DiyMobileAudio.com Car Stereo Forum banner

181 - 200 of 208 Posts

·
Registered
Joined
·
530 Posts
Without TDA you are just on the halvway towards a good sound in your car. It might be less important in your office as distance to L/R loudspeakers normally is the same.

Sendt fra min GT-I9505 med Tapatalk
 

·
Premium Member
Joined
·
97 Posts
Can any one help me analyse what my TA should be base on these measurements?

I have a 3 way active without subs. Crossovers at 250/3500 L-R 24

My mids are the furthest away from me base on measurements. Measurements seem to say that my underseat woofers are further away then my mids or tweeter so they should instead be put as Zero delay and around 3-5ms delay can be added to my mids and tweeters ?
 

Attachments

·
Registered
Joined
·
5,098 Posts
Set everything based on physical distances and then do your measurements. Don't try and out think the poor tape measure :).
 

·
http://*****aplaudio*****/
Joined
·
122 Posts
Can any one help me analyse what my TA should be base on these measurements?

I have a 3 way active without subs. Crossovers at 250/3500 L-R 24

My mids are the furthest away from me base on measurements. Measurements seem to say that my underseat woofers are further away then my mids or tweeter so they should instead be put as Zero delay and around 3-5ms delay can be added to my mids and tweeters ?
Right channel has discontinuity of delay at 3.5kHz.
A 0.7 ms delay (2.5 cycles of 3.5kHz) for tweeter should be good.
Crossover points at 250 Hz are ok.
You should observe 3 curves - TDA, AFR FFTq and GDR with "subtract minimum phase" ON for final decision.
 

·
Registered
Joined
·
5,669 Posts
Right channel has discontinuity of delay at 3.5kHz.
A 0.7 ms delay (2.5 cycles of 3.5kHz) for tweeter should be good.
Crossover points at 250 Hz are ok.
You should observe 3 curves - TDA, AFR FFTq and GDR with "subtract minimum phase" ON for final decision.
That's good stuff.. I need to crack out TDA and try it again once my system is in again ready for a retune. From memory, it was quite the iterative process when I did it, trial and error. You could really tell when it aligned up the 3D plot suddenly looked lovely, smooth and lined up on the left.
 

·
Premium Member
Joined
·
97 Posts
Very Strange. I was sure i saw Raimonds and some other replies yesterday. But they seem to have disappeared? Raimonds was saying my Right Tweeter was 0.7ms off? Cant find it anymore.

Anyway I made that adjustment and tried to bring the understead woofer closer to zero. But now it seems like things are a little more messed up. It could be my mic measuring position is a little off. I now have saved the seat position in a memory. The left right should be correct. But the up and down or forward and back may be a little different then the first measurement. This new position however, is the correct position.

I am still a little unsure how to read the charts to make the appropriate adjustments. Seems like my tweeters are a little ahead of my mids which is a little ahead of my woofers.
 

Attachments

·
Registered
Joined
·
5,669 Posts
Very Strange. I was sure i saw Raimonds and some other replies yesterday. But they seem to have disappeared? Raimonds was saying my Right Tweeter was 0.7ms off? Cant find it anymore.

Anyway I made that adjustment and tried to bring the understead woofer closer to zero. But now it seems like things are a little more messed up. It could be my mic measuring position is a little off. I now have saved the seat position in a memory. The left right should be correct. But the up and down or forward and back may be a little different then the first measurement. This new position however, is the correct position.

I am still a little unsure how to read the charts to make the appropriate adjustments. Seems like my tweeters are a little ahead of my mids which is a little ahead of my woofers.
Yeah posts are kinda showing up and kinda disappearing.. The site's got bugs and gremlins right now I think after a software upgrade.

Yeah I remember reading Raimond's response and it was a rather largish .7ms or so adjustment recommended.. Delaying out the tweeter to meet the mid. Trick is trial and error getting all those drivers to line up. Keep in mind then, when you're fixing delay times between sides, you'll then need to re-assess and adjust between pairs (left compared to right).

Theoretically you should be able to nail everything if you adjust left tweet/mid together to right tweet/mid. Helix allows this so easily by grouping drivers in the TA screen so I can move my whole left side over to right without disturbing timing between drivers on the left side tweeter on down.
 

·
Registered
Joined
·
530 Posts
Yeaa, it is nice feature of Helix. But you can make this very simple with apl-soft by adding time delay to one of the channels.

Sendt fra min GT-I9505 med Tapatalk
 

·
Registered
Joined
·
530 Posts
For doing this it is possible to delay fir-filter. And more over, you can do it very precisely .

Sendt fra min GT-I9505 med Tapatalk
 

·
Registered
Joined
·
5,669 Posts
Here’s an interpretation question for you guys on TDA... I’ve not listened to this yet but, here’s what I was able to achieve in TDA mainly by a lot of trial and error starting with sides then overall. Very very odd delay setting for right midbass to achieve this. Also very low delay amounts on IB subs.

Thoughts?



Edit: So I listened with skepticism that was well founded I suppose. It wasn’t even remotely close, on all fronts. So either I’m just doing something terribly wrong in assessing TDA, or something else.


Sent from my iPhone using Tapatalk Pro
 

·
http://*****aplaudio*****/
Joined
·
122 Posts
Hi,

You can use such tune if crossover points are pretty much good in terms of AFR - no deep holes, for minimum phase filters (ordinary IIR)
It is absolutely correct if you have linear phase filters.
You are correcting min. phase crossover's GD such way also in your example.

But, if try to be accurate as much as possible you should go to not "flat" as in your example
but to a crossover's group delay curve, for minimum phase filters.
Let's take such crossover:



Its Group delay curve is such:



And its TDA picture:



Then, it is possible to use time correction with such GD:



And get this:





If summarize, it is semi solution to correct crossover's GD using band's delays
that can be used in some cases with caution, keeping look on AFR.
 

Attachments

·
Premium Member
Joined
·
97 Posts
I know this thread is rather quiet. But I was hoping for help on getting my time alignment correct using TDA. Please see my measurements for both sides at drivers seat with TA only via tape measurements.

And then both with extra delay on mid by 10cm and woofer by 20cm as well as reversing phase for mid and woofer to see if it improves.

My system is a 3 way active with tweeter, mids and woofer. No subs. crossovers at 200 and 2.5k

If I were to check the first measurement, it would seem like the mids are around 1ms slower then the tweeter and the woofers are at least 7ms slower then the mids. So the correct way to adjust for it is to increase the distance/delay on mid by 1ms and woofer by 7ms so that they arrive earlier and in time with the tweeter?

How do you determine from the graph if say the Right tweeter signals is coming faster then the Left tweeter? Do you run a Left only measurement and see the graph and adjust accordingly?
 

Attachments

·
Registered
Joined
·
101 Posts
I am having a similar problem.

Here is the graph i get with TA set by tape measure:







When add almost 15ms delay to the midbass and tweeter, the graph looks something like this:



It is a Car PC. I am using AudioMulch for processing. Plugin called LS_Filter is used as a crossover. Some PEQ to fight with peaks. ConvolverVST is applying a FIR filter generated by Python based thingy called PORC (Python Open source Room Correction).

I have a huge sub in the trunk of a Range Rover Sport. It is a 21" PA sub. Have Beyma 8G40 in ported enclosures in my doors. And full body horns under the dash.

Delay on the sub is 0, midbass drivers and horns a delayed in the range of 0-2,5ms. So, nothing huge.

Frankly, the system sounds terrific :) Imaging is nice, tonality is fantastic, dynamic is from the outer world. Sometimes sub bass sounds like it is lagging a bit. But maybe it is me now trying to rationalize the graph :)

Why TDA "thinks?" the sub starts acting 15ms later than the rest of the system?

Why the graph becomes crazier after the "needed" adjustment?

P.S. Measurements are taken with the engine running. Measuring it with engine off gives the same basic picture.
 

·
Registered
Joined
·
1,999 Posts
This is without any FIR applied right?
Also engine running might not change a lot but I would measure without it just for clarity, it could disturb the reading near 80Hz.

I imagine it’s your ported midbass delay around 100Hz (but that’s really high).
I wouldn't try to get it flat, but just the smoothest possible, even if 20Hz plays 20ms later.
 
181 - 200 of 208 Posts
Top