Do you use the Grid? (Grid discussion)

[slo_poe]

I think the grid is very clever and very well implemented, but I don’t use it…I just can’t seem to find a purpose for it. With Max at least people who are far clever than me can make things that are stand alone devices with GUIs etc that I can just use, but with the grid (at least at this stage) it’s a bit more DIY.

These are my exact thoughts about the Grid as well. It's a pretty cool tool and I enjoy seeing what people can create with it, but I've personally just never really found it to be particularly useful or fun/inspiring in comparison to my other plugins. I find it to be a bit inefficient for my needs and overall just not that enjoyable if I'm honest.

That said, I could see myself messing around with it a lot more if Bitwig expanded it to the point where people can make & share entire devices with customizable UIs and so on like mentioned here and in other discussions around the web.

[lokomotiv]

I think the grid is very clever and very well implemented, but I don’t use it…I just can’t seem to find a purpose for it. With Max at least people who are far clever than me can make things that are stand alone devices with GUIs etc that I can just use, but with the grid (at least at this stage) it’s a bit more DIY.

I think it depends on how you approach music. I generally already know what sound I want and then try to create it with a synth. And if a regular synth can't do what I'm looking for, I jump to Grid. Like I was trying to do wavetable FM in Zebra, but you can't use OSCs to modulate each other and the FMOs are very limited in wave shape. In Grid it was easy.

I do agree that being able to create GUIs and save them as separate devices would be great, like we already have Polymer and Filter+. For some devices remotes or macros can be enough, but oftentimes not.

[pdxindy]

I think the grid is very clever and very well implemented, but I don’t use it…I just can’t seem to find a purpose for it. With Max at least people who are far clever than me can make things that are stand alone devices with GUIs etc that I can just use, but with the grid (at least at this stage) it’s a bit more DIY.

I think it depends on how you approach music. I generally already know what sound I want and then try to create it with a synth. And if a regular synth can't do what I'm looking for, I jump to Grid. Like I was trying to do wavetable FM in Zebra, but you can't use OSCs to modulate each other and the FMOs are very limited in wave shape. In Grid it was easy.

I do agree that being able to create GUIs and save them as separate devices would be great, like we already have Polymer and Filter+. For some devices remotes or macros can be enough, but oftentimes not.

I enjoy modular, but not the usual tedious workflow in software. The Grid is easily the fastest software modular I've tried. It's almost as fast as a regular fixed architecture synth.

I think the tendency for Grid videos to be massive generative structures turns some people off and makes it seem complex and esoteric. I never use it that way myself. Most of my Grid patches are simple structures but which allow me a flexibility very few synths have. Even a basic sample based patch becomes something different when I can use the Segments MSEG as the Amp Env. It's one of the best software synths available today.

The Grid is also constantly useful for utility purposes. I can use it add a pressure expression curve to synths that don't have it, create a different kind of latch device, transform CC data, etc., etc.

[Hez]

Like some others in this thread I use the Grid extensively, on pretty much every single track in every project. Sometimes as a sound source, sometimes as a note source or note processor, occasionally as a pseudo-modulator (by nesting devices in the pre or post FX slots) and very frequently as a mixing/routing utility.

Since the introduction of the Grid I've noticed that my projects have started to contain fewer individual tracks, as I can achieve a lot more 'stuff' on a per-track basis, or by creating cool & creative interactions between tracks.

In some sense for me the Grid isn't a distinct 'feature', it's more like another view (or mode) in the way that we have Mix/Clip/Arrange views etc. It's a distillation of my favourite aspects of Bitwig into a slightly more sandboxed environment than the other views, which are more akin to traditional DAW. I would love to see 'Modulator Grid' (build your own modulators, save them, drop them in any modulator slot) and 'Project Grid' (a sort of Master track Grid which has access to all track/group modulators, volumes, mutes etc.) as extensions of this philosophy.

[foosnark]

So here's an example of the kind of thing I get up to in Grid sometimes, how it evolves from something simple to something that... is still pretty simple when you look at it piece by piece.

In another thread someone was asking about a delay with different times on the L and R channels:

patch1.jpg

But delays should have feedback.

patch2.jpg

DIGRESSION: many Grid modules don't allow this directly -- Long Delay is an exception. If we wanted to use the regular delay module we'd have to work around it using Modulator Out assigned to a knob of some kind. This isn't my favorite aspect of the Grid, but it works.

sidequest.jpg

(Edit: looking at this the next day, I realize I could have simplified and just used the one merge off the two Bias knobs for the output too. Oh well.)


Back to the main story...

What if the repeating delays moved toward the center for a little more interest in the stereo field?

patch3.jpg

The material I'm using this with is pretty bright, maybe the echoes should be filtered. I tried a couple of different filter options, and tried placing it in different places. I settled on XP since it combines high and lowpass, and put it where the first echo and all the feedback repeats are routed through it:

patch4.jpg

It already sounds pretty cool. But let's go deeper. The notes coming in have quite a bit of sustain... what if I were to apply an envelope to just the echoes to make each repeat more distinct?


Let's say we can't assume there is MIDI associated with the incoming audio, so we have to extract a trigger from the input level to trigger an envelope. We want to make sure the envelope only applies to the incoming audio, not to our repeats:

patch5.jpg

Maybe we want the envelope to modulate the filter cutoff while we're at it. (That probably means adjusting our feedback level, which we could do via the filter drive or the feedback mixer or both.)

But also, maybe we want to restrict how often the envelope is triggered. Because the Pluck envelope doesn't care about the length of the gate, just the rising edge, we could have our detector hold the gate high until something else resets it. Let's use a gate sequencer for that for a little rhythmic interest.

patch6.jpg

Now let's get slightly ridiculous: let's say we want to alternate from this plucky envelope, to a tempo-synced rising ramp. On every other trigger, we will switch between modulation sources, and then use the selected source to control both the input volume and the filter.

For our rising ramp we can use the Transport phase output. And since we're not relying on the audio path in Pluck anymore, we'll need a VCA -- the Multiply module under Math is fine for this.

patch7.jpg

And that's usually how "big" patches happen. (The same is true in hardware modular.) The latch thing and the switching are a bit more on the advanced side, and using phase in Bitwig is kind of a unique thing (most modular clocking happens with trigger pulses). But still, the patch can be broken down into distinct sections that each serve a particular role. If you just look at the complete patch it looks like a much bigger mess than it really is.

[Trancit]

nevermind

[xbitz]

...

And that's usually how "big" patches happen. (The same is true in hardware modular.) The latch thing and the switching are a bit more on the advanced side, and using phase in Bitwig is kind of a unique thing (most modular clocking happens with trigger pulses). But still, the patch can be broken down into distinct sections that each serve a particular role. If you just look at the complete patch it looks like a much bigger mess than it really is.

Polarity's autolever is also a very good example of the power of the Grid

https://www.youtube.com/watch?v=MsWodF8 ... j&index=10

[Alexander137]

Tell me, please: FM synthesizers have a feedback parameter on each operator. How to implement this in grid? I can't achieve the same effect as in FM4 or Fem.

[foosnark]

Tell me, please: FM synthesizers have a feedback parameter on each operator. How to implement this in grid? I can't achieve the same effect as in FM4 or Fem.

Most modules in the Grid don't allow feedback. But you can work around it using a Modulator Out linked to a Bias knob:

grid_feedback.jpg

If you wanted another operator to affect this one, you'd connect it into the input on Bias.

fm2.jpg

[Alexander137]

Thank you )
The effect is not exactly the same, but it is close.

[BobDog]

I'm guessing changing the buffer size to as low as possible might get you a bit closer.

[duanosforde]

Like some others in this thread I use the Grid extensively, on pretty much every single track in every project. Sometimes as a sound source, sometimes as a note source or note processor, occasionally as a pseudo-modulator (by nesting devices in the pre or post FX slots) and very frequently as a mixing/routing utility.

Since the introduction of the Grid I've noticed that my projects have started to contain fewer individual tracks, as I can achieve a lot more 'stuff' on a per-track basis, or by creating cool & creative interactions between tracks.

In some sense for me the Grid isn't a distinct 'feature', it's more like another view (or mode) in the way that we have Mix/Clip/Arrange views etc. It's a distillation of my favourite aspects of Bitwig into a slightly more sandboxed environment than the other views, which are more akin to traditional DAW. I would love to see 'Modulator Grid' (build your own modulators, save them, drop them in any modulator slot) and 'Project Grid' (a sort of Master track Grid which has access to all track/group modulators, volumes, mutes etc.) as extensions of this philosophy.

This is a great explanation, i've never thought of it that way. I should try the grid more.

[Tj Shredder]

Tell me, please: FM synthesizers have a feedback parameter on each operator. How to implement this in grid? I can't achieve the same effect as in FM4 or Fem.

The Phase-1 oscillator has that built in. The feedback over several operators need to be done with a one block delay like explained. Often people say FM and PM is different, but the result is the same. The original DX7 had PM. The difference you hear when comparing FM of a DX7 with FM of a modular is due to different curves for the modulation inputs. In modulars its V/oct, which is exponential, in a DX7 its linear…
I made a little collection of all 32 DX7 algorithms for the Grid. When you rebuild sounds of a DX with it you can get pretty close when tuning it by ear…
Unfortunately we still get no envelopes like you have them in a DX...

GriDX7.bwproject.zip

I have also thrown in the free Dexed into the project to compare...
Btw. the new Bite oscillator does exponential FM internally...

[foosnark]

The difference you hear when comparing FM of a DX7 with FM of a modular is due to different curves for the modulation inputs. In modulars its V/oct, which is exponential, in a DX7 its linear…

Without writing several pages about it:

With PM, the depth of modulation can be varied dynamically with no consequences to pitch.

With thru-zero linear FM (TZFM), the above is true IF the modulation signal is balanced positively and negatively (no DC offset).

With standard linear FM, the above is true only within a limited range of modulation depth. If the depth is greater than the core carrier frequency, the oscillator "stalls" at 0Hz during negative modulation, so the overall pitch shifts upward as the depth increases. (TZFM solves this by reversing the oscillator's direction to allow "negative frequencies".)

And with exponential FM, increasing the modulation depth also increases the resulting frequency. This is because it's shifting pitch rather than frequency -- one octave up is double the frequency but one octave down is half; if you average 2.0 and 0.5 you get 1.25, not 1.0. So you can't use this method for DX-style FM.

Generally, digital oscillators can do any of them. Dr. John Chowning worked with TZFM on computers without ever really thinking about it. Yamaha implemented it as phase modulation because it's simpler to implement. You'll see a mix of PM and TZFM in digital Eurorack oscillators without necessarily distinguishing between them, and a few which do both.

For analog oscillators, TZFM used to be a relative rarity, requiring some tricky calibration and with quite different behavior (turning the oscillator into almost more of a waveshaper, with the output frequency determined by the modulation signal). But in the last couple of years the SSI2130 chip was introduced and at least a dozen manufacturers have released analog TZFM VCOs that use it, and it behaves pretty much like a digital oscillator.

Phase modulation in analog is fun and is almost exactly what Yamaha did with the DX7. Take the core oscillator's sawtooth or triangle wave output, mix in the modulation signal with it (wrapping it around when it overflows), and then feed that to a saw-to-sine waveshaper. Presto, phase modulation.

[j wazza]

Phase modulation in analog is fun and is almost exactly what Yamaha did with the DX7. Take the core oscillator's sawtooth or triangle wave output, mix in the modulation signal with it (wrapping it around when it overflows), and then feed that to a saw-to-sine waveshaper. Presto, phase modulation.

I havent heard of this type of phase modulation before by mixing and sending to a waveshaper. I thought phase modulation was only possible digitally, and i thought the dx7 was digital. It also seems like if the modulator and carrier are the same and you mix them before the waveshaper, this would have the same sound as just the carrier into a waveshaper, but twice as loud. Its interesting though, do you have any links about this?