Audio Mixing Platform (AMP)

Please login or register.

Login with username, password and session length
Advanced search  

News:

Pages: 1 [2] 3

Author Topic: Starting on my own control surface...  (Read 16670 times)

rhyde

  • Guest
Re: Starting on my own control surface...
« Reply #15 on: September 10, 2015, 03:44:43 PM »


I would really hate to see a one-off only come as a result.  I think this has tremendous opportunities if it is done well.
Well, as mentioned in other posts, if I get it all working well I'll be more than happy to let someone else run with it. However, I know what it costs to turn something like this into a real product and I don't have anywhere near that kind of money laying around, particularly considering the limited market for the product.

Quote
My suggestion would be to look at the A&H GLD as a model for a control surface.  They do a nice job.  They have one rotary encoder that can be Pan, HA Gain, or two user-defined possibilities, selected by another little button on the side.  Having two would be an extra bonus... Could definitely have an aux on one of those (like when using aux-fed subs).  Not sure you need all those selection buttons, as you can accomplish all the assigning stuff with just the selects.

Things to ponder...

Bob

Definitely the A&H iLive board has been the template I am using for ideas. I like that board's layout and functionality a LOT.
Cheers,
Randy Hyde
Logged

admin

  • Administrator
  • Hero Member
  • *****
  • Posts: 956
    • View Profile
Re: Starting on my own control surface...
« Reply #16 on: September 10, 2015, 06:51:02 PM »

Go Randy, Go! :-)

Yes, cheap displays are definitely the way to go.  And yes, you definitely want to have something that can do graphics, and be responsive.  Perhaps you could map it into a USB-to-VGA adaptor?

Bob
Logged

RBIngraham

  • Guest
Re: Starting on my own control surface...
« Reply #17 on: September 10, 2015, 08:15:58 PM »


At least six aux sends (matching SAC, haven't gotten that far into AMP to know if this is enough, too few, too many [doubtful], etc.) with pre/post and in/out buttons. Aux returns aren't channel specific and would be mapped to the normal fader strips.


Ummmm..... not quite... with AMP there can be up to 32 Aux Sends.  It's a very different animal than SAC Randy.  Actually it's more like most other consoles except that it has more Aux Sends.  There is no 24 separate consoles for monitor mixing.  It's one big console rather than 25 separate simple consoles.

Also you can have a variable number of them so frankly if it were me I would not bother with a rows of knobs for Aux Sends.  To me that is what sends on fader if for.  You do it on the surface with the same faders.  Put in some buttons to select a mix and let the same faders do the work.

I would concentrate on the EQ and Dynamics for knobs as those are the things that best tweaked with a knob.

That's my opinion anyway.  But then, even putting cost asside, what you're describing wouldn't be my cup of tea.
Logged

rhyde

  • Guest
Re: Starting on my own control surface...
« Reply #18 on: October 16, 2015, 01:52:03 PM »



Ummmm..... not quite... with AMP there can be up to 32 Aux Sends.  It's a very different animal than SAC Randy.  Actually it's more like most other consoles except that it has more Aux Sends.  There is no 24 separate consoles for monitor mixing.  It's one big console rather than 25 separate simple consoles.
For AMP (versus a SAC implementation), my thought was to grab the first six mix busses and use them as the aux sends.
I was also planning on grabbing a mix bus for the Center channel (for LCR) and one for the subs channel (aux-fed subs).

I have never used more than 12 monitor mixes (some were stereo). I'm fairly certain I'd be okay with 16 mix busses dedicated to monitor mixes. That leaves 8 mix busses for use as groups (I assume that's how groups are implemented here).


Quote
Also you can have a variable number of them so frankly if it were me I would not bother with a rows of knobs for Aux Sends.

Guest engineers expect this...

Quote
To me that is what sends on fader if for.  You do it on the surface with the same faders.  Put in some buttons to select a mix and let the same faders do the work.
I am trying to avoid overloading controls. Can't be avoided entirely, but if someone selects a channel to do work with EQ, Aux sends, dynamics, etc., on a single channel, I still want to be able to quickly grab a fader for a different channel and make adjustments in a hurry, w/o having to clear the "mode". Again, this is a user interface issue that I'm concerned about when dealing with guest engineers. It's bad enough that the board will have four layers on the faders (which for my rig, would rarely be used).

Quote
I would concentrate on the EQ and Dynamics for knobs as those are the things that best tweaked with a knob.
Certainly that is one area where I do have a whole lot of controls available (trying to simulate SAC's "wide view" in that area of the board). OTOH, this is one area where I will be overloading the faders. I've chosen 36 channel strips as the size of my board so that I can implement a 33-band GEQ on the faders (just like the Yamaha boards, for example); that's too nice a feature to pass up and I think the compromise to the UI is justified by this nice feature.

I intend to provide 3-4 rotary encoders to control compression, 2-3 to control the gate (with an appropriate number of additional pushbuttons, keeping in mind that each RE has a push button, too).

I intend to provide four rotary encoders for the PEQ section (freq, BW/Q, gain, type) plus appropriate push buttons (in particular, at least seven push buttons to select one of the seven possible bands plus various other buttons as needed by the PEQ).  At first, I was thinking of providing 28 rotary encoders, but I cancelled that idea because a 4x7 bank of RE could be rather confusing and you'll probably only be adjusting one frequency ever at a time so selecting the band via buttons seems like the best compromise here.

I'm also going to add a set of buttons (24? 32?) to direct an input channel's output to one of the mix busses.

There will also be miscellaneous controls (e.g., phase, channel disable, etc.).


Quote
That's my opinion anyway.  But then, even putting cost asside, what you're describing wouldn't be my cup of tea.
Well, like I said, I'm building one for me. It's not a product. Cost isn't that much of an issue (parts alone will probably run between $2,500 and $5,000); the labor is going to be the real killer; I've just spent the better part of a week (full-time, off work right now) wiring up the two RE boards for the first bank. That's okay, it's a hobby endeavor and it's a lot of fun...


Update on the progress:
Parts are still coming in, but I have wired up the rotary controllers and displays for the 1st 12 channels. The system will consist of three banks of 12 channels. Each bank is controlled by a Teensy 3.2 micro controller (ARM cortex m3 CPU) with I2C and 74595 expansion to 128 bits of input and 128 bits of output. I'm currently waiting for some red LED buttons to arrive so I can complete the button inputs for the 12 channels.  Currently, the channel strip is arranged thusly:

|-----------Display1--------|
  RE1          RE2        RE3
  RE4          RE5        RE6

    B1            B2         B3
    B4            B5         B6
    B7            B8         B9
    B10          B11       B12

    F1             F2         F3

(repeat the above four times for a total of 12 channel strips per bank).

RE = rotary encoder. Consists of a rotary encoder (virtual pot), push button, and RED+GREEN LEDs (produces four colors: black [off], red, green, and yellow).

B = button with LED. Top row is white (select), second row is red (solo), third row is blue (group/DCA), and fourth row is amber (mute).

F = fader. Includes a capacitive touch sensor so the software knows when someone is touching a fader.

On the rotary encoders, I'm programming three different functions.
On the bottom row the three functions per RE are "pan/balance" (LED=yellow), "Center" (LED=red), and "subs" (LED=green). Pressing the button cycles between the three modes. The LEDs are off (black) if the channel is disabled.

For the top row of encoders, Yellow=Attenuator. I've not completely decided on the red and green modes yet, but they will probably correspond to FX feeds (reverb and delay) unless I come up with something more pressing. Again, pressing the button will cycle between the three modes.

RE parameter values will appear on the LCD displays along with channel strip info, digital VU meters, and other (as yet undefined) information.

Oh well, back to wiring...
Cheers,
Randy Hyde

Logged

admin

  • Administrator
  • Hero Member
  • *****
  • Posts: 956
    • View Profile
Re: Starting on my own control surface...
« Reply #19 on: October 17, 2015, 01:11:38 AM »

Hey Randy,

Sounds like you're making some good progress!  Personally, I don't like the 3 or 4 knob EQ - I like having all the EQ bands there and visible.  I also don't see an issue with overloading a bunch of knobs.. If you've seen how a Soundcraft VI 4 (and the like) overload knobs, it's very intuitive.

Bob
Logged

rhyde

  • Guest
Re: Starting on my own control surface...
« Reply #20 on: October 17, 2015, 02:09:18 AM »

Hey Randy,

Sounds like you're making some good progress!  Personally, I don't like the 3 or 4 knob EQ - I like having all the EQ bands there and visible.  I also don't see an issue with overloading a bunch of knobs.. If you've seen how a Soundcraft VI 4 (and the like) overload knobs, it's very intuitive.

Bob

They will certainly all be visible. The EQ section will have its own 480x320 (half VGA) color display. It will show the values for all the PEQs (in seven columns). There will be a select button under each column to make the knobs control the selected frequency range.

The rotary encoders aren't that expensive (about $6 each), but multiply this by 28 and it starts to add up. Not to mention, it takes about a day to wire up 12 rotary encoders on the channel strip (so figure two days for 28 in the EQ section) and 28 of those puppies will take a *lot* of board space. Like I said, using buttons (under the display) to select the channel to affect seems like a very reasonable compromise.

Also note that various digital boards (such as the Allen & Heath boards) don't provide encoders for each frequency band -- they also require some user interface magic to share a single set of controls amongst all the bands -- so I'm in good company there.

Note that my board is going to be relatively compact. 6 channel strips are only a little over 4-" wide. So 36 channels will fit into a little over two feet wide (call it three feet by the time other stuff is added like layer selection buttons and gaps between the banks). That doesn't leave a whole lot of room at the top of the board for output assignments, aux sends, dynamics (with its own display), PEQ (with its own display), and a master controller display (running off a Raspberry Pi 2 B, probably a 5"-7" display).

Based on my experiences with touch screens in the field, I am avoiding touch screen designs like the plague, though I may break down and put a resistive touch screen on the RPi (resistive isn't as sensitive as capacitive touch screens so it's less likely a bug will affect them). The RPi screen will mainly exist for system utilities, saving data to flash cards, configuring the system, and a main screen that shows board status.

The system will probably have 5-6 micro controllers in it: three teensy 3.2 boards (about the size of a stick of gum), one small board (haven't decided on the type, but probably a Teensy just to be consistent) that will provide a MIDI interface, and a Raspberry Pi for general control and Ethernet interface.

Just working on the first 12-channel strip right now, so lots of time to figure everything else out.
Cheers,
Randy Hyde

P.S. Ive attached pictures of the RE/Display boards (with and without displays attached) and the button board (currently waiting for red LED buttons to arrive, so one row of buttons is missing).
« Last Edit: October 17, 2015, 02:23:29 AM by rhyde »
Logged

admin

  • Administrator
  • Hero Member
  • *****
  • Posts: 956
    • View Profile
Re: Starting on my own control surface...
« Reply #21 on: October 17, 2015, 10:43:36 AM »

Wow, that's a great start you have!  WIll you allow direct screen access via a host PC?

Just a point of clarification: Allen & Heath mixers *do* have separate controls for each EQ band.  Thats why I like them. Even the little QU has 12 EQ knobs.

How much are those screens you are using?  How are you interfacing all of this to a host computer?

Bob
Logged

RBIngraham

  • Guest
Re: Starting on my own control surface...
« Reply #22 on: October 17, 2015, 11:55:45 AM »

Actually many of the more recent console actually have individual knobs for all the bands of EQ with no switching between bands.  Yamaha has done it that way for years now.  You have to go back to the 01V96, which is very long in the tooth but still available to find that banking between bands for the knobs.

In a very brief search of the el cheapo digital desks only the X32 seems to use the old 4 knobs with buttons to select the band approach.  You don't want to be in that group do you?   :D

The difference of course is that most consoles have settled on 4 bands of parametric plus a low cut, rather than AMP's craziness of 7 bands plus a low cut.   :)  So it takes fewer knobs to do it.
Logged

RBIngraham

  • Guest
Re: Starting on my own control surface...
« Reply #23 on: October 17, 2015, 11:58:16 AM »


For AMP (versus a SAC implementation), my thought was to grab the first six mix busses and use them as the aux sends.
I was also planning on grabbing a mix bus for the Center channel (for LCR) and one for the subs channel (aux-fed subs).

I have never used more than 12 monitor mixes (some were stereo). I'm fairly certain I'd be okay with 16 mix busses dedicated to monitor mixes. That leaves 8 mix busses for use as groups (I assume that's how groups are implemented here).


Quote


What are these "Mix Buses" that you speak of? 

AMP has DCAs, Outputs and Aux Masters.   :)
Logged

rhyde

  • Guest
Re: Starting on my own control surface...
« Reply #24 on: October 17, 2015, 04:48:51 PM »

Wow, that's a great start you have!  WIll you allow direct screen access via a host PC?
Which screen? :)
Right now, I'm planning on a total of 15 or 16 480x320 displays, a larger display for the main console controller (Raspberry Pi) and possibly one smaller screen for other purposes.

I can't see any reason to allow the PC to have direct access to the 12 displays that form my digital scribble strip (36 channel strips, three channel strips per display) -- at least, not on a pixel-by-pixel basis.  However, each channel strip will contain space for around 12 characters (the digital scribble strip), two digital VU meters (stereo operation), and the associated information obtained from the rotary encoders for each channel strip. Certainly there will be some protocol to have the PC specify the characters and VU meter settings. Likewise, the Rotary encoders (that control things like pan/balance, center feed, sub feed, attenuator level, and two other functions) will have their values set by the PC program (the knobs won't control the values on the screen directly, the assumption is the the knobs send the data to the PC and the PC sends back the actual values to display; the console software will simply interpret the return values and display them in an appropriate fashion.

The EQ display will operate in a similar fashion. Internal to the console software I'll maintain all the individual parameters for the EQ knobs. Whenever one changes I'll transmit that value to the PC. When the PC responds with new values, I'll update the display in an appropriate fashion.  Same for dynamics.

I may very well implement scenes and cues directly on the board (on the Raspberry Pi) rather than using the host PC for this function. It depends on how easy it is to reconcile the control between SAC and AMP (I do have a requirement for the board that I can use it with both systems, even if SAC operation is slightly limited by SAC's inability to effectively support control surfaces).

DCAs are another interesting area. SAC doesn't have them; I can build (true) DCAs for SAC directly into the control surface itself; just need to make sure I can easily reconcile this with AMP's DCAs.


Quote
Just a point of clarification: Allen & Heath mixers *do* have separate controls for each EQ band.  Thats why I like them. Even the little QU has 12 EQ knobs.
Yeah, but they're still overloaded; and only four bands to boot.
SAC has five and AMP has seven. For SAC I'd need 17 rotary encoders (5 bands * 3 plus HP and LP), for AMP (assuming I use an RE for filter type selection) it takes 28. That's starting to look like an old analog monitor mixer console :)

My approach has a dedicated display (480x320) for the EQ, so it is always visible (unlike most other mixers that have a single display and multiplex screens depending on the function).

In my experience (YMMV), I've never tried adjusting two bands simultaneously, so I'm not sure of the utility of having separate knobs for each band. [P]EQ isn't something you need to rapidly change on the fly (unlike, say, adjusting faders or hitting mute buttons) so I can easily live with the extra step of pressing a button before making an adjustment on a particular band.


Quote
How much are those screens you are using?  How are you interfacing all of this to a host computer?

Bob
As I noted above, I'm probably using about 15-16 different displays.  The 480x320 are amazingly cheap (about $11 each). The three Teensy 3.2 boards (picture attached) will control 12 channel strips (four displays) each. They connect to the Raspberry Pi via USB. The Raspberry Pi will have an Ethernet port to connect to the PC.

I'm also planning on using a small single board computer (probably a Teensy, we'll see) to serve as a USB peripheral device programmed to look like a MIDI port. This will get me up and running with SAC and AMP a whole lot faster, though the Ethernet port is a better solution in the long run.

Ultimately, what I'd like to have is an AMP "graphical user interface" API via ethernet. Rather than running a graphical interface program that talks to a control surface, I'd like my control surface to *be* the "graphical interface".  Hopefully, you'd be more amenable to this than Bob Lentini (I've been asking for an API for *how* many years?).

Obviously, this will be a whole lot easier to interface with AMP than with SAC (especially given Lentini's reluctance to embrace control surfaces). Ultimately, I will force this to work with SAC one way or another, even if I make it look like a combination of Behringer and Mackie control surfaces (running two copies of SAC Remote) -- Behringer to get 32 fader support and Mackie to get control of the effects sections. I can attach a keyboard to the Raspberry Pi and enter digital scribble strip information directly on the console (bypassing SAC); this won't let me save the info with the mix files, but I rarely reuse mix files anyway (and I can save that data in a separate file on the Raspberry Pi, if I have to).

BTW, one (optional) SAC feature that I use all the time is Bob L's Frequency/Spectrum Analyzer plug-in (really great for ringing out monitors). Are you ever going to add something like that? If so, I might add another screen to display a spectrum analyzer.

Cheers,
Randy Hyde


Logged

rhyde

  • Guest
Re: Starting on my own control surface...
« Reply #25 on: October 17, 2015, 04:59:40 PM »


What are these "Mix Buses" that you speak of? 

AMP has DCAs, Outputs and Aux Masters.   :)
And 240 mix busses, according to the AMP web page.

Now granted, I've not used AMP yet (other than to play with the graphical interface component) so I don't know how to use all the features yet, but I'm pretty sure AMP will be flexible enough to do what I want.

As for the monitors, other than the limitations on EQ, I could make those Aux Sends look a whole lot like SACs 24 monitors if I really wanted to; though, as I said, I doubt I'd set up my control surface to support more than 16 channels of monitor mixes.

When you have complete control over the surface (i.e., when you're the one writing the software for it), it's easy enough to reconfigure everything on the fly so that I *could* have a monitor mixer layout on the board (granted, with AMP there would be limited channel strip signal processing when doing this).
Cheers,
Randy Hyde
Logged

RBIngraham

  • Guest
Re: Starting on my own control surface...
« Reply #26 on: October 17, 2015, 06:48:16 PM »


In my experience (YMMV), I've never tried adjusting two bands simultaneously, so I'm not sure of the utility of having separate knobs for each band. [P]EQ isn't something you need to rapidly change on the fly (unlike, say, adjusting faders or hitting mute buttons) so I can easily live with the extra step of pressing a button before making an adjustment on a particular band.




It's not about adjusting more than 1 band at a time.  It's about reaching up instinctively and starting to twirl the knob only to realize... AWWWW Shit... that's the wrong band.... now I have to try and put that one back, then switch to the band I wanted to adjust and tweak it.

I can not tell you how many times I've done that over the years in the small format consoles that operated that way.

I know you'll disagree but that is why I find it faster to tweak on the screen.  My screen is often right in line of sight to the stage, I don't have to look down or take my eye off what's happening to make a tweak and it's easier to not muck up the wrong band... especially now with the rubber band handles on the EQ graph.  I used those for the first time this week while showing some colleges AMP and everyone enjoyed them.  It was so damn fast.
Logged

RBIngraham

  • Guest
Re: Starting on my own control surface...
« Reply #27 on: October 17, 2015, 07:03:41 PM »



And 240 mix busses, according to the AMP web page.


Now granted, I've not used AMP yet (other than to play with the graphical interface component) so I don't know how to use all the features yet, but I'm pretty sure AMP will be flexible enough to do what I want.

As for the monitors, other than the limitations on EQ, I could make those Aux Sends look a whole lot like SACs 24 monitors if I really wanted to; though, as I said, I doubt I'd set up my control surface to support more than 16 channels of monitor mixes.

When you have complete control over the surface (i.e., when you're the one writing the software for it), it's easy enough to reconfigure everything on the fly so that I *could* have a monitor mixer layout on the board (granted, with AMP there would be limited channel strip signal processing when doing this).
Cheers,
Randy Hyde

I was just being a smart ass...  you should recognize that by now..   ;D

But in all seriousness, I think before you spend a bunch of time laying out your surface and doing all this work, I would probably make sure I understand the software I'm making a surface for.

The so called "mix buses" are just a virtual bus to transport audio data.  You don't have any control over them at all other than assigning them as the source to a particular channel (typically an output or Aux Master).  In fact the vast majority of them now are hard assigned to a particular task.  For example there are 64 buses hard assigned for Aux Masters, another 96 for outputs (since any of those might be an L-C-R Output master) and a few more for things like tone generators and the media player.  There are a few left for user assignable tasks but really those serve only a few tasks, primarily to take the output of a particular channel and route it to the source or input of another.  (can be any channel really as long as the sending channel is a lower channel number than the receiver)

 In other words you don't really want to spend much time thinking about those virtual buses.  Think of the console as variable architecture mixer with up to 32 Outputs (that can act like a submaster or group output and 32 stereo/mono aux master outputs.  And up to 32 DCAs but obviously those are just a virtual control.  You can have up to 255 total channels and and any that you don't use up for DCAs, Outputs or Aux Masters can be an input. 

I personally think it's best to think of the Aux Masters as a combination of Monitor Sends and Matrix Outputs and that is certainly how I use them.  The Outputs really work best as submasters in my opinion.  Of course an output can send it's audio to an Aux Master and an Aux Master can be assigned to an Output, again as long as the sending channel is lower than the receiving channel.

These are all things you should take into consideration and understand if you're going to build a truly integrated surface for it, at least in my not so humble opinion.

« Last Edit: October 17, 2015, 09:54:46 PM by Richard B. Ingraham »
Logged

rhyde

  • Guest
Re: Starting on my own control surface...
« Reply #28 on: October 17, 2015, 10:05:48 PM »



It's not about adjusting more than 1 band at a time.  It's about reaching up instinctively and starting to twirl the knob only to realize... AWWWW Shit... that's the wrong band.... now I have to try and put that one back, then switch to the band I wanted to adjust and tweak it.

I can not tell you how many times I've done that over the years in the small format consoles that operated that way.

I know you'll disagree but that is why I find it faster to tweak on the screen.  My screen is often right in line of sight to the stage, I don't have to look down or take my eye off what's happening to make a tweak and it's easier to not muck up the wrong band... especially now with the rubber band handles on the EQ graph.  I used those for the first time this week while showing some colleges AMP and everyone enjoyed them.  It was so damn fast.

Well, the bottom line is that if there is physical room on the board, I'll certainly consider it.  Or maybe do it for some subset of the PEQ channels (perhaps four).  Right now, physical layout space is my most precious resource; I'm loathe to make the board physically wider than it absolutely has to be. That, plus the fact that I still think that a 7x4 matrix wouldn't be quite that intuitive (I remember my days on an A&H MixWiz monitor board and tracing the knobs with my fingers along the X and Y axes in order to locate the knob to turn...).
Cheers,
Randy Hyde
Logged

RBIngraham

  • Guest
Re: Starting on my own control surface...
« Reply #29 on: October 17, 2015, 10:32:02 PM »



Well, the bottom line is that if there is physical room on the board, I'll certainly consider it.  Or maybe do it for some subset of the PEQ channels (perhaps four).  Right now, physical layout space is my most precious resource; I'm loathe to make the board physically wider than it absolutely has to be. That, plus the fact that I still think that a 7x4 matrix wouldn't be quite that intuitive (I remember my days on an A&H MixWiz monitor board and tracing the knobs with my fingers along the X and Y axes in order to locate the knob to turn...).
Cheers,
Randy Hyde

Well you're not building this monster for me, so you should do whatever you like best.  I'm just telling you why I prefer to have one set of knobs per EQ band.

One most of the consoles I've worked with that have this, it's not like hunting for the right knob on vast columns of channel strips.  In those days you're as much looking for the correct channels' knobs as you are the function you wish to tweak.

The fastest surface I have personally used for tweaking channels I think is either the Allen and Heath T112 iLive and I still have fond memories of the DM2000's interface even though it's architecture was never really meant for live use and as such it was kind of a pain in the ass for live sound.

I liked the DM2K because if memory serves the settings were displayed on tiny LCDs right next to the knobs.

And the T112 has what I think of as the best channel strip layout.  One channel in a horizontal orientation across the top of the desk with pretty much 1 knob per function.  So you select a channel and then tweak it.  And I really liked how it was on the slanted surface so you could easily read it.  Granted that probably makes it very bulky if you're lugging it around on a tour or one off events. 
Logged
Pages: 1 [2] 3