DiyMobileAudio.com Car Stereo Forum banner

161 - 174 of 174 Posts

·
Registered
Joined
·
187 Posts
Remind me where your target crossover point is?

At any rate, I think we’re talking about the difference between electrical crossover (that which is set in the DSP) and the acoustic crossover point.

The acoustic crossover point and slope (that which is measured with a microphone) are the only thing that matters. You can set your electrical crossover any way you like. Doesn’t have to be linkwitz Riley or 24 dB/octave. Electrically, you can do whatever you like as long as it gets you to your acoustic target.

This occurs because of the acoustics of the environment. Certain frequencies get altered by the car environment so that it’s actually unusual (in my experience) that an electrical crossover point and slope measures perfectly in REW.

It has nothing to do with the particular DSP. But a great DSP, which a helix is, has all sorts of tools to help you get that target acoustic crossover point and slope.


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
734 Posts
Different crossover types accomplish slightly different things. LR sacrifices phase alignment to get better frequency response. Bessel sacrifices frequency response to maintain phase alignment. Throwing EQ on top further adjusts phase response.
 

·
Registered
Joined
·
187 Posts
Different crossover types accomplish slightly different things. LR sacrifices phase alignment to get better frequency response. Bessel sacrifices frequency response to maintain phase alignment. Throwing EQ on top further adjusts phase response.

Opekone, you measure phase a lot which is awesome.

If you had two crossover slopes that when measured with a mic look exactly the same, would the phase of those curves be different if a LR filter was used versus some other crossover type?

Let’s assume everything else is the same. The slopes are identical, the shape is identical, everything is the same, but LR and some EQ was used in one version and a Butterworth and some EQ was used in another. The effect on phase and frequency would be the same...no?

At least I think they would be as you’re starting with the same original signal and applying the overall same filters (in sum) to get the same end result acoustically. Right? Or am I wrong...


Sent from my iPhone using Tapatalk
 

·
Registered
2020 WRX
Joined
·
18 Posts
Different crossover types accomplish slightly different things. LR sacrifices phase alignment to get better frequency response. Bessel sacrifices frequency response to maintain phase alignment. Throwing EQ on top further adjusts phase response.
Thank you for explaining!
 

·
Premium Member
Joined
·
4,711 Posts
Opekone, you measure phase a lot which is awesome.

If you had two crossover slopes that when measured with a mic look exactly the same, would the phase of those curves be different if a LR filter was used versus some other crossover type?

Let’s assume everything else is the same. The slopes are identical, the shape is identical, everything is the same, but LR and some EQ was used in one version and a Butterworth and some EQ was used in another. The effect on phase and frequency would be the same...no?

At least I think they would be as you’re starting with the same original signal and applying the overall same filters (in sum) to get the same end result acoustically. Right? Or am I wrong...


Sent from my iPhone using Tapatalk
Does the impulse response look the same?
 

·
Registered
Joined
·
312 Posts
Discussion Starter #167
Remind me where your target crossover point is?

At any rate, I think we’re talking about the difference between electrical crossover (that which is set in the DSP) and the acoustic crossover point.

The acoustic crossover point and slope (that which is measured with a microphone) are the only thing that matters. You can set your electrical crossover any way you like. Doesn’t have to be linkwitz Riley or 24 dB/octave. Electrically, you can do whatever you like as long as it gets you to your acoustic target.

This occurs because of the acoustics of the environment. Certain frequencies get altered by the car environment so that it’s actually unusual (in my experience) that an electrical crossover point and slope measures perfectly in REW.

It has nothing to do with the particular DSP. But a great DSP, which a helix is, has all sorts of tools to help you get that target acoustic crossover point and slope.


Sent from my iPhone using Tapatalk
Crossover points are <--60, 80 - 400, 400 - 4000, 4000--->.

A kind I've started over again today. I took new baseline measurements and then went through would I call another "round one" of REW EQ tuning. I will post some screenshots later tonight.

I was originally using 80 - 300, 300 - 3000, and 3000--->, but shifting everything to the right seemed like it was working a little better for me.
 

·
Registered
Joined
·
187 Posts
Does the impulse response look the same?
Are you asking if the IR is the same when two crossovers are set on the same driver but by different means (like LR Plus EQ vs Butterworth plus EQ) to achieve the same acoustic response? Hmmm, I would think it would be the same but I haven’t tried this. Is it?


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
312 Posts
Discussion Starter #169
Ok, here we go:

Left and Right Tweeters:

274823


Left and Right Midrange

274824


Left and Right Midbass

274825


Subwoofer

274826


Midrange and Tweeters

274827


Midbass and Midrange

274828


Whole system left and right:

274829


Subwoofer and Midbass (Normal sub and reverse polarity Sub)

274830


All drivers:

274831


All drivers with summing:

274832


I've definitely got some more work to do. I need to level match a bit and then flatten out some of those responses.
 

·
Premium Member
Joined
·
4,711 Posts
Are you asking if the IR is the same when two crossovers are set on the same driver but by different means (like LR Plus EQ vs Butterworth plus EQ) to achieve the same acoustic response? Hmmm, I would think it would be the same but I haven’t tried this. Is it?
Usually they are not.

Then the passband ripple and the steepness of the rolloff are tradoffs.
And different filter have different passband ripple.

I do not know if the passband ripple is causing the ripple in your response, but it looks more rippled than I would have imagined.
Maybe someone else smarter, or more experienced knows?
 

·
Registered
Joined
·
312 Posts
Discussion Starter #171
Here is a plot with JBL (Andy) individual driver overlays.

274892


So I'm obviously not hitting all of my crossover points and I doubt I will ever get there with just EQ (mainly the right midrange and both of my midbasses). It takes a lot of 6dB boosting to move the slopes, so would you suggest changing crossover setting on individual drivers to get closer to target? For example, would moving my right midrange back to 320 - 3920 while keeping the left one 400 - 4000 be a better option? For the midbass, should I change my crossover pints from 80 to 70? I'm not quite sure that will help. I've read people rave that the GB60's are midbass monsters. Would not being able to reach target rolloff at 80Hz be an indication that I need a more rigid mounting solution for the driver? Aside from the top of the door where the window goes up and down (which gets sealed up when the door card goes back on), the door is pretty well sealed with sound deadener.
 

Attachments

·
Registered
Joined
·
312 Posts
Discussion Starter #172
Changed sub and midbass crossover points to 63 and 85. Fooled around with reverse polarity on the sub. Seems to work well.

274918


I also changed the right midbass lowpass crossover to 3920Hz and adjusted gains . I may go to 360 on the highpass next time. Seems to have matched rolloff a bit better. Now I just have to work on flattening them out a bit.

274919
 

·
Registered
2020 WRX
Joined
·
18 Posts
Changed sub and midbass crossover points to 63 and 85. Fooled around with reverse polarity on the sub. Seems to work well.

View attachment 274918

I also changed the right midbass lowpass crossover to 3920Hz and adjusted gains . I may go to 360 on the highpass next time. Seems to have matched rolloff a bit better. Now I just have to work on flattening them out a bit.

View attachment 274919
Have you tried the auto time alignment feature? I had to mess with the sub-phase. When I used auto time alignment It was fine the subs measured slightly more away, after that No need for any phase adjustments.

Might be worth a shot.Record your info before trying.
 

·
Registered
Joined
·
312 Posts
Discussion Starter #174
Have you tried the auto time alignment feature? I had to mess with the sub-phase. When I used auto time alignment It was fine the subs measured slightly more away, after that No need for any phase adjustments.

Might be worth a shot.Record your info before trying.
I did try the auto-TA and it didn't seem to work for me.
 
161 - 174 of 174 Posts
Top