that would be why most of the programs require some pretty hefty processors.
do you happen to have that link to the driver haven discussion?
It sounded like you could string the 5 band parametric EQ together. (Or you could use the 31 band VST pluging with Winamp.) Its not one complete software soultion but at least I am rethinking my use of the Alpine processor again. Dammit!
Wow damn durwood, I give you the noob award of the year. Since you got here youre full of information, im lovin it. That VST plugin and that 31 band EQ just made my day, its so much better than the 4front eq which was like 10band 1 octave. This winamp eq will do all my equalizing. I wont use any artificial boosts or anything. Just to get it as flat as possible (link to tune) as I know I have some overlapping frequencies with my component sets... I havent even heard it yet and I know im gonna need some tuning. But once is all said and done its gonna be awesome. Plus with that line driver you sold me, I should have excellent efficiency
:whs:, guess all n00bs aren't bad n00bs:rofl: jk ;) nice to have your input for sure:)
:buddy: Thanks and cheers to everyone here. I just wanted to help others as much as I can. I remember when I got into car audio 10 years ago and carputers 3 years ago. I struggled through both eras trying to gather as much info by learning, researching, experimenting, and asking questions. ..and just when I think I got it all figured out a whole set of new ideas comes around (thanks mostly to RED and scott_fx ;) ). I just want to pass it along because there are so many out there who don't share...but hey...I am a car audio/computer/electronics junkie. I need a support group. :D
heh - THIS is one hell of a support group.
Sounds like we're one step closer to an all in one solution! And kudos if we can use winamp for it all!
my biggest single concern with a total setup like this is going to be processor usage... how much will it take to run this behind the scenes, & how heavy of a performance hit can we expect on a carpc.... that's got to be the main variable that will make this either very practical or very impractical..., anyone have any idea on this??
durwood, I really appreciate your input, as well as Red and scott_fx and everyone in this forum. People with a "real" knowledge under their belt.
I know that a Parametric EQ is more computationally efficient than Graphical EQ. See the latest apple IPOD's patent, which basically converts 10-band graphical EQ into single Parametric EQ - is 10 times "lighter" on processing power.
This is why you see more of Parametric EQ's on digital mixers (such as Yamaha Digital mixers). In your stand along GEQ, it is composed of passive RCL or active OpAmps for each freq band, so # of band does not really matter. But that is why they are not very precise (or the precise GEQ's are very, very expensive).
But in all-digital domain, the PEQ is better. I am sure that somebody somewhere has worked out the math of converting a Graphical EQ curve into n # of PEQ's with appropriate parameters, and wrote some codes for it. Personally, I think 4 PEQ's can create whatever the 31 band GEQ can give you (in Practical Audio Application), only 10 times faster in processing speed.
I have a DEQ2496 Behringer processor/GEQ/PEQ with RTA that I use in my Recording Studio. It has a Microphone input. It generates Pink/White noise, measure the overall frequency response at the position of the mic, draws the curve, inverts it, and makes a "suggestion" of GEQ that makes the overall frequency response flat. I'll have to go look at the manual to see if it can also give a "suggestion" on PEQ settings, but I'm sure this is possible with a software and a soundcard. May be this one can do it?
Too bad their free version can only do 1 oct. We need 1/3 octave to be at least useful.
Whoa. Aren't we havin' fun yet?
There's no doubt that something along these lines will be very processor intensive.
however, in my eyes, at this stage of the game, if you want to be able to do this, it's something you'll have to accept.
The way I see it, until the need is made known, there's no reason to code it any differently. I'm not sure if there would be a way to decrease the processor load on something like this...but then again, I'm not a programmer by any stretch of the imagination.