• Hi Guest: Welcome to TRIBE, the online home of TRIBE MAGAZINE. If you'd like to post here, or reply to existing posts on TRIBE, you first have to register. Join us!

My Mixer design thesis.....

ADT

TRIBE Member
So I completed my design thesis, and have gotten official credit for the concept..

so I can now talk about it at least..

essentially the idea was to redesign the crossfader using digital filtering instead of analog filtering.

thereby allowing the dj have band-specific mixing curves..

i.e. you could have a sharp curve on the lows if you just wanted to cut between basslines while leaving the mids and the highs fully in the mix..

you could also move the curves left and right if you say wanted overlap in the middle (or not)..

i got it to work for low frequencies, but i didnt really have the time or $$$ to build a real audio capable version..

that, and my stupid partner bought the wrong ADCs.. %^+*&$

anyway.. we got an A.. which is a relief

what do you think? worthwhile design?

v..abstractdatatype
 

Adam Duke

TRIBE Member
Cool idea - kind of like kill switches but with the crossfader, right? I assume that would be the most obvious approach to the tool anyways..

Can you post any pics now?

@m.
 

sentence

TRIBE Promoter
I forget who makes it but there is a mixer on the market that has separate cross faders for the lows, mids, and highs.

It's a good concept.
 

ADT

TRIBE Member
i had to return the main FPGA(field programmable gate array) to the school.. and didnt bother with a photo..

actually the truth is that i didnt get the f%$$@$% thing working until the night before the presentation because i couldnt figure out why i couldnt get any decent HF response, since my partner told me his ADCs were good for audio..

it wasnt till i wrote up the report that i realized what the problem was..

but it wasnt much to look at anyway.. all the controls were simulated on my computer so there was actually no physical crossfader or nobs or anything! for that matter except a bigass breadboard and an fpga..

up1_bd.gif


this is the board i used.. basically hooked up to my computer with a couple of ADCs attached to it (analog digital converters) and a DAC (dig. analog. conv.) for the output..

and thats basically it..

v..abstractdatatype

ironically hard trance ( :D ) gave me the idea because of the driving lows and melodic mids.. people never cut it quickly like techno djs do back and forth because the melodys get all fucked..

with this design you could.. :)
 

ADT

TRIBE Member
this concept is better than that..

A - you can also move the curves left and right to control overlap

B- 3 faders are a bitch..

1 fader that you can have presets for is so much better


the advantage of digital over analog!
 
Subscribe to Cannabis Goldsmith, wherever you get your podcasts

ADT

TRIBE Member
It is my 4th year Elec. Eng. design thesis.. we all have to do one.. basically anything we want as long as it incorperates digital design...
 

C-NiCE

TRIBE Promoter
Originally posted by sentence
I forget who makes it but there is a mixer on the market that has separate cross faders for the lows, mids, and highs.

It's a good concept.

I think this is the fader you're talking about. Made by Red Sound.

infader2.gif


Chris
 
Subscribe to Cannabis Goldsmith, wherever you get your podcasts

C-NiCE

TRIBE Promoter
^^^I think they go for around 300-350 pounds. Never seen one here though. Is this similar to what you were trying to build ADT?

Chris
 

ADT

TRIBE Member
yes and no..

instead of having 3 faders.. you have one fader, but you can preset the curves to whatever shape you want for each band..

so if you want the bass-lines to swap instantaneously without fucking with kills but to keep the melodys mixed you set a sharp curve on the lows with no overlap and then just move the fader across..
 

labRat

TRIBE Member
why didn't you use a DSP instead - you could've set up a bunch of FIR and IIR filters and change the number of taps or the coefficients to produce whatever kind of curve you wanted.

you could've easily modelled it in matlab and then write it in ansi c and compile to a dsp.

but you're done now and got an A for it. i remember you mentioning this a while back and you wanted to market it or somethin' - is this still the case?

--craig
 

ADT

TRIBE Member
i would have liked to use a dsp, but part of the project requirments were that the core functions had to be coded in VHDL, thus the use of the FPGA
 
Subscribe to Cannabis Goldsmith, wherever you get your podcasts
Top