LSSP melody

[basa333]

Hi Colin or another VM-fans,

I have the following problem. I have built the following melody line, but in the current version the last two notes of the melody sequencer are not played. My knowledge is now at an end - need I to connect one more rhythm sequencer (of course then I would have to connect another time splitting, or time flow changer? or something else?). advice?

thanks in advance for the advice tom

[ColinP]

Hi Tom,

There are 13 notes in the rhythm provided by the Rhythm Sequencers for bar 3 nd 4, so the Melody Sequencer is only playing 13 notes even though it has a 15 note pitch sequence programmed.

TomRhythmSeq.png (182.46 KiB) Viewed 4384 times

The numbers in brackets underneath the Melody Sequencer's LINK IN 1 and LINK IN 2 sockets show the number of notes coming from the two Rhythm Sequencers as being 9 and 4 so 13 in total while in the top right of the Melody Sequencer the 13 of 15 text indicates that there are 15 steps to the pitch sequence.

If you add a couple more notes to the rhythm (by clicking on the gate patterns of the existing Rhythm Sequencers) then those pitches will sound.

I hope this helps. The disentanglement concept in LSSP is not familiar territory in the sequencer world so things like this take a little getting used to, but do make sense eventually

TomMelodySeq.png (140.8 KiB) Viewed 4384 times

Also note it's always best to connect the PITCH OUT of the final Pitch Sequencer in a chain back to the MERGE input of the first as this stops the pitch CV dropping to zero on hand over from one pitch sequencer to the next when the first step of the Rhythm Sequencer is off. It also means that the pitch CV holds its state when a song ends.

This feedback loop is a bit fiddly and difficult to understand but LSSP sequencer outputs have to drop to zero volts when inactive to make the merging work. After you've done it a few times it just becomes something automatic that you do without having to think too deeply about why. If you ever get a mysterious drop to zero just try connecting the final output in a chain back to the first MERGE socket and it should fix the problem.

[basa333]

Hi, Colin,


thank you for your reply.
My goal was to transform the attached composition (that I competed with in OSC126) into VM.
Does this mean that transformation is not possible like this and I would have to make changes to the melody?
Would that mean, that such projects are only possible via mighty piano roll?

thanks and greets tom

[ColinP]

Hi Tom,

I don't quite understand your point. You should be able to map any melody on to LSSP (even extremely complicated ones if you use Groove and perhaps shift up to 32 steps per bar sequencing).

A melody consists of a certain number of notes. The rhythmical sequence supplies the timing information (and dynamics, but this information isn't used in your patch) and the pitch sequence supplies the pitch information.

In the final two bars of your patch there are 13 notes in the rhythmical sequence but 15 notes in the pitch sequence.

LSSP looks at the rhythmical sequence and plays it back selecting the pitch for each note from the pitch sequence.

If the number of notes in both the rhythm sequence and pitch sequence are exactly the same then each note in the rhythm maps to its own pitch in the pitch sequence.

If the pitch sequence is shorter than the rhythm sequence then the pitch sequence wraps around and repeats but if the pitch sequence is longer than the rhythm sequence then the extra pitches in the pitch sequence aren't reached.

A slight complication is when the TIE button is engaged on a Rhythm Sequencer as then adjacent steps that are on are joined together to form a single note.

One aspect of the philosophical idea behind LSSP disentanglment is that you can work on the rhythm and the pitch sequence independently. So you can find a nice pitch sequence of say five notes and then explore multiple ways of mapping this to a rhythmical pattern - experimenting with rythmical phrasing and repetition yet keeping the same five note sequence intact.

Or one can work in the other direction - creating a nice rhythmical pattern first and then explore different pitch sequences that fit with the rhythm.

In practice it's possible to move fluidly between these two approaches to home in on a melody.

In addition the pitch sequence can be automatically adjusted to conform to a particular scale or the current chord in a chord progression and this quantization can be customised step by step so that you can build melodies that intelligently adapt to the chord progression or don't (allowing things like passing notes).

It's all a bit left-field but the partnership between the Melody Sequence, Rhythm Sequencer, Progression and various scale and chord modules is meant to provide a deep creative framework for exploring compositional possibilities. It's a very different workflow than entering a fixed (entangled) set of notes into a piano roll. Although one can actually do fixed programming if one desires. But if all you want is fixed programming then something like the Mighty Piano Roll or a DAW piano roll would be more appropriate.

But by using a piano roll you no longer have any freedom to experiment with multi-level form, different chord progressions, scales and interactions between melopdy and rhythm. Instead you need to map out an entire piece of music note by note. This is what I call entanglement as form, harmony, melody and rhythm are all tied together in a piano roll rather than being fluid and independent.

[UrbanCyborg]

I think the problem arises here because of a written-music convention regarding ties and slurs. If you've got, say, a triplet of different pitches under a phrase mark, the tendency is to play them legato, instead of distinctly articulating each. I think the original poster is thinking of the three pitches specified in the Melody Sequencer as such a legato group, with the single Rhythm Sequencer note being the phrase mark. Of course, each note must have a rhythmic value in LSSP, or it won't play. The legato-ness just has to be specified in some other way. In other words, the Melody Sequencer and Rhythmic Sequencer combined define the information contained in a written-music note, but no phrasing information. Have I got that right, Colin?

Reid

[ColinP]

Thanks Reid.

That makes a lot of sense. I hadn't considered that Tom was wanting a three note legato phrase.

This can be achieved by modulating gate time so that the gate length becomes 100% for those three notes.

In the image below I use the unutilized velocity output of the final Rhythm Sequencer to modulate its own gate time but in general one would use velocity for velocity and so one might want to use a parallel chain of Rhythm Sequencers to modulate each note's gate length if one wanted to play with legato or one could use a spare track of a Drum Sequencer chain instead. CV Sequencers could also be used but they wouldn't be able to track micro-timing adjustments if any Groove modules were involved.

Another approach would be to modulate decay/release rather than gate length.

The CV Watcher traces show the gate in red, the gate time modulation in yellow and the pitch in green. As you can see when the gate modulation goes high the three step gate merges into one but the pitch information still contains three different pitches.

TomLegato.png (189.41 KiB) Viewed 4339 times

[basa333]

Hi, Colin, hi Reid,

thank you for your time and inventiveness to solve my problem. you have given me so much informations, that now I have to try to understand and process it. this is for me a very specific problem, that I have not seen (yet) in LSSP tutorials and the whole topic is quite challenging for me. I will work on it though. The reality is, that I was really trying to 100% transform the melody without any other aspects, and maybe that's not the full goal of LSSP, which is all about easy changes to certain original concepts.
Thank you and I'll be sure to get back to you!

greets tom

[ColinP]

Hi Tom,

Yeah, LSSP's main goal is to provide innovative modular ways to compose/generate new music rather than to capture pre-existing music.

It would probably be easier for your purposes to use something like the Mighty Piano Roll as this can record live MIDI, or the MIDI File Player if you have a MIDI file of the melody already or a DAW piano roll editor.

LSSPs workflow is very different to these tools. Even though it can be used to do fixed programming it would make more sense to use conventional tools for this task.

----

Having said that, in your patch you could just switch on two more steps of the bar 3 or bar 4 Rhythm Sequencer and try turning the GATE TIME knob up to maximum.

[basa333]

Hey, Colin,

is there any way to change the key in the middle of a song, e.g. verses in C and choruses in G?

Thanks and greets tom

[ColinP]

Hi Tom,

There are several ways to achieve this depending on how you are sourcing things.

Here's a basic example if you are using diatonic triads...

keyChange.png (1 MiB) Viewed 4234 times

If this doesn't help then please let me know a few more specifics as there are quite a few options depending on your patch.