Software options for Tuning - Page 4

durwood · 10-26-2006, 11:26 AM

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.

RedGTiVR6 · 10-26-2006, 01:26 PM

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!

turbocad6 · 10-26-2006, 01:35 PM

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??

jasonsjwou · 10-26-2006, 01:36 PM

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.

http://www.macsimumnews.com/index.ph..._ipod_and_more

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?

http://www.trueaudio.com/

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?

RedGTiVR6 · 10-26-2006, 01:45 PM

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.

turbocad6 · 10-26-2006, 01:48 PM

my opinion on parametric VS 1/3 octave... parametric is fine for a studio or home theater, I mean your finetunning the curve, by pushing up or down at specific spots, & varying the rolloff or slope of your tuning (Q)... this is fine when your dealing with a reference system & reference drivers in a relatively controlled environment...., or at least an environment that is not wreaking havoc on the response...

a car is a whole different animal.... even if your reference system & your reference drivers were spot on, the vehicle itself will show so many anomolies & add & alter so much compared to the actual response, that in my opinion, a parametric just won't be adiquate to really fine tune, unless your talking many bands...

there are almost always instances in a vehicle that it will be necissary to boost a freq while cutting an adjacent freq... this is not what parametrics do best..., yes a parametric can help, & yes, I'm sure that a parametric can tune a vehicle well, but I think 1/3 ovtave tunning is going to be superior in a vehicle...

a car & a studio/home are 2 totally different environments....

jasonsjwou · 10-26-2006, 01:58 PM

Excuse my insufficient knowledge in Car Audio environment, but it it really that severe of an acoustic space it is? (such that you have to give such a steep slope just over 1/3 oct?)

I've measured numerous churches, home theaters, and recording studios and small open park concert stages... The compensating curve in those situation looks like a SideWinder trail...

I've never "measured" the car environment, nor have I ever "seen" a RTA actually being measured in the car cabin. Do you have a website or a reference of such information? - just so that I may get some Idea of how "harsh" it really is?

turbocad6 · 10-26-2006, 02:25 PM

not offhand, but considering the off-axis configuration, combined with the multitude of close reflective surfaces, close absorbsion materials, & odd resonances, etc... it's bad

even the worse cases of what your talking about will be significantly less harsh, simply because of the size... throw a tennis ball in a large room, & compare that to throwing a tennis ball in a car, to give you a little idea of what I mean...

durwood · 10-26-2006, 02:31 PM

First-I have and use the RTA program (TrueRTA) you listed with an inexpensive behringer mic and powered mic mixer. Not an Audiocontrol RTA ($2000) but a $200 RTA solution. Works great for my needs as I use it at the very end to verify what I hear.

The car audio environment is crazy. There are mechanical aspects and electrical aspects of tuning. If the mechanical part isn't right (i.e. driver placement, mounting materials) then you need a whole slew of electronics to "fix" it. Some people frown on this. But electronics have gotten better over the years and more options are available.

In the SQ scene, there is a whole battle between the old folks and new folks or what I like to call battle of kick panels vs. no kick panels. I am a no kick panel person so you NEED everything to make it work-time delay, crossover, EQ. (I use minimal EQ'ing because if everything else is right it should be damn close without it)

Software can do this for us, because most audio software is based on math, and there is a lot of math behind sound waves. So as long as the software is programmed correctly it should work. I have a feeling winamp could do it all with the right plugins.

turbocad6 · 10-26-2006, 02:34 PM

oh, don't even get me started on time delay

this has to be one of the most overused & misundestood ends of car audio... I mean there are actually guys who will insist that delaying the left side can push the drivers back, & equalize the pathlengths some, thereby making the system "better".... yeah, maybe... if it is a one person car

durwood · 10-26-2006, 02:37 PM

Quote: Originally Posted by turbocad6

oh, don't even get me started on time delay

this has to be one of the most overused & misundestood ends of car audio... I mean there are actually guys who will insist that delaying the left side can push the drivers back, & equalize the pathlengths some, thereby making the system "better".... yeah, maybe... if it is a one person car

Yep It is...who else is driving it

..you must be a kick panel person...

You must like the "rainbow effect" of kicks. My ears hear better than my ankles.

turbocad6 · 10-26-2006, 02:39 PM

no, I'm a hlcd type person

jasonsjwou · 10-26-2006, 02:39 PM

May be we should start slappin' one of them "Diffusers" and "Absorbers" in our cars... (or have it build into the interior... like, plastic pieces with ripples, headliner with acoustically transparent materials with sound diffusers (like ribs and fresnels) underneath, etc... I see the glass can be a major issues...

Personally, I hate salesman talk of somtehing like "oh you put this in the car it'll sound incredible because it has XXXXX".. I'd much rather see (even a ill-represented ones) data/chart/graph of lot of before/after and with/withouts...

Well. I'll go home tonight and pull my RTA and ECM8000 out of the rack and get into my car and start measurin'

durwood · 10-26-2006, 02:43 PM

Quote: Originally Posted by turbocad6

no, I'm a hlcd type person

Hmm.. don't know your type.

jasonsjwou · 10-26-2006, 02:44 PM

Personally, I want the system to sound good to MY EARS. I don't give a damn about people sittin' next (or behind) me. Are they making my car payment? Then I'll program it so that the sweet spot moves to wherever they are seating.

But for now, I'm making my car payments so I am steering that sweetspot right on my head