Jump to content

MIDIbox SEQ V4 Release + Feedback


Recommended Posts

Posted

It's true what k2z3k0 says, SELECT + CLEAR doesn't clear Layer A if it's a Note layer. I tried setting Layer A to Roll, that was cleared without problems, and also other note layers (Layer E etc.) cleared fine.

Posted (edited)

The latest beta that allows Record, Follow and Move to be set with the F keys is great and has been a huge help.

I had one feature request that I couldn't find using search: Would it be possible to move a step forward and backward by ticks when a modifier button is used (like 'select' or 'fast')  and only when Move step is enabled? So really it would just be another state of that Move mode. There are a few other requests for "micro-shifting" like that, and I know the Delay layer is already there for shifting forward in time, but it seemed like an appropriate place to allow that functionality.

in the display, Step would change to Tick (and the Tick value).

Edited by v4
clarifying request. further clarification.
  • 4 weeks later...
Posted (edited)

I have a bug (maybe) and also a request.

On drum tracks can we have (2*64/2*128), so 64 steps long, 2 parameter layers + 2 trigger layers with 8 instruments?

Also, when I change a drum track (128/2*128) parameter LayerA to CC, GPK10 doesn't let me select the CC number. It just shows "---"

I'm being selfish here as I have an LXR, with 7 instruments and I'd like to throw 2 layers of CC at it.

Edited by mongrol
Posted
Quote

On drum tracks can we have (2*64/2*128), so 64 steps long, 2 parameter layers + 2 trigger layers with 8 instruments?

ok, I added this to the wish list

Quote

Also, when I change a drum track (128/2*128) parameter LayerA to CC, GPK10 doesn't let me select the CC number. It just shows "---"

CCs are actually not supported for drum tracks, it's an error in the UI that it allows to select this.

The pattern structure has no free space to store the CC numbers (because the appr. slots are already used to store Note number, accented and normal velocity)

Therefore I added to the wish list, that the UI shouldn't display the CC selection

Only solution that I see at your side: use another track to play the CCs
I know it's cumbersome, but there won't be a quick solution to solve this

Best Regards, Thorsten.

 

Posted

I realized that "force to scale" is forcing all "non matching" notes 1 semitone UP. This leads to non expected chords when I work with  chord progressions:

The minor chords end up being 1st 4th 5th instead of 1st 3rd 5th.

Forcing has to be one semitone DOWN to solve this.

For example a Cmaj scale with a root chord of CEG and the progression vi | V | iii | IV  gives me ADE | GBD | EAB | FAC instead of ACE | GBD | EGB | FAC .

(I do this with a note track, but I think its the same with a chord track)

If I choose a root chord of DFA and transpose it 2 semitones I can fool the SEQV4.

But it would be nice if this could be changed! Thanks!

 

 

Posted
11 hours ago, workspace said:

The minor chords end up being 1st 4th 5th instead of 1st 3rd 5th.

Forcing has to be one semitone DOWN to solve this.

For example a Cmaj scale with a root chord of CEG and the progression vi | V | iii | IV  gives me ADE | GBD | EAB | FAC instead of ACE | GBD | EGB | FAC .

Small music theory understanding question... 

If notes do not match during the force-to-scale process, Is an option of "transpose a halftone up" / "transpose a halftone down" always sufficient? Doesn't it depend on the input notes? And aren't there any cases, where some notes should be transposed up and some down?
Because, technically, such a switch could probably be implemented in the options without much pain, but I am wondering, if it will solve all "scaling" problems (or only solve some problems for certain source scales/chord progressions)?

Many greets,

Peter

Posted (edited)
5 hours ago, Hawkeye said:

Small music theory understanding question... 

If notes do not match during the force-to-scale process, Is an option of "transpose a halftone up" / "transpose a halftone down" always sufficient? Doesn't it depend on the input notes? And aren't there any cases, where some notes should be transposed up and some down?
Because, technically, such a switch could probably be implemented in the options without much pain, but I am wondering, if it will solve all "scaling" problems (or only solve some problems for certain source scales/chord progressions)?

Many greets,

Peter

I agree, this is a request that should be looked at carefully. My example is just for the canonical chords on a Cmaj scale.

I looked at other software how the approach the problem:

Abletons "scale" effect allows fully customization. on a 12x12 input matrix every input semitone (x-axis) has an output semitone (y-axis). A chromatic scale would have the marked (orange) dots on the diagonal. As you can see the Cmaj presets is forcing the semitones down (but this can be changed by the user) :

Scale.jpg.0f04e6334b5c64af52787fc37981aa

This would obviously be the most excellent solution. Instead of predefined scales we could have a table for input/output notes with presets for defined scales.

Another approach is this one: http://six4pix.com/arpie/manual.html#span

It lets you choose if you want to sharpen (force up) or flatten (force down) the notes.

Regards Michael

 

Edited by workspace
  • Like 1
Posted

Thanks for the feedback, and it is always nice to learn something new!

I sometimes had similar problems with transposing and FTS sounding strange, so maybe more users would benefit from such a configuration option and vote for this feature?

In the end, TK. needs to weigh this feature request against the other open requests and the wishlist and all other time constraints. If everything fails, I'd be willing to have a look at it (in a few weeks) - the option to sharpen or flatten if out-of-scale seems to be quite standard (also from the six4pix manual), so this would probably be a good (and easy) start...

Many greets!
Peter

Posted
2 hours ago, Hawkeye said:

Thanks for the feedback, and it is always nice to learn something new!

I sometimes had similar problems with transposing and FTS sounding strange, so maybe more users would benefit from such a configuration option and vote for this feature?

In the end, TK. needs to weigh this feature request against the other open requests and the wishlist and all other time constraints. If everything fails, I'd be willing to have a look at it (in a few weeks) - the option to sharpen or flatten if out-of-scale seems to be quite standard (also from the six4pix manual), so this would probably be a good (and easy) start...

Many greets!
Peter

Likewise! I'm always happy to give something back to this great community.

An an additional feature would be nice.  I in the meantime I just work with this small change to the code:

Changing the Major Scale in seq_scale.c line 63 from

{{            0,      2,      2,      4,      4,      5,      7,      7,      9,      9,      11,     11,     "Major               " }},

to

{{            0,      0,      2,      2,      4,      5,      5,      7,      7,      9,      9,     11,     "Major               " }},

I havent checked it out but I think this should solve the problem for me.

Regards Michael

 

Posted

Diatonics chords are quite tightly linked to the scale being used.  A standard major or minor chord has a 1st, 3rd, 5th.  in C major (Ionian)  C has 1=C 3=E 5=G  and becomes major , D has 1=D 3=F and 5=A, so it is minor.  An example for 7th:  C: 1=C 3=E 5=G 7=B(H)  which is major seventh.  G: 1=G; 3=B(H); 5=D; 7=F which is dominant; D: 1=D 3=F 5=A 7=C which is minor 7th.  This applies for all chords without tensions.   6th, 9th, 11th 13th and all the other jazz stuff. Tension chords have non-diatonic 5th (#5 or b5) and 9th (b9 and #9)./

So a  force-chord-to-scale function should look rather at the 1-3-5 principle than  per-note settings for up/down forcing.  in the code chords are specified in half tone steps, but the 135 principle makes far more evident what is going on in the relation between chords and scales.

 

Posted
10 hours ago, EsotericLabs said:

So a  force-chord-to-scale function should look rather at the 1-3-5 principle

Yes that's exactly my point. But rather than in 1-3-5, I prefer thinking in intervals of semitones 3:4 / 4:3, and if the mapping is correct we don't even need a "intelligent" function:

Let me explain: (although I suppose you know all this. but then it might be also of interest for others ;)

Every (classical) chord is made by playing every second note of the scale. Because our scale has steps of semitones ( mi-fa + ti-do or EF + BC in C-maj) and whole tones, we end up with a stack of major 3rds (4 semitones) and minor 3rds (3 semitones).

A major chord is defined by 3 notes with the intervals 4:3 (semitones) , a minor chord has the revers sequence: 3:4

With a major scale we get major chords on the I IV , V minor chords on the ii, iii, vi and the diminished chord (3:3) on the vii

We can extend it by a fourth tone and we get Cmaj7(4:3:4), Dm7 (3:4:3), Em7 (3:4:3) Fmaj7(4:3:4) G7 (4:3:3) Am7 (3:4:3) and Bø7 (3:3:4)

The beautiful thing is that all these chords sound harmonic together, and almost every song follows one of the classical patterns. Here's a nice database https://www.hooktheory.com/trends

Now, if we want to play one of these patterns we can use a chord track and transpose the chords. But then we have to know if the base tone will be one that has a Major(7) chord or a Minor chord. If we pick the wrong one or if we want to change the pattern on the fly we are in trouble.

But if we take a major (or maj7) chord transpose it and force it to scale we will end up with the correct chord all the times IF the mapping is correct (that means if the notes that are not on scale are forced down. And we can change the pattern on the fly.

This will work for every case on a maj and minor scale. the program doesn't need to be "aware" that it is playing a chord (1-3-5 or even 7):

00 01 02 03 04 05 06 07 08 09 10 11 00 01 02 03 04 05 06 07 08 09 10 11 00
 C     D     E  F     G     A     B  C     D     E  F     G     A     B  C 

 |           |        |           |                                           CEGB (maj7)
       |        |< |        |        |< |                                     DFAC (m7)
             |        |< |        |        |< |                               EGBD (m7) 
                |           |        |           |                            FACE (maj7)
                      |           |        |        |< |                      GBDF (7)
                            |        |< |        |        |< |                ACEG (m7)
                                  |        |< |     |< |        |< |          BDFA (ø7)
                                        

|< | : notes forced down

Right now as I explained in my previous post we get a chord with the intervals 5:2 for every minor chord. that's a perfect forth and a major second. both having a quite different harmony than what we want.

This all of course, only applies to all the western scales. But then I don't know about the harmonic of the Indian and Asian ones.

Regards Michael

 

Posted (edited)

Yes, I agree.  The 135 system is another way of presenting the 4 and 3 semitones structure, that is true. And I now see your point that notes must be forced down to stay in diatonic harmony instead of up.

On the other hand, many songs do not perfectly follow diatonic harmony and that gives them their spice. Twelve bar blues has I IV and V dominant.  Not to mention jazz progressions jumping through many ii-IV movements and changing root note every few bars.  Tension chords deliberately break diatonic harmony. How to cope with that?

But maybe I digress, who plays complex jazz on the seq after all? if there's interest, us theoryheads might think of something to help the programming.   Indian and middle eastern music are not very chords progression oriented. Klezmer harmony is strongly harmonic minor oriented (i, iv, V7) with some occasional twists.  But that's probably too far out for now.

 

Edited by EsotericLabs
Posted (edited)

Haha maybe yes. Let the seq improvise with the random function.  I'd love to hear your tune...

But now I come to think about it..  Is there a way to have the random note generator follow the rhythmic patterns in the Jam page? That would probably make seriously cool seqjazz..

 

Edited by EsotericLabs
Posted

A scale editor is still on the Wishlist:

----------------------------------------------------------
http://midibox.org/forums/topic/19701-blm-feature-requests/?do=findComment&comment=171522
Scale Editor
----------------------------------------------------------

it will probably be provided for MBSEQ V4+, since it will cause some additional RAM consumption.

I guess that this will solve the request.

Alternatively some additional scale entries could be provided for "flattened scales", but this approach would be less flexible.

Best Regards, Thorsten.

Posted
2 hours ago, TK. said:

A scale editor is still on the Wishlist:

Great! I did not see this.

Until then I can change the scale in the source code for my custom firmware.

regards Michael

  • 4 weeks later...
Posted

dear Midi Freaks

while working a lot lately with seq4 i have two ideas

1.

could be Jam Recording Single and Multi channel ?

SINGLE CHANNEL MODE:incoming data is recorded to Selected Active track

MULTI CHANNEL MODE:incoming data is recorded on same channel that is coming from

 

if not-would be possible to make Selection of Active track remotely controlled by midi and sensitive to current midi channel of incoming data?

2.

idea for midi BPM to PITCH converter 

 

I notice if loop is on 170 bpm and pitch bend range of that loop is 12(up&down), and seq is also 170 bpm then if i have track layer PItch bend set to 32,64 or 96 its gonna slow or speed up loop exactly double

so i thougt maybe there could be a function included in each track that

if on checks bpm and accordingly to ,calculate and change Pitch bend value

so loop is always on exact pitch(speed) to match bpm of seq

 

bests

 Kazik

 

 

 

 

 

Posted

Hi Kazik,

to 1) I added this request to the MBSEQ V4+ wish list - to make this possible, more RAM will be required which is only available for the MBHP_CORE_STM32F4

to 2) this would require an interpolation algorithm, added this to the MBHP_CORE_STM32F4 wishlist as well (it will consume more compute power which should be available for this core - let's try it out!)

Best Regards, Thorsten.

Posted (edited)
11 hours ago, mongrol said:

What is mbseq v4+? Is it an upgraded core? Is there a specificv4+ wishlist or just the normal wishlist in the repo?

MBseqv4+ is a firmware version which will only work on the newer core (STM32F4) because the old one (LPC1769) doesn't have the resources to run it. Basically, all requests for resource-intensive features that are likely to be too much for the LPC1769 based core have just been put into the MBseq4+ basket for some time now (a year or something?). Already some features (like 'undo') have been disabled on the LPC1769 firmware because of the LPC1769 hardware limitations.

If your seq is already running on an STM32F4 core, I think you can just update the MBseqv4+ firmware on it (when it comes out).

Edited by jjonas
typo

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...