My life with an RTA
05-16-2004, 09:21 AM
#1
1000 Watt CAFz'r
Thread Starter
Join Date: Oct 2003
Posts: 1,783
With the sheer number of SPL people seeming to own their own "meter", I find it strange that more "SQ" guys don't have their own RTA.
There are lots of simple to use computer based RTAs for under $1000. And if you like to do a little digging, there are some measurement programs that can be integrated with your computer's sound card for under $300.
The bulk of my RTA experience is from high end home speaker design. Using an RTA to tune an EQ in a car is a new experience to me. So to start, I was pretty excited about the concept of "fixing" peaks and valleys in my system's response. I was also excited to see how the car environment interacts with speakers.
Over the past few months that I've had the RTA, I've learned a LOT about car interiors. Car's are HORRIBLE for putting a stereo into. Every time I think that I've got a good sense of something in the car, I get thrown a curveball.
The Tests:
To start, I have measured every driver's response from both sides of the car. In doing so,the mic was pointed both at and away from the driver being tested. I then started adjusting X.O. points for individual drivers and measuring the response from both sides of the car. On average, I have used 10 different High-pass X.O. points and 10 different Low-pass X.O. points for a driver. (the tweeter used just high pass points) All of the X.O. points were measured from both sides of the car. I then started measuring ALL of the above X.O. points with the drivers in pairs. For example, I would connect the tweeters and midranges and adjust the X.O. points. I have done this now for each driver overlap of my 3 way front stage, and my midbass to sub overlap.
Each test sweep performed has had a resolution of between 50 to 100 points per octave. I have done HUNDREDS of sweeps, with sweeps taking anywhere from 1 to 4 minutes to setup and exicute. And.....I'm still not done!
All of the above was just to establish a baseline for when I do start to tune. Imagine spending a morning doing some listening and tuning in your car. At the end of the morning you throw the car on the RTA to see what it looks like. I'll assume that there have been 4 small tuning changes performed over the morning. It would be hard to quantify the 4 small changes unless you have a baseline. And...our ears don't have memory. (no mater how often you try to convince yourself that you "remember how it sounds")
I mentioned before that I've had some "funny stuff" going on in the car. Last night I added 3db of gain at 6.3khz to compensate for a dip in response and the car responded with a 9 db increase. And...it works just the opposite when you have phase cancellation.( 3 db boost might give you 1 db in the car) The baselines that I've laid down allow me to see how the driver interacts with the car. It also shows a driver's limits. For example, how high can you play your midrange driver.
As a few of you know, I've been struggling to tame my Focal TN51 tweeter. I've had other tweeters in the car that sound better, but they were HUGE and couldn't be mounted to give better imaging response than the TN51s on the A-pillars. What I've found, is that that TN51 is a really flat tweeter out to 20khz outside of the car. It's flater than many of the "home" tweeters that I've had in the car. It does however, interact horribly with the windshield. I'm sure a few of you experienced "PROS" will understand the peak in the 5khz region. While this can be easily tamed with my EQ, it sounds worse afterwards. In fact, I've found that where-ever I've tried to EQ a problem area, it sounds less life like afterwards. As a result, I'm finding that working on X.O. points and gains are netting me far great results than EQing.
Note: I'm not saying that my EQ sounds bad, I'm saying that the changes I'm making with it sound bad.
I'm really enjoying the in depth view of my system that the RTA privides me. It's a really good tool to anyone with the patience and knowledge to use it properly. I'm finding more and more that it relies on the user's knowledge of audio and sound in general to generate good results. Like anything, the more you do it, the better you get at it. For me, it's a valuable tool for learning about sound in general, and hopefully, for improving the sound of my car.
Shortly, I'll be posting lots of graphs of my work. The aim will be the sharing of the knowledge that I've gained from RTA, and hopefully encouraging a few others to get out there working with one too.
Adam
There are lots of simple to use computer based RTAs for under $1000. And if you like to do a little digging, there are some measurement programs that can be integrated with your computer's sound card for under $300.
The bulk of my RTA experience is from high end home speaker design. Using an RTA to tune an EQ in a car is a new experience to me. So to start, I was pretty excited about the concept of "fixing" peaks and valleys in my system's response. I was also excited to see how the car environment interacts with speakers.
Over the past few months that I've had the RTA, I've learned a LOT about car interiors. Car's are HORRIBLE for putting a stereo into. Every time I think that I've got a good sense of something in the car, I get thrown a curveball.
The Tests:
To start, I have measured every driver's response from both sides of the car. In doing so,the mic was pointed both at and away from the driver being tested. I then started adjusting X.O. points for individual drivers and measuring the response from both sides of the car. On average, I have used 10 different High-pass X.O. points and 10 different Low-pass X.O. points for a driver. (the tweeter used just high pass points) All of the X.O. points were measured from both sides of the car. I then started measuring ALL of the above X.O. points with the drivers in pairs. For example, I would connect the tweeters and midranges and adjust the X.O. points. I have done this now for each driver overlap of my 3 way front stage, and my midbass to sub overlap.
Each test sweep performed has had a resolution of between 50 to 100 points per octave. I have done HUNDREDS of sweeps, with sweeps taking anywhere from 1 to 4 minutes to setup and exicute. And.....I'm still not done!
All of the above was just to establish a baseline for when I do start to tune. Imagine spending a morning doing some listening and tuning in your car. At the end of the morning you throw the car on the RTA to see what it looks like. I'll assume that there have been 4 small tuning changes performed over the morning. It would be hard to quantify the 4 small changes unless you have a baseline. And...our ears don't have memory. (no mater how often you try to convince yourself that you "remember how it sounds")
I mentioned before that I've had some "funny stuff" going on in the car. Last night I added 3db of gain at 6.3khz to compensate for a dip in response and the car responded with a 9 db increase. And...it works just the opposite when you have phase cancellation.( 3 db boost might give you 1 db in the car) The baselines that I've laid down allow me to see how the driver interacts with the car. It also shows a driver's limits. For example, how high can you play your midrange driver.
As a few of you know, I've been struggling to tame my Focal TN51 tweeter. I've had other tweeters in the car that sound better, but they were HUGE and couldn't be mounted to give better imaging response than the TN51s on the A-pillars. What I've found, is that that TN51 is a really flat tweeter out to 20khz outside of the car. It's flater than many of the "home" tweeters that I've had in the car. It does however, interact horribly with the windshield. I'm sure a few of you experienced "PROS" will understand the peak in the 5khz region. While this can be easily tamed with my EQ, it sounds worse afterwards. In fact, I've found that where-ever I've tried to EQ a problem area, it sounds less life like afterwards. As a result, I'm finding that working on X.O. points and gains are netting me far great results than EQing.
Note: I'm not saying that my EQ sounds bad, I'm saying that the changes I'm making with it sound bad.
I'm really enjoying the in depth view of my system that the RTA privides me. It's a really good tool to anyone with the patience and knowledge to use it properly. I'm finding more and more that it relies on the user's knowledge of audio and sound in general to generate good results. Like anything, the more you do it, the better you get at it. For me, it's a valuable tool for learning about sound in general, and hopefully, for improving the sound of my car.
Shortly, I'll be posting lots of graphs of my work. The aim will be the sharing of the knowledge that I've gained from RTA, and hopefully encouraging a few others to get out there working with one too.
Adam
05-16-2004, 10:37 AM
#2
Yankee
Join Date: Oct 2003
Posts: 3,599
If I were serious enough I would purchase the same mic/RTA/software as the sanctioning body I would be competing with. The variances are there with the hardware.
Mechanical damping and diffusion are tools you can employ to tame the hot tweeter. I have just purchased L-pads for my tweeters and midrange drivers to work with my modular concept.
Mechanical damping and diffusion are tools you can employ to tame the hot tweeter. I have just purchased L-pads for my tweeters and midrange drivers to work with my modular concept.
05-16-2004, 11:00 AM
#3
1000 Watt CAFz'r
Thread Starter
Join Date: Oct 2003
Posts: 1,783
I think the sanctioning bodies use the Audio Control 1/3 octave RTA. 3 points per octave just isn't enough to really see what's going on. I'm able to graph as high as 200 points per octave, but in reality 50 points per octave gives you enough info to make good decisions.
I haven't started reflective tuning yet. Still working on angles.
Adam
I haven't started reflective tuning yet. Still working on angles.
Adam
05-16-2004, 12:45 PM
#4
50 Watt CAFz'r
Join Date: Oct 2003
Posts: 146
Originally posted by JohnVroom:
If I were serious enough I would purchase the same mic/RTA/software as the sanctioning body I would be competing with. The variances are there with the hardware.
If I were serious enough I would purchase the same mic/RTA/software as the sanctioning body I would be competing with. The variances are there with the hardware.
And as PEI330Ci said, those RTAs usually don't show a high enough resolution for you to fine tune the system. 1/12 oct or higher should be better, but then you need an EQ with that scale too. <_<
I've been playing around with the idea of measuring the averaged transfer function of the entire car at the headrest, inversing it, then running the audio signal through the function to see how it would sound. I don't have the equipment to do that though... Has it been done before?
05-16-2004, 12:57 PM
#5
1000 Watt CAFz'r
Thread Starter
Join Date: Oct 2003
Posts: 1,783
^^I have the equipment to do that.
There is a transfer function that I can use to show a car's sensitivity at a specific point in the car. But, it changes as you move the mic around. As you said "averaged" is the way to go, I just have to figure out how to do that with my software. Got to go read the freaking manual again.....
How would it sound? Like crap! I've already EQ'd my car pretty flat across a few frequency ranges, and it sucks the life right out of the music.
I wish I could have more people hear the things that I experiment with...sometimes it boggles all reason the things that I hear.
Adam
There is a transfer function that I can use to show a car's sensitivity at a specific point in the car. But, it changes as you move the mic around. As you said "averaged" is the way to go, I just have to figure out how to do that with my software. Got to go read the freaking manual again.....
How would it sound? Like crap! I've already EQ'd my car pretty flat across a few frequency ranges, and it sucks the life right out of the music.
I wish I could have more people hear the things that I experiment with...sometimes it boggles all reason the things that I hear.
Adam
05-16-2004, 05:04 PM
#7
Yankee
Join Date: Oct 2003
Posts: 3,599
There shouldn't be any variances in the hardware, if it's good hardware. Also, systems can be calibrated.
05-16-2004, 06:32 PM
#9
1000 Watt CAFz'r
Thread Starter
Join Date: Oct 2003
Posts: 1,783
Here is what I have:
http://www.linearx.com/products/...LMS/LMS_01.htm
Here is what I'd recommend if you have the bucks:
http://www.mlssa.com/
Here is a great bang for the buck setup that offers phase response measurement as well:
http://www.cliowin.com/
These are the biggies, there are a few more but I don't have any experience with them. The above are serious RTAs capable of higher than 1/12 octave resolution. I have discounted the Audio Control units as they offer less than desireable resolution.
Adam
[ May 16, 2004, 07:41 PM: Message edited by: PEI330Ci ]
http://www.linearx.com/products/...LMS/LMS_01.htm
Here is what I'd recommend if you have the bucks:
http://www.mlssa.com/
Here is a great bang for the buck setup that offers phase response measurement as well:
http://www.cliowin.com/
These are the biggies, there are a few more but I don't have any experience with them. The above are serious RTAs capable of higher than 1/12 octave resolution. I have discounted the Audio Control units as they offer less than desireable resolution.
Adam
[ May 16, 2004, 07:41 PM: Message edited by: PEI330Ci ]
