Six Sines - a New Synth

[baconpaul]

I also don't see the TX81Z waveforms. An additive waveform editor would be nice.

The wave selector is in the screen for each operator

[gentleclockdivider]

Yes they are based on that manual not the scope. although I watched that scope video and it seems i wasn't that far off. They are the same patterns we have in surge basically, with a dc blocker for the asymmetric ones

why , if you want tx1z waveforms you should change waveforms tx2 ,tx4
They are wrong (the slopes of the quarter sine )

[baconpaul]

Thanks!

[baconpaul]

Oh and it’s just 2/4? The rest check out?

If so I can fix them for 1.1 (and leave the old ones as an option for legacy patches with a new name)

[surreal]

Nice!!

[baconpaul]

Yes they are based on that manual not the scope. although I watched that scope video and it seems i wasn't that far off. They are the same patterns we have in surge basically, with a dc blocker for the asymmetric ones

if you want tx1z waveforms you should change waveforms tx2 ,tx4
They are wrong (the slopes of the quarter sine )

OK this was super useful.

I just pushed a fix to this which will be in 1.1. I kept the 1.0 waveforms with name "Spiky TX2" (2, 4, 6 and 8 were impacted) and did patch migration so 1.0 patches using the too-sharp TX2 keep that same sound.

You can grab the 'nightly' (which is really just an 'every commit') where I'm pushing the fixes to 1.1. (There's one I promised Chris, one I promised Kinsey, and a few which are coming in now it is getting more use) https://github.com/baconpaul/six-sines/ ... ag/Nightly - the waveform fix should be there in about 20 minutes - it's building now.

Note that if you use the nightly your 1.0 patches will all work, but if you use nightly features going back to 1.0 won't work, obviously. But at this point it really is the fix to these waveforms.

Thanks for the feedback! Super appreciated.

[gentleclockdivider]

Yes they are based on that manual not the scope. although I watched that scope video and it seems i wasn't that far off. They are the same patterns we have in surge basically, with a dc blocker for the asymmetric ones

if you want tx1z waveforms you should change waveforms tx2 ,tx4
They are wrong (the slopes of the quarter sine )

OK this was super useful.

I just pushed a fix to this which will be in 1.1. I kept the 1.0 waveforms with name "Spiky TX2" (2, 4, 6 and 8 were impacted) and did patch migration so 1.0 patches using the too-sharp TX2 keep that same sound.

You can grab the 'nightly' (which is really just an 'every commit') where I'm pushing the fixes to 1.1. (There's one I promised Chris, one I promised Kinsey, and a few which are coming in now it is getting more use) https://github.com/baconpaul/six-sines/ ... ag/Nightly - the waveform fix should be there in about 20 minutes - it's building now.

Note that if you use the nightly your 1.0 patches will all work, but if you use nightly features going back to 1.0 won't work, obviously. But at this point it really is the fix to these waveforms.

Thanks for the feedback! Super appreciated.

You're welcome and thanks for the hard work ( surge , sines etc.. )
It seems that we can't have feedback loops with mulitple operators
3-->2-->1-->back into 3
It's really a monolithic top-bottom approach

[baconpaul]

It seems that we can't have feedback loops with mulitple operators
3-->2-->1-->back into 3
It's really a monolithic top-bottom approach

yes that's correct.

The thing I didn't share at start of thread was: This entire synth start to finish was a 30 day sprint (that stretched to 40 in beta). So there are some 'hard' things (like 'wide feedback loops') which I didn't implement. (The motivation for the sprint was 4 fold; to un-slump me from a dev slump; to test out some of our libraries; newfangled doing obliterate in a month; and a fourth person who made a joke about fm synths to me on discord).

So the intent with six sines really is: It does what it does and that's what it does. Unlike surge with constant evolution I would be surprised if I ship a six sines 1.4 (but if someone here wants to collaborate on expanding it i'm all ears).

But yes the FM matrix is topologically sorted no cycles.

[baconpaul]

Oh and if you want to see the git log of how it came together, hear's a neat visualization using 'gource'

[IvyBirds]

The thing I didn't share at start of thread was: This entire synth start to finish was a 30 day sprint (that stretched to 40 in beta). So there are some 'hard' things (like 'wide feedback loops') which I didn't implement. (The motivation for the sprint was 4 fold; to un-slump me from a dev slump; to test out some of our libraries; newfangled doing obliterate in a month; and a fourth person who made a joke about fm synths to me on discord).

Thank you for taking this 30 day journey over the holidays and for giving this away for free

We can never have enough FM plugins and the more obtuse the better

[muki]

unfortunately it (AU) does not validate in logic (crashes AU validation)
that's v10.5.1 under OSX mojave

crashlogs are saved in case of interest....

[baconpaul]

unfortunately it (AU) does not validate in logic (crashes AU validation)
that's v10.5.1 under OSX mojave

crashlogs are saved in case of interest....

Would love the logs yeah - I made my test track in logic 11 so I know it can validate but a log would help me find why it didn’t for you thank you (and maybe others)

[edit]: I just installed the version from the 1.0 release with the dmg and it validated fine here (Sonoma). So as well as the crash log if you dont mind can you open a terminal and type the following and send me the output

auval -l | grep "Six Sines"

(i see this: aumu sxSn BcPL - BaconPaul: Six Sines (file:///Library/Audio/Plug-Ins/Components/Six%20Sines.component/) [AUv2])

and then

auval -v aumu sxSn BcPL

and send me the output. One easy way to do that is

auval -v aumu sxSn BcPL | pbcopy

and then you should have the output on your clipboard to paste.

Probably all best as PM tho. And thanks!

[onerob]

Validated fine for me in Logic 10.6.3.

[baconpaul]

Yeah my guess is i'm trying to read documents folder and wasn't careful and muki may not have logic -> documents perms on. A stack would help me figure it out tho

[gentleclockdivider]

Oh and if you want to see the git log of how it came together, hear's a neat visualization using 'gource'

Damn that's cool , I see Evildragon jumps in at the end
I didn't know you gave yourself a 40-day deadline .
Great job