Mixbus as a composition tool : MIDI loops

[jonetsu]

You do not use Midi channels ?
Mixbus has lots of miditools to filter.

Instead the 'dialog' with Bitwig goes like this:

- Hello Bitwig, you know, the Prokeys88...
- Yes, I know it
- Well, can you make it play the Biotek synth ?
- Sure
- And also, you know, the Axiom 25 ...
- Yes, I know it and I have a configuration for it too.
- Good. Can you make it play the Hive synth ?
- Sure.

And then I'm ready to play 'Rick Wakeman' style (or 'Keith Emerson' if you prefer, but since Rui replied that my guitar on my last piece "Summer Theme" made him think of Steve Howe, I'd stick with Wakeman. He's a funny guy too, this Wakeman) with one hand on one keyboard for Biotek and the other hand on the Axiom. Who needs MIDI and channels, and filtering and all that ?

and think of users with several keyboards ? ( My setup has more than seven devices)

Perhaps you should take a bit more time to read. Above I wrote that I have 3 keyboards and two other devices. And the Faderport. And a Novation Launchkey Mini with a broken key. Not made to be played.

The Midid recieve button will not be of use in such a case. And having more than one device is rather common. I carry three or four in my backpack

Makes me think, that's a long time I haven't carried a backpack. Used to put my shelter and food for 7 seven days and go out on lakes in my canoe.

I suggest you use midi channels.

It could be a way to circumvent the problem of Mixbus having no session memory whatsoever of what MIDI device was connected to which track. It's worth a try in this case since I have to use Mixbus as part of the creation process. This said, I consider there's creation also in mixing and mastering, but you know what I mean.

Cheers.

[Frank]

Yes i can play with one hand on Biotek, the other at Waverazor.

and/Or when i change the midi channel it will play Reaktor or Kontakt or Razor.

The i have midi input on this channel is not much of value if one has several devices.

Program changes changes the type of sound as well

I do read

[madmaxmiller]

I see. I used some soundfonts years ago. Then for whatever reason dropped their use. I liked some, especially trombone and violin. Maybe I should try them again. What are your sources for excellent soundfonts ? I add "excellent" because there are also not-so-good ones out there.

Sonatina Symphonic Orchestra is quite good. And there is Virtual Playing Orchestra. These are the free "full" orchestras with all sections. Then there are a lot of "themed' free sound fonts around, piano, brass, strings, whatever... and collections which are compilations of the others and reflect the taste of the respective author.
And then there are the commercial ones, EastWest and the like. They mostly are not in sfz/sf2 format and require Kontakt or their own player. Some of them in combination with the native player and their modulation options are what you probably mean with "excellent".

For me personally Sonatina and the like will do, until I can really say I now need the extended capabilities of a commercial one...
Here's an example for Sonatina (not by me )
https://youtu.be/lBhYNXFWI9M

(if you listen to more there you will know the sound font alone is nothing, it's what you make of it )

MMM

P.S. for a quick mockup the FluidGM which comes with every distro is not bad

[jonetsu]

https://youtu.be/lBhYNXFWI9M

Sounds pretty good. The Sonatina main page 'sounds' like the project is closed and refers to two other ones, one you mentioned and the VSCO. On East West after a few minutes I could not find a download version of their Play 6 software (nor any further details if it's a VST, if it requires Kontakt, etc) so it's not worthwhile from that point of view. Some people also mention to save money and go free instead since it seems the quality is there anyways.

In those times when I used some soundfonts I used qsynth. I don't know if it handles sfz files too. I guess it's still around. I reckon it's a front end for fluidsynth. I have heard of a free sfz player by Plogue, sforzando. Requires linVST though. I never had a good experience with Linux Sampler. Sforzando runs in Bitwig. I would have liked to quickly tried a sound and I downloaded a baritone sax from Plogue but, this is another world not related to synths presets at all. It seems sfz relies on a hardcoded directory structure. Eg. there's some manual reading to be done before trying a sound. The Table Warp sounds that comes with sforzando are not really convincing.

Anyways, thanks for mentioning soundfonts and Sonatina/others !

Here's a quick Sakura example that I just did now using a cello. Sakura is basically modeling a string, although giving sound config options.

Sakura Cello

   

---

Yes i can play with one hand on Biotek, the other at Waverazor.

The important difference is that in one case the DAW itself knows and remembers (and loads a config when there's one) every single MIDI controller. Whereas in the other case it uses anonymous numbers of MIDI channels.

[madmaxmiller]

Sounds pretty good. The Sonatina main page 'sounds' like the project is closed and refers to two other ones, one you mentioned and the VSCO.

In those times when I used some soundfonts I used qsynth. I don't know if it handles sfz files too. I guess it's still around. I reckon it's a front end for fluidsynth. I have heard of a free sfz player by Plogue, sforzando. Requires linVST though. I never had a good experience with Linux Sampler.

Yes qsynth is a front end for fluidsynth which plays sfz.
Sfz has a samples directory and it has the actual sfz scripts which refer to these samples. This way yes it's "hardwired" but lets you edit all aspects with a simple text editor - very Linux like I find.
Sforzando never convinced me and its Windows VST anyway and I don't want to add another layer of complexity just to make a thing sound.
Linuxsampler actually runs good and stable and is available as LV2 instrument. If you load it into a MB track and open qsampler you will find a corresponding track there where you can load your sound font. Plus it knows sfz and sf2 and gig! Plus it's client / server! That's the future for me, sfzero was only thought for bridging the time...
Here's a guy who makes full use of this. Scroll to the middle of the interview - "What is your typical workflow when making music?":
http://libremusicproduction.com/articles...rian-bador

Here's a quick Sakura example that I just did now using a cello. Sakura is basically modeling a string, although giving sound config options.

Sakura Cello

I love the idea of physical modeling, with a little fantasy you can create impossible instruments or create some which you later can build physically and see if they sound like predicted
I don't have much desire to exactly imitate existing instruments while I can have excellent samples available.

Anyway, nice to chat... back to work
MMM

[jonetsu]

Here's a guy who makes full use of this. Scroll to the middle of the interview - "What is your typical workflow when making music?": http://libremusicproduction.com/articles...rian-bador

Come on. 4 computers to make music. Funny though, at the beginning, how he goes from seeing God in everyone to everyone pays what he wants for music on his Trust Music web site where one can only get in upon invitation.

I love the idea of physical modeling, with a little fantasy you can create impossible instruments or create some which you later can build physically and see if they sound like predicted
I don't have much desire to exactly imitate existing instruments while I can have excellent samples available.

So far, I could see they offer different approaches. With samples, it's possible to derive something very, very different. I did some experiements using Redux/Renoise in which the result had nothing to do with the original sample. With modeling OTOH, the expression can be very variable. The expressiveness can be modulated by a lot because the sound does not already exists.

I was reading about additive synthesis and saw that the person who created Discovery, Vertigo, Phantom, and other DiscoDSP products, Aguru, died inb 2007 in a car accident. He also created Direct Wave (Image Line) which I have tried yesterday. There's a SSO version adapted transient-wise for DW. I got a saxophone sample which is amazing. But then again, samplers are pretty much new to me.

Cheers.

[madmaxmiller]

Here's a guy who makes full use of this. Scroll to the middle of the interview - "What is your typical workflow when making music?": http://libremusicproduction.com/articles...rian-bador

Come on. 4 computers to make music. Funny though, at the beginning, how he goes from seeing God in everyone to everyone pays what he wants for music on his Trust Music web site where one can only get in upon invitation.

Come on. A computer to make music?
(I never got into this religious stuff anyway, was just interested in the technical things - NetJack and Linuxsampler as a server, X terminal all was in my mind for a while before I read this.)

I actually plan to have a fanless (noiseless) X terminal in the studio and the big beast in my server rack outside. And "outsourcing" the sampler makes sense to me, too, given that I have a soundfont for a grand piano (only) which is 1GB big and others will add.
Machine #4 can have office programs, browsers and the like - and can be used from another X terminal on my office desk and one in the kids' study as well.
So the music machines can be completely off the interwebs as a bonus

I love the idea of physical modeling, with a little fantasy you can create impossible instruments or create some which you later can build physically and see if they sound like predicted
I don't have much desire to exactly imitate existing instruments while I can have excellent samples available.

So far, I could see they offer different approaches. With samples, it's possible to derive something very, very different.

Sure. Look at Borderlands: https://ccrma.stanford.edu/~carlsonc/256a/Borderlands/

Cheers,
MMM

[madmaxmiller]

If not, I could open a thread about why is it that MIDI keyboards must always be reconfigured for each track when reloading a saved session ?

Back to topic
I just had some blunt time and thought I could as well do some tests. So I loaded Mixbus with 2 MIDI tracks and connected my Keyrig49. Closed Mixbus, reopened, keyboard still connected. Closed MB and stopped Jack, then restarted, reopened - KB still connected. Closed MB, stopped Jack, closed Qjackctl, rebooted computer, restarted all - MB NOT connected.
Did several stop starts of Mixbus/Jack/Qjackctl - it never failed to reconnect.
Then even did the whole machine restart again - KB connected this time no fuss.

I believe it's not necessarily Mixbuses fault, it's more that Jack or a2j sometimes represent the same device differently.
No problem with Bitwig as this refuses to accept keyboards while Jack (a2j) is running and takes care of them itself. Btw, after the Bitwig start you can safely restart a2j and make the keyboards available to the rest of the system again...
So I believe if you create a Jack patchbay along with your project you'll be at least connected. Still have to toggle "IN" per channel though...

MMM

[Frank]

Now these keyboards are they pure midi
or are they midi over USB?

I do not own or use Bitwig
But the procedure which was described before, that Bitwig detects keyboard automagically, will it detect plain Midi keyboards? Ketboards from a midi controller ?

The Jonetsu described the way Bitwig interacts with keyboards would make me crazy. I think.
I have no standard way to connect: a keyboard can connect to several tracks at once,
Or a track at the time.

Prefer to determine at each time how it connects, which channel goes where.
Most of the time the plugin determines what it does with the input.

regards

[jonetsu]

I just had some blunt time and thought I could as well do some tests. So I loaded Mixbus with 2 MIDI tracks and connected my Keyrig49. Closed Mixbus, reopened, keyboard still connected.

This will work. It works here too. The key there is and connected my keyboard.

After being connected, and until the computer is shut off (or the MIDI config not altered) it will work.

The second part is not the problem. The problem is the first part. Eg:

1) Start the computer
2) Connect the keyboard

And in a practical case when creating/composing music:

3) Repeat the connection of a keyboard 35 times in order to have all synths ready to create on all tracks. Or do them one by one as you go from one synth to the other.

And that is the problem.

This is totally unknown in Bitwig per se. Yes, the keyboards are either automatically recognized by Bitwig (going by their USB signature) or a generic interface is connected to a keyboard. Once this is done, there is no more need to connect any keyboards forever, for every existing or new session. Bitwig remembers them.

Click on a track REC button and immediately play a note. Nothing else. Click on another track REC button, play another synth immediately. Ctrl-click, play two synths. So on so forth [l]right after starting the computer and loading a 35-track project[/i].

Or with any new session. Drop in a synth, start immediately to play with any of your keyboards/device Bitwig remembers. No need to configure again, to connect again, and again, and again.

In essence this is the basis for the observations regarding this lack of functionality in Mixbus. Since I'm using Bitwig for about 2.5 years now as a musical creation tool, and Mixbus32C as a mixing/mastering tool, when I had to use Mixbus32C very recently with some linVST plugins I immediately noticed the missing user friendliness in the musical creation use case.

Cheers.

---

Now these keyboards are they pure midi
or are they midi over USB?

What ? You are still using actual MIDI cables ??

I do not own or use Bitwig But the procedure which was described before, that Bitwig detects keyboard automagically, will it detect plain Midi keyboards? Ketboards from a midi controller ?

It's actually not magic. A USB device has an ID. Any software can query this ID and see if it knows about it. If it knows about it, it might then a have a config. Think Faderport in Mixbus. Once the Faderport is connected, Mixbus will use the (user-modified) configurationb it has about this device.

Same with MIDI keyboards. Because after all, this is MIDI, traditionnal MIDI cable or not.

The Jonetsu described the way Bitwig interacts with keyboards would make me crazy. I think. I have no standard way to connect: a keyboard can connect to several tracks at once, Or a track at the time.

Exactly. Click on the REC button of a track and the keyboard will work with that track. Ctrl-click on several tracks on the keyboard will layer several synths at once. This said, Bitwig also offers specific per-track options. Like use my Prokeys88 for that track. Use my Launchpad for that other track. All devices go by names since Bitwig works with actual device names.

Prefer to determine at each time how it connects, which channel goes where. Most of the time the plugin determines what it does with the input.

I also do not want to end up playing a growling hip hop bass when I want a mellow cello. And I want this selection to be transparent as possible, easy-flowing, and ready to go on 35 tracks right after the compiter has started and an existing DAW session opened.

On the other hand, a Bach piece with a growling hip hop bass ....

Cheers.