Nava2021 (post v1.028beta)

Neuromancer wrote:it has become some kind of Nava editor.....

... you might just have opened Pandora's box with the above statement! :-p
F.

Neuromancer wrote:Noted...currently working on the SysEx.
In order to test it I created a windows program...

does it mean that you do not own a nava and test your firmware build with a program?

General_Midi wrote:
Neuromancer wrote:Noted...currently working on the SysEx.
In order to test it I created a windows program...
does it mean that you do not own a nava and test your firmware build with a program?

No it, means there are 2 sides on sysex communication. Sending and receiving, that is the part I need testing for.
There's only so much MidiOX will show you.
It was just that I needed a way to confirm the correctness of the received data. Also I coulnd't find anything in de eeprom's docs about it also reading in 64 byte pages.

I can asure you my Nava is running on it's normal ATMega1284 microcontroller and on real hardware.

recent build is working fine so far...

thanks

I just created a very beta build with MIDI System Exclusive dumps.

Also included in the release is the installer for the Nava Tool.
The Nava Tool can handle pattern banks, edit patterns, handle and edit tracks and change settings in the Config pages.
When editing patterns you can use Shift and Control to change the velocity level and add flam.

https://github.com/BenZonneveld/Nava-20 ... o-20210711

Neuromancer wrote:
FlavioB wrote:
Neuromancer wrote: IMO I should be able to create a pattern group of any size up to 16 patterns and edit that group in step mode as it is one single large pattern.
So for my understanding: you'd like to group some patterns, which, I believe, have to be sequential (like A1-2-3-4), and then you'd like to enter step mode and be able to actually change let's say the SD active steps in a way that you can go through all of the grouped patterns? Like if the patterns are 16 steps long, you'd like to be able to just go to step 38 (which would be step 6 of pattern A3) and punch in an SD?
Or am I mixing up pattern group with pattern sequence?
F.
As I understand pattern groups to be working you can "chain" any sequential group of patterns.
Current situation is:
Step mode with running sequencer, at the moment the switching to another pattern in the group happens all unsaved changes are lost.
I'd like to be able to see that pattern group as a single longer pattern, with no lost changes at a pattern switch.
Then when saving the pattern you would save all patterns in that group.

I noticed people asking for longer patterns (32.64...etc) and using pattern groups is the only way for that to happen.

Big +1 for this one. I really miss that too. That feature even works on a "simple" TR606...

FlavioB wrote:
Neuromancer wrote: As I understand pattern groups to be working you can "chain" any sequential group of patterns.
Yes, that's what I also was trying to describe: A1-2-3-4 could be B6-7-8-9-10-11 - it doesn't matter how many patterns you chain, but you cannot group patterns as such. Grouping would be like A3+A6+A7 - and afair there is only the chaining feature implemented (which is the only one making sense, IMO).
Current situation is:
Step mode with running sequencer, at the moment the switching to another pattern in the group happens all unsaved changes are lost.
Yes, that's how it's intended to be - or at least, how the software is working (I'm not discussing pros&cons of this).
I'd like to be able to see that pattern group as a single longer pattern, with no lost changes at a pattern switch.
Then when saving the pattern you would save all patterns in that group.

I noticed people asking for longer patterns (32.64...etc) and using pattern groups is the only way for that to happen.
I understand what you mean and I can relate somehow to it. But IMO it'd be better to implement longer patterns, instead of having people deal with this workaround to get where they want to. I mean: if I'd like to use 64-step patterns and I would have to chain four 16-step patterns, that would be awkward and it would mean losing patterns in a bank. You'd be sort of ending up with only 4 patterns (4 chained 16-step patterns), instead of eventually being able to set the pattern length to 64 steps and have 16 patterns à 64-steps at your fingertips within the same bank.
What do you think?

F.

The problem on current FW is that you have to hit ENTER live for each bar to save when chaining, so when the sequencer leaves the first bar, all edits are lost. It would be more logical to use enter to save the whole chain before leaving the chain.

Although I would prefer to ditch the save-by-enter alltogether.

Oortone wrote: The problem on current FW is that you have to hit ENTER live for each bar to save when chaining, so when the sequencer leaves the first bar, all edits are lost. It would be more logical to use enter to save the whole chain before leaving the chain.

Maybe a combination like "shift+enter"?

Although I would prefer to ditch the save-by-enter alltogether.

Agreed on this one as well. I'm more used to the TR-808/TR-606 way of doing patterns...
F.

It's been a long ago since I used a real x0x drum machine. The first ever was a TR606 and later on I had a TR707. But it must have been over 25 years since I sold the last of my Roland machines.

The drum machine I own the longest is a SCi Drumtraks which I own since 1990.
Also an MPC2000 or different samplers/computer.

Short list of things I noticed:

Bugs:
* In mute mode the tempo display is not updated.
* In track write the pattern IMO should not be reset to A1 after pressing enter. Just advance the position and stay on the last selected pattern.
* EXT instrument should be able to have 2 velocity levels.

Planned features:
* SYSEX import and export
* Make the startup mode configurable.
* Make the sound levels user configurable. Keep in mind the patterns store the current instrument levels.
* Temporary change last step in a playing pattern (Pattern play only)

Things I'm not implementing:
* Sending swing over MIDI clock. A device should not abuse MIDI clock to send swing. A clock has to be a source of a stable timing, swing conflicts with stable timing.