Cytomic "The Scream" stomp box distortion plugin

[andy-cytomic]

A fuzz would be a lot of fun.

Okay, you know that weird 1.0.10 crash bug that was in windows 7? I think it's on up-to-date Macintoshes as well.

Mac OSX Mojave 10.14.2
Reaper 5.965
The Scream 1.0.10

Opening the "mods" pop-up in either the AU or the VST version will crash reaper instantly, close to 100% of the time. If it doesn't crash it, just twiddle a few controls or type a number into one of the full mod text boxes and that'll do it. 1.0.8 doesn't do this at all.

https://pastebin.com/VmNx8Gkq Here's the crash log.

I recommend sticking to the officially supported v1.0.8 release.

[sleepcircle]

Oh no I most definitely am using 1.0.8 for all my stuff. I just wanted to see what was up with 1.0.10, and when it crashed, I thought maybe the fact that it crashed on OSX as well might help you triangulate the original problem a bit better. Cause I thought you said originally it only crashed on windows 7, so I thought to myself, 'hey, here's something else it crashes on!'

if it doesn't actually help at all--like if you've sidestepped that problem entirely in the interim, then i'm sorry for confusing the issue.

[andy-cytomic]

News update: Progress is going well on the final release. I've been working on the warmup code and have re-visited the whole process using my new solver and I'm happy to report that it now starts up with zero thump I measured this on Ableton's meters and it shows -inf when I start up The Scream with a variety of settings. This is up and running for the full solver, and it's now time to get it working for the fixed CPU one and continue and get back to doing the randomisation panel.

[sqigls]

i have NO idea what you just said Andy, but thank you

[andy-cytomic]

i have NO idea what you just said Andy, but thank you

Sorry bout that, let’s try again: you know when you plugin a stomp box to a power supply how it makes a loud thumping sound? People normally avoid this by turning down the gain of the input for that channel, plug in the stomp box, then turn it up again. Exactly the same thump happens for a detailed analog model, switch it on, massive thump that settles down to silence. In a DAW people find the identical workflow objectionable, having to mute the channel you add the plugin to, then wait 10 seconds, then unmute it again people don’t really like that much. So how do you get the same result as processing silence for 10 seconds?

Unsurprisingly actually processing 10 seconds of silence would do the trick, but that would mean you have to wait 10 seconds before you hear anything, which isn’t ideal. Alternatively I could max out the cpu and crunch 10 seconds of silence as soon as the plugin is added, which would probably cause massive dropouts of your DAW, also not ideal.

I’ve sorted out the maths to process 1 sample of silence and get the desired “warmed up” stage of all the energy storage devices like capacitors, so no cpu spike, no waiting 10 seconds, and no thump. My previous goes at this didn’t work so well. My new circuit solver has taken things in its stride, and it’s all working brilliantly.

[michaeltn86]

it's so interesting to read cytomic's posts. a lot of developers should care about this "thump" and cpu spike. even in the very $$$ price category, and really really respected developers. I admire this kind of attention so much that I'm kinda forced to give praise haha looking great

[dermage]

I'm thinking of buying the Scream, but some questions regarding the final release of "The Scream" came up. Although I see these questions are or should be mostly obvious I'd still like to know the official standing.:

1. Will the price change, once the final is released?
2. Will the sound change, once it reaches a non-beta stage?
3. Will final open older project files correctly once released?
4. Will automation break once final is released? (Not the most important, but would like to know if it could happen)

[andy-cytomic]

it's so interesting to read cytomic's posts. a lot of developers should care about this "thump" and cpu spike. even in the very $$$ price category, and really really respected developers. I admire this kind of attention so much that I'm kinda forced to give praise haha looking great

Thanks! I'm really looking forward to having this perfect warmup code rolled out to all our current products as well.

[andy-cytomic]

I'm thinking of buying the Scream, but some questions regarding the final release of "The Scream" came up. Although I see these questions are or should be mostly obvious I'd still like to know the official standing.:

1. Will the price change, once the final is released?
2. Will the sound change, once it reaches a non-beta stage?
3. Will final open older project files correctly once released?
4. Will automation break once final is released? (Not the most important, but would like to know if it could happen)

All good questions

1) The reason you can buy the beta is because customers that can handle a bit of rough and tumble requested it so they get to have fun using it as soon as possible. I don't like encouraging people to buy the beta, but it would also be unfair to actually charge more for it, so the price will remain the same. Of course you get the final version once it's ready.

2) The sound of the HD model will remain pretty much the same, although I did add a little more detail in the input and output stage clipping (reverse saturation). The MD model will sound close to the full HD, but obviously some detail is left out, but the cpu is much lower. The current beta only has HD mode.

3) The full release will open beta presets just fine

4) Automation will work just fine, the order of parameters won't change

[Funkybot's Evil Twin]

Andy, will the randomisation panel just be for creating some stereo difference between right and left channels on stereo operation, or will it also allow you to slightly randomise component tolerances and save any favorites? Kind of like allowing each user to have their own slightly unique version if they so desire.

Thinking out loud about how UA claims to model "Golden units" when they do some plugins. "This Pultec was the best Pultec of over twenty examples we listened to" (making that up to produce an example). Obviously that might not be the biggest thing in the world for a distortion pedal, but for compressors or something, I could see people hitting randomise until they found something they liked and saving it as their own personal Golden unit that no one else has.

[dermage]

Andy thanks for the answers, I bought it already . Hope there will be more presets (got the beta v2 zip) and possibly some hint which components to tweak in the mod section for what character. Really liking the acid synth preset so far using abl3.

[andy-cytomic]

Andy, will the randomisation panel just be for creating some stereo difference between right and left channels on stereo operation, or will it also allow you to slightly randomise component tolerances and save any favorites? Kind of like allowing each user to have their own slightly unique version if they so desire.

Thinking out loud about how UA claims to model "Golden units" when they do some plugins. "This Pultec was the best Pultec of over twenty examples we listened to" (making that up to produce an example). Obviously that might not be the biggest thing in the world for a distortion pedal, but for compressors or something, I could see people hitting randomise until they found something they liked and saving it as their own personal Golden unit that no one else has.

Yes, all this will be possible. There will be four tolerance groups, and each parameter will belong to a group, and you can change which group it belongs to if you want. Each tolerance group has it's own % scaling, and there is a global mono and global stereo scaling to adjust that on top.

Each preset will store both the root value of each parameter, and also two random numbers that will alter this parameter. The two random numbers get scaled by the group the parameter is in, and the global mono and stereo scalings. This way you will be able to share presets you come up with exactly, but someone receiving this preset could fine tune the mono or stereo variation, or change the tolerance group variation, or even the group each parameter belongs to, all non-destructively.

There will be two buttons "randomise" and "mutate" which will change things, and add a bullet point to a scroll bar so you can go back to previous values easily. Randomise will leave the root value alone, but re-generate all the random numbers, and mutate will apply the current scaling of the random numbers to the root value and then generate a new set of random numbers. Using mutate repeatedly will ultimately lead you a long way from your initial preset.

[acousticglue]

In the beginning with Scream I made bunches of presets then later noticed I missed the entire point of some settings lol. Later made much better settings on what I had learned. Once again this and the Drop are Deeep. I hope the Glue gets many added components and mid sides etc then I will buy it.

[Funkybot's Evil Twin]

Andy, any chance of adding a "lock" to the knobs? Or maybe "Mod Presets" that only impact the mod panel, leaving the top section alone? Here's the use case: I've setup a pretty good tone with the stock setting, dialed in roughly the right amount of drive, tone, and level, now I want to hear what the TS9 mod sounds like. If I go to the TS9 preset, I've get to move all knobs back to the same neighborhood. Would be faster if I could right-click to lock certain settings (even if only Drive, Tone, Level, and Mix) from preset changes. That way, my starting points are in the same ballpark. Or if I could just select between a few different "mod presets" that might be even better.

[waltercruz]

i have NO idea what you just said Andy, but thank you

Sorry bout that, let’s try again: you know when you plugin a stomp box to a power supply how it makes a loud thumping sound? People normally avoid this by turning down the gain of the input for that channel, plug in the stomp box, then turn it up again. Exactly the same thump happens for a detailed analog model, switch it on, massive thump that settles down to silence. In a DAW people find the identical workflow objectionable, having to mute the channel you add the plugin to, then wait 10 seconds, then unmute it again people don’t really like that much. So how do you get the same result as processing silence for 10 seconds?

Unsurprisingly actually processing 10 seconds of silence would do the trick, but that would mean you have to wait 10 seconds before you hear anything, which isn’t ideal. Alternatively I could max out the cpu and crunch 10 seconds of silence as soon as the plugin is added, which would probably cause massive dropouts of your DAW, also not ideal.

I’ve sorted out the maths to process 1 sample of silence and get the desired “warmed up” stage of all the energy storage devices like capacitors, so no cpu spike, no waiting 10 seconds, and no thump. My previous goes at this didn’t work so well. My new circuit solver has taken things in its stride, and it’s all working brilliantly.

beautiful explanation.