MIDIbox SEQ V4 Release + Feedback

[grizz]

I updated to V4.0BETA21 successfully today! :-)

The only issue I had was these messages:

[101205.130] [sEQ_FILE_C] ERROR: unknown parameter: MIDI_IN_Channel

[101205.132] [sEQ_FILE_C] ERROR: unknown parameter: MIDI_IN_Port

[101205.134] [sEQ_FILE_C] ERROR: unknown parameter: MIDI_IN_TA_Split

[101205.136] [sEQ_FILE_C] ERROR: unknown parameter: MIDI_IN_TA_SplitNote

Is this a problem? I'm thinking it has to do with settings that I have that are no longer compatible or relevant with the new version?

Cheers,

Grizz

[TK.]

Is this a problem?

No - don't panic! ;)

I'm thinking it has to do with settings that I have that are no longer compatible or relevant with the new version?

Yes, and once the MBSEQ_C.V4 file will be written again, these parameters will be replaced by

MIDI_BUS_Channel 0 1

MIDI_BUS_Port 0 0

MIDI_BUS_Lower 0 0

MIDI_BUS_Upper 0 127

MIDI_BUS_Options 0 0x00

MIDI_BUS_Channel 1 2

MIDI_BUS_Port 1 0

MIDI_BUS_Lower 1 0

MIDI_BUS_Upper 1 127

MIDI_BUS_Options 1 0x00

MIDI_BUS_Channel 2 3

MIDI_BUS_Port 2 0

MIDI_BUS_Lower 2 0

MIDI_BUS_Upper 2 127

MIDI_BUS_Options 2 0x00

MIDI_BUS_Channel 3 4

MIDI_BUS_Port 3 0

MIDI_BUS_Lower 3 0

MIDI_BUS_Upper 3 127

MIDI_BUS_Options 3 0x00

[/code]

and the error messages will disappear.


Best Regards, Thorsten.

[vcfool]

To your initial request: since each step can only store 128 values (0..127), there is no possibility to disable a CC without reducing the original value range.

Ah ok, now I get it.

I was noticing some odd behaviour when trying to use menu-->trigger to set a drum tracks trigger layers. I wanted to to select a different trigger layer for my drum track instead of accent and I was getting stuck. When I hold menu and select trigger GP, I am given the same things as holding trigger layer C. BUT I was able to at some point use menu-->trigger to select the layers such as accent, roll and nofx on the same drum track. Switching through track groups and tracks then switching back produced some results but I am still unsure of what is going on. I haven't dl'd the 21 beta yet.

So, what should I say to this "issue description"? ;)

I think I know what Echopraxia means because I have noticed it too. I will try to explain:

When you enter to the Trigger Assign page, (sometimes) the page will not be displayed correctly. You get the Trigger Layer Select page instead.

If you then press (and release) Trigger Layer C button the Trigger Assign page will show.

Pressing the Trigger Layer C button on a Drum track will do nothing. You have to switch to a Note track first. But once the button has been pressed, you can go back to the Drum track and it will also show the Trigger Assign page correctly.

Hope I made myself clear.

Thanks again.

[strophlex]

Are you a Mac or Windows user?

Could you please check with a MIDI Monitor that only USB1 sends the MIDI clock?

Best Regards, Thorsten.

I didn't have time to test until now and I installed beta 21 before trying this out more. Now it works. Thanks!!!

[Gridracer]

Hello Thorsten,

I noticed that the chaselight is no longer working propperly with beta21 on my singlecolour 4x16 BLM:

Active steps get inverted (LED goes off) when the chaselight passes -> OK

But inactive steps are not inverted (LED stays off) -> NO CHASELIGHT VISIBLE.

(On the GP LEDs the chaselight works well, also on the BLM if i upload beta20)

Here if required my relevant settings in the config:

BLM_ENABLED     1

BLM_DOUT_CATHODES_INV_MASK 0xf0

BLM_DOUT_DUOCOLOUR  0

By the way, i really enjoy having 4 transposer tracks since beta 21 :)

Best Regards,

Gridracer

Edited May 26, 2010 by Gridracer

[grizz]

I have noticed something strange that is happening with beta 21 on my SEQ. When I try and create a new session I get a SD card error. Unfortunately I don't have my sequencer here with me as its at my practice space because I have a show tonight. It gives some sort of FAT32 message. I can however open and save already existing sessions. I will edit this post tomorrow with the exact error message that I get.

Cheers,

grizz

[TK.]

Alright, a new release is available now:

MIDIboxSEQ V4.0beta22

~~~~~~~~~~~~~~~~~~~~~


  o running status optimisation can now be optionally disabled in

    the MBSEQ_HW.V4 file for MIDI devices that don't fully comply

    to the MIDI specification.

    Search for "RS_OPTIMISATION" in the MBSEQ_HW.V4 templates for

    further informations.


  o chaselight issue for mono colour 4x16 BLM fixed


  o trigger assignment page now works as intended in drum mode


  o new MIDI Bus parameters now loaded correctly after session change


  o another fix for crash if preset name is longer than 7 characters

[/code]


grizz: I need to know the error number if this happens



Hint for Windows7 users: currently we (Phil and me) are very sure that the Windows7 USB driver provided by Microsoft cannot handle SysEx messages correctly in MIOS Studio if MIOS32 is configured for more than one USB MIDI port, because we have a counter example where SysEx transfers are perfectly working under Win7 with an alternative USB MIDI Driver.


But I found a workaround that helps to get SysEx running again: open MIOS Studio, select the MIDI I/O ports of your MBSEQ V4 and type "help" into the MIOS Terminal multiple times until the complete help page is displayed:




This allows you to upload a new firmware via USB w/o manually selecting bootloader mode.



Btw.: virtual BLM16x16+X under windows anyone? :)



Best Regards, Thorsten.

[Gridracer]

WOW!

TK, your speed in development is unbelievable!

I am so happy to have the chaselight back, and it's also nice to be able to upload code via USB now... so much faster :)

Thank you very much!

Just tried creating a new session, works fine here.

Best Regards,

Gridracer

[grizz]

Okay here is the error that I get. It flashed really quick so it took some time...

SD Card Error Fat32

E 3 (Fat FS: D1)

Cheers,

grizz

[daleong]

Beta22 is acting very strange for me. My sequencer has only been completed for a few weeks, and I have only created 1 session that contains a drum track, but it plays incorrectly now (the "note" tracks play as they should). Although the length is set at 16, stages 9-16 don't play, and every time it loops through stages 1-8, it skips stage 6! The stage LEDs are moving at half the speed that the track is playing. I checked the clock divisor and it displays the correct value. It seems as though there are also some random drums thrown in sometimes too. Has anybody else experienced anything like this?

[Echopraxia]

Great work :thumbsup: Beta22 was just what I needed.

I was transposing a drum track and like what semitone was doing for me on that particular track. I want to be able to always go to semitone first when I select the transpose page but only want it to change this setting per track, per pattern. Can we select the semitones first per track (instead of octave) without having to press the select button? I feel this would make using different keys on different tracks more instinctive because of lack of pressing select.

As an alternative to this I was also thinking using the last known placement of oct-semi from the trigger page.

so what ever you had selected last will still be active on when you come back to that page ( per track of course.

Regards,

echo

[Echopraxia]

I was also thinking about rocket travel.

and being able to record the transpose movements from the panel buttons or from ext keyboard :rolleyes:

I would take a simple one note chord and use the transpose function on the front panel to make melodies. I guess I can always lengthen the track out and copy paste and change the bars accordingly.

Regards,

echo

[TK.]

Okay here is the error that I get. It flashed really quick so it took some time...

SD Card Error Fat32

E 3 (Fat FS: D1)

E3 means that the SD Card wasn't mounted correctly.

I've currently no explanation why this error could only occur while creating a new session, and why other file operations are still working. :-/

Do you have an alternative SD Card to try out if sessions can be created there?

Beta22 is acting very strange for me. My sequencer has only been completed for a few weeks, and I have only created 1 session that contains a drum track, but it plays incorrectly now (the "note" tracks play as they should). Although the length is set at 16, stages 9-16 don't play, and every time it loops through stages 1-8, it skips stage 6! The stage LEDs are moving at half the speed that the track is playing. I checked the clock divisor and it displays the correct value. It seems as though there are also some random drums thrown in sometimes too. Has anybody else experienced anything like this?

The reason for this "incompatible" behaviour is, that trigger layer assignments now also take effect on drum tracks.

As you can read in the history of this thread, people were confused that this didn't happen before - I already feared that the result could be new confusion. ;)

However, there is a simple solution: press MENU+TRIGGERS and assign "Gate" to trigger layer A, assign "Accent" to trigger layer B (if available), and disable the layers ("-") for all other trigger functions.

I was transposing a drum track and like what semitone was doing for me on that particular track. I want to be able to always go to semitone first when I select the transpose page but only want it to change this setting per track, per pattern. Can we select the semitones first per track (instead of octave) without having to press the select button? I feel this would make using different keys on different tracks more instinctive because of lack of pressing select.

As an alternative to this I was also thinking using the last known placement of oct-semi from the trigger page.

so what ever you had selected last will still be active on when you come back to that page ( per track of course.

Would it be sufficient to keep the last setting regardless of the track selection?

Because I fear new error reports such as "strange behaviour of Octave/Semitone selection when switching between tracks".

There is no other page that would behave this way

and being able to record the transpose movements from the panel buttons or from ext keyboard :rolleyes:

I would take a simple one note chord and use the transpose function on the front panel to make melodies. I guess I can always lengthen the track out and copy paste and change the bars accordingly.

So, you mean a possibility to play notes directly with GP buttons like in the BLM16x16+X video, with optional "force-to-scale", but w/o velocity?

How to enter this mode - special button combination (which one?) or with a dedicated button?

Best Regards, Thorsten.

[daleong]

Thanks Thorsten - that did it! I read those posts, but it just didn't click for me...

I'm really loving this sequencer! I had been shopping for a good step sequencer for some time, and had made up my mind to buy an Octopus, but before I had saved the required funds, I found the MIDIbox. Saving money was nice, and the feeling of accomplishment you get from completing a project of this scope is great, but the real reward is the sequencer itself. The feature set compares quite favorably with anything else out there. It is a joy to work with and has quickly become the centerpiece of my studio! Thank you and the whole MIDIbox community for making this possible!

- Dale

[grizz]

I used another SD card, and here are the steps I took:

I have two SD adapters and two micro SD cards at my disposal, both of which are soldered correctly and worked prior to upgraded from Beta20 to Beta21.

Tests:

1. I removed the microSD card (SD1) from the adapter attached to J16 and inserted the second (SD2) microSD card.

Result: Same error

2. I removed the adapter from J16 put the other adapter in place and inserted SD1 into it.

Result: Same error

3. I inserted SD2 into the adapter.

Result: It successfully created a new session.

4. I switched adapters again, and this time put SD1 in place (this is the original configuration that was giving me problems)

Result: It successfully created a new session.

So it seems to be intermittent. Its possible that the IDC connector got jingled around moving it between my house, practice space, and shows... But I would think that it wouldn't switch between already created sessions if that was the case. I'll mess around with it a bit more, but I just got a new resistor network for the AOUT_NG jumper, and I want to try that out.

Cheers,

grizz

Edited May 29, 2010 by grizz

[Gridracer]

I have a question regarding the arpeggiator (and transposer)

I noticed that the arpeggiator (and transposer) behave different when they are contorlled from external eg. keyboard or from an internal track:

- If the contorl notes come from external, the settings HOLD and RESTART work as they are described in the manual

- If the contorl notes come from an internal track, the arpeggiator (and transposer) always behaves like if HOLD is ON and RESTART is OFF,

no matter to which value they are actual set.

Is this behaviour intended?

Best Regards

Gridracer

[Gridracer]

Sorry to disturb again,

but I noticed anotehr issue with beta22 and the 4x16BLM:

onyl on step 1-16 the LEDs for active steps get lit, on step pages 17-32 and higher the LEDs stay off

whether the step is active or not. Chaselight and buttons work correct, also the GP LEDs

Best Regards,

Gridracer

[TK.]

I will check this soon!

Best Regards, Thorsten.

[stuartm]

Finally, I finished upgrading my SeqV3 to V4!

I'm still running it in the "old" wooden case to find out which additional buttons are useful for my workflow, then it will get a nice Schaeffer frontpanel.

I did a few testruns, especially the Live Recording mode, work very smoothly.

I'm quite happy to be able to play in more-than-3-note-chords with velocity and length :smile:

As soon as I'm familiar with it and having done the new casing, I will definately do more demo videos!

Oh, and in terms of feedback:

Building the Core32 was a real joy!

When I browsing through the "About this MIDIbox" menu, somehow the button assignments got scrambled (i.e. "exit" was "Follow mode", "Select" became "F1"), i rebooted and it was all perfect again, I couldn't reproduce the error.

Also I once had a very brief SD card error displayed, but it did seem to affect anything.

Edited June 7, 2010 by stuartm

[Echopraxia]

Would it be sufficient to keep the last setting regardless of the track selection?

Because I fear new error reports such as "strange behaviour of Octave/Semitone selection when switching between tracks".

There is no other page that would behave this way

I would say if the transpose page stayed on octave or semi, whichever being the last setting used globally, would not do much good. I would tend to forget that I set it to semi and make a mistake during jamming. Anyone else?

So, you mean a possibility to play notes directly with GP buttons like in the BLM16x16+X video, with optional "force-to-scale", but w/o velocity?

How to enter this mode - special button combination (which one?) or with a dedicated button?

I was not thinking of that. I was thinking about being able to record stuff like cc movements or transpose switches into the track. Say record cc messages from the mixer map to the track and have it play them back automation style. Elektron machine drum does this I think. For transpose I was thinking that you would record enable track for live recording maybe and press the transpose keys and watch your transpose keys automate on the second pass in the pattern since it was recording the changes. This may be a big effort and hassle and is not a priority at all.

Regards,

echo

[stuartm]

I would say if the transpose page stayed on octave or semi, whichever being the last setting used globally, would not do much good. I would tend to forget that I set it to semi and make a mistake during jamming. Anyone else?

Also referring to your first post about this, I would also prefer to have Semitone selected first.

How about a setting in the config file where you could select if you'd have Octave or Semitone selected first, as a global setting?

On another issue, TK:

I think the external Restart is not working as perfectly as it was on the SeqV3. I have the V4 now in the same MIDI setup as the V3 before.

When hitting Start, everthing is fine and in sync. Stop, Start. Again in Sync.

Now while the Seq is running, i hit the external restart ("ext. restart request" appears), the start signal is sent, but the slave machines are just slightly out of sync.

It might be just a few ms, but I can hear it immideately.

This happens when a stopped slave machine is started again as well as a running slave machine is "re-synched" (tried with different machines).

Also, when I request an ext. restart while the SeqV4 is stopped (message appears) and then hit start, the slaves are slightly out of sync.

This slighty discrepancy stays, and the only way to get rid of it was to hit Stop and Start.

I guess that due to the MIDI handler in the Core, the function is implemented differently in the V4 than in the V3.

If you need more info or some test data, I'll be glad to provide.

This is THE killer feature of the MBSeq for me, so let's hope we can get this sorted out! :thumbsup:

[stuartm]

Here's an MP3 demo of what I mean:

synctest_bd.mp3

two layered kick drums, one triggered directly via the Seq,

the other from the Machinedrum's internal sequencer running as slave of the MBSeq clock.

First 8 kicks are more or less perfectly in sync, on the 9th kick the ext. restart happens and you can hear that they are slightly out of sync.

I've tested different ports on the core, direct connections and different machines. All the same.

I'll have a look at the source code, C is a bit more readable for me :blush:

[TK.]

Thanks for the input! :)

I just came back home and won't have the time today to check firmware changes by myself.

However, I think that I know from where the delay is coming from: MIDI clock start was sent directly after the MIDI clock event, but it should be sent before.

Could you please try following temporary release if it works at your side?

If it still doesn't work, I will test this with various sequencers at my side... later.

Track transpose page now always starts with semitone selection.

Due to conceptional reasons it isn't possible to store the last setting, nor to make it configurable in a config file.

The issues reported by Gridracer are not fixed yet.

-> http://www.ucapps.de/mios32/midibox_seq_v4_0beta23_pre1.zip

There are some preparations for OSC which hopefully won't conflict with existing functions.

They are currently only working on an iPad, MBHP_ETH support will be added this weekend.

Best Regards, Thorsten.

[stuartm]

Perfect!

That fixed the sync issue!

I think I owe you a beer :thumbsup:

[Gridracer]

Hi,

I would like to propose some improvement of the handling in the case when you have assigned more than one layer to"NOTE":

At the moment, when you have assigened several layers to "Note"

and you have assigend note numbers on the same step on those layers,

(Which is really great for programming chords)

it occures that when you decrease the note number on ONE layer to "___" in order to deactivate that note,

the WHOLE STEP will be deactivated, with all the remaining notes on the other layers.

Ok, you can bring them back in by turning the encoder forth and back one dent on those layers.

But it would be really nice if the step would stay activated as long as ANY Layer contains a valid (note) value.

That would be a great wish from my side as i use multiple note layers a lot :)

Finally just to mention I noticed a cosmetic thing (may be already solved in beta 20+)

-For "probability" the encoder works inverted: CW increases, CCW decreases values

Best Regards

Gridracer