DiyMobileAudio.com Car Stereo Forum banner

1 - 18 of 18 Posts

·
Registered
Joined
·
19 Posts
I think it’s best to set TA before you EQ, because if you do EQ first then you could be adjusting the EQ to address issues that were introduced because your time alignment is off. Then, when you set the TA, you’d have to redo your EQ (because changing the TA will definitely mess with the center image you just worked so hard to achieve with the EQ)


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
7,852 Posts
It's hard to argue with Mauian's logic.

TA is an absolute, it will be the same regardless of your crossovers, or EQ. By setting it first you know that you've done your best to get the arrival times from each speaker as close as possible, and keep the speakers in phase with each other. Then, you can EQ the left and right sides knowing that the dips and valleys in the response that you see on the RTA are not phase related.

If you set EQ first, you may find yourself compensating for frequency differences that are simply caused by the speakers being out of phase, since TA hasn't been set to put them in phase. If you then set the TA, which will eliminate the destructive interference, then you will still have to re-EQ in order to un-do some of the EQ you already did but was simply phase related in the first place.

So, even thought the same amount of TA will be used regardless of what you do first, setting it first prevents you from needing to EQ twice.
 

·
Registered
Joined
·
7,852 Posts
So would the rule of thumb be?
1- Gain matching
2- Time calibration
3- Equalization
Yes, you should definetly "rough in" your gains first. Before doing any further tuning you should have a relatively decent balance between your subs, mids, and tweets. It should sound mostly balanced. But, you're just roughing the gains in.

So, the steps would be:

  • Rough in the gains by ear; subs, mids, and tweets should sound fairly well balanced, without one set of speakers standing out among the rest. Your ears are all you need at this point, so if your tweeters are obviously too bright, turn their gain down. If the sub is too loud, turn it down. You don't need to pull out any tools (except a flat head to turn your gain knobs), just listen to a few songs, and rough in the gains to a reasonably listen-to-able level.
  • Set time alignment. Measure carefully with a measuring tape, and you're done.
  • Take some RTA measurements, and fine tune the gain matching to match your final target curve. All you're doing here is fine tuning the rough gain match you did earlier, you're just getting the relative output of each set of drivers to be close to the target curve (overall) as possible. You'll have peaks and valleys, but adjust the gain so that you get each set of speakers (subs, mids tweets) to line up with your curve.
  • EQ. Cut and boost as needed to match your target curve.
There are some great sources out there for more specifics related to each of these steps, but this is a simplified outline.
 

·
Registered
Joined
·
591 Posts
I’ll await the posts saying my centre isn’t centre’d and I did time alignment by tape... the fact is time alignment is a set number... what people miss is that the install is critical in getting individual speakers to be perceived to be coming from the actual speakers location and not 6 or 10” inside the pillar across the screen, that is why the centre is diffused and not coming from between the speakers in lots of cases (normally the passenger side driver that causes issues, but the drivers side can also)

adjusting levels is a bit of a cure, however this can then lead to bunching of the off centre positions and left or right, to get tracks like the IASCA drums correct it counts on very careful aiming and adjustments of drivers...

and this goes against the audiofrog and it’s very good guide to a degree, it states as long as a speaker is below beaming it doesn’t matter if it’s on or off axis... this is accurate for its frequency response, however it doesn’t take into account reflections and where the speaker actually sounds like it’s coming from, this is why andy also recommends upper door panel mounting (everyone misses the reasons for this and that’s where the guide doesn’t get followed as andy intended!) as it reduces the reflection issue and the speaker to the brain appears to come from where it’s mounted to a greater degree than for example pillars (I see lots of installs with high dollar kit on Facebook... and regularly have conversations with guys at the top of the rung about them... and we can all see massive install issues purely so a build is easy to trim, Ive seen lots on here also, but it’s not my place to say anything)

I will at some point be putting a rather deep and meaningful document together about what I’ve learned over the years about sq and what I’ve gleaned from guys who do make rather excellent sounding cars from the U.K. and the world 👍🏼
 

·
Registered
Joined
·
19 Posts
I’ll await the posts saying my centre isn’t centre’d and I did time alignment by tape... the fact is time alignment is a set number... what people miss is that the install is critical in getting individual speakers to be perceived to be coming from the actual speakers location and not 6 or 10” inside the pillar across the screen, that is why the centre is diffused and not coming from between the speakers in lots of cases (normally the passenger side driver that causes issues, but the drivers side can also)

adjusting levels is a bit of a cure, however this can then lead to bunching of the off centre positions and left or right, to get tracks like the IASCA drums correct it counts on very careful aiming and adjustments of drivers...

and this goes against the audiofrog and it’s very good guide to a degree, it states as long as a speaker is below beaming it doesn’t matter if it’s on or off axis... this is accurate for its frequency response, however it doesn’t take into account reflections and where the speaker actually sounds like it’s coming from, this is why andy also recommends upper door panel mounting (everyone misses the reasons for this and that’s where the guide doesn’t get followed as andy intended!) as it reduces the reflection issue and the speaker to the brain appears to come from where it’s mounted to a greater degree than for example pillars (I see lots of installs with high dollar kit on Facebook... and regularly have conversations with guys at the top of the rung about them... and we can all see massive install issues purely so a build is easy to trim, Ive seen lots on here also, but it’s not my place to say anything)

I will at some point be putting a rather deep and meaningful document together about what I’ve learned over the years about sq and what I’ve gleaned from guys who do make rather excellent sounding cars from the U.K. and the world
Awesome! Looking forward to the guide. Sounds like it will have some great stuff in it.


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
7,852 Posts
thank you all, i was just confirming my methods were right... had a local shop tell me i should T.A. last which didn't make sense to me...
It's fine, but not ideal. It adds extra steps, since TA will affect the frequency response, so you may end up needed more EQ work after setting TA.
 

·
Premium Member
Joined
·
3,672 Posts
It's hard to argue with Mauian's logic.
...
That won't stop me from trying...

If one EQs each speaker independently, then it is unclear whether they need attention after they are TA and playing together.
It is possible, but they should be pretty close ignoring the fact that the reflections may be what is the loudest signal.
 

·
Registered
Joined
·
491 Posts
That won't stop me from trying...

If one EQs each speaker independently, then it is unclear whether they need attention after they are TA and playing together.
It is possible, but they should be pretty close ignoring the fact that the reflections may be what is the loudest signal.
As I recently learned, just because you EQ each speaker individually, that doesn't mean that you won't need to do more EQ later after analyzing the response when multiple speakers are playing together. Playing speakers together (L+R, whole-side, etc) causes phase issues that simply don't exists when playing speakers individually - and you really shouldn't analyze multiple-speaker responses until after TA is set.

So yes, the individual speaker EQ will be the same regardless of whether it's before or after TA, but since you still need to do more EQ work after taking measurements when more than one speaker is playing, you might as well just set TA first and then do all EQ work afterwards - there is really no benefit to doing any EQ work before TA - but there is a benefit in doing EQ work after TA is set and multiple speakers are playing. :)

To me, it just makes sense to set TA at the beginning so it's done and out of the way. The speaker distances aren't going to change - and since TA is the direct result of speaker distances, it shouldn't change either.

At least that is my understanding (obviously, I'm still learning though).
 

·
Registered
Joined
·
19 Posts
As I recently learned, just because you EQ each speaker individually, that doesn't mean that you won't need to do more EQ later after analyzing the response when multiple speakers are playing together. Playing speakers together (L+R, whole-side, etc) causes phase issues that simply don't exists when playing speakers individually - and you really shouldn't analyze multiple-speaker responses until after TA is set.

So yes, the individual speaker EQ will be the same regardless of whether it's before or after TA, but since you still need to do more EQ work after taking measurements when more than one speaker is playing, you might as well just set TA first and then do all EQ work afterwards - there is really no benefit to doing any EQ work before TA - but there is a benefit in doing EQ work after TA is set and multiple speakers are playing. :)

To me, it just makes sense to set TA at the beginning so it's done and out of the way. The speaker distances aren't going to change - and since TA is the direct result of speaker distances, it shouldn't change either.

At least that is my understanding (obviously, I'm still learning though).
Agreed. Thanks for explaining that in more depth.


Sent from my iPhone using Tapatalk
 

·
Premium Member
Joined
·
3,672 Posts
As I recently learned, just because you EQ each speaker individually, that doesn't mean that you won't need to do more EQ later after analyzing the response when multiple speakers are playing together. Playing speakers together (L+R, whole-side, etc) causes phase issues that simply don't exists when playing speakers individually - and you really shouldn't analyze multiple-speaker responses until after TA is set.

So yes, the individual speaker EQ will be the same regardless of whether it's before or after TA, but since you still need to do more EQ work after taking measurements when more than one speaker is playing, you might as well just set TA first and then do all EQ work afterwards - there is really no benefit to doing any EQ work before TA - but there is a benefit in doing EQ work after TA is set and multiple speakers are playing. :)

To me, it just makes sense to set TA at the beginning so it's done and out of the way. The speaker distances aren't going to change - and since TA is the direct result of speaker distances, it shouldn't change either.

At least that is my understanding (obviously, I'm still learning though).
Well the options are:
1)
EQ individually
TA
EQ together

Or 2)
TA
And then do we EQ individually or just go straight to EQ together?

If it is individually first, then it doesn't matter if it is TA/EQ or EQ/TA...
If it is EQ just once, then it has to be TA/EQ.

If it is Dirac or other some phase (FIR) EQ, then it should be individually ... so that can be before or after TA.
But there is still an opportunity for chin scratching when playing the ensemble together.
 

·
Registered
Joined
·
19 Posts
Ah, those are good options. I always assumed the question was about should the OP EQ everything (individually and together) before or after TA. But yes, if you save the EQ for pairs of speakers and speakers altogether after TA then that works too.


Sent from my iPhone using Tapatalk
 

·
Registered
Joined
·
491 Posts
Well the options are:
1)
EQ individually
TA
EQ together

Or 2)
TA
And then do we EQ individually or just go straight to EQ together?

If it is individually first, then it doesn't matter if it is TA/EQ or EQ/TA...
If it is EQ just once, then it has to be TA/EQ.

If it is Dirac or other some phase (FIR) EQ, then it should be individually ... so that can be before or after TA.
But there is still an opportunity for chin scratching when playing the ensemble together.
You need to EQ both individually and together. My point was simply that from a "procedure" standpoint, it just makes more sense (to me at least) to have a step to do TA and a step to do EQ (instead of EQ individual, TA, EQ together, ...). To me, It just makes more sense from a procedural standpoint to do all of the EQ work in the same step instead of putting the TA in the middle of the different EQ processes...

It doesn't really matter if you do TA or individual speaker EQ first, but I would rather just group all of the EQ work together from a process point of view. To me, it just complicates things to put TA in the middle of the EQ steps - especially since there is no benefit of doing it that way - just creates more "steps" in the overall process flow. :)

More of a preference, I guess.
 

·
Registered
Joined
·
7,852 Posts
Jtrosky pretty much summed up my approach. You can EQ individual speakers, first, but you'll still want to go back and check them once you turn the other speakers back on. If you EQ your left mid, and left tweeter individually, then TA, you still need to look at what's going on around the crossover point with both the mid and tweet playing.

Either approach can get you to the same place, but I'd prefer to set TA right away, then move on and not have to question it. Then I can focus strictly on the crossovers, and EQ.
 

·
Premium Member
Joined
·
472 Posts
Discussion Starter #17
glad to know i was doing things somewhat rite.. i set my cross overs. then T A then i check each speakers levels and come up with the average and match them... then match the level from left and right.... then eq each separate, then the tweet and mid-range combo, then mid-bass. then do the other side... then eq all together..
 

·
Premium Member
Joined
·
3,672 Posts
Jtrosky pretty much summed up my approach. You can EQ individual speakers, first, but you'll still want to go back and check them once you turn the other speakers back on. If you EQ your left mid, and left tweeter individually, then TA, you still need to look at what's going on around the crossover point with both the mid and tweet playing.

Either approach can get you to the same place, but I'd prefer to set TA right away, then move on and not have to question it. Then I can focus strictly on the crossovers, and EQ.
This sort of implies something...
If they were EQ'ed individually, and then TA'ed then one starts with a solid fundation.

When they are then combined and there is a null, that extra step up front can become an ah-ha moment where one can know that something like crossover slope or phase is happening. One could EQ it out, but it becomes, in mind, more clear that maybe something with the phase would be a better approach to the cancellation rather than overpowering the band.

But I am no expert in it. So take it as a thought or question.
 
1 - 18 of 18 Posts
Top