DiyMobileAudio.com Car Stereo Forum banner

441 - 460 of 954 Posts

·
Premium Member
Joined
·
4,514 Posts
Thanks Erin...
...
With that said, I would shape your midbass target for a little steeper incline between 100-200hz (not a lot) so that you can overcome road noise and keep the sub at its highest level while maintaining that cohesion between subs and highs. And create a target just for playing loudly (where he wanted more gain)
...

...
I keep getting drawn toward band pass boxes, and I think it conceptually along the same concepts...
 

·
Registered
Joined
·
515 Posts
The problem that I have with a lot of what I am reading on this thread is that there is no adherence to a process that is grounded in what objectively works. You can't EQ to a house curve and then change the levels of the drivers relative to each other without having a change in the acoustic crossovers. When you do that you are creating either overlap or underlap in the crossover region. That is objective fact. That doesn't mean that it won't sound good because it might. A great example would be that I tend to like a little cut that happens to be between where I crossover my mid and tweet. I could eq the the mid and tweet response to a flat line and then lower the level of the tweeter. What I would likely end up with would be a dip in the response between tweet and midrange because of the change to the acoustic crossover. Granted, if I only drop the level of the tweets a couple of db's and I am using LR24, there isn't a huge change in the crossover point. But you need to UNDERSTAND why it does what it does. My point is that level adjustment in any legitimate tuning process should happen prior to eq to a house curve, especially when using Dirac which is autotuning. It doesn't make sense to autotune and then change levels because when doing so you are not only changing levels, you are changing the acoustic crossover points between driver bands. I made my post only because I wanted to be sure that people understand that point. Again, it isn't just a level change, it is a crossover change as well and that impacts the response AFTER EQ.

In the prior example I agree 100% that you don't want to set your targets far below the average response. That is just throwing signal gain away for no reason. It makes much more sense to measure drivers in REW and get crossovers and levels set as close as possible to the target curve in REW. Then run Dirac to EQ everything. If you don't like the sound, start over with a difference house curve and do it again until you get the sound that you like. Bouncing around in the process will lead to less predictable results unless we fully understand what effects are going to happen when we make changes.
 

·
Registered
Joined
·
4,394 Posts
Discussion Starter #443 (Edited)
Edit

Bass knobs have no place in sq

We should tell anyone that uses one
To just throw it away because

1. We don’t want anyone to know we actually do use them

2. Don’t know how to use a bass knob

3. Shouldn’t have subs because it makes spectral balance off

4. Don’t ever ever ever ever ever ever ever deviate from my target
 

·
Registered
Joined
·
1,661 Posts
The problem that I have with a lot of what I am reading on this thread is that there is no adherence to a process that is grounded in what objectively works.
I agree... I think most of us are still trying to figure out that process to adhere too :)

BTW, have you had a chance to install yours yet & tinker with it?... Looking forward to your impressions too.
 

·
Registered
Joined
·
515 Posts
I get what you are saying Oabeieo. You are correct that you can do whatever floats your boat. And, I totally agree that in the past the amount of bass that I want sitting at a contest getting judged can be different than what I want driving down*the road. However, I would say that what I have learned recently is that if you get it right you can be happy with the same level setting going down the road as what you run in the lanes. Understanding what it takes to get it right is key.

*

My problem is that people that don’t have a lot of experience tuning will not be helped by statements of “do whatever you want”. There really needs to be some grounding in a process that is based on objectivity to get someone on the right*track and again, we have the freedom to tune to whatever house curve floats our boat. You can call that your artistic input towards creating something slightly different than what the recording artist and the producer intended if you like. But if that is*what you want, go back to mounting an EQ on your dash and have at it, LOL.

*

Just for the sake of folks who come across this thread that are learning to tune I want to illustrate with REW what I am talking about. I am not an expert, but I have learned quite a bit from people whom I believe are indeed expert. They*have engineering degrees and a lot of trophies to validate their process so I pay attention when they offer advice.*

*

This first slide is showing a black line that is the “JBL” house curve that Andy Wehmeyer has shared and that many people use as a tuning reference. I am not promoting this curve and I don’t personally use it although it works well. The*slide also shows the individual driver response curves that will blend to create that house curve in a 3 way plus subwoofer system with crossover points at 60, 200 and 2000. Those curves are imported text files that were created with a spreadsheet that was created*by DIYMA member Jazzi. They are a fantastic help with a solid tuning process. As you can see, the drivers overlap at 60 hz, 200 hz and 2000 hz. All crossovers are LR 24 db/octave.

*

The next slide shows what happens if I increase the output of the subwoofer by 10 db without changing anything else. As you can see, now the subwoofer output crosses with the midbass at 80 hz even though the electronic crossover on the*subwoofer remains set at 60hz. Now there is a much greater level of output from the subwoofer at 80hz and above. From a competition standpoint, if you have your sub mounted in the rear of the car it is much more likely to be localized to the rear, which*is not what you want. The way to fix it would be to set the lowpass for the subwoofer at around 40 or 45hz (not stopping to do the math but you get the idea). That would give you the increased output but still blend well with the midbass and have proper phase relatiinship and possibly not localize to the rear of the car. More importantly though,*is that the response is more predictable and phase relationship between drivers more consistent.

*

The next slide is showing the subwoofer back at the original level and now we have dropped the tweeter output by 4db. Now the midrange and tweeter response intersects at 2242 hz instead of 2000 hz. They will not sum flat since the crossovers*are not symmetrical. The crossover point has been changed and the phase relationship between drivers has changed.

I still don’t think it makes sense to make a level change that yields both a crossover and phase change as unintended consequence. Better to*set those levels prior to EQ and crossover settings.



Sent from my LM-V405 using Tapatalk
 

·
Registered
Joined
·
515 Posts
I agree... I think most of us are still trying to figure out that process to adhere too :)

BTW, have you had a chance to install yours yet & tinker with it?... Looking forward to your impressions too.
I've not installed mine yet. I was having all kinds of issues with the Helix Director wanting to load presets at random and the display not working correctly. I've given up on it and disconnected it. I have a couple more things I want to work on with the manual tune with the Helix before I let Dirac do its thing.
 
  • Like
Reactions: Truthunter

·
Registered
Joined
·
37 Posts
I am ready to order a MiniDSP C-DSP 8x12 DL. Has anyone purchased this model or the non-Dirac model since May 10, 2019? If so, I want to know if your order was subject to a 25% duty (tariff) and if so, was it the shipping carrier that was collecting that?
 

·
Registered
Joined
·
4,394 Posts
Discussion Starter #448 (Edited)
I get what you are saying Oabeieo. You are correct that you can do whatever floats your boat. And, I totally agree that in the past the amount of bass that I want sitting at a contest getting judged can be different than what I want driving down*the road. However, I would say that what I have learned recently is that if you get it right you can be happy with the same level setting going down the road as what you run in the lanes. Understanding what it takes to get it right is key.

*

My problem is that people that don’t have a lot of experience tuning will not be helped by statements of “do whatever you want”. There really needs to be some grounding in a process that is based on objectivity to get someone on the right*track and again, we have the freedom to tune to whatever house curve floats our boat. You can call that your artistic input towards creating something slightly different than what the recording artist and the producer intended if you like. But if that is*what you want, go back to mounting an EQ on your dash and have at it, LOL.

*

Just for the sake of folks who come across this thread that are learning to tune I want to illustrate with REW what I am talking about. I am not an expert, but I have learned quite a bit from people whom I believe are indeed expert. They*have engineering degrees and a lot of trophies to validate their process so I pay attention when they offer advice.*

*

This first slide is showing a black line that is the “JBL” house curve that Andy Wehmeyer has shared and that many people use as a tuning reference. I am not promoting this curve and I don’t personally use it although it works well. The*slide also shows the individual driver response curves that will blend to create that house curve in a 3 way plus subwoofer system with crossover points at 60, 200 and 2000. Those curves are imported text files that were created with a spreadsheet that was created*by DIYMA member Jazzi. They are a fantastic help with a solid tuning process. As you can see, the drivers overlap at 60 hz, 200 hz and 2000 hz. All crossovers are LR 24 db/octave.

*

The next slide shows what happens if I increase the output of the subwoofer by 10 db without changing anything else. As you can see, now the subwoofer output crosses with the midbass at 80 hz even though the electronic crossover on the*subwoofer remains set at 60hz. Now there is a much greater level of output from the subwoofer at 80hz and above. From a competition standpoint, if you have your sub mounted in the rear of the car it is much more likely to be localized to the rear, which*is not what you want. The way to fix it would be to set the lowpass for the subwoofer at around 40 or 45hz (not stopping to do the math but you get the idea). That would give you the increased output but still blend well with the midbass and have proper phase relatiinship and possibly not localize to the rear of the car. More importantly though,*is that the response is more predictable and phase relationship between drivers more consistent.

*

The next slide is showing the subwoofer back at the original level and now we have dropped the tweeter output by 4db. Now the midrange and tweeter response intersects at 2242 hz instead of 2000 hz. They will not sum flat since the crossovers*are not symmetrical. The crossover point has been changed and the phase relationship between drivers has changed.

I still don’t think it makes sense to make a level change that yields both a crossover and phase change as unintended consequence. Better to*set those levels prior to EQ and crossover settings.



Sent from my LM-V405 using Tapatalk

Okay I can agree with you. You make good sense and speak honestly and are articulate.


I really like that you don’t just arbitrarily say “oh it’s your sub” or some other “remark” that is aimed at shut down a good argument.


And like, some old zeppelin or maybe old bob segar or hank Williams, sorry +30db is perfect to my ears


I really appreciate the time you spent on this post. It’s well thought out


I like how you pointed out lowering the sub crossover point.

That’s a excellent point and definitely contributes to my argument.




Again I’m in no way saying everyone should just crank up there subs and rock out and forget SQ. I am saying there is a way to have both under certain conditions. That’s all , it’s a compromise indeed , but it’s worth it when you can’t hear your sub at all



And I think my point being , you’ll be surprised how well Dirac keeps it together if you fiddle with things enough to get the bass response you want under those conditions while still maintaining a good portion of that spectral balance

Geesh
Don’t everyone go install a epicenter at once now :D
 

·
Registered
Joined
·
515 Posts
That should really be the big advantage with this processor. It should be fairly easy to have an extra bass preset letting dirac do a lit if the time consuming heavy lifting for you. You should ne abke to have a +20 dvc bass preset with the best possible blend without a great deal if the eq work that would normally be required, right?

What I am hoping to get is the ability to play with different target curves and/or crossovers without having to do hours and hours of tuning with eq. If it does that well it really changes the time commitment.

If course it still takes sime knowledge to know your vehicle, system and specific challenges in your setup and not expect dirac to magically make all your troubles go away. But it would be very nice and worth the price if it cuts your tuning time in half.

Sent from my LM-V405 using Tapatalk
 

·
Registered
Joined
·
3,348 Posts
That should really be the big advantage with this processor. It should be fairly easy to have an extra bass preset letting dirac do a lit if the time consuming heavy lifting for you. You should ne abke to have a +20 dvc bass preset with the best possible blend without a great deal if the eq work that would normally be required, right?

What I am hoping to get is the ability to play with different target curves and/or crossovers without having to do hours and hours of tuning with eq. If it does that well it really changes the time commitment.

If course it still takes sime knowledge to know your vehicle, system and specific challenges in your setup and not expect dirac to magically make all your troubles go away. But it would be very nice and worth the price if it cuts your tuning time in half.

Sent from my LM-V405 using Tapatalk
It cuts my tuning time in like.. 1/5th if not more.

Its seriously some impressive tech.

Sent from my XT1710-02 using Tapatalk
 

·
Registered
Joined
·
3,348 Posts
Did a tune tonight. Brought my acoustic crossovers closer to my imported curve is so its not correcting as much.

Put the center mic position dead on where my head is and just kinda..cocked my head side ways to do it. Ended up having to give my left drivers anywhere from .5 & .7ms of delay but it all dialed in nicely.

Im not sure how i feel about measuring the whole cabin(couch) when it comes to 500hz or less. My passenger side doesn't have some of the larger dips that my driver side does and i think its skewing some of the results.

I left my sub above 0db(most of it) but made the curve follow its response perfectly as i could.

In the Dirac tab of the plug in. Its showing my sub channel gain is down -16 but rew shows differently. It shows everything is playing the way i tuned it. Kinda strange.

Sent from my XT1710-02 using Tapatalk
 

·
Premium Member
Joined
·
4,514 Posts
The problem that I have with a lot of what I am reading on this thread is that there is no adherence to a process that is grounded in what objectively works. You can't EQ to a house curve and then change the levels of the drivers relative to each other without having a change in the acoustic crossovers. When you do that you are creating either overlap or underlap in the crossover region. That is objective fact...
...
Except if the band pass filter go to a brick wall, like one can do with a FIR... then there's is nor cancellation in the overlap, as the overlap goes to zero.
 

·
Registered
Joined
·
4,394 Posts
Discussion Starter #455
That should really be the big advantage with this processor. It should be fairly easy to have an extra bass preset letting dirac do a lit if the time consuming heavy lifting for you. You should ne abke to have a +20 dvc bass preset with the best possible blend without a great deal if the eq work that would normally be required, right?

What I am hoping to get is the ability to play with different target curves and/or crossovers without having to do hours and hours of tuning with eq. If it does that well it really changes the time commitment.

If course it still takes sime knowledge to know your vehicle, system and specific challenges in your setup and not expect dirac to magically make all your troubles go away. But it would be very nice and worth the price if it cuts your tuning time in half.

Sent from my LM-V405 using Tapatalk

You know what’s wierd , is it didn’t make some of what I thought were problems go away, it fixed problems I didn’t realize were problems, afterwards listening I bought a clue what it was doing.
The problem I had before I was eventually able to solve it compermise but with other problems fixed those other problems weren’t as bad and were manageable

Especially in the suband midbass goodness yes
 

·
Registered
Joined
·
4,394 Posts
Discussion Starter #456 (Edited)
So , to be clear so no body’s confused

Like rockinridgine said he dosent like that there no process,

Well there is , and I think what I’ve been advocating is just that, but also mixing up my super sub ideas with it.



So, in regards to that, (process) what everyone is doing is ok
That’s the point , define a target and use it.

What I was saying was a way to help get your gain structure set to compliment your targets
Instead of massive cuts by using the target.

I have advocated staying under the avg spectrum line in the normalized response
However not way way way below it.

In my experience with Dirac I have found the way the eq responds against the actual response works best if you set your gains to compliment the target (like you would any eq)

However it’s a little diffrent critter we all know by now

So by drawing the shape of the response (not doing anything with the target in regards to the crossover) so basically drawing a straight (ish) line than after go to rew and make the response you want (like the stair-step) by using gin, it dosent have to be perfect just get in the ball park.

Than next go back and re measure in Dirac and the avg spectrum will be much closer to your desired response , thus using much less eq , and achieving a much more repeatable target and the target will reflect in REW much more faithful to your target.

That’s all i was saying. A way to get your gains to compliment the way Dirac will create your target because you set your gains with a Dirac correction active.

How it measures and applies the correction is a bit different gang. I have found doing this allows me to dump just about any target I want and the crossovers are where I want them, the response matches better in REW.

The only thing is if you have massively different targets you may want tiny bit of gain on each of those configurations , but I would be willing to bet that you use targets that are very similar.


I don’t think we’re wanting to cut so much gain that there’s no volume,
Just getting the gains a little closer to the target


By using a Dirac correction to set with REW it will be much much easier to see what’s what,
Maybe the auto target , set gains semi close to desired target , than run it with your targets, I be willing to bet your before is much closer to what you want and a lot less eq, again more repeatable , etc


I’m not saying the rew part has to be dead on balls , just get a little closer in the ballpark than 10db and get the crossovers to be where you want them acoustically before using the target to cutoff entire octaves.

I don’t see how that is not good practices
 

·
Registered
Joined
·
2,581 Posts
FU forum.... timed out while writing a post so now I have to write it again.....

Can anyone give me pointers on setting the initial output in DLCT? I reset my unit as it was having a weird issue with the tweeters (setting them at -40dB no matter what). Started putting a new tune together and I cannot for the life of me get a solid output level.

I first used REW, set my levels and slope to look roughly what I wanted. Tweeters about -5dB down from the mids. Open DLCT, check the outputs, take the first measurement and it's all good. Move to the 2nd measurement, it tells me the level is too low (looks like both tweeters). Go back to the output tab, increase the output and take the 2nd measurement.

Get to the 3rd measurement. Front lower right. Here the subwoofer level clips, despite it being nowhere near clipping on the previous 2 measurements.

Back to the output tab, lower it a little, take the measurement again.

Onto the 4th measurement. Front lower left.

The fucking right tweeter is now reading as too low.

Back to the output tab. Increase the level a little.

Now the damn left mid/midbass are clipping.

Smash my head into the steering wheel multiple times.

Give up.

I don't know if I am either doing something completely wrong in setting the levels, I don't know what, I had a pretty good manual tune before getting the DL unit, so it's not like I have no idea what I am doing. Or, my car has stupidly annoying reflections causing the sweeps Dirac makes to be wildly different depending on the microphone location. Either way, it is driving me mental.

Anyone have any suggestions?
 

·
Registered
Joined
·
3,348 Posts
FU forum.... timed out while writing a post so now I have to write it again.....



Can anyone give me pointers on setting the initial output in DLCT? I reset my unit as it was having a weird issue with the tweeters (setting them at -40dB no matter what). Started putting a new tune together and I cannot for the life of me get a solid output level.



I first used REW, set my levels and slope to look roughly what I wanted. Tweeters about -5dB down from the mids. Open DLCT, check the outputs, take the first measurement and it's all good. Move to the 2nd measurement, it tells me the level is too low (looks like both tweeters). Go back to the output tab, increase the output and take the 2nd measurement.



Get to the 3rd measurement. Front lower right. Here the subwoofer level clips, despite it being nowhere near clipping on the previous 2 measurements.



Back to the output tab, lower it a little, take the measurement again.



Onto the 4th measurement. Front lower left.



The fucking right tweeter is now reading as too low.



Back to the output tab. Increase the level a little.



Now the damn left mid/midbass are clipping.



Smash my head into the steering wheel multiple times.



Give up.



I don't know if I am either doing something completely wrong in setting the levels, I don't know what, I had a pretty good manual tune before getting the DL unit, so it's not like I have no idea what I am doing. Or, my car has stupidly annoying reflections causing the sweeps Dirac makes to be wildly different depending on the microphone location. Either way, it is driving me mental.



Anyone have any suggestions?
I set my dlct to -30 and then mic sensitivity to 0. I use the plug in and not dlct to set actual levels because dirac locks levels of drivers if you do it via dlct. Bring levels linearly back up after your done doing dlct. Like if your MB is at -8, your mr is -5 a d your tw are at -3. Bring everything up by 3 db so you're back at a 0db output.

I haven't scoped this dsp to see if it puts out a clean signal above 0db.

As for clipping in certain measurements and not others. The "window" is only a few db wide from what i can tell so if one spot is peaky, it'll clip it. Generally its anywhere really close to boundaries, windows, doors, ext. I found moving the mic around to a slightly different spot usually fixes it and if not, do a sweep in Rew of the driver that clips alot and find out where and bring it down via peq. I had to bring down 400hz by 6db and i know Oab had to do some pre peq work with his horns(im not running horns yet) as well.

When you're done and saving a dlct setting to the box(export menu), bring the output level from -30 back up to 0.

That almost has enough output for me but my gains are still way lower than they need to be.

You probably know all that but just throwing out what i can.

Sent from my XT1710-02 using Tapatalk
 

·
Registered
Joined
·
2,581 Posts
I set my dlct to -30 and then mic sensitivity to 0. I use the plug in and not dlct to set actual levels because dirac locks levels of drivers if you do it via dlct. Bring levels linearly back up after your done doing dlct. Like if your MB is at -8, your mr is -5 a d your tw are at -3. Bring everything up by 3 db so you're back at a 0db output.

I haven't scoped this dsp to see if it puts out a clean signal above 0db.

As for clipping in certain measurements and not others. The "window" is only a few db wide from what i can tell so if one spot is peaky, it'll clip it. Generally its anywhere really close to boundaries, windows, doors, ext. I found moving the mic around to a slightly different spot usually fixes it and if not, do a sweep in Rew of the driver that clips alot and find out where and bring it down via peq. I had to bring down 400hz by 6db and i know Oab had to do some pre peq work with his horns(im not running horns yet) as well.

When you're done and saving a dlct setting to the box(export menu), bring the output level from -30 back up to 0.

That almost has enough output for me but my gains are still way lower than they need to be.

You probably know all that but just throwing out what i can.
Thanks, appreciate the tips. I went back out for another try, I did essentially what you are saying and changed the levels in the plug in, then would go back to DLCT, check the output. Move onto the next speaker, repeat for each until they were all hitting around the same spot on the graph.

Just as an example, this first picture is the "sweet spot":



This second is from the lower left rear position:



The tweeter levels and subwoofer are so much lower in this position, annoyingly when I then did the upper rear measurements the rear speakers were clipping slightly. Had to lower the output a little and try again.

I had turned the microphone gain up a little for these, but will set it back to zero again and instead just increase the overall output slider. But, at least for now I have semi decent music playing. The sub level is lower than I would like now, but I just used the auto target in Dirac in order to get somewhere. I'll definitely look into using PEQ if needed to cut certain frequencies before Dirac if needed.

At least the tweeters are no longer -40dB!!!
 
441 - 460 of 954 Posts
Top