Dante, Ethersound, ultranet,aes/ebu, adat,etc,etc, are all digital audio transfer protocol.
May i ask why you want your mixer to be digital and what is the task you plan for it? If this is typical live application (mixing of analog source - typical microphone, instruments, dj gear) i agree with Turk182 analog is the way to go imho. If you need recall and a lot of treatment power, you know most of your source will be digital, then yes digital mixer are great.
Diy one digital desk is not going to be easy nor cheap: if you want to connect microphone or other analog source you'll need analog preamp at least ( including +48v psu), then ad converters and da ( for analog out for main, groups and send,...). This will be a lot of work only for analog in/out. Then digital in/out cards. Then you'll have to develop software for treatment in digital domain (including matrix for bus/group, mains,sends, eq, comp,...) and interface to drive them ( faders, knobs, button,switch,display...). All taken into account, an already made system will be cheaper, probably much more stable and as already said capable to withstand conditions of use without falling appart or have bug.
May i ask why you want your mixer to be digital and what is the task you plan for it? If this is typical live application (mixing of analog source - typical microphone, instruments, dj gear) i agree with Turk182 analog is the way to go imho. If you need recall and a lot of treatment power, you know most of your source will be digital, then yes digital mixer are great.
Diy one digital desk is not going to be easy nor cheap: if you want to connect microphone or other analog source you'll need analog preamp at least ( including +48v psu), then ad converters and da ( for analog out for main, groups and send,...). This will be a lot of work only for analog in/out. Then digital in/out cards. Then you'll have to develop software for treatment in digital domain (including matrix for bus/group, mains,sends, eq, comp,...) and interface to drive them ( faders, knobs, button,switch,display...). All taken into account, an already made system will be cheaper, probably much more stable and as already said capable to withstand conditions of use without falling appart or have bug.
Last edited:
Hamsa to give you some perspective i once got to the prototype stage of an 8x8 mix matrix that used U-pots to mix analog audio (no A2D or D2A conversion of audio) that used all the I/O of a PIC and 16meg of memory to run basic routines with simple fade and crossfade routines, and was still a long way from a fully working device, project cost close to 4 thousand dollars at that price you can get a Yamaha 01v (didn't Krivium mention that!!)
the other aspect of building a digital mixer is you must be a Jedi level software/code master to do even simple things like write "fade ramp routines" otherwise you can make zipper noises across your audio...even the mechanics of PCB layout is a challenge unless your using a service that can make a multilayer board for you (not cheap) sometimes despite limitations, ready made is the way to go.
although i have not looked into that sort of thing in a while i do wonder if any of the firewire adapters could bussed to create a crude mixer...?
the other aspect of building a digital mixer is you must be a Jedi level software/code master to do even simple things like write "fade ramp routines" otherwise you can make zipper noises across your audio...even the mechanics of PCB layout is a challenge unless your using a service that can make a multilayer board for you (not cheap) sometimes despite limitations, ready made is the way to go.
although i have not looked into that sort of thing in a while i do wonder if any of the firewire adapters could bussed to create a crude mixer...?
What about just using a powerful laptop with a big USB interface?
Do all EQ/processing (could even be linear-phase) in the laptop, and just use the interfaces as DAC/ADC and I/O.
It'd be a load of programming, but not impossible by any stretch. I'd probably want to do it within Linux or similar - I wouldn't trust Windows with that.
Alternatively, these guys have done a 16-channel desk in 1RU: StudioLive 16R | PreSonus
Chris
Yes i thought about a solution like that too ( after all protools had a dedicated control surface for live act iirc). BUT i wouldn't trust a system like that... too much chance of bug! If this is to be used for real ( live act) only something known as being stable should be allowed imho.
Even more if money is involveld ( if you get paid for the job).
If for the fun of build, so why not if you are ready to be the next 'jedi' programmer.
Then yes why not use a pro (or semi pro) soundcard and focus on the digital domain ( enough work just in there to have something simple and usable).
Even more if money is involveld ( if you get paid for the job).
If for the fun of build, so why not if you are ready to be the next 'jedi' programmer.
Then yes why not use a pro (or semi pro) soundcard and focus on the digital domain ( enough work just in there to have something simple and usable).well, it is assigned to us (Me and Hamza) as Final Year Project for our Bachelor Degree. So. basically we have to build it from scratch.
Ok you have to build a whole console or only the digital part of it? A whole console seems a bit overoptimistic from your teacher as it seems Hamza didn't know of digital transfer protocol which is a obvious prerequisite.
If i was you (and if teachers allow) i would focus on a software developpement at first. Choose a digital platform ( maybe a raspberry pi) learn the code used ( python for the pi) and make a simple application with volume control, pan, mute and a main bus.
From there you could then use some of the raspberry i/o bus protocol to implement ad/da cards. The Pi will have the advantage you've got a whole lot of already developped function at your disposal ( this is open source) and implementation of digital i/o and control ports should be easier as there is 'plug boards' offer for it. This will be the easiest way imho.
If i was you (and if teachers allow) i would focus on a software developpement at first. Choose a digital platform ( maybe a raspberry pi) learn the code used ( python for the pi) and make a simple application with volume control, pan, mute and a main bus.
From there you could then use some of the raspberry i/o bus protocol to implement ad/da cards. The Pi will have the advantage you've got a whole lot of already developped function at your disposal ( this is open source) and implementation of digital i/o and control ports should be easier as there is 'plug boards' offer for it. This will be the easiest way imho.
Yep, and most mixing desks are expected to be able to do something useful when microphones are connected.
You can probably find a simple preamp online. Don't bother adding a +48v supply unless you want to make a commercial product. Most mics are dynamic anyway.
Chris
Have a close look at THAT corps recently introduced single chip digitally controlled mic preamp & ADC driver, solves most of the analogue input bit for you (They are two channel parts, so you will want 4).
AKM make some sane looking 8 channel ADC parts, stuff into a modest DSP (Actually, for 8 channels a modest ARM microcontroller would probably get it done), do the maths and stuff the results back out a DAC.
A really workable control surface and software is actually the elephant in the room, by far the hardest part of all this.
Clocks should be an explicit crystal osc and not based on the integrated processor timing hardware.
Regards, Dan.
AKM make some sane looking 8 channel ADC parts, stuff into a modest DSP (Actually, for 8 channels a modest ARM microcontroller would probably get it done), do the maths and stuff the results back out a DAC.
A really workable control surface and software is actually the elephant in the room, by far the hardest part of all this.
Clocks should be an explicit crystal osc and not based on the integrated processor timing hardware.
Regards, Dan.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.