Pollyanna posted:

For example, for a bass guitar, knowing that there's a beginning percussive "chk" part when the bass is first struck, then a relatively quick decaying high-resonance part (what is resonance?), then a long-term sustaining resonant part that stays at roughly the same level, then finally an ending part that decays quickly from medium resonance to silence. (Interestingly enough, his assessment matches up with the ADSR of a bass guitar. hmmmm )

Pollyanna posted:

I am so stoked right now, you have no idea.

RYMCast lets you rip any patch from a Sega Genesis track. So naturally, I had to pull the bass patch from Green Hill Zone. It's one of my favorite bass instruments ever. It has just the right mix of bassiness, brightness, and punch, and I love it to death.

When I ripped it and tried it out, though, it sounded...wrong. It felt muddy and muted, and I wasn't sure what went wrong. I tried adding a high pass filter and all sorts of fuckery with a 5 EQ, but nothing got close, and I still had no idea what the difference actually was.

So I clipped the ripped patch and a snippet from Green Hill Zone in Audacity, and compared their spectrums.

Mine:


Theirs:


That didn't make any sense. This was the exact patch used in Green Hill Zone, and it sounded perfectly normal in RYMCast (and I've certainly never emulated the game before but I'm told it's normal there too). How come it sounded different when I used it?

Looking at the spectrum, I noticed that there was more emphasis on frequencies around 2kHz in the original patch, and more around 1kHz in the ripped patch. I was reading FM Synthesis of Real Instruments earlier today, and according to that text, if frequencies other than the fundamental harmonic show up in the spectrum, that's because a modulator put them there. Considering that the patch sounded muddy and not bright or "spiky" enough, what if the problem was with one of the modulator's ratios?

Every ratio was set to 1.0 except for the topmost operator, the one with feedback. That one was set to 10. I bumped it up to 15, and wouldn't you know, it started sounding very familiar. Spectrum analysis backed me up:

Mine, higher ratio:


Theirs:


But it wasn't quite there. Clearly, the ratio was somehow loaded incorrectly when importing the patch, so just bump up the ratio for more brightness, right? Problem is, the maximum ratio on the YM2612 is 15 - you can't go any higher. But there's got to be a way to fix it, and here I got curious. Since this is a matter of ratios, what are they supposed to be?

I would be hosed if it wasn't for the fact that RYMCast's patches are just plain XML, thank god. That let me check what the ratios are in the file, and:

code:

OP1 MULTIPLIER: 0.666667
OP2 MULTIPLIER: 0.0333333
OP3 MULTIPLIER: 0.0333333
OP4 MULTIPLIER: 0.0333333

Dividing 1 by 0.0333333, we get 30. Multiplying all these numbers by 30, we get our expected ratios:

code:

OP1 MULTIPLIER: 20
OP2 MULTIPLIER: 1
OP3 MULTIPLIER: 1
OP4 MULTIPLIER: 1

Waaaaait a minute.

Yup, RYM2612 doesn't properly interpret ratios from its own patch format! The operators are supposed have a ratio of 1:1:1:20, but it gets imported as 1:1:1:10 instead. I have no idea why.

So, I need to get an effective ratio of 1:1:1:20 while still working within the 15-ratio maximum. That's simple: just drop it by half instead, and set a ratio of 0.5:0.5:0.5:10.

Mine, corrected:


Theirs:




This patch has been my white whale for years. I've always wanted to play around with it. And now, I have it. _{and I spent way too much effort trying to get it}

Here's a comparison of the different patches: original, raw ripped, 1:1:1:15, 0.5:0.5:0.5:10.

Pollyanna posted:

Intro synth question snipped

Pollyanna posted:

It's pretty good, actually. Video Guy's advice pretty much breaks down to the following:

1. Identify the different parts, which he calls "layers", of the sound you're trying to make. This benefits from becoming intimately familiar with the sound of whatever instrument you're trying to replicate. For example, for a bass guitar, knowing that there's a beginning percussive "chk" part when the bass is first struck, then a relatively quick decaying high-resonance part (what is resonance?), then a long-term sustaining resonant part that stays at roughly the same level, then finally an ending part that decays quickly from medium resonance to silence. (Interestingly enough, his assessment matches up with the ADSR of a bass guitar. hmmmm )

2. Come up with your own algorithm based off of the parts you identified. For example, for a bass guitar, you'd need to combine at least four oscillators:

• The first, "chk" part would need to be percussive in some way, so that suggests at least one modulator and one carrier, with the modulator getting feedback. (Don't ask me how I know this, Video Guy just said that's the case for percussion.)
  
• The second, decaying resonant part would need to be resonant with some brightness, which Video Guy says also calls for feedback. (I disagree, you can totally get brightness without feedback.) That also implies one modulator and one carrier, with the modulator optionally getting feedback.
  
• The third, long-lasting resonant part is pretty constant and takes the form of any typical bass patch, so anything other than a simple sine wave is enough. One modulator, one carrier.
  
• The final, quickly-decaying sound apparently doesn't need to be anything more complex than a sine wave, which I don't understand - isn't it meant to go from somewhat resonant to not-resonant? That would imply a resonance matching part 3. Am I making it too complicated? Regardless, he says it only needs one carrier.
  

Which results in this algorithm:



3. Pare your algorithm down to something reasonable. That algorithm above, for example, isn't available on a 6op let alone a 4op. So, you'll have to condense it somehow. I'm not sure how to do this, myself, but Video Guy does this:

• Part 1 ("chk") and part 2 are nearly identical, so they can be combined if you're willing to meet halfway on the desired feedback. So, combine "chk" and part 2 into one stack. That brings us to 5op, which is doable on the DX7 but not the DX11 or OPN line.
  
• Part 4 is very simple, being characterized entirely by a quick decay on a static, non-evolving sound. It doesn't actually add much to the sound other than it representing its quick decay once you let go of the key. This can be rolled into part 3 by just defining a quick release decay on part 3. That brings us to 4op, which basically any FM synth can handle.
  

Which results in Algorithm 5, which I'm sure you've seen before:



And indeed, if you load up Algorithm 5 and set all the operators to an initial state, it does sound bass-y right out of the box. It seems like a good fit!

So now we have one percussive, highly resonant 2op, and one moderately resonant, quick-release 2op. Easy enough.

4. Determine the envelopes for your modulators and carriers. The envelopes for the carriers are straightforward, since they characterize the volume of the sound:

• Carrier 1 (op1) needs to be punchy and therefore both immediately come online and reach max, and quickly reach a moderate resonance. It also needs to remain at roughly the same level before quickly silencing on key-off. op1 gets a quick attack rate, a relatively quick decay rate, a moderate sustain level, and a quick release rate.
  
• Carrier 2 (op3), due to its role as both a percussive sound (old part 1) and a long-lived moderate resonance (old part 2), will need to evolve in an identical manner to op1 over time. op3 gets the same envelope as op1.
  

...but I'm less sure about the envelopes for the modulators, which characterize the evolution of the resonance of each part over time.

At this point, we'd have to go back and think about the sound of a bass guitar again. If we listen to it, it starts from dull (for a split second), then quickly reaches spikiness, then mellows out over a period of time before completely going away once the sound fades out. I like to do something silly and actually try and replicate the sound with my own voice, and a bass guitar sounds something like "puh-ehhh-ahhhhhhhh-ooo".

So it's kind of like the carrier waves' envelope, with a slightly longer attack rate, longer decay rate for sure, and what appears to be a similar release rate. Video Guy concurs, but extends the release rate a little in comparison.

• Modulator 1 (op2) needs to be somewhat punchy and therefore both come online relatively quickly and reach max, then moderately taper off to a moderate resonance. It also needs to remain at roughly the same level before somewhat quickly silencing on key-off. op2 gets a somewhat quick attack rate, a moderate decay rate, a moderate sustain level, and a somewhat quick release rate.
  
• Nothing about Modulator 2 (op4) suggests that it should behave any differently - at least, as far as I can tell - so we can just use the same envelope as op2.
  

Which leaves us with these envelopes:



5. Figure out your feedback (on op4 in our case). I have no idea what feedback even does or why we use it, so I'll just defer to Video Guy here. In the case of the bass patch, the feedback will provide a percussive element to stack 2. We want a decent amount of percussion for the "chk", but not so much that we lose out on all the resonance after it in stack 2. Video Guy splits the difference and sticks feedback at something like 50%.

6. Finetune. Now you can sit down at your synth and press buttons.

This step differs from patch maker to patch maker, and it's where you get to be picky about envelopes and levels and such. You're basically improvising this step. The patch by this point is pretty good already, and just needs extra bells and whistles and personal touches. This is where having a clear mental picture of the sound you're looking for, as well as a clear understanding of FM synth basics, becomes particularly important.

With your algorithms, envelopes, and feedback programmed, tweak:

• Frequency ratios (think about the harmonics of the sound and how you want it to look on the spectrum view, then explore )
  
• Detunes (tune each operator slightly differently to try and even out the coverage of the harmonic range - this is where the spectrum view is REALLY handy)
  
• Total levels (carriers should be straightforward but the modulators are a mystery to me, sorry)
  
• Envelope parameters (finicky bullshit)
  
• Whatever else you feel like (as long as it doesn't deviate too far from the plan)
  

---

And then, you'll have a decent-sounding patch! I think.

Some questions I still have:

- What exactly is "resonance" in this case? How do we characterize it?
- What exactly does detuning do? IME it seems to even out the harmonic range from a few high peaks to a lot of moderately sized peaks, and is what provides "full" sound.
- How do total levels matter on modulators, and what am I trying to achieve with them?

Agreed posted:

I freaking wish there was. You can make your own from a touchpad but I didn't, I just use the Nanopad 2. To use it as an X/Y I MIDI-assign the X and Y separately to the appropriate parameters, I don't know if there's an easier way than that but it works great for X/Y pad usage this way in whatever application I've wanted to use it for.

McCoy Pauley posted:

Are there any options for a midi USB XY pad to use with Ableton, I guess like the Korg Nanopad, but just the XY pad part? There seems to be the Nanopad, and stuff like the Kaosspad, and I guess some midi controllers with XY joysticks, but does anybody make something that is just an XY touchpad with midi over USB?

If there's nothing like that, anybody have hands experience with the Nanopad2? Does the XY touchpad work well in Ableton?

petit choux posted:

Been really enjoying this NTS-1 a lot over the past couple days since I assembled it, boy is it a hoot. I've tried using the librarian but it wasn't working, but it was with an old Windows 7 computer that has problems with its USB connectors anyway, so I'm holding out hope that when I actually connect it to a Windows 10 machine it will properly work. As much as I've enjoyed it it won't be right without that.

But I've been Floyding out over here bigtime, just with this tiny box.

Martytoof posted:

I paid mpcstuff a bunch of bucks for the upgraded pad stack and it’s SO MUCH BETTER. It feels like my MPC One in feedback now, tbh, which I always liked. Also replaced a bunch of the sloppy plastic buttons and now everything feels super tight.

Since it’s October I thought it would be fun to go with a spoopy halloween themed setup but it’s not really jiving with me as much as I thought. Whenever I can justify spending a few more bucks I’ll grab something less “wacky”, or maybe I’ll forget about it and just stick with what I have, who knows.

Did a DIY upgrade to the tired green LCD screen too. Didn’t pay mpcstuff for this but ordered a screen from china for thirty bucks and made the cable myself. It was super easy and the difference is striking. Light night and day. Feels like a brand new machine.




Anyway, hardware swaps aside this thing is SO MUCH FUN. It’s EXACTLY what I needed to get my workflow in order. I’ve been experimenting with tracks so much more than I ever did with the One, mainly because there’s not a metric ton of stuff for me to menu dive. It’s not lightweight or basic, but it doesn’t feel like everything I touch is leading me astray into a deep dive of options or possibilities. That’s not a bad thing per se, unless your personality is the kind that will take that detour and just stop making music for the next two hours.

I want to say it’s perfect. Even the fact that it’s CF and USB1 is turning out to be good. It’s keeping me from loading up a hundred sample packs and idly flipping through them, and making me work with a more limited subset, both because I can’t fit a ton on a 2gb card and because it takes a year and a day to load them over USB (and I can’t find my CF adapter at the moment).

I dunno, all around this has to be one of the best investments I made in my creativity in a long time.

Google Butt posted:

After you spend some more time with it, you should return to the one and see if it's changed your workflow to the point that the one makes more sense

Martytoof posted:

So Math posted:

There's the ROLI Lightpad. I've even written some apps for it. Be aware that the company is a startup; they are rebranding focusing on more beginner hardware instead.

B33rChiller posted:

Here's the reminder to install the korg driver, have as little else plugged in to USB as possible, and to plug the nts1 directly into your computer (avoid hubs).
Both going through a hub, and having a midi keyboard plugged in caused me issues separately while trying to use the librarian.
Hopefully that works for you. They're great little boxes, even just for the fx.

A MIRACLE posted:

I think I’m going to re buy the TR-8

I really miss it. The gated reverb and side chain effects are really good and I need an 808 again

echinopsis posted:

what do people think about the digitone. I am extremely intrigued by it.