Dear Native Instruments (Open Letter)

[tekki]

Hoping Hotshot has also dropped an email to Native Instruments, in stead of just hoping for them to miraculously picking this message up telepathically through the intarwebzzzz.

[nem0nic]

@nem0nic, Most people, especially me, don't care about lighting up LED's. We just want to knobs on the controller to be mapped directly to software features most of the time. I agree that Traktor mappings system provides more functionality than Ableton's, short of Python and Ableton API (and, I'm sure a bag of Max's "features") But that doesn't negate the fact that what most common use-case should be easy, quick, intuitive = user-friendly.

Here's the thing. Even the simple task of mapping an encoder to a function in Live can be a huge mess. Encoders are a problem for Live, because there's no way to define their behavior. You can't adjust the "sensitivity" of the mapped function to the resolution of the encoder. So you end up with a knob that moves too slow or too fast - which is why hardware manufacturers (like Akai) have had to set the output of their encoders to absolute (instead of relative), and do their own feedback generation and variable tracking in their firmware.

I'm not arguing that the Controller Manager is beyond criticism. There are lots of issues I would love to see addressed, and I totally agree that it could be MUCH easier for a new user to create a mapping. I've been saying that there needs to be an abstraction layer above the existing controller manager - think of this as a virtual controller - that could be mapped to easily via traditional MIDI learn. The Controller Manager would be used to create the "virtual controller". Ideally, the virtual controllers would be the NI controller offerings. The interface would be an image of an NI controller, and the user would click a control on the virtual NI controller, then move a control on their own controller - mapping done. For those who want to get their hands dirty, the existing interface would be available as well.

But using Live as your benchmark is a problem, because it's terrible. LED feedback is a big part of the Live controller workflow. Not being able to define output messaging without writing a script is pretty awful.

[Karlos Santos]

Just wanted to point out that the title of this thread is " dear native instruments" and this is clearly the DJ Tech Tools forums.

Mind boggling. Why on earth this has been posted on this forum is a mystery to me.

NI have a Features Suggestion sub forum.


It's a good topic for debate but an "open letter"? Really?

[deevey]

which is why hardware manufacturers (like Akai) have had to set the output of their encoders to absolute (instead of relative), and do their own feedback generation and variable tracking in their firmware.

Not quite right - Akai's "encoders" are not encoders (with the exception of "cue" on the APC and data entry on the MPK's / MPD's), they are regular pots albeit endless ones which is why they function just like a regular pot.

I bloody wish they were proper encoders!

In fairness at least NI gave us midi learn for led outputs a few versions back

[nem0nic]

Not quite right - Akai's "encoders" are not encoders (with the exception of "cue" on the APC and data entry on the MPK's / MPD's), they are regular pots albeit endless ones which is why they function just like a regular pot.

No, they're standard encoders. The same encoders any other product with encoders uses. The encoder is the actual physical part - the output of that part doesn't take away from the fact that it's actually still an encoder.

In the APC firmware, that encoder output is changed from relative to absolute. When the encoder reaches the end values in either direction, it stops sending output. We have an absolute mode built into the CMD controllers as well, and accomplish it in exactly the same way. The host application can update the value of the encoder by sending it a new value (same CC as the output message). This updates the LED feedback AND the encoder output. So, if encoder 1 is currently at 64 and the software sends that CC a value of 128, then the encoder and it's LED ring will snap to the full clockwise position, and any new movement will be relative to that new value.

What's more, the bottom group of encoders (the Device Control section) is modified by the TRACK SELECTION buttons. Each TRACK SELECTION button stores the values of all 8 of the Device Control encoders in the device firmware. It's not the Live software doing the variable tracking and conditional operations and pushing those state changes to the controller. The controller is doing it in the firmware.

[deevey]

No, they're standard encoders. The same encoders any other product with encoders uses. The encoder is the actual physical part - the output of that part doesn't take away from the fact that it's actually still an encoder.

Are you sure ? .... I've been attempting to get hold of replacement "Encoders" that would fit the bill which was almost impossible until I came across this on Akai's message board by a tech rep regarding MPD's lack of encoders and managed to find the parts (same as APC/MPK):

Yes, very well thank you. Sorry for the delay.
This is what I got from the akai development team.

The MPD24 has endless rotary POTS just like the other MPD/MPKs.
An encoder usually has 24 or 32 points in a circle, and it senses its position by an indicator passing over one of these points
A circle gives you 32 points of reference on a pot. This means, if you want to go 0-127 (standard midi controller), you have 2 choices:

1. interpolate 32 points as 0-127 (effectively sending 0,3,7, 11 etc) because you only have 32 points, or
2. count 32 points as 0-31. This means you need to turn the knob 4 times to get 0-127.

Most encoders will use a hybrid of these, and count accelleration to act somewhat like a real pot, but its never effective.

A POT has no steps. It uses a wiper and electrical resistance to sense the position of the pot. Pots can be then interpolated however you want. This means you can have (if you want) 128 steps, which is what we do. What our pot has is 2 wipers, and we use the relative position between the two to give us the benefit of an encoder moving in a 360 degree motion, but allow us as many points as we want.

The real benefit for 360 pots on a controller is banking

The term "endless rotary encoder" does not directly associate with NRPN functionality although most devices do. Keep in mind that all details are subject to change regarding trade shows/NAMM. Yes, the Project Manager did refer to them as "endless rotary encoders" but that could have been changed afterwards to improve knob functionality, like described above.

It certainly might explain the 4th Pin on the pots too and why they aren't "standard" parts that you can just order from radionics/mouser etc ...

[nem0nic]

They look like encoders to me, and none of what is said in that explanation you posted makes much sense. All of the knobs and faders I've tested on both units send standard 7bit absolute messages. The CC numbering starts at 1 for fader one, and goes from left to right and from bottom to top (ending in B0 37 ** at Track Control encoder #8). It's clear looking at the controllers that the numbering was done in an arbitrary way, and not to specifically choose NRPNs (of which they would only use one bit of the output).

Also, that explanation of an incremental encoder is only one of many. If I were looking for a higher resolution encoder and didn't want to rely on the firmware to interpolate (which can actually work amazingly well), I might specify something like a gray code encoder. There are other cool encoders out there (I'm currently looking at a hall effect (magnetic) encoder that has some interesting physical properties).

My point is that there isn't any secret sauce in the APC40 components - only it's firmware.

[Ambionic Being]

I think most people will admit that when they first want to start mapping their own controllers in Traktor they get a little intimidated or annoyed at it's seemingly not so user-friendly layout. But once you get into and just browse through the menus for a bit you realise it's awesome potential.

When you first opened Ableton the same could be said for that with it's completely individual take on the DAW, but perseverance and a desire to make boss music made you stick with it.

As far as I'm concerned I would rather use Traktors controller manager over Ableton's because it forces the user to think about the mappings in a much more logical way.

[tekki]

Mind boggling. Why on earth this has been posted on this forum is a mystery to me.

NI have a Features Suggestion sub forum.


It's a good topic for debate but an "open letter"? Really?

Yeah, the thread was initially called "Dear Native Instruments", as that this is posted on our forums, I decided to rename it.

[Karlos Santos]

Yeah, the thread was initially called "Dear Native Instruments", as that this is posted on our forums, I decided to rename it.

Ah so it's ALL your fault.