Roland Supernatural Questions

Discussion in 'Keyboards' started by RMosack, Mar 12, 2019.

  1. RMosack

    RMosack Supporting Member

    Messages:
    2,958
    Joined:
    Mar 21, 2003
    Location:
    Oceanside, CA
    Looking at the Juno DS, the VR-730 and the FA Series. As I understand it, only the FA synths have the Supernatural sounds.

    1) How are those sounds different thant the rest of the samples and waves?

    2) Are they worth the price difference from DS to FA? And why?

    Sorry if these are basic questions. I'm a guitar player dipping my toes in the synth pool!
     
  2. stevel

    stevel Member

    Messages:
    13,255
    Joined:
    Apr 6, 2008
    Location:
    Hampton Roads, Virginia
    I think you are correct, only the FA has the Supernatural sounds.

    You could probably get a better and/or more technical answer from Roland or some of the experts on Keyboard Corner forum.

    Let me give you a brief rundown of synthesis techniques and see if that helps.

    Sound is produced by a vibrating or "oscillating" object. Luckily for us, the generators downtown produce "alternating current" that "oscillates" from postive to negative, and through circuitry this can be multiplied or divided to produce all the frequencies we want, which get sent to a speaker which oscillates itself to produce sound waves of those frequencies.

    So what makes sound in a traditional analog synthesizer is an "oscillator".

    Sound waves in the real world are not only made up of a single oscillation, but many actually are multiple oscillations - your guitar string vibrates not only the length of the string, but in half, in 3rds, in 4ths, in 5ths, etc. as it vibrates (when you play a harmonic, you are forcing the string to divide into these "partials").

    In order to emulate real instruments (one desire of early synthesists) you have to have as many oscillators as you have partials - so if you want a fairly complex sound wave, like that of a Violin or Trumpet, you need like at least 16 partials - and thus 16 oscillators to get even close.

    This is called "Additive Synthesis" because you start with a basic oscillator, then add more to it.

    But this is costly and was impractical in the early days.

    So instead, they came up with "Subtractive Synthesis" which instead starts with a complex wave (which is actually easy to create electronically with diodes and junk) and they use a Filter to "filter off" or "subtract" Partials from the more complex wave to make the simpler ones.

    While the results don't produce highly authentic acoustic instrument sounds, it's more cost-effective.

    Most of the synthesizers you know from the past were Subtractive (the Moog being the big deal) but they often included 2 or 3 oscillators so there was a slight bit of additive synthesis as well.

    This basically continued through the transition from Analog to Digital, just using digital circuitry to produce the sounds.

    FM (Frequency Modulation) Synthesis was slightly different (and owned by Yamaha and prevalent in a lot of their synths including the famous DX7) and actually did a better job of recreating Brass like and "Metal like" sounds (Vibes, the famous DX7 Electric Piano) and "wire string" things like Harpsichords and Clavichords. However, they still didn't sound like the real counterparts too much.

    The Mellotron was an attempt at recreating sounds more accurately and it literally used Tape Recordings that were triggered to play back when a key was pressed.

    When we entered the digtial era, people tried to recreate that by recording digital samples and playing them back. The advantage was you didn't have to record a separate tape for every single note (that had to be done so you could play chords on the Mellotron) and you could digitally increase or decrease it to produce additional notes - though if you raise or lower it too far it sounds "chipmunky" and obviously moved, so the better sounds sampled as many notes per octave as they could.

    but that too is expensive - especially at the time. Memory space was huge back then (same size today, but we have stuff that can handle it) so they weren't too practical.

    At some point, some scientists had done some studies that showed people's brains only uses the first couple of milliseconds to determine what a sound was, and then just "accepted" the continuation was the same sound.

    Some smart guys said, "well, what we can do is make the first few milliseconds a sample, and then the rest a traditional oscillator and people will base their opinion of the quality of the sound on the short sample". And the advantage there is the short sample - which doesn't take up so much memory, and in many cases doesn't suffer from the pitch shift difference so fewer samples can be taken per octave (which again saves memory).

    This is the rise of the "Rompler" (they're also called PCM too because the audio we use in digital audio is Pulse Code Modulation and that's actually what your AIFF from Apple or your WAV files from Windoze is - they're PCMs wrapped in Apple's or Microsoft's "containers" - you'll also see them on your BD player as an audio format a lot of times).

    The way my old D5 worked is it had 4 "partials" as Roland called them. Up to 2 could be samples and up to 2 could be traditional (though digital) oscillators. Roland called this form of synthesis "Linear Analysis" or "Linear Arithmetic" or something like that but it's also called "Wavetable Synthesis".

    This way you could mix the "tu" sound a flute player makes when they start a note, with a simple sine wave oscillator (thus only using 2 partials total) and make a far more realistic sounding flute sound than either Subtractive or FM synthesis was capable of (but a Mellotron could be better!).

    Romplers vary in quality because of how the samples are taken - what their audio quality is, how well-played and well-recorded they were, the quality of the instruments and gear they were recorded on, the talent of the engineers recording them, the talent of the editors and programmers putting them in, and how many were taken per octave - the more samples per octave, the more realistic the sound because they didn't need to be pitch shifted.

    The famous Alesis HR-16 drum machine is famous because it was the first to use full CD quality samples - 16 bit. So the drums sounded really real - they were - they were recordings of real drums (and since some of those sounds could be short, there was not as much memory space needed).

    Having an 88 Key sampled piano was the holy grail for most companies, and by the time they reached it, we were already on to the next technologies!

    Most modern synthesizers still use this Rompler technology. Some sounds will be created out of samples only, some a mix of samples and digital oscillators, and some out of just digital oscillators.

    And while that part has "additive" aspects of it, aside from Yamaha's FM engines, most of it is all still just basic Subtractive Synthesis (which is why that's such an important thing to learn).

    Roland's new technology is Modelling.

    What it does is give you "more bit depth" than with traditional digital as I understand it. This allows for smoother transitions in pretty much everything - timbre, volume, attack, etc. So IOW, when you play piano, the timbre can more smoothly change over time and with pressure much more like a real piano does.

    But I suppose the real plus is, it doesn't take the kind of memory and processing it does with Samples - with samples, you need one for each volume (as a piano gets "brighter" as you play it louder, so the timbre has to change with volume). The problem is, you can't possibly account for all the variations and even if you could it would make it so expensive because of the memory (digital audio at CD quality is still 5mb per minute per sample - so even with little samples it adds up). Also, a sample has to be "looped" in the middle during the sustain phase and poorly looped ones are noticeable. Roland does an excellent job with them but you can hear them on some of the sound effect kinds of sounds (which is OK because they're supposed to sound like that).

    Modelling alleviates all that to my knowledge.

    So "Supernatural" is a form of modelling technology.

    Because the modelling is "specific" for different sounds - just like Kemper probably uses a different kind of modelling from Line 6, the "Acoustic" sounds have one Modelling Engine, and the Drums use another, and the Synths use another.

    Roland (and others who do this) call the Supernatural Synth sounds "Virtual Analog" because they profiled real Analog synths.

    I've actually put the waves on an oscilliscope and looked them up online because they don't necessarily name them "moog" or "oberheim" but by finding some visual samples online I was able to compare them and the Roland waves scope in my DAW match the orignal instruments' waveforms pretty much right on (people do this to see how accurate a VST is so I'd say the Roland is pretty accurate).

    So it's like an Analog Synth using Subtractive Synthesis - a real Moog or a real Prophet, except the waveforms are Modelled (and live in the engine as "software" and are thus "Virtual").

    So not only are the Acoustic Sounds (Piano, EP, Organ, Strings, Guitar, and Bass) modelled on real-world waveforms, so are the Drums, and then so are the Synths, but the synths are going one deeper and starting with the actual waveshapes the analog oscillators produce.
     
    BPlexico and RMosack like this.
  3. stevel

    stevel Member

    Messages:
    13,255
    Joined:
    Apr 6, 2008
    Location:
    Hampton Roads, Virginia
    Now, the 64,000,000 question - is it better?

    I'll be honest: I am coming from an "old school" background. The Piano, the EP, the Strings, they sound good as F$%K. I've seen videos online comparing them and other brands that don't use modelling and the general consensus is the Supernatural is just really, really, really good. The Synths - well I've never owned the originals, but they sound great. And they are great fun. The editing is more like traditional synths, and not so "menu diving" like my old D-5. Now, you still have to scroll through screens but it's tab oriented and very easy to work and understand, and the controls are large, they can be controlled remotely, and so on. Now this is true about the PCM sounds as well, but since the engines are different in the Supernatural Sounds, they really only have the controls that "matter" - you're not going to get in and edit the envelope of the piano sound - but you can change things like the "character" of the sound - some of this boils down to stuff like how much of the soundboard of the piano you hear (which is partially a result of miking) but it's kind of the difference between having your head in the lid, versus sitting at the keyboard, versus being in the audience. That's not really a great explanation, but that's the kind of like what's happening. How "tine-y" the electric piano is.

    The Hammond organ - a "clone wheel" organ, will never satisfy a purist of course. But, I think it sounds great. And it's WAY easier to edit than doing it in a PCM based environment because in the "software environment" of this kind of modelled engine, you've got sliders on screen (that look like Hammond bars) you can slide in and out. You can't really control them in real time easily (which is a major complaint of many, but the problem is there are 9 of them and only 6 knobs so how do you pick?).

    In fact, I should mention that with the Supernatural sounds, the display looks more like a VST screen on a computer - it's more "interactive" and you highlight knobs and turn them with the wheel, etc.

    The Synth sounds are only missing one traditional control I can find, which is Sync, but it has a sync modelled waveform if you want to use it. But the cool thing is, it just doesn't have one Filter when you call up a Moog waveform (like a Moog would) but you can choose different types of filters for any wave you pull up. I mean, to be honest, it's a shame people don't know about this stuff because the SN-S engine is EXTREMELY POWERFUL and rivals and even surpasses many VSTs I've seen

    The PCMs in the FA are an entirely different engine - still basically Wavetable Synthesis that's Subtractive in nature. It has more parameters, and less intuitive ones and honestly, it takes a pretty good understanding of Analog Subtractive synthesis to get around the PCM architecture in the Rolands. At least this one - the DS may be better becuase you have the software tool for it.

    Come to think of it, that's probably the main reason Roland didn't feel the need to make an editor for the FA because I bet they felt the Supernatural sounds wouldn't need it because you could already do it right on screen and the PCM sounds are just basically a "bonus".

    Now as I started to say above, I'm old school, and I'm used to my D-5 and my Sound Canvas.

    The Supernatural sounds sound really good, but maybe TOO good! Too real. I mean it's like, they're supernatural or something!!!!

    I've found in a lot of cases I actually prefer the PCM based sounds.

    Now, this is most likely a bias - I'm just used to the older stuff. I'm also used to hearing old recordings where a piano never really sounds like a real piano - I mean I rarely hear a great recorded piano sound on a pop album - in fact, I hear a lot of bad ones on Classical solo piano recordings! I've recorded some myself (9 foot steinway grand) and been praised for them by people I feel would know, so I was very happy with the result as were the clients.

    I'll go through all the pianos in the unit and find I like the XV one the best (actually, there are some in the EXPs that are better even).

    And as I was saying before, many people don't want a piano that sounds like a piano - they actually want a piano that cuts through a band mix and sounds like a piano that was recorded very "up front" and EQ'd to cut through the mix - so again the Supernatural may not be the best choice for that - they want one that actually sounds rather "processed".

    But I've been playing a gig and using the Piano #2, the Strings, and the Electric Piano "raw" from the Supernatural set. No one has said they sounded bad or anything! I use the Organ from one of the PCMs as note of the presets sounded like what I wanted from the SN set (though I could edit one with the drawbars but just didn't feel like messing with it).

    So I would say, the Supernatural Sounds sound amazing, but I have a tendency to prefer sounds that are what I feel more accurate to what I'm trying to cover, which tend to be PCM based sounds. However, I'm sure if I tweaked the SN sounds more (which I just haven't done yet) they could be just as good if not better. But many of the sounds I'm trying to cover are either highly processed or don't sound that "real" to begin with, so the PCMs cover that.

    If I could compare it to anything, I would say that the FA is like a great PCM- based synth coupled with a 3 great VSTs (like you'd use on your computer, but inside the FA), one for Acoustic Instruments, one for Synths, and one for Drums.

    But again, the comparisons I've seen with SN versus other sounds - especially with Piano and keys, the SN tends to be winning in the videos I've seen.

    So cons - more expensive to get into the SN level. You're not going to get the most out of the SN Piano without a great weighted 88 key keyboard (but that's really true of ANY piano sound no matter how it's made) but the FA-06 is barely useable for it, and you may just not need those sounds if the PCM sounds do it for you.

    Pros - sound exceptional. Easy to edit and even fun and inspiring to play with - especially the synth engine - you can play with it for hours without menu diving more than a single level and if you want a couple of tabs. Honestly, for me, without having a VST that does it, it's the next best thing to having some real analog synths in the house. That alone is worth the price of admission for a non-purist who's an interested hobbyist to semi-serious synthesist. And I'm sure you could get more out of the "stock" presets than I've bothered to mess with, so don't take any of my dismissal of the presets as negative - those who are serious keyboardists will almost aways assume a default patch will need to be tweaked and do so ruthlessly - sort of like I do with Line 6 models - I'm GOING to change it from the presets because the presets don't suit me. I just haven't done it with the Supernatural sounds because many of the PCMs (and there are many, especially with the EXPs) suit me OK so I don't need to.

    One more thing: You mentioned the D-05 earlier. I saw them compare that to a real D-50. The D-05 sounds are sampled from (or modelled on, not sure) the D-50 or use the same architecture. The D-05 was pretty much right on with almost every patch and any differences present in the very few patches that had them were minimal. So they were extremely successful at re-creating the D-50. I checked the D-50 patches I downloaded - same sounds. EXACT. I mean EXACT. And that's on the SN-S engine so they are modelled on the D-50.

    Now I don't know that you can edit them to get every single sound you could edit into on the D-50, but the starting points are identical, as are the patches they gave you. It's pretty amazing. It's not hundreds of patches or anything, but they freakin sound just like the D-50 - as to the Juno, Jupiter, and TB etc. sounds!

    So you get a "taste" of those, in addition to all the default sounds.

    If you're into that, that's another Pro. The PCM versions of those sounds are also very good though so again the DS might be pretty capable in that regard.
     
    RMosack likes this.
  4. stevel

    stevel Member

    Messages:
    13,255
    Joined:
    Apr 6, 2008
    Location:
    Hampton Roads, Virginia
    Oh, I'll just answer your 2nd question now :). I think so, absolutely, if you either "need" or "want to learn more about analog synthesis" because again, the SN-S stuff - even though you have to edit on screen and scroll through some tabs, is worth the price of admission IMHO, especially if you don't have access to any other traditional analog synths. Now, me personally, I like having 2 EXPs - 1 would be almost an insult. The Sequence I needed too. That the FA could act as an interface was another bonus (not sure if the DS does). The dealbreaker would be if you really wanted mappable samples - then the DS becomes the only choice (though I might go for the 76 or 88 key models to pay the same as an 06 or 07). But I think the price difference is justified by the SN Synth alone, and everything else is just a bonus. If not, the SN-S alone is worth 50% and then all the other features combined the other 50% of the price difference.

    But as always, depends on what you really need, so YMMV of course.

    I hope that helps you out.
     
    RMosack likes this.
  5. RMosack

    RMosack Supporting Member

    Messages:
    2,958
    Joined:
    Mar 21, 2003
    Location:
    Oceanside, CA
    @stevel. That was awesome. So much information. Thanks so much.
     
  6. Devnor

    Devnor Member

    Messages:
    3,234
    Joined:
    Jul 17, 2006
    The old Jupiter 80 videos explain supernatural. FA uses the same sources they are implemented in a different way and there are less of them than whats inside the Jupiters and Integra.



     
    RMosack likes this.
  7. anotherscott

    anotherscott Member

    Messages:
    36
    Joined:
    Apr 25, 2010
    VR-730 and FA series both have SuperNatural SYNTH (better emulating an analog synthesizer), the big difference being that synth editing is built in to the FA but uses an iPad app on the VR. The iPad approach is much better but less convenient, and of course requires that you have an iPad.

    VR-730 and FA series both have SuperNatural Tonewheel Organ, but the implementation in the VR is much better.

    The FA also has the following SuperNatural ACOUSTIC tones, which are not in the VR: piano, tine and reed EPs, clav, acoustic and electric bass, acoustic guitar, ensemble strings.

    The DS has numerous other advantages over the FA:

    * as you mention, ability to load custom samples as full keyboard-playable instruments
    * vocal processor (i.e. pitch correction)
    * ability to change performances (equivalent of FA's Studio Sets) without your previous sound cutting out
    * ability to put 2 or 3 insert (MFX) effects on a single sound (but you cannot put MFX on 16 sounds at once like you can on the FA... 3 is the max if you assign one MFX to each. Effects on 16 channels is more important for sequencing than live performance, so it's a difference that makes sense since the FA has the 16 part sequencer and the DS doesn't.)
    * when playing live with a 2-way split, you can easily adjust the volumes and octaves of each part on the fly with the front panel sliders and octave buttons (it's great to be able to quickly switch the octave of your RH sound when playing LH bass, for example... can't do that on the FA)
    * the numeric keypad function repurposes the Favorites/Selection buttons (instead of the pads), so it doesn't get in the way of your using the pads for other purposes. And on the DS, as soon as you finish entering a number, the buttons revert to their previous function, whereas on the FA, once you use the pads for numbers, you have to use a multi-button sequence to manually switch out to get back to your normal pad function. So overall, the numeric entry just works more smoothly.
    * Someone may find the DS basic sound set preferable. Apart from the SuperNatural sounds, the FA sounds are from the XV-5080, while the Juno DS sounds are newer (Fantom and up). Plus axial even gives DS owners a way to add the FA's XV-5080 sounds if you need them, but you can't add the DS/Fantom sounds to the FA.
    * DS has a pattern sequencer (of course FA has its own sequencers, but each is better for different things)
    * Roland has a Mac/PC editor/librarian for DS
    * DS has a front panel MFX control

    Of course, the FA has a whole bunch of advantages over the DS as well. You mentioned some. Another nice one is the sub out which adds flexibility for sending individual sounds to different external amps/processors/mixer channels.

    Also, while it's true that the FA lets you load two of the expansions instead of one, the DS ability to load custom samples means that if you just want a couple of sounds from some second expansion, you may be able to sample them and load them into the DS that way. In fact, you could grab sounds from multiple additional expansions this way... meaning that, unlike the FA, you could actually conceivably have sounds available from more than two of the expansions. (Though there are also significant limitations compared to having the actual expansions loaded.)

    Getting back to the OP: The VR730 is the most limited of the three boards, but also the simplest in operation, by far the best for organ, and probably has the best feeling action. Other moderately priced non-hammer boards with at least 73 keys to consider could be Yamaha MODX7, Kurzweil Artis7, Numa Compact 2X, Korg Krome, each with their own pros and cons vs. the Rolands.
     
    RMosack likes this.
  8. stevel

    stevel Member

    Messages:
    13,255
    Joined:
    Apr 6, 2008
    Location:
    Hampton Roads, Virginia
     
  9. Devnor

    Devnor Member

    Messages:
    3,234
    Joined:
    Jul 17, 2006
    This sounds really awesome until you understand

    A. Grabbing sounds from multiple expansions means sampling, normalizing and editing every single sample you want to use. Then you'll setup the patch TVF, TVA, EFX, ect.
    B. DS only holds one sample at one time. Soon as the patch changes, the new one is not going to sound correct until you menu dive to load in the correct sample from USB

    The reality is nobody works like this on the DS. The DS sampler was intended for beats/loops or 1 shot stuff drums vocals effects. This suggestion on recent threads that the DS has any real sampling capability is misleading to those that don't understand this tech.
     
    RMosack likes this.
  10. stevel

    stevel Member

    Messages:
    13,255
    Joined:
    Apr 6, 2008
    Location:
    Hampton Roads, Virginia
    @Devnor - so let me get this clear.

    Can you, or can you not, do something like record your voice saying "Fame", and then with that one sample, play it up and down the keyboard so it changes pitch?

    That's what "mappable" sampling means to me. The Fantoms did it, the FA does not. I wasn't sure that the DS would but Scott seems to be saying it does. Your post here makes it hard to tell.

    My impression has been that the sampler on the DS was basically the same as on the FA (with fewer pads) which is for "triggering a single sample" like a loop or one shot thing. I think it may have been Scott above who says the update allows "mappable" samples on the DS.

    Even if that's true, and it's only one, that's not so great - it'd be like having to pull out all the tapes and put another set in a Mellotron to get a different sound ;-)
     
  11. anotherscott

    anotherscott Member

    Messages:
    36
    Joined:
    Apr 25, 2010
    No, the FA update added the ability to seamlessly switch among groups of sounds within a Studio Set, but you still can't seamlessly switch from one Studio Set to another.

    You can automate the sampling process using software like SampleRobot, Samplit, Extreme Sample Converter, or the Autosampler in Mainstage. Then you can finesse them in the DS Tone Manager app.

    ...depending on your need to alter what you sampled. But for minimal editing (like attack/release, you can even just use the front panel knobs (as illustrated in the video at the bottom of his post).

    DS is not limited to one sample at a time. And you create saved patches that will automatically use their saved custom samples when you select them.

    I think you have it backwards. You seem to be describing the FA sampler, which is intended to be used as you describe (with the sounds assigned to the pads). The DS pitches a variety of playable samples across the keys, pitch-stretched as needed.

    You can definitely do it if you record your voice saying "Fame" into the computer and bring it to the DS via Tone Manager or on a USB stick. I don't know whether or not you can record your voice directly into the Juno.

    Not at all. They are almost opposite. The FA lets you assign samples to pads and does NOT stretch samples across the keyboard. DS lets you stretch samples across the keyboard and does NOT directly assign them to pads (though the pads can be used to play non-overlapping samples directly from USB, and I think they may also be able to play keyboard-assigned samples by using the function that allows the pads to play notes).

    Here's a video showing where someone sampled the piano sound from a Motif XS and brought it into the DS...



    and here's a video of the older (pre-2.0) version of the DS, before Tone Manager existed, showing that, even then, you could sample a sound, play it across the keyboard, and recall it with a preset (demonstraion starts at about the 7 minute mark). These sounds were done just by bringing in a single wave file and tweaking them only with the front panel controls (illustrated earlier in the video).
     
    strike3 likes this.
  12. stevel

    stevel Member

    Messages:
    13,255
    Joined:
    Apr 6, 2008
    Location:
    Hampton Roads, Virginia
    Thanks for clarifying all this!

    Yep, I checked the FA last night and sure enough, it's only within one Studio Set.

    I should have remembered because it will also reset everything to the stored version if you leave a SS and return - which I sometimes do when I'm experimenting with stuff and get too far in to go back to where I started!

    An earlier SW version didn't even allow that IIRC correctly - called "Tone Remain".

    I find it interesting that there's no input to record a sample into the DS though - but if you're going to work with mappable ones it is better do tweak it all up in software (maybe normalize it, etc.) and then import it.

    Damn I wish they'd give us this on the FA.
     

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice