sequencers
Posted: Thu Feb 06, 2020 2:00 pm
I am a bit puzzled about the apparent lack of sequencers for scope. (And I don't mean for the modular, but for the regular environment). Or am I wrong?
Page 1 of 4
Re: sequencers
Posted: Thu Feb 06, 2020 3:13 pm
Again, depends on whether youre running 23 or 64 bit. Some sequencing stuff doesnt work on 64 bit.
Re: sequencers
Posted: Thu Feb 06, 2020 5:58 pm
Modular 3/4 + Flexor + BlackBox II/III + etc
The advantage with the above is that you can use SAW ramps from Flexor to sequence, and have the option of everything working at audiorate (sync on your dsp meter) and so staying sample accurate.
Most Midi in Scope is Async, and utilizes the host cpu... and while Scope midi is fairly tight compared to some of my USB Midi interfaces, it won't be sample accurate. So there's no real advantage over Scope sequencing versus software using it's software midi i/o drivers. And there's a LOT of sequencing options out there in software, including your typical DAWs, Plogue Bidule, tracker style interfaces and so much more.
The advantage with the above is that you can use SAW ramps from Flexor to sequence, and have the option of everything working at audiorate (sync on your dsp meter) and so staying sample accurate.
Most Midi in Scope is Async, and utilizes the host cpu... and while Scope midi is fairly tight compared to some of my USB Midi interfaces, it won't be sample accurate. So there's no real advantage over Scope sequencing versus software using it's software midi i/o drivers. And there's a LOT of sequencing options out there in software, including your typical DAWs, Plogue Bidule, tracker style interfaces and so much more.
Re: sequencers
Posted: Fri Feb 07, 2020 8:21 am
Yeah, I've got my choice of Windows sequencers, but I was just wondering. Seems like a thing that would just be there, to drive all those wonderful synths in Scope.
Re: sequencers
Posted: Fri Feb 07, 2020 11:24 am
@valis
"Midi in Scope is async which utilizes the host cpu" this is kind of wrong.
In scope there are 3 types of signals:
1. Sync (computed on the DSPs)
2. Async "A" type (computed on the DSPs but on a lower rate). As the name implies these signals are created asynchronously which means computed on demand but on a sample accurate basis. (an interrupt serves the demand inside the audio block)
3. Async "B" type (computed on the host CPU) (When you move a panel these signals stop updating...)
To identify the different async modules in the sdk, use the "module attributes" window. If a module displays dsp usage it is computed in the DSP's. The async modules included in the "logical" folder (some math modules) are the type-B i.e. for panel usage.
You can find the type-A async modules in the DSP modules list. There are plenty of these for math, conditions and more.
Midi internally in Scope is async "A" type and computed on the DSP's. It is sample accurate to the lower rate that asynchronous control signals are computed.
Async B type signals are mainly used (if used correctly) for panels/graphics or anything that is not timing sensitive.
The jitter that exists with the midi sequencer modules is due to the communication with windows. If you create a sequencer with the correct async modules it is sample accurate. If you send midi created internally in Scope to the physical midi outputs it is extremely tight similar to any digital hardware sequencer.
Regards,
John
"Midi in Scope is async which utilizes the host cpu" this is kind of wrong.
In scope there are 3 types of signals:
1. Sync (computed on the DSPs)
2. Async "A" type (computed on the DSPs but on a lower rate). As the name implies these signals are created asynchronously which means computed on demand but on a sample accurate basis. (an interrupt serves the demand inside the audio block)
3. Async "B" type (computed on the host CPU) (When you move a panel these signals stop updating...)
To identify the different async modules in the sdk, use the "module attributes" window. If a module displays dsp usage it is computed in the DSP's. The async modules included in the "logical" folder (some math modules) are the type-B i.e. for panel usage.
You can find the type-A async modules in the DSP modules list. There are plenty of these for math, conditions and more.
Midi internally in Scope is async "A" type and computed on the DSP's. It is sample accurate to the lower rate that asynchronous control signals are computed.
Async B type signals are mainly used (if used correctly) for panels/graphics or anything that is not timing sensitive.
The jitter that exists with the midi sequencer modules is due to the communication with windows. If you create a sequencer with the correct async modules it is sample accurate. If you send midi created internally in Scope to the physical midi outputs it is extremely tight similar to any digital hardware sequencer.
Regards,
John
Re: sequencers
Posted: Fri Feb 07, 2020 11:32 am
Do you mean something like cubase?
But we have a few step sequencers!
Check my Xite tips. I mention most there. But there is also now kron from spacef plus cwm did a cool one.
But we have a few step sequencers!
Check my Xite tips. I mention most there. But there is also now kron from spacef plus cwm did a cool one.
Re: sequencers
Posted: Fri Feb 07, 2020 11:38 am
Re: sequencers
Posted: Fri Feb 07, 2020 11:40 am
Does only sync work correctly on Xite?fra77x2 wrote: ↑Fri Feb 07, 2020 11:24 am @valis
"Midi in Scope is async which utilizes the host cpu" this is kind of wrong.
In scope there are 3 types of signals:
1. Sync (computed on the DSPs)
2. Async "A" type (computed on the DSPs but on a lower rate). As the name implies these signals are created asynchronously which means computed on demand but on a sample accurate basis. (an interrupt serves the demand inside the audio block)
3. Async "B" type (computed on the host CPU) (When you move a panel these signals stop updating...)
To identify the different async modules in the sdk, use the "module attributes" window. If a module displays dsp usage it is computed in the DSP's. The async modules included in the "logical" folder (some math modules) are the type-B i.e. for panel usage.
You can find the type-A async modules in the DSP modules list. There are plenty of these for math, conditions and more.
Midi internally in Scope is async "A" type and computed on the DSP's. It is sample accurate to the lower rate that asynchronous control signals are computed.
Async B type signals are mainly used (if used correctly) for panels/graphics or anything that is not timing sensitive.
The jitter that exists with the midi sequencer modules is due to the communication with windows. If you create a sequencer with the correct async modules it is sample accurate. If you send midi created internally in Scope to the physical midi outputs it is extremely tight similar to any digital hardware sequencer.
Regards,
John
Re: sequencers
Posted: Fri Feb 07, 2020 11:59 am
What do you mean correctly?
Everything works correctly if used in a correct scheme.
The signals differentiation is not specific to Scope. It is applied and works correctly in Scope as to other digital devices.
Pehaps some 3rd party developers or some free devices may have used some modules in a wrong manner.
The midi jitter is not specific to the scope midi driver. It is due to the implementation in windows (and to mac os in a way) and similar in performance to any other "external" midi device. Only vst's are sample accurate or anything that is computed internaly in DSPs. An easy way to overcome these limitations that are apparent to specific sounds that need to be very tight (percussive stuff), is to use audio triggers from the daw and use gates or envelope followers to create envelope curves. i.e. Use simple midi for pitch information and audio triggers for vca triggering/gating.
Everything works correctly if used in a correct scheme.
The signals differentiation is not specific to Scope. It is applied and works correctly in Scope as to other digital devices.
Pehaps some 3rd party developers or some free devices may have used some modules in a wrong manner.
The midi jitter is not specific to the scope midi driver. It is due to the implementation in windows (and to mac os in a way) and similar in performance to any other "external" midi device. Only vst's are sample accurate or anything that is computed internaly in DSPs. An easy way to overcome these limitations that are apparent to specific sounds that need to be very tight (percussive stuff), is to use audio triggers from the daw and use gates or envelope followers to create envelope curves. i.e. Use simple midi for pitch information and audio triggers for vca triggering/gating.
Re: sequencers
Posted: Fri Feb 07, 2020 12:08 pm
The last paragraph was my intention, but thank you for correcting me in regards to Async "A", which is basically what we would call "K-rate" in Csound (with which Scope shares some roots) or control rate in other software. I had neglected to mention this in as much depth as you havefra77x2 wrote: ↑Fri Feb 07, 2020 11:24 am @valis
"Midi in Scope is async which utilizes the host cpu" this is kind of wrong.
In scope there are 3 types of signals:
1. Sync (computed on the DSPs)
2. Async "A" type (computed on the DSPs but on a lower rate). As the name implies these signals are created asynchronously which means computed on demand but on a sample accurate basis. (an interrupt serves the demand inside the audio block)
3. Async "B" type (computed on the host CPU) (When you move a panel these signals stop updating...)
To identify the different async modules in the sdk, use the "module attributes" window. If a module displays dsp usage it is computed in the DSP's. The async modules included in the "logical" folder (some math modules) are the type-B i.e. for panel usage.
You can find the type-A async modules in the DSP modules list. There are plenty of these for math, conditions and more.
Midi internally in Scope is async "A" type and computed on the DSP's. It is sample accurate to the lower rate that asynchronous control signals are computed.
Async B type signals are mainly used (if used correctly) for panels/graphics or anything that is not timing sensitive.
The jitter that exists with the midi sequencer modules is due to the communication with windows. If you create a sequencer with the correct async modules it is sample accurate. If you send midi created internally in Scope to the physical midi outputs it is extremely tight similar to any digital hardware sequencer.
Regards,
John
In practice however, I don't recall encountering a module that sequences externally and doesn't use the host cpu, are you aware of devices you could mention for others to utilize that stay entire on DSP with async "A" and do sequencing? When flexor was created, I seem to recall some timing variation even with modular devices, though I will admit I don't own every 3rd party device and haven't checked this since years ago.
Correct, hence my suggestion:fra77x2 wrote: ↑Fri Feb 07, 2020 11:59 am The midi jitter is not specific to the scope midi driver. It is due to the implementation in windows (and to mac os in a way) and similar in performance to any other "external" midi device. Only vst's are sample accurate or anything that is computed internaly in DSPs. An easy way to overcome these limitations that are apparent to specific sounds that need to be very tight (percussive stuff), is to use audio triggers from the daw and use gates or envelope followers to create envelope curves. i.e. Use simple midi for pitch information and audio triggers for vca triggering/gating.
Re: sequencers
Posted: Fri Feb 07, 2020 12:14 pm
Of course using audio signals to drive sequencers is the correct procedure. Jitter comes from the midi sequencer module and the communication with the host. So midi clocks and similar signals suffer. (when there is communication) If you create midi clocks internally they are perfect (or similar to other digital devices).
Re: sequencers
Posted: Fri Feb 07, 2020 2:30 pm
I understand that, but I'm curious... are there any that you can think of though that would be entirely on dsp? I'm not even sure if the synths that have internal sequencers are entirely on the dsp cards to be quite frank.
Re: sequencers
Posted: Fri Feb 07, 2020 2:33 pm
I checked your website, but cannot find those tips. I found some nice pictures, thoughLiquid EDGE wrote: ↑Fri Feb 07, 2020 11:32 am Do you mean something like cubase?
But we have a few step sequencers!
Check my Xite tips. I mention most there. But there is also now kron from spacef plus cwm did a cool one.
Re: sequencers
Posted: Sat Feb 08, 2020 12:43 am
Do you have an Xite?fra77x2 wrote: ↑Fri Feb 07, 2020 11:59 am What do you mean correctly?
Everything works correctly if used in a correct scheme.
The signals differentiation is not specific to Scope. It is applied and works correctly in Scope as to other digital devices.
Pehaps some 3rd party developers or some free devices may have used some modules in a wrong manner.
The midi jitter is not specific to the scope midi driver. It is due to the implementation in windows (and to mac os in a way) and similar in performance to any other "external" midi device. Only vst's are sample accurate or anything that is computed internaly in DSPs. An easy way to overcome these limitations that are apparent to specific sounds that need to be very tight (percussive stuff), is to use audio triggers from the daw and use gates or envelope followers to create envelope curves. i.e. Use simple midi for pitch information and audio triggers for vca triggering/gating.
This is the issue!
type of bug: Sequencer settings unresponsive until dsp reload(reallocation or sample rate switch)
Re: sequencers
Posted: Sat Feb 08, 2020 12:45 am
Sorry, I meant here on planetz.Joey wrote: ↑Fri Feb 07, 2020 2:33 pmI checked your website, but cannot find those tips. I found some nice pictures, thoughLiquid EDGE wrote: ↑Fri Feb 07, 2020 11:32 am Do you mean something like cubase?
But we have a few step sequencers!
Check my Xite tips. I mention most there. But there is also now kron from spacef plus cwm did a cool one.
http://forums.scopeusers.com/viewtopic.php?t=34658
Re: sequencers
Posted: Sat Feb 08, 2020 2:11 am
@Liquid EDGE
Yes I use a Xite-1.
I think the "sequencer" modules in the sdk has started to display this behaviour after scope 5 release. In my old PCI cards with 3.1c they work fine.
@Valis
I have many years to play with this midi sequencing stuff in scope, I program in C these days.
By my memory:
In sdk there are some "sequencer" building blocks. I have never been able to use these. These are the ones that have stopped working correctly.
Other than this I have made sequencers with async switches and stuff like this.
A purely audio rate sequencer can be made (I have made one) with comparators in the sdk.
To create the midi message you have to use a module named "sync to message" if I remember correctly.
In any case it is not worth the strungle perhaps only for educational purposes.
Why do you need this sequencing stuff in scope? I mean is their any benefit? The UI of that stuff made in scope is kind of basic.
15 years ago when I tried to find a solution to the jitter nuisance I had made several different types of "midi sequencers" in scope using only basic modules but I have solved the issue by a different procedure 10 years ago that allows me to use native stuff for sequencing.
Yes I use a Xite-1.
I think the "sequencer" modules in the sdk has started to display this behaviour after scope 5 release. In my old PCI cards with 3.1c they work fine.
@Valis
I have many years to play with this midi sequencing stuff in scope, I program in C these days.
By my memory:
In sdk there are some "sequencer" building blocks. I have never been able to use these. These are the ones that have stopped working correctly.
Other than this I have made sequencers with async switches and stuff like this.
A purely audio rate sequencer can be made (I have made one) with comparators in the sdk.
To create the midi message you have to use a module named "sync to message" if I remember correctly.
In any case it is not worth the strungle perhaps only for educational purposes.
Why do you need this sequencing stuff in scope? I mean is their any benefit? The UI of that stuff made in scope is kind of basic.
15 years ago when I tried to find a solution to the jitter nuisance I had made several different types of "midi sequencers" in scope using only basic modules but I have solved the issue by a different procedure 10 years ago that allows me to use native stuff for sequencing.
Re: sequencers
Posted: Sat Feb 08, 2020 2:24 am
Yeah, it’s an Xite thing fra! As scope 5 on a pci card doesn’t display that behaviour. It’s the chips, assign that stuff to dsp2 and sequencer modules work fine. As it’s a old style shark chip like on the pci cards.
Re: sequencers
Posted: Sat Feb 08, 2020 2:29 am
Yes I remember this trick but it is not consistent.
In Xite there are 6 older sharks. Async modules (the dsp computed ones) work on any dsp old or new.
If you like arrange a skype meeting with me (with a symbolic fee), I can show you 3-4 different procedures to create suquencing stuff in the sdk.
As I mentioned the sequencers building blocks stopped working right after scope 5 release. It is a software issue that perhaps remain unsolved due to luck of any benefit. Why use this archaic way of sequencing nowadays?.
In Xite there are 6 older sharks. Async modules (the dsp computed ones) work on any dsp old or new.
If you like arrange a skype meeting with me (with a symbolic fee), I can show you 3-4 different procedures to create suquencing stuff in the sdk.
As I mentioned the sequencers building blocks stopped working right after scope 5 release. It is a software issue that perhaps remain unsolved due to luck of any benefit. Why use this archaic way of sequencing nowadays?.
Re: sequencers
Posted: Sat Feb 08, 2020 3:58 am
Cool, if I ever get round to installing and booting up the SDK I’ll get in touch.fra77x2 wrote: ↑Sat Feb 08, 2020 2:29 am Yes I remember this trick but it is not consistent.
In Xite there are 6 older sharks. Async modules (the dsp computed ones) work on any dsp old or new.
If you like arrange a skype meeting with me (with a symbolic fee), I can show you 3-4 different procedures to create suquencing stuff in the sdk.
As I mentioned the sequencers building blocks stopped working right after scope 5 release. It is a software issue that perhaps remain unsolved due to luck of any benefit. Why use this archaic way of sequencing nowadays?.
it’s something I intend and want to do! It would be nice if it was fixed tbh, as the sequencer element on Solaris and within modular etc are more about sequencing sound parameters and movement of filters etc. With ofcourse notes and it being used in a more arpeggiator way!
Re: sequencers
Posted: Sat Feb 08, 2020 8:40 am
Joey,
Something like the G.O.S.T. MATTOMAT? If yes, l'm very surprised it's not already been mentioned. Did it have a few teething problems? It reminded me of a VST called Atomic by Algomusic which has a built in synth. I really liked that one.
Something like the G.O.S.T. MATTOMAT? If yes, l'm very surprised it's not already been mentioned. Did it have a few teething problems? It reminded me of a VST called Atomic by Algomusic which has a built in synth. I really liked that one.