Question about live recording / quantize and unquantized

[flyweight]

I read in the manual that

Quantize: than lower the percentage value, than lower the quantization (-> events not shifted to the next step)

Does this mean that the MB seq can do unquantized recordings or semi unquantized recordings?

Are step delay's used for placing the notes on the right location?
I sure hope its the case as its mandatory for my workflow.

Last time I checked MB was not able to do this.

Thanks 

 

Tim

[latigid on]

LoopA is good for unquantised recording :)

[flyweight]

Ah yeah I have see,  looks cool. But taking account step delays might offer semi unquantized recording in Seqv4 as well. 

Although I have no idea what the scope of this is and if Seqv4 has enough resources to handle it.

Cheers

 

Tim

[srmaietta]

I’m thinking that if LoopA can do it then seq4 can too.. Aren’t they based on the same core code?

mot sure of the unquantized (less quantized I guess) recording options but I’ll give t a try tonight.  

Steve

[Hawkeye]

@srmaietta LoopA has a different, non-step-sequencer approach - its code was mostly written to extend the features of the SEQ v4+, so unquantized polyphonic recording would be easily possible. It also adds a few other things, like the beatloop mode, named user instruments, high-resolution CC recording/editing/generation and graphical noteroll views, while it also can't do a lot of things its bigger brother does easily - so they compliment each other nicely :).

Best regards,
Peter

[srmaietta]

@flyweight

I just tried some non quantized recording now. (Nope..)

I set the track divider relatively slow so it would be easier to notice if the notes were falling on or off the beat perfectly. (Longer steps, more room for the notes to be off the grid)

It definitely seems like unquantized recording is not possible.  I tried again after setting a parameter layer to delay which allows one to manually shift notes forward in 96th step increments. But recording, even with quantize set to zero does not actually put the notes on the exact spot they come in.  They either land in the beginning of the step that you were in when you hit the note, or the beginning of the next note. 

The quantize percentage just controls whether or not the note will be put into the step after you actually play the note (100% quant) or before the note was received.(0% quant.   The fact that quantize mode is doing some calculations in real time based on where the note actually fell, in my mind makes it seem that it might be possible to set a number of ticks in the delay layer at the same time. This way the note would fall on the step you recorded it and have the ticks of delay set at the same time. 

May be possible to add in the future, Hawkeye would know more about the mechanics of the code and whether or not this is feasible without a major rewrite.  Sure would be nice.  

The workaround is to set a longer pattern with a quicker timebase. Although it’s definitely not the same thing.

-Steve

Edited December 19, 2020 by srmaietta

[slo]

On the Jam page under Live recording there is this from the manual. http://ucapps.de/midibox_seq_manual_m.html
Have not tried it yet but it looks like what you want.

Quote

• Quantize: than lower the percentage value, than lower the quantization (-> events not shifted to the next step)

 

[srmaietta]

20 hours ago, slo said:

On the Jam page under Live recording there is this from the manual. http://ucapps.de/midibox_seq_manual_m.html
Have not tried it yet but it looks like what you want.

 

It seems that the quantize percentage is used to decide wether to put the note-on into the step it was received or the next step.  I’m guessing based on how far it is from either.

The question (solution?) is, (for the coders) why can’t the note-on  be put where it falls based on step number plus added “delay parameter” ticks.

not that I would use unquantized recording much!  But seems possible, and may be desirable for users..

-Steve

Edited December 20, 2020 by srmaietta

[flyweight]

On 12/21/2020 at 12:30 AM, srmaietta said:

The question (solution?) is, (for the coders) why can’t the note-on  be put where it falls based on step number plus added “delay parameter” ticks.

Hi Steve,   

Yeah that is exactly what I would like.

To be honest I did not know the Seq V4 could not do this then started to build and find this out.

I have not tried to record at very high grid settings.

But realtime recording for me is a big desire atm.
I record often without quantize for beats with nice feel and swing.  

LoopA looks interesting but I don't want to abandon the seq v4 and build something else.
It took me long to complete and it should have become the center peace of my live setup but as it looks like now if it can really only do step's unquantized I made a big mistake.
So in a way I am stupid to not check this before starting this building journey.  

I mean building as fun but its kind of disappointing that in the end I put all this hours to later sell it. 

Please note this is not a rant.  I really enjoyed building the seq v4 and think its an amazing project where I learned a lot about digital logic and electronics.
Its just my own stupidity. 
Lets hope the high resolution grid setting will work acceptable.  Or even one day the step delas will be recorded.

 

Cheers

 

Tim

 

[Hawkeye]

@flyweight hopefully there's no need to be disappointed - when you use the SEQ v4/v4+ with long pattern lengths (e.g. 256 steps instead of 16) and a clock divider set up to play back the individual steps quickly, this is very close to unquantized recording. If unquantized recording is your top item, i would recommend to additionally add a LoopA to your sequencer setup - it is optimized for that, but can't do a lot of things the bigger v4/v4+ can do. Our "dream" sequencer setup would be one v4/v4+, one LoopA and one MatriX (BLM) connected to both, for 16 tracks of step-sequencing plus 6 tracks of unquantized note-roll sequences plus hands-on-control with a large button matrix.

@srmaietta can't really speak for Thorsten, but i believe the v4/v4+ was basically developed as a step sequencer with rigid timing, basically for electronic music. With the availablity of long sequences AND clock dividers, it is absolutely possible to record "human" sequences that are not consisting of 16th notes or 16th note positions. This method avoids introducing the "delay-per-step" parameter layer, which is difficult to interpret - i.e. what happens when swing is applied, what happens with other fx (like echo) and non-linear sequence progression modes and so on.

Best regards,
Peter

[srmaietta]

@Hawkeye    Thank you for the thoughtful replies.  It is easy to think of future wishes and add ons but at the same time forget about how it would effect everything else.  I love the Seq4+, very little I would change!  And the fact that I can get into the code and try things myself is pretty amazing.  That alone is a "feature" that very few seq's can boast.

@flyweightIf I was looking for more humanized recordings I'd use the longer tracks with higher divisions OR the humanizer feature!  Also, you can specify your own shuffle parameters, you may find something that gives you YOUR feel..   if you could analyze how your beats groove, you may just be able to create that groove as a template and there ya go.  May be worth digging into.  At the same time, your experience and learning from building the Seq is not lost if you sell it.    You lost time, yes. I'm sure if you put it on the market it will sell pretty quickly.

 

~Steve 

Edited September 6, 2021 by srmaietta