Thorn: Dmitry Sches' new synth!

[EnochLight]

Arrow keys don't work in either.

Well that's it, then. Dmitry clearly doesn't think this is the sort of thing that is necessary in modern virtual instruments, which is troubling.

Thanks for letting me know!

Uninstalling.

[chk071]

It's your decision, but, i'm sure he'd consider it, if you asked for it.

[lnikj]

They work in the sense that you can click to bring up a preset browser, then arrow up and down through a list, press enter and then the preset is loaded. That is, the same as they work in Thorn.

I'm not sure what you are expecting? What behaviour do you want exactly?

"Archaic" synths like Serum and U-He synths behave just like Thorn.

I assume you are not expecting that the arrow keys will be hard mapped to the preset arrows?

[recursive one]

Arrow keys don't work in either.

Dmitry clearly doesn't think this is the sort of thing that is necessary in modern virtual instruments

TBH, neither did I before your posts Actually I hardly see how it may be necessary or useful, I don't even know if any of my synths has this feature. It seems that should this be possible the chances of accidentially switching a preset would be too high.

I'm not much of a preset user though.

[Sound Author]

I am a little curious about how to go about creating custom "frames" for Thorn's spectrum table using WAV files...at least in a way that produces a desired result. Otherwise, importing WAV content feels like a crapshoot, which is fine I guess. But it would be VERY cool to create WAV files intended for very specific purposes.

[BlitBit]

[...]
The manual says that the osicllators have 128 harmonics organized into 16 frames. Whatever it means

If Thorn really uses 128 harmonics to represent a wave then a low tone at 40 Hz would cut out at around 5 kHz.

Here's how to arrive at this result: The fundamental is at 40 Hz. The next partial is at 40 Hz + 40 Hz = 80 Hz. The next partial is at 80 Hz + 40 Hz = 120 Hz. We can do this 128 times because that's the number of partials in this example. So we end up at 40 Hz * 128 = 5120 Hz.

You can also calculate it the other way around. If you want to know how many harmonics you need to fully represent the spectrum of a tone up to 20kHz then you just divide by the fundamental frequency of that tone: 20000 Hz / 40 Hz = 500. So we need 500 harmonics to fully represent a tone of 40 Hz up to 20 kHz. One thing to take away from that is that lower tones need more partials than higher tones in an additive engine. That's also why Harmor sounds fat and bringht even in the low registers because it can use the 516 harmonics to fully represent the spectrum of such a low tone.

However, perhaps the 16 frames can be grouped somehow internally by Thorn's engine. If it's possible to group 4 frames then this will give you 4*128 = 512 partials which would be enough to fully cover a tone with a fundamental at around 39 Hz up to 20 kHz.

In the end it should be easy to check with a spectrum analyzer. What does it show for just one raw oscillator at a low tone, e.g. 40 Hz?

[chk071]

Wow, that's a great explanation even a not so tech savy person like me understands. I'll give it a try with a spectrum analyzer.

[JonSolo]

Superb synth. Will buy. Most of my concerns other people have addressed:

The polyphony really seems too low, especially if you're doing lots of pads/chords with long releases. This synth is already being marketed as an EDM machine, but obviously that's just marketing - it's versatile and so much more than that.

I think this is the second time I have read this here. What on earth are you talking about? If it were 4, or even 8 I would probably agree with you. But 16 voices for a powerful synth like this should be enough. Especially when stacking pad sounds it can break your CPU, heh.

16 voices is just fine for a synth like this.

[audiosabre]

The manual says that the osicllators have 128 harmonics organized into 16 frames. Whatever it means

128 harmonics/partials & 16 frames (comparable to a cycle of a waveset/wavetable). Makes more sense if you open the waveform editor; harmonics in the main box, frames along the bottom:

Thorn_WF_edit.png

I only count 127 though (as seen in 'bin' readout)

[chk071]

In the end it should be easy to check with a spectrum analyzer. What does it show for just one raw oscillator at a low tone, e.g. 40 Hz?

Looks like you were right, there is a frequency cut at 5 kHz. It actually shows some harmonics above that, which i'm not sure where they originate from, see the picture below.



... and here's a saw with the same frequency in Spire, for comparison:



I must admit that this is quite a bummer for me, as it makes Thorn quite hard to use for sounds in the lower registers. I can see where the technical limitations come from, but, it'd be really great if there was a less severe cut for those sounds.

[lnikj]

The manual says that the osicllators have 128 harmonics organized into 16 frames. Whatever it means

128 harmonics/partials & 16 frames (comparable to a cycle of a waveset/wavetable). Makes more sense if you open the waveform editor; harmonics in the main box, frames along the bottom:

Thorn_WF_edit.png

I only count 127 though (as seen in 'bin' readout)

It says in the manual that the first is fixed and cannot be edited.

[recursive one]

The manual says that the osicllators have 128 harmonics organized into 16 frames. Whatever it means

128 harmonics/partials & 16 frames (comparable to a cycle of a waveset/wavetable).

But then does it mean that when you press a key the synth produces only 128 harmonics? If we assume that a frame is comparable to an individual position in a wavetable (which seems to be somehow true as there is a POS knob which morphs between the adjacent frames) shouldn't this mean that when you press a key only one of the 16 frames produces a sound?

And the chk071's pics seem to confirm this. It this a clean saw? (i.e. are the filter and whatever comes after it disabled?)

I'm not at my DAW right now. Definitely should dive deeper into this and also educate myself about what spectral synthesis actually is.

[chk071]

And the chk071's pics seem to confirm this. It this a clean saw? (i.e. are the filter and whatever comes after it disabled?)

Yep. I completely disabled the filters, harmonic filters, boost, and limiter. Should be just the pure oscillator's signal.

[S0lo]

In the end it should be easy to check with a spectrum analyzer. What does it show for just one raw oscillator at a low tone, e.g. 40 Hz?

Looks like you were right, there is a frequency cut at 5 kHz. It actually shows some harmonics above that, which i'm not sure where they originate from, see the picture below.



... and here's a saw with the same frequency in Spire, for comparison:



I must admit that this is quite a bummer for me, as it makes Thorn quite hard to use for sounds in the lower registers. I can see where the technical limitations come from, but, it'd be really great if there was a less severe cut for those sounds.

I'm not sure I see this as a limitation, more rather a flavor. Diversion already gives you full spectrum saw tooth (just like spire). So it's reasonable to think that Dmitry doesn't necessarily wana give you the same thing again. Besides, such low harmonic content is easier to mix in with the highs without further EQing, specially when doing base.

[chk071]

Sounds very lo-fi though.