Still Struggling with Quantizing HALP

[Karlo]

As a potential alternative, I believe (but have not double-checked the exact mechanics) you could clock the table from FX. Might be worth a look if you want to get fancier and/or can’t afford to tie up an automator, etc.

Well, clocking the table is not giving me the desired or expected result... now if I can trigger a table with quantization/allowed notes that only has one row/step and it will trigger it for that step and hold the pitch cv voltage until the table is triggered again... I can waste the automator.... because I can barely use half the features NerdSEQ offers anyways!

[Karlo]

Hello, you could set your table with an automator as source. 
Then use this automator in envelope mode and trigger it with the gates of the track you are using.

This gonna make the table to go one step every time you trigger a note.

Another way to do the same thing is to use a cv input instead as a source and just feed it with the trigger output of the track you use.

So you have choice to spend either a cv in or one envelope and one automator

I could not get this to work at all. So frustrating. It didn't step every time a new note was gated. It would step every repetition of the pattern. So on 00 of my pattern it would advance one step in the table, not when the note gated the envelope... AND the table didn't quantize the pattern pitch to the allow notes in the table. 

Which brings me to something else....

The envelopes are terrible, they're noisy in my VCA. I can hear little "rattle sounds" regardless of attack and decay settings. I think it's in the decay stage. it's doing something stupid, don't know what. I don't have a scope, I just use my ears and it sounds like shit. I find it worthless.

I'm starting to develop the opinion that some of these features; envelopes, pitch quantize, ratcheting, clock dividing, are shit. The manual sucks too.

[mgd]

I could not get this to work at all. So frustrating. It didn't step every time a new note was gated. It would step every repetition of the pattern. So on 00 of my pattern it would advance one step in the table, not when the note gated the envelope... AND the table didn't quantize the pattern pitch to the allow notes in the table. 

Which brings me to something else....

The envelopes are terrible, they're noisy in my VCA. I can hear little "rattle sounds" regardless of attack and decay settings. I think it's in the decay stage. it's doing something stupid, don't know what. I don't have a scope, I just use my ears and it sounds like shit. I find it worthless.

I'm starting to develop the opinion that some of these features; envelopes, pitch quantize, ratcheting, clock dividing, are shit. The manual sucks too.

Maybe you could either describe in detail what you do so that others can reproduce it exactly or share an minimal example project file, possibly with some short explanation. Currently it is difficult to assess what actually goes wrong.

I don't understand technically how an envelope can be noisy in a VCA (the "worst" thing that can happen is some AM due to oscillation) but I have a scope and am willing to look at it

[XORadmin]

Hello, you could set your table with an automator as source. 
Then use this automator in envelope mode and trigger it with the gates of the track you are using.

This gonna make the table to go one step every time you trigger a note.

Another way to do the same thing is to use a cv input instead as a source and just feed it with the trigger output of the track you use.

So you have choice to spend either a cv in or one envelope and one automator

I could not get this to work at all. So frustrating. It didn't step every time a new note was gated. It would step every repetition of the pattern. So on 00 of my pattern it would advance one step in the table, not when the note gated the envelope... AND the table didn't quantize the pattern pitch to the allow notes in the table. 

Which brings me to something else....

The envelopes are terrible, they're noisy in my VCA. I can hear little "rattle sounds" regardless of attack and decay settings. I think it's in the decay stage. it's doing something stupid, don't know what. I don't have a scope, I just use my ears and it sounds like shit. I find it worthless.

I'm starting to develop the opinion that some of these features; envelopes, pitch quantize, ratcheting, clock dividing, are shit. The manual sucks too.

Het Karlo,

thanks for your kind words.
Starting with the tables to quantize your notes.
I would: Use Up or Down as a quantize setting. Because then the selected allowed note will always stay the same. If you choose coin, the table would choose either the next higher or lower note with every table step and this you can hear. Also select a faster speed like 1 or maybe also Internal time as source. Your notes will quantize well and you should only hear your 3 notes D, F, G#.
Range is not relevant here and does nothing. 
No need for any work arounds with timings or automators or whatever.

As for the envelopes. I'm sorry that they are so shitty for you and to be honest, you are the first to mention it. Of course they are not perfect analogue envelopes, but I never claimed so. They work fine and good for many applications
How do you use them? Are you using them correctly?

What is your issue with ratcheting?

What is your issue with clock dividing?

What is your issue with the manual exactly?

[Karlo]

Hello, you could set your table with an automator as source. 
Then use this automator in envelope mode and trigger it with the gates of the track you are using.

This gonna make the table to go one step every time you trigger a note.

Another way to do the same thing is to use a cv input instead as a source and just feed it with the trigger output of the track you use.

So you have choice to spend either a cv in or one envelope and one automator

I could not get this to work at all. So frustrating. It didn't step every time a new note was gated. It would step every repetition of the pattern. So on 00 of my pattern it would advance one step in the table, not when the note gated the envelope... AND the table didn't quantize the pattern pitch to the allow notes in the table. 

Which brings me to something else....

The envelopes are terrible, they're noisy in my VCA. I can hear little "rattle sounds" regardless of attack and decay settings. I think it's in the decay stage. it's doing something stupid, don't know what. I don't have a scope, I just use my ears and it sounds like shit. I find it worthless.

I'm starting to develop the opinion that some of these features; envelopes, pitch quantize, ratcheting, clock dividing, are shit. The manual sucks too.

Het Karlo,

thanks for your kind words.
Starting with the tables to quantize your notes.
I would: Use Up or Down as a quantize setting. Because then the selected allowed note will always stay the same. If you choose coin, the table would choose either the next higher or lower note with every table step and this you can hear. Also select a faster speed like 1 or maybe also Internal time as source. Your notes will quantize well and you should only hear your 3 notes D, F, G#.
Range is not relevant here and does nothing. 
No need for any work arounds with timings or automators or whatever.

As for the envelopes. I'm sorry that they are so shitty for you and to be honest, you are the first to mention it. Of course they are not perfect analogue envelopes, but I never claimed so. They work fine and good for many applications
How do you use them? Are you using them correctly?

What is your issue with ratcheting?

What is your issue with clock dividing?

What is your issue with the manual exactly?

I'm just another asshole with a negative opinion on the internet. Don't take it personally. I'm actually quite fond of NerdSEQ and I'm impressed with your effort on this.

This issue I had with the pitch quantize is it quantizes per tick. If I have a quarter note and each quarter note is comprised of 6 ticks I hear the table quantize the quarter note to 6 different pitches instead of just holding one. It's infuriating when you have whole notes, some eight notes, sixteenths, etc. in one pattern. Infuriating as in inferior... like my NerdSEQ skills in getting this to work. LOL.

What I figured out, is make the table with allowed notes 1 step in length, and set to play once (like XORAdmin mentioned in my "make next note random" thread). Then just put that table on every step that has a note you want quantized. This worked but took me forever to figure out last night. 

So yes, I eventually managed to get it to work, but it wasn't well documented any where on how to make it work for a simple application. That's why the manual isn't very good. I want to sequence, not experiment for 6 hours listening to the same pattern over and over again trying to figure out how to quantize the pitch to C maj. LOL. Also, please please please, put all the expanders info in a separate manual or an addendum at the end. It's painful scrolling through pages and pages of information for something that is an elective add-on to the NerdSEQ and not relevant to it's core functionality. Put it at the end of the manual. Please.

The envelops - All my modular envelops are digital so I'm not sure your comparison to analog envelopes is relevant. For comparison, I have an Intellijel Quadrax and/or IME Kermit mkIII (both digital and buttery smooth) and neither of them contaminate my signal like the NerdSEQ envelopes. I think what MGD said - it might be oscillating and making a short ringmod type sound to my output signal on the VCA - but I don't know why it would oscillate. I don't have the envelope set to cycle, it is set to one shot or once. I would argue Digital envelopes are superior to analog... just not the NerdSEQ envelops. Something you may want to look into.

- clock dividing happens on the following pattern and is not instantaneous, I'm hopeful for the next release that clock divide is instantaneous and set from the pattern you want divided but how it's set up now, I just don't really understand the intent really. Example: Pattern 00 with the clock divide, then on 01, the pattern is now slower... great, but what if my sequence goes 00, 01, 00, 02... but I don't want 02 to be divided? I have to clone 00 to 03 jut to remove the clock divide FX. The clock divide should happen on the pattern you want clock divided...

Ratcheting... again, I'm hopeful for the next release that this is addressed and can be set to musically division in the trigger column instead of all the button pressing trying to find the right combinations of repeats to trigger length as seen in the 1.23 firmware. As others have stated... if you later decide to change the BPM of your pattern, the repeats to trigger lengths will be "off" and you have to go back and go through the the massive amount of button presses to find the perfect combination of repeats to trigger length.

[mvdirty]

I'm just another asshole with a negative opinion on the internet.

Good day fellow internet asshole.

This issue I had with the pitch quantize is it quantizes per tick. If I have a quarter note and each quarter note is comprised of 6 ticks I hear the table quantize the quarter note to 6 different pitches instead of just holding one. It's infuriating when you have whole notes, some eight notes, sixteenths, etc. in one pattern. Infuriating as in inferior... like my NerdSEQ skills in getting this to work. LOL.

This sounds like either coin-flipping and/or a table being clocked at the wrong speed, no? Though, in fairness, it could be cool if coin could (optionally?) flip again only when its source note changes. And/or when gated. Or something like that.

Also, please please please, put all the expanders info in a separate manual or an addendum at the end. It's painful scrolling through pages and pages of information for something that is an elective add-on to the NerdSEQ and not relevant to it's core functionality. Put it at the end of the manual. Please.

If we’re putting in our two bits: As an owner of all expanders, please don’t. Instead, please use whatever documentation tooling is needed to have each expander section cover all of the supported FX for that track type without just pointing the reader back to the modular track type FX section of the documentation. Yes, I know it is duplication, but this is a PDF and not a printed manual.

The envelops - All my modular envelops are digital so I'm not sure your comparison to analog envelopes is relevant. For comparison, I have an Intellijel Quadrax and/or IME Kermit mkIII (both digital and buttery smooth) and neither of them contaminate my signal like the NerdSEQ envelopes. I think what MGD said - it might be oscillating and making a short ringmod type sound to my output signal on the VCA - but I don't know why it would oscillate. I don't have the envelope set to cycle, it is set to one shot or once. I would argue Digital envelopes are superior to analog... just not the NerdSEQ envelops. Something you may want to look into.

Could this perhaps be a 12 bit vs 16+ bit DAC thing?

- clock dividing happens on the following pattern and is not instantaneous

I for one would appreciate this if optional.

Ratcheting... again, I'm hopeful for the next release that this is addressed and can be set to musically division in the trigger column instead of all the button pressing trying to find the right combinations of repeats to trigger length as seen in the 1.23 firmware. As others have stated... if you later decide to change the BPM of your pattern, the repeats to trigger lengths will be "off" and you have to go back and go through the the massive amount of button presses to find the perfect combination of repeats to trigger length.

I admit that I’ve been holding off on the pre-release firmware so I’ve not messed with them yet, but are you perhaps using repeats when you should be using ratchets?

[mgd]

What I figured out, is make the table with allowed notes 1 step in length, and set to play once (like XORAdmin mentioned in my "make next note random" thread). Then just put that table on every step that has a note you want quantized. This worked but took me forever to figure out last night.

So yes, I eventually managed to get it to work, but it wasn't well documented any where on how to make it work for a simple application. That's why the manual isn't very good. I want to sequence, not experiment for 6 hours listening to the same pattern over and over again trying to figure out how to quantize the pitch to C maj. LOL.

I agree in so far as I think the manual lacks a couple of simple "snippets" as in "When you want to do this, then do it like that" type of examples. Really the basic stuff. That would flatten the learning curv.

Also, please please please, put all the expanders info in a separate manual or an addendum at the end. It's painful scrolling through pages and pages of information for something that is an elective add-on to the NerdSEQ and not relevant to it's core functionality. Put it at the end of the manual. Please.

That's probably a matter of personal preference. I like it the way it is.

The envelops - All my modular envelops are digital so I'm not sure your comparison to analog envelopes is relevant. For comparison, I have an Intellijel Quadrax and/or IME Kermit mkIII (both digital and buttery smooth) and neither of them contaminate my signal like the NerdSEQ envelopes. I think what MGD said - it might be oscillating and making a short ringmod type sound to my output signal on the VCA - but I don't know why it would oscillate. I don't have the envelope set to cycle, it is set to one shot or once. I would argue Digital envelopes are superior to analog... just not the NerdSEQ envelops. Something you may want to look into.

On what data do you base your claim "Digital envelopes are superior to analog"?

Anyway, if you provide a simple project file showing the problem I'm happy to look into it on a scope.

Kind regards,
Michael

---

The envelops - All my modular envelops are digital so I'm not sure your comparison to analog envelopes is relevant. For comparison, I have an Intellijel Quadrax and/or IME Kermit mkIII (both digital and buttery smooth) and neither of them contaminate my signal like the NerdSEQ envelopes. I think what MGD said - it might be oscillating and making a short ringmod type sound to my output signal on the VCA - but I don't know why it would oscillate. I don't have the envelope set to cycle, it is set to one shot or once. I would argue Digital envelopes are superior to analog... just not the NerdSEQ envelops. Something you may want to look into.

Could this perhaps be a 12 bit vs 16+ bit DAC thing?

I very much doubt that. 12bit is enough resolution for decent envelopes.

[Karlo]

I'm just another asshole with a negative opinion on the internet.

Good day fellow internet asshole.

This issue I had with the pitch quantize is it quantizes per tick. If I have a quarter note and each quarter note is comprised of 6 ticks I hear the table quantize the quarter note to 6 different pitches instead of just holding one. It's infuriating when you have whole notes, some eight notes, sixteenths, etc. in one pattern. Infuriating as in inferior... like my NerdSEQ skills in getting this to work. LOL.

This sounds like either coin-flipping and/or a table being clocked at the wrong speed, no? Though, in fairness, it could be cool if coin could (optionally?) flip again only when its source note changes. And/or when gated. Or something like that.

Also, please please please, put all the expanders info in a separate manual or an addendum at the end. It's painful scrolling through pages and pages of information for something that is an elective add-on to the NerdSEQ and not relevant to it's core functionality. Put it at the end of the manual. Please.

If we’re putting in our two bits: As an owner of all expanders, please don’t. Instead, please use whatever documentation tooling is needed to have each expander section cover all of the supported FX for that track type without just pointing the reader back to the modular track type FX section of the documentation. Yes, I know it is duplication, but this is a PDF and not a printed manual.

The envelops - All my modular envelops are digital so I'm not sure your comparison to analog envelopes is relevant. For comparison, I have an Intellijel Quadrax and/or IME Kermit mkIII (both digital and buttery smooth) and neither of them contaminate my signal like the NerdSEQ envelopes. I think what MGD said - it might be oscillating and making a short ringmod type sound to my output signal on the VCA - but I don't know why it would oscillate. I don't have the envelope set to cycle, it is set to one shot or once. I would argue Digital envelopes are superior to analog... just not the NerdSEQ envelops. Something you may want to look into.

Could this perhaps be a 12 bit vs 16+ bit DAC thing?

- clock dividing happens on the following pattern and is not instantaneous

I for one would appreciate this if optional.

Ratcheting... again, I'm hopeful for the next release that this is addressed and can be set to musically division in the trigger column instead of all the button pressing trying to find the right combinations of repeats to trigger length as seen in the 1.23 firmware. As others have stated... if you later decide to change the BPM of your pattern, the repeats to trigger lengths will be "off" and you have to go back and go through the the massive amount of button presses to find the perfect combination of repeats to trigger length.

I admit that I’ve been holding off on the pre-release firmware so I’ve not messed with them yet, but are you perhaps using repeats when you should be using ratchets?

Good day.

Quantize works as expected using a bunch of different note lengths in my pattern with out it quantizing per tick, it had nothing to do with the quantize up/down/Coin you just have to set it up a very specific way not outlined in the manual.

I don't have any expanders, and I have a small system with no space or use for all the expanders. Yes I can use PDF search, but I still have to click through the search results of what seems like a thousand pages of expander info to get to the patch screen or the table screen. Just put it the expander info at the end... then you know exactly where you're precious More CV expander info is instead of embedded half way through the manual somewhere.

12 bit should suffice for smooth envelopes, I've had 12bit midi converters and never had a problem with them adding "artifacts"

Clock dividing on some other non-relevant upstream pattern as an option is fine by me, still doesn't make any sense tho. LOL

Yes, I would love to use ratchets, but where are the ratchets in 1.23? I've only found/used the repeat triggers in the pattern screen.

[XORadmin]

This issue I had with the pitch quantize is it quantizes per tick. If I have a quarter note and each quarter note is comprised of 6 ticks I hear the table quantize the quarter note to 6 different pitches instead of just holding one. It's infuriating when you have whole notes, some eight notes, sixteenths, etc. in one pattern. Infuriating as in inferior... like my NerdSEQ skills in getting this to work. LOL.

There is no issue with that. You can have a note over several patterns long and it would not change the quantized pitch.
It is the way you use the quantization which doesn't work for your application. If you follow my suggestions it will work just as you intended.
Your one step purpose might work, but I wouldn't do it like this in the first place, especially because you need to fill in a table with every note. That is not necessary here for the result.

The envelops - All my modular envelops are digital so I'm not sure your comparison to analog envelopes is relevant. For comparison, I have an Intellijel Quadrax and/or IME Kermit mkIII (both digital and buttery smooth) and neither of them contaminate my signal like the NerdSEQ envelopes. I think what MGD said - it might be oscillating and making a short ringmod type sound to my output signal on the VCA - but I don't know why it would oscillate. I don't have the envelope set to cycle, it is set to one shot or once. I would argue Digital envelopes are superior to analog... just not the NerdSEQ envelops. Something you may want to look into.

Of course beside comparing it to analogue modules it is the same thing comparing it to dedicated modules. They do only one thing, but the nerdSEQ does 1000 things at the same time. Unfair comparison. Again my question, how do you use them exactly (settings etc)? 

- clock dividing happens on the following pattern and is not instantaneous, I'm hopeful for the next release that clock divide is instantaneous and set from the pattern you want divided but how it's set up now, I just don't really understand the intent really. Example: Pattern 00 with the clock divide, then on 01, the pattern is now slower... great, but what if my sequence goes 00, 01, 00, 02... but I don't want 02 to be divided? I have to clone 00 to 03 jut to remove the clock divide FX. The clock divide should happen on the pattern you want clock divided...

When the current pattern executes the FX command to change the clock...then the pattern is running on it's clock already. That's why...that's the intent. If you are sitting in the bus and the door closes, then you got to wait for the next station until you can change to another one.
However, as you know of course, there is a force mode with the next firmware. If used well, it will do what you want it to do.

Ratcheting... again, I'm hopeful for the next release that this is addressed and can be set to musically division in the trigger column instead of all the button pressing trying to find the right combinations of repeats to trigger length as seen in the 1.23 firmware. As others have stated... if you later decide to change the BPM of your pattern, the repeats to trigger lengths will be "off" and you have to go back and go through the the massive amount of button presses to find the perfect combination of repeats to trigger length.

Retriggering is not ratcheting. No one ever said this and it is clear described in the manual and tutorials what it does..all but not (bpm) clock based ratcheting.
You can use table presets to create some ratchets as you know.
And also here as you surely saw already, there have been ratcheting added in multiple variations in the coming firmware.

Last but not least. The manual. As most of you know I am not a native english speaking guy and the developer of it all. And writing the manual is the worst thing to do as a developer. I try to put all the information in there. But all the use cases...sorry the manual would be 200+ pages big and the firmware would be only half evolved. Not in your interrest and I'm not even talking about my motivation. But the fact is, there is always a up-to-date manual with new versions (and you don't have to wait for months or if it's ever getting updated at all) and functions are described. All the tricks and use-cases is something to explain or show here on the Forum, to share on videos etc. And I am not even aware of many tricks people use. (In fact, the quantization of a main sequence with tables is also a trick and was never planned. But it works.)
With every new version I spend a few days, re-reading the manual, updating things, adding stuff. This is for me harder work than adding function X/Y to the sequencer. But I still do it...and I'm still happy to do it.

That all said, I get back to work, else there won't be any release.

[Karlo]

What I figured out, is make the table with allowed notes 1 step in length, and set to play once (like XORAdmin mentioned in my "make next note random" thread). Then just put that table on every step that has a note you want quantized. This worked but took me forever to figure out last night.

So yes, I eventually managed to get it to work, but it wasn't well documented any where on how to make it work for a simple application. That's why the manual isn't very good. I want to sequence, not experiment for 6 hours listening to the same pattern over and over again trying to figure out how to quantize the pitch to C maj. LOL.

I agree in so far as I think the manual lacks a couple of simple "snippets" as in "When you want to do this, then do it like that" type of examples. Really the basic stuff. That would flatten the learning curv.

Also, please please please, put all the expanders info in a separate manual or an addendum at the end. It's painful scrolling through pages and pages of information for something that is an elective add-on to the NerdSEQ and not relevant to it's core functionality. Put it at the end of the manual. Please.

That's probably a matter of personal preference. I like it the way it is.

The envelops - All my modular envelops are digital so I'm not sure your comparison to analog envelopes is relevant. For comparison, I have an Intellijel Quadrax and/or IME Kermit mkIII (both digital and buttery smooth) and neither of them contaminate my signal like the NerdSEQ envelopes. I think what MGD said - it might be oscillating and making a short ringmod type sound to my output signal on the VCA - but I don't know why it would oscillate. I don't have the envelope set to cycle, it is set to one shot or once. I would argue Digital envelopes are superior to analog... just not the NerdSEQ envelops. Something you may want to look into.

On what data do you base your claim "Digital envelopes are superior to analog"?

Anyway, if you provide a simple project file showing the problem I'm happy to look into it on a scope.

Kind regards,
Michael

---

The envelops - All my modular envelops are digital so I'm not sure your comparison to analog envelopes is relevant. For comparison, I have an Intellijel Quadrax and/or IME Kermit mkIII (both digital and buttery smooth) and neither of them contaminate my signal like the NerdSEQ envelopes. I think what MGD said - it might be oscillating and making a short ringmod type sound to my output signal on the VCA - but I don't know why it would oscillate. I don't have the envelope set to cycle, it is set to one shot or once. I would argue Digital envelopes are superior to analog... just not the NerdSEQ envelops. Something you may want to look into.

Could this perhaps be a 12 bit vs 16+ bit DAC thing?

I very much doubt that. 12bit is enough resolution for decent envelopes.

You like the manual the way it is because you probably have the expanders. LOL

Digital evelopes are better because they can be upgraded/improved through firmware... that's my data. Analog... get out the soldering iron.. no thanks. 

I'll see what I can do about the project file.