Modular Digitally Controlled Analog Console Format

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
MIDI is a good choice but maybe limited by the 7 bit standard values, then you can add double precission but bandwith will be limited then for a big project... As control surface will work ok because you don't have all the controls running at one, you say this parameter this value... no more than a dozen of parameters at a time (I only have 10 fingers to move faders at once so maybe 10 parameters is enought) so double precision wouldn't harm... To send the parameters to the desk is other story, because everithing could be automated and there are a lot of parameters there... even if not all the automation is happening in at the same time when you stop and go back to start many of them may change and this could be a problem, I don't know the exact message capability of MIDI but I'd guess won't make a job with the speed we may want.

JS
 
Hi!

After each NAMM, it happens to me the same: I start dreaming on new gear... ideas.. etc...
But I want to  get back to the subject of the thread...given the market today.  It was 2013 ( I think) when SSL released the SIGMA summing mixer... a digitally controlled analog summing mixer....and Bettermaker has more digitally controlled gear....so....¡There is some ¿limited? interest!.... 8)....¿Right? ;D Also the modern chips today allow a certain "simplicity" or less complicated circuitry than early 2000's... Just some thoughts.

Jay x
 
JAY X said:
Hi!

After each NAMM, it happens to me the same: I start dreaming on new gear... ideas.. etc...
But I want to  get back to the subject of the thread...given the market today.  It was 2013 ( I think) when SSL released the SIGMA summing mixer... a digitally controlled analog summing mixer....and Bettermaker has more digitally controlled gear....so....¡There is some ¿limited? interest!.... 8)....¿Right? ;D Also the modern chips today allow a certain "simplicity" or less complicated circuitry than early 2000's... Just some thoughts.

Jay x
Odd that this shows up after all these years.
A friend has an idea for a digitally-controlled compressor. The compressor design is unique; whether it's good, bad or indifferent is beside the point. But his idea is that it should be controlled by a DAW. DAWs store automation, and the automation can send MIDI messages to connected equipment. This way, you can insert this analog compressor into your project using DAC and ADC channels on your interface, and the project recalls the settings from the automation and then sets up the compressor.

To that end, I implemented USB-MIDI in a Silicon Labs EFM8UB2 micro. It appears as four MIDI ports to the computer, two in and two out. One in/out pair just implement USB-to-serial MIDI, so it'll talk to the standard DIN jacks.

The other pair of ports talk to the compressor control (MIDI OUT from host to device) stuff and status (MIDI IN to host). When you send the correct control-change messages, you can set the threshold, ratio, makeup gain. The messages are interpreted and it sets those hardware controls with digital pots or DACs or whatever. It will also set front panel indicators. On the status side, the design will have front-panel encoders and buttons, and when adjusted, the associated parameters are adjusted immediately and MIDI control messages are sent back to the computer. I think it'd be cool to also use the micro's ADC to monitor the gain-reduction voltage and then send that back to the DAW for display in the channel, although I haven't delved enough into how to configure a DAW to use it.

I ported the code to one of SiLabs' 32-bit ARMs, an EFM32GG, but I haven't tested it too much.

I am going to use this USB-MIDI interface block in a studio monitor controller design, replacing the RS-232 port I'm using now. (My friend also suggested adding Ethernet to the studio monitor design, so it can be controlled by a computer or even a smartphone app. I'm actually working on that now.)
 
Hi!

Andy, as you have explained, I think also this is the way today: Some midi/daw control and perhaps a vst plugin/panel for the daw automation...So you can write/read automation sent to the hardware compressor...Today people want RECALL!! Ethernet also is interesting. SSL Sigma uses it for control volume/pan etc... from any internet browser, Explorer or Chrome.

Jay x

 
JAY X said:
Hi!

Andy, as you have explained, I think also this is the way today: Some midi/daw control and perhaps a vst plugin/panel for the daw automation...So you can write/read automation sent to the hardware compressor...Today people want RECALL!! Ethernet also is interesting. SSL Sigma uses it for control volume/pan etc... from any internet browser, Explorer or Chrome.

Jay x

Also, you asked, "Also the modern chips today allow a certain 'simplicity' or less complicated circuitry than early 2000's... Just some thoughts."

And that's completely true. The push to the "Internet Of Things*" has meant that all of the processor vendors (TI, ST, SiLabs, Microchip) make highly-intergrated microcontrollers with Ethernet and USB and other good stuff. Doing Ethernet and USB in an embedded design ten years ago was painful and expensive. Now it's a $5 chip with vendor-provided libraries, and free development tools.
 
I am curious; would one replace the rotary switches and gain pots with rotary encoders? If so how is the value determined, e.g. low frequency uses 47k, mid, 100k, etc....?
 
digitally controlled analog is not a new concept and it has been tried before to various degrees of success and price tag.

Euphonix consoles  up until the system 5 were digitally controlled analog. I never liked how they sounded but to each their own. I used them anyway as that is what some of the studios had at the time.  They never really caught on.

Harrison had the system 10 which was digitally controlled analog.  these sounded good but didn't really catch on as you would have thought.

One of the current digital controlled analog is the mcdsp box.  you get 16 instances of a plug in your daw that is controlling analog circuits. it's pretty slick and sounds good. but it also has a 7K price tag.
 
I just read this cool conversation between Abbey and John Roberts where John mentions the use of DPOT. Very cool stuff. Analog Devices makes an affordable circuit that uses SPI to set the value of the DPOT to anywhere from 600r to 100k.
 
Rocinante said:
I am curious; would one replace the rotary switches and gain pots with rotary encoders? If so how is the value determined, e.g. low frequency uses 47k, mid, 100k, etc....?
You use a "digital pot," a little chip that implements exactly that. It has a top and bottom and a wiper. It has a digital control port (SPI or I2C usually) that lets you load a digital number that corresponds to the wiper location. (There are variants that use a clock and direction control, so on the clock edge if direction is up the pot changes value by one step in one direction and if down it goes the other way.) Like their mechanical counterparts, the absolute resistance of the pot is generally ±10% but the steps (256 for an 8-bit pot, 1024 for a 10-bit pot) are quite accurate. You can get them in linear and log taper in a variety of total resistances.

The three pot ports go in your circuit where a standard mechanical pot would go. You need to mind your board layout so the digital control signals don't put noise onto the analog signal, but that's not difficult.

Basically, your microcontroller scans the encoders and determines when one was moved and in which direction, and based on that, it updates the pot. You can make the pots jump values "quickly," say by increasing the number loaded into it by ten instead of one, but you might want to listen and make sure that doing so doesn't have audible artifacts. You could just increase the number by one several times to get to the desired position.

The cool thing about this is that you can synchronize a computer display to the hardware and vice versa. When the user adjusts the parameter on the computer, a message is sent to the hardware which changes the pot as if the encoder was turned. And when the encoder is turned, the pot is updated and a message is sent back to the computer with the new value.
 
Andy Peters said:
You use a "digital pot," a little chip that implements exactly that. It has a top and bottom and a wiper. It has a digital control port (SPI or I2C usually) that lets you load a digital number that corresponds to the wiper location. (There are variants that use a clock and direction control, so on the clock edge if direction is up the pot changes value by one step in one direction and if down it goes the other way.) Like their mechanical counterparts, the absolute resistance of the pot is generally ±10% but the steps (256 for an 8-bit pot, 1024 for a 10-bit pot) are quite accurate. You can get them in linear and log taper in a variety of total resistances.

The three pot ports go in your circuit where a standard mechanical pot would go. You need to mind your board layout so the digital control signals don't put noise onto the analog signal, but that's not difficult.

Basically, your microcontroller scans the encoders and determines when one was moved and in which direction, and based on that, it updates the pot. You can make the pots jump values "quickly," say by increasing the number loaded into it by ten instead of one, but you might want to listen and make sure that doing so doesn't have audible artifacts. You could just increase the number by one several times to get to the desired position.

The cool thing about this is that you can synchronize a computer display to the hardware and vice versa. When the user adjusts the parameter on the computer, a message is sent to the hardware which changes the pot as if the encoder was turned. And when the encoder is turned, the pot is updated and a message is sent back to the computer with the new value.

Awesome explanation, thank you. So keeping the digital traces on one side of a pcb, and the analog on the other could be good enough? I'm assuming digital ground and analog ground would also need to be seperated?
And I assume the i2c gets translated to midi via an arduino or teensy.
Yeah I would implement this mostly for DAW recall as others have said. EQ  and compression settings on recall would be outstanding. I know Bettermaker has an eq out that does this. Its pricey though.
 
Michael Tibes said:
Are these devices suitable for high-end Audio?

Michael
That depends on what you mean by high end audio...

modern DPOTs are quite good.

The same TG technology used for the dpot wiper switches are used inside modern digitally controlled mic preamps, so good enough for many.

JR

 
Rocinante said:
Awesome explanation, thank you. So keeping the digital traces on one side of a pcb, and the analog on the other could be good enough? I'm assuming digital ground and analog ground would also need to be seperated?

Yes and there are digital isolator chips like Si8602AC which are used to isolate the parts properly.

And I assume the i2c gets translated to midi via an arduino or teensy.

These days I'd say use OSC (open sound control), MIDI 2.0 maybe later. By embedding OSC into JSON to create RESTful API's would be it easy to implement the front ends using modern JS frameworks and libraries like React, Express (nodeJS) and so on.
Sennheiser has already done that partly using JSON on their protocol:
https://assets.sennheiser.com/global-downloads/file/9541/TI_1093_v2.0_Sennheiser_Sound_Control_Protocol_ew_D1_EN.pdf
Behringer digital mixers also use OSC:
https://behringerwiki.musictribe.com/index.php?title=OSC_Remote_Protocol

Yeah I would implement this mostly for DAW recall as others have said. EQ  and compression settings on recall would be outstanding. I know Bettermaker has an eq out that does this. Its pricey though.

There already the FMR's Really Nice Compressor which does have digital (microprosessor) side channel with analog VCA. RNC though can't be controlled externally.

I2C, SPI or even CAN bus can be used at lower level to control the DAC's and such parts. Maybe I2S as well. There are already such products with use VST plugin to implement LFO's and such for modular synths and DC-output ADAT converters to send the CV's , gate and trigger signal to synth modules.
 
Back
Top