Recreating the D-50's "Soundtrack" patch

[desmond]

Yes yes I know, I don't care whether you hate it or not

I really like this patch despite it being used on everything, and I'm trying to recreate this as closely as I can (on an XV5080 principally, but possibly also on softsynths as well, depending on how "Roland-ey" it is).

If anyone can help out, I'd be grateful. I'm after the D-50 sysex data for the patch (so I can load it into an editor as see the parameters).

Also, if any of you D-50 owners can do some small recordings of each tone solo'd, and recordings of the raw PCM samples used (not for sampling purposes, for matching the waves as close as possible on the pool of waves available on the XV), that would be super-helpful.

I don't think the samples are particularly critical (it's not like DigitalNativeDance or anything) but it would be helpful to hear the raw tones.

(Yes, the XV5080 has a "Soundtraque" patch which has some similar characteristics - a slow 5th interval pad) but it's very different in tone.)

Cheers if you can help. I'll happily make the patch available if I can get somewhere close...

[Yoozer]

http://www.synthmania.com/d-50.htm

Sounds highpass-filtered to me, but you'd probably need a pre-(lowpass)-filtered sawtooth.

[Bitexion]

There is only a lowpass filter in the D-50, so you can strike that immediately.

[Synthaholic]

I did a Soundtrack-like patch on my Microkorg. A PWM wave and a triangle wave tuned to a fifth, with a lowpass resonant filter sweep, a slow attack on the amp EG and a slight pitch modulation via LFO.

I don't recall Soundtrack using any PCM waves on the D-50, just synth waveforms (which are either saw or pulse/PWM).

[Bitexion]

The reason it may be difficult to get the timbre right on other synths, is that the D50 sawtooth isn't really a sawtooth at all when you look at it in a scope. It looks more like a rounded shark fin.

[desmond]

http://www.synthmania.com/d-50.htm

Sounds highpass-filtered to me, but you'd probably need a pre-(lowpass)-filtered sawtooth.

I'm aware of audio demos of the complete patch - I was hoping to get the sysex data, and recordings of the individual tone and the individual waves, to cut down on some of the guesswork...

[desmond]

I don't recall Soundtrack using any PCM waves on the D-50, just synth waveforms (which are either saw or pulse/PWM).

Yeah, that's what I thought, which is why recreating the patch should be doable.

Again, if some D-50 owner could list the waves used (and better still do a quick raw recording of them), or provide the sysex info, that would be super-helpful...

[Paolo Di Nicolantonio]

It's not an easy task... as usual with a lot of D-50 patches, it uses 4 partials, so it can be complex to recreate perfectly, but I'll write down the most important parameters and most of all - provide audio examples (single note at useful range for this patch - C3, C4, C5, C6) - so you can try to emulate it on another synth.

Keep in mind that this is just an approximation, because there are several other parameters that would take forever to analyze... for instance, the envelopes of Lower and Upper Partial 1 change dramatically by velocity release... Pitch envelope and pitch modulation... slightly different tunings per tone... there are different LFO rates... there are EQ and Chorus settings... there is slight filter aftertouch etc. etc... AND there is the on-board reverb processor, which modern processors or plug-ins are way too hi-fi to reproduce... I'd suggest getting a $50 older unit such as the Yamaha REX50 - You'll have to 'eyeball' these parameters by ear as best as you can - and you really are going to need to see the patch on an editor to see what's really going on. Anyway, here it is - To help a bit more while synthesizing, I recorded dry and reverbed versions.

Patch: 37 Soundtrack (without reverb: Soundtrack no rvb)

Key Mode: Dual
Split: C4
Output mode: 1 (Upper and Lower in stereo passing through reverb)

Reverb balance: 63%

Upper Tone name: MelloTones
Lower Tone name: BriteFifth

Balance (between Lower and Upper tones): 50%

Lower Tone Structure: 1 (S + S) (synthesis only)
Lower Tone Balance: 70

Upper Tone Structure: 1 (S + S) (synthesis only)
Upper Tone Balance: 70

Lower Tone Partial 1 waveform: Square (these are a fifth apart)
Lower Tone Partial 2 waveform: Square

Upper Tone Partial 1 waveform: Square
Upper Tone Partial 2 waveform: Square

Let me know if this helps...

[desmond]

Paolo - excellent, thank you for your info and detailed efforts - they are much appreciated!

That should get me started nicely. If anyone can also do a sysex dump of the patch, that would also be helpful - I've found afew banks on the net but they are all in obscure formats or unreadable...

Cool beans...

> I'd suggest getting a $50 older unit such as the Yamaha REX50

Actually, I *do* have an SPX50, which is a REX50 in a rack

Still, I should be able to get close to the essence of the patch in the XV - it will be interesting to see whether the increased fidelity in the XV makes the sound even better, or whether I have to grunge it up a little to be a bit more old-school...

[desmond]

Interestingly, the XV contains as part of it's ROM wave quite a few waves from other Roland gear. It has three variations of the D-50 sawtooth, which *is* in fact much more grungy than the more hifi waves - much less top end etc.

It doesn't unfortunately have the D-50 square wave, so I might need to use a regular square and try and "de-fi" it somehow.

Paolo - by any chance could you do an mp3 of the D-50 square wave on it's own? That would be helpful too...

[Synthaholic]

AFAIK there's nothing "lo-fi" about the D-50's square/pulse, it's a true representation of a square. What dirties it up is the low-pass filter which just is a digital alteration of the waveform as it leaves the oscillator.

The saw on the D-50 is unusual because it's created by some sort of internal algorithm that combines the square/pulse with a sine wave. This, along with the filter gives the D-50 synth tones its unique character.

[desmond]

Ok cool. Yeah, perhaps the D-50 saws in the XV simply have the filter closed a little during the sampling.

Yeah - I noticed in SoundDiver the D-50 saw has a kind of pulse width setting to vary the saw waveshape output.

In any case, thanks for the info about the square wave... good stuff!

[Mixolydian]

Paolo Di Nicolantonio's comments are excellent.

This is sort of a general response, but the D-50 uses two different sound generation methods: the attack which is a PCM sample, and the decay which is LAS generated (a non-sample based digital circuit.) By combining two synths, one with bright attack, and one with a soft or pad sound, one can create D-50 type sounds. The Soundtrack patch may be impossible to duplicate exactly because, as Paolo stated, it uses 4 very unique partials.

Also, wasn't the Soundtrack patch on the D-50 a copy of the same patch on the Roland JX-10? I wonder if Roland sampled their own synth. Anyway, it was on the D-50 where it came to fame.

Edit: I just remembered, the Roland MT-32 has a fairly decent Soundtrack patch. It uses the same sound generation methods as the D-50, although a budget version. MT-32s sell cheap on ebay - around $40.

[desmond]

By combining two synths, one with bright attack, and one with a soft or pad sound, one can create D-50 type sounds. The Soundtrack patch may be impossible to duplicate exactly because, as Paolo stated, it uses 4 very unique partials.

Well kind of. The D-50 basically has a variety of "structures" which determine which tones are PCM and which are synthesis in a given patch. So you could have all four partials as PCM only, all four as synthesis only, or various other combinations thereof.

In this particular patch, it's basically a layer of four synthesis tones, all using a square wave. So it doesn't really require any features unique to the D-50. Hence why I'm trying...

Also, wasn't the Soundtrack patch on the D-50 a copy of the same patch on the Roland JX-10? I wonder if Roland sampled their own synth. Anyway, it was on the D-50 where is came to fame

No, the JX8P/10 soundtrack was a warm pad I'm prety sure (and yes, there are variants of this type of patch of virtually all synths).

The D-50 soundtrack was a fifths pad but with a very unique texture to it, so it's a different kind of patch entirely...

[Paolo Di Nicolantonio]

Interestingly, the XV contains as part of it's ROM wave quite a few waves from other Roland gear. It has three variations of the D-50 sawtooth, which *is* in fact much more grungy than the more hifi waves - much less top end etc.

It doesn't unfortunately have the D-50 square wave, so I might need to use a regular square and try and "de-fi" it somehow.

Paolo - by any chance could you do an mp3 of the D-50 square wave on it's own? That would be helpful too...

Here you go, Desmond:

Roland D-50 SQUARE waveform (first at 60% TVF {initialized}, then 100%)