[yoho]

Hi all,

I checked the source out to have a look, but things aren't compiling/linking. Can somebody explain?
I added ../mois32 to the searchpath, but I'm still left with a couple of link errors.

Cheers!

[TK.]

Hi,

I updated the project setup; try it again

Best Regards, Thorsten.

[sineSurfer]

Just downloaded the svn files again and compiled the seq with iOS 4.2, all ok.

Any updates on the midi connection kit front TK?

Btw, just in case anyone else hates white icons, "proper" Default screen + app icons:


 MIDIbox_ipad_default-icons.zip (461.22K)
Number of downloads: 11

Laters!

[yoho]

Thanks TK.
I can compile again... and again. But there no installing on an iPad, because I don't have an developer account. Any chance somebody could include the installable app here, or is impossible install it that way?

[Narwhal]

yoho, on 23 June 2011 - 12:08, said:

Thanks TK.
I can compile again... and again. But there no installing on an iPad, because I don't have an developer account. Any chance somebody could include the installable app here, or is impossible install it that way?

Sorry yoho, it doesn't work like that. You need security keys from both Apple and from a paid developer account to install on an iPad. You should be able to run it in the simulator without those keys though.

[nangu]

Would this help?

http://www.buildingi...ete-review.html

Someone would still have to get the keys, but it sounds like TestFlight would make it a lot easier to share the app with other people. It's free, too..

[Narwhal]

nangu, on 23 June 2011 - 17:59, said:

Would this help?

http://www.buildingi...ete-review.html

Someone would still have to get the keys, but it sounds like TestFlight would make it a lot easier to share the app with other people. It's free, too..

That's a pretty good idea nangu. But someone would need to put together a build machine that gets SVN updates, builds the code with keys that are kept private, then publishes to a group of people with TestFlight. Then there is a limit (100 devices) to the number of devices that that TestFlight and debugging builds can run on. So this is not a good solution to the general public being able to install the app. For that someone will need to actually publish the code as an app on the app store. The other downside is that it really doesn't let someone do quick builds and test, everything checked in would become a live TestFlight automatically.

On the flip side it's only $99/yr to be able do build anything you want on your iPads.

This post has been edited by Narwhal: 23 June 2011 - 18:54

[electrofux]

Hi there,

i am new to the midibox world and found this ipad app thread. A friend of mine is a App developer and i might get him to try this out. Where exactly can we find the files that are to be compiled?

Besides any news on the status of the ipad app?

Cheers
Markus

This post has been edited by electrofux: 30 January 2012 - 18:01

[TK.]

I canceled this plan because when you look on existing iPad sequencers today, you will quickly notice that the user interfaces are superior compared to that what I could ever achieve.
It isn't really worth the effort anymore.

Instead I'm now going the Lemur route and design user interfaces with this App, see also: http://midibox.org/f...ipad-via-lemur/
Since the sequencer kernel is running on a MBHP_CORE_STM32 or MBHP_CORE_LPC17 module, we can expect best MIDI timings - and this is one topic which makes the MBSEQ (and could never be achieved with an iPad)

Best Regards, Thorsten.

[findbuddha]

TK, I have been thinking about the idea of tablet interfaces for MBSeq..... How possible is it to create an entire remote interface? Especially I wonder if it's possible to do remote display (no LCDs on hardware Seq) via MIDI or OSC and still have acceptable performance.

I don't expect you to code these features, just looking for expert advice

[monokinetic]

findbuddha, on 31 January 2012 - 01:54, said:

TK, I have been thinking about the idea of tablet interfaces for MBSeq..... How possible is it to create an entire remote interface? Especially I wonder if it's possible to do remote display (no LCDs on hardware Seq) via MIDI or OSC and still have acceptable performance.

I've also been thinking about this. My main motivation is to be able to take iPad + a simple CORE / SDcard for a live set, so that I don't have to take out my whole MBSeq and risk it being damaged etc. I'd hate for all those hours of hard work soldering and building to be lost.

Anyway, I spent an evening looking at the Sysex / CC remote files in the svn/docs directory. My conclusion was that the majority of the settings that we can edit using a front panel could be edited using Lemur / iPad and some clever sysex. The main challenge would be to actually create the Lemur template, something that I won't have time for until summer holidays....

If anyone makes some progress on such a template in the meantime I'd happily be a tester

Oh and one other crazy idea I had was for a web interface to MBSeq. This could be interesting from a cross platform perspective. Lemur is iPad only at the moment, and I'm not so sure that they will port to Android. You could have nice features like a Ableton-esque grid view of all patterns showing their names, faders for the MixerMap settings etc.
TK: I'm thinking out loud here, but would such a thing be theoretically possible?

[findbuddha]

Interestingly Reaper has recently added a web interface plugin - it seems highly configurable. My suspicion is that other protocols for using MIDI / OSC will offer less latency though.

http://charlie-roberts.com/Control/ Seems to be a promising solution for crossplatform use, but it's very much less polished than Lemur. I asked the Lemur dev (Liine) if they were porting to Android and they suggested they would consider it but can't promise anything.

[TK.]

findbuddha, on 31 January 2012 - 01:54, said:

TK, I have been thinking about the idea of tablet interfaces for MBSeq..... How possible is it to create an entire remote interface? Especially I wonder if it's possible to do remote display (no LCDs on hardware Seq) via MIDI or OSC and still have acceptable performance.

I don't expect you to code these features, just looking for expert advice

Mainly there are two possibilities:

1) a 1:1 emulation of the MBSEQ frontpanel. Button/Encoder events have to be sent to MBSEQ, and MBSEQ sends LED/LCD changes to Lemur.
In fact I already implemented this into the MBSEQ V4 firmware two years ago to allow remote-control of a second "core-only" MBSEQ, but it's undocumented and well hidden in a configuration file ;)
This mechanism could also be used to control the MBSEQ remotely from an iPad once Lemur supports SysEx messages with more than 128 bytes (this limitation should be fixed in a future update)

2) creating a completely new control surface which is optimized for the iPad, because the usage of virtual encoders is too cumbersome, and at many places you would like to control certain functions with dedicated buttons (on dedicated pages) instead of clicking through the menu pages.
This is feasible as well, it could be implemented on the same way like I did for the MBCV V2, and the implementation at the MBSEQ firmware side is easy - but creating a perfect Lemur control surface for controlling all the parameters is a challenge.
If somebody with Lemur skills would like to do this, I could take over the MBSEQ firmware part.

monokinetic, on 31 January 2012 - 13:26, said:

The main challenge would be to actually create the Lemur template

Exactly!

Quote

Oh and one other crazy idea I had was for a web interface to MBSeq. This could be interesting from a cross platform perspective. Lemur is iPad only at the moment, and I'm not so sure that they will port to Android. You could have nice features like a Ableton-esque grid view of all patterns showing their names, faders for the MixerMap settings etc.
TK: I'm thinking out loud here, but would such a thing be theoretically possible?

No, this isn't possible, the HTML protocol isn't made for this, and there are no RAM resources free for a webserver anyhow.

findbuddha, on 03 February 2012 - 02:33, said:

My suspicion is that other protocols for using MIDI / OSC will offer less latency though.

Haha, totally wrong! ;)
MIDI (over USB) has the best performance, OSC is a bit slower since it requires additional string based parsing at the receiver side, and all other protocols are not compatible to MIDI synthesizers or DAWs anyhow, and would require some kind of proxy/bridge which will add latency.

Best Regards, Thorsten.

[findbuddha]

TK., on 05 February 2012 - 20:11, said:

Mainly there are two possibilities:

1) a 1:1 emulation of the MBSEQ frontpanel. Button/Encoder events have to be sent to MBSEQ, and MBSEQ sends LED/LCD changes to Lemur.
In fact I already implemented this into the MBSEQ V4 firmware two years ago to allow remote-control of a second "core-only" MBSEQ, but it's undocumented and well hidden in a configuration file ;)
This mechanism could also be used to control the MBSEQ remotely from an iPad once Lemur supports SysEx messages with more than 128 bytes (this limitation should be fixed in a future update)

2) creating a completely new control surface which is optimized for the iPad, because the usage of virtual encoders is too cumbersome, and at many places you would like to control certain functions with dedicated buttons (on dedicated pages) instead of clicking through the menu pages.
This is feasible as well, it could be implemented on the same way like I did for the MBCV V2, and the implementation at the MBSEQ firmware side is easy - but creating a perfect Lemur control surface for controlling all the parameters is a challenge.
If somebody with Lemur skills would like to do this, I could take over the MBSEQ firmware part.

Hadn't seen the MBCV V2 development.... looks great!

My suggestion is a combination of 1 and 2. Have a main interface similar to MBSEQ frontpanel, but keep a space for a 'pop up menu'. Press and hold on the virtual encoder (or other button) and all available options pop up in the menu area. Then you could have direct selection of notes etc.

I will investigate the purchase of a second hand iPad. If I get one I'll help with the Lemur stuff.

[findbuddha]

No iPad yet (I will wait till release of iPad 3 in early March), but here are some early tinkerings:

 Clipboard01.jpg (113.22K)
Number of downloads: 43

The encoders work with vertical touch movements - I think this would be comfortable to use but I haven't actually tried it myself. For some actions quick selection options could come up in the container at bottom right.

[findbuddha]

TK, I've bought an iPad and after much hassling with wireless network configs I've gotten the BLM Lemur app working.....

Do you have any thoughts about the Seq remote layout and its communication protocol?


Cheers!

[TK.]

Get some inspirations from the MBCV Lemur template: http://svnmios.midib...v_v2%2Flemur%2F
It has nothing to do with the future hardware control surface, but entering parameters is optimized for touchscreen usage.
You could try out the communication with the core by uploading the MBCV app - just contact it via Ethernet or USB-MIDI->Lemur Daemon

The protocol is based on NRPNs, because they are easy to decode within Lemur.
Basically we've get and set functions for each parameter, see also: http://svnmios.midib...ents%2FMbCv.cpp

The routing to the CV channels is done here:
http://svnmios.midib...Environment.cpp

Also look into the NRPN dump function, because it's an essential feature: whenever a certain CV channel is selected on the Lemur app, MBCV dumps all parameters to the Lemur template in order to update the control elements. This ensures, that all knobs are up-to-date, even if the user loads a new patch.

For a similar MBSEQ approach, replace CV channel by SEQ track -> the same concept can be used!

Best Regards, Thorsten.

[findbuddha]

Thanks TK,

I've put a few hours in (noob here).

I can successfully set and retrieve track CC info via SEQ_CC_MIDI_Set() and SEQ_CC_Get(), and it's also easy to utilize SEQ_UI_REMOTE_MIDI_Keyboard() for simple button functions like Play, Stop etc.

I'm wondering how best to set and retrieve other data for which there is no established interface method. For example I tried to use SEQ_UI_Button_DirectTrack() to directly switch tracks but got strange results. (Also I feel nervous littering my code all through yours ...)

Cheers

[findbuddha]

^^BUMP^^



I've been looking at some other step sequencers for workflow ideas (five12 Numerology and Sequencomat (a Lemur controller for a Max/MSP sequencer). I like particularly Numerology's ability to directly see/edit several 'parameter layers' from the same screen.

Here's a rough idea how it might look in midibox/lemur land:

 Clipboard01.jpg (78.25K)
Number of downloads: 23

[findbuddha]

Managed to accidentally delete my source containing my NRPN implementation. At least I can reverse engineer from my Lemur patch.

Been working on a different tangent. I found the Sysex remote feature already present in the Seq. I've got working button and encoder input from the Lemur, along with partly working LCD feedback. The Lemur app doesn't have some of the required characters for Seq display so I'll have to find a work around for that.

So far I haven't been able to decode the LED stream coming from the Seq. It also seems to put out enough data to dramatically increase CPU usage on my laptop. I guess I'll need to slow down its output rate.


Here's a copy of my current Lemur patch (caution, WIP, hacking needed for most functions to function ):  MBSeq_Combined5.zip (11.37K)
Number of downloads: 1

And an export of TK's BLM patch (in a container) as a .jzlib for easy importing into other Lemur patches:  blm.zip (5.23K)
Number of downloads: 0