clearly intelligible in Play mode, but even if you step through the syllables at exactly the same rate in Note mode, it’s hard to achieve the same clarity, because consonant sounds are softened or obscured. My guess is that the vocal synthesis engine can articulate syllable transitions in Play mode because it can anticipate their timing; the same is obviously not true in Note mode, because it doesn’t know when you’re going to press the next key. Having said that, there are various editing parameters that can be adjusted to improve matters, and raising the ‘attack balancer’ setting on one of the shadow menus can bring back some consonant articulation if needs be.
Round the back we find a micro ‘To Host’ USB port and a ‘To Device’ USB A port, 3.5mm headphone and audio in sockets, quarter‑inch audio outputs and a pair of quarter‑inch pedal inputs.
Creator Of Words
If all you could do was play back the preset Lyrics, the CT‑S1000V’s Vocal Synthesis would be no more than an amusing novelty. Fortunately, it can do a lot more than that, as long as you have a smartphone capable of running Casio’s Lyric Creator app, and the means to have it communicate with the keyboard. The latter point can be a bit of a stumbling block: for full functionality, you’ll need a physical cable that connects to the Type‑B micro‑USB port on the rear of the CT‑S1000V. There is an optional Bluetooth dongle for the keyboard, but inexplicably, this only deals with audio and can’t perform data transfer. If you don’t have a physical cable, Lyrics can be saved to a USB drive and loaded onto the CT‑S1000V from there, but creating a Sequence of Lyrics isn’t possible without connecting the phone directly.
The Lyric Creator app is essential for creating your own Lyrics and Vocalists.Individual Lyrics can be somewhat longer than the presets that come with the CT‑S1000V, but there is an upper limit, and you’ll do well to fit more than a couple of lines into a single Lyric. Nor would you particularly want to, because you’d then be painting yourself into a corner when it came to playing that Lyric in either Note or Play modes. So, if you want to store an entire song, you’d be better off turning each line into a separate Lyric and then creating a Sequence which strings them together, a feature added in the v1.02 firmware. It’s also possible to step through Lyrics using a connected sustain pedal, which is ideal from a performance point of view. However, you’ll use up the 150 Lyric memories pretty fast if you start storing entire songs.
Lyrics can be created in three ways. You can import a MusicXML file, you can type in your Lyric, or you can simply speak, whereupon the app will use your phone’s speech‑to‑text function to transcribe the words. I didn’t try the first approach, but both of the others work pretty well, the main fly in the ointment being that auto‑complete seems to generate duplicate entries when used. Editing Lyrics, adding extra syllable delimiters and changing the musical duration of each syllable is likewise easy and intuitive.
What the Lyric Creator app doesn’t have, though, is any means of auditioning the results. In order to do that, you’ll need to transfer them to the keyboard — another good reason to use a wired connection, since there’s a certain amount of trial and error involved, and repeatedly dumping files onto a USB stick to transfer them gets old fast. Even with a physical connection, the transfer process is pretty counterintuitive; the default file location shown on your phone is not the keyboard, so you need to navigate each time to find the CT‑S1000V.
The Lyric Creator app is also the only place where you can define and name a new Vocalist. This involves setting familiar parameters such as Gender and Age, but the key option that’s not available on the keyboard itself is the ability to import a short WAV file to use as the basis for the vocal synthesis. As far as I can tell, this has to be a mono 16‑bit file under 10 seconds in duration. The temptation is to use a snippet of human speech or singing, but this won’t really work: what’s needed is something with a prominent broadband noise component as well as a pitch element with plenty of harmonic content. I generated a few different sounds using soft synths, and although some of them sounded interesting, none of them really produced intelligible vocals. Again, the convoluted process of transferring audio files from a computer into a phone and thence into the keyboard will probably limit people’s willingness to experiment, as will the fact that there appears to be only one slot available in the keyboard for user Vocalists.
Play Time
The addition of hardware controls and a tactile user interface often seems to unlock the sonic potential of a synth engine, and I’m sure we’ve all had the experience of playing around with a soft synth and wishing it existed as a hardware instrument. The Casio CT‑S1000V is the first instrument I’ve ever played that has produced the opposite response. With only a mod wheel and a handful of assignable knobs, it offers little more performance control than you’d get from a basic controller keyboard, while the obscure user interface and need to handle Lyric editing from a smartphone make it hard to fully exploit the potential of Vocal Synthesis. It’s not hard to imagine how the same technnology in software format would give you built‑in Lyric creation, the ability to easily audition changes, and a more intuitive interface without shadow menus and the like.
In saying that, I’m conscious that the CT‑S1000V is not conceived as a ‘professional’ instrument. Design conventions are different in the world of arranger keyboards, and as I’ve mentioned, it also boasts numerous features that won’t be relevant to most Sound On Sound readers. Yet, despite the frustrations involved in getting to grips with Vocal Synthesis, I do think it has something to offer in a music production context.
The CT‑S1000V may not be able to generate a convincing human lead vocal, but it has a sound all of its own, which makes a fascinating alternative to vocoders, talkboxes and the like.
That something is not realism. No matter how you play or process it, Casio’s Vocal Synthesis is unlikely to produce something you’d mistake for an actual human vocalist. You can usually understand what the CT‑S1000V is singing, at least in Play mode, but its output is obviously synthetic. However, I absolutely don’t mean that as a criticism. The CT‑S1000V may not be able to generate a convincing human lead vocal, but it has a sound all of its own, which makes a fascinating alternative to vocoders, talkboxes and the like. As such, its obvious application is within electronic music, and in a house or techno track where you just need a couple of memorable lyrical phrases, it’s particularly easy to come up with something that is both ear‑catching and fresh. And the CT‑S1000V’s ‘home keyboard’ origins are reflected in its relative affordability. It might not have the same sex appeal as a vintage analogue synth or a boutique effects box, but it has the same sort of potential to deliver something unique that will help your track stand out. What else is there that allows you to actually play vocal lines from the keyboard?
Pros
- Allows lyrical phrases to be played in real time from the keyboard.
- Creating and editing Lyrics in the Lyric Creator app is straightforward.
- Has a unique and distinctive sound that’s akin to, but not the same as, vocoders and talkboxes.
- Relatively affordable.
Cons
- The user interface is confusing and not very intuitive.
- You can’t audition Lyrics or Vocalists from the app, and transferring them to the keyboard is tedious.
Summary
Although the CT‑S1000V might not be the perfect implementation of it for studio use, Casio’s Vocal Synthesis offers something unique that is genuinely new, allowing you to play vocal parts from the keyboard in real time.