DIYMobileAudio.com Car Stereo Forum banner
1 - 7 of 7 Posts

· Super Moderator
2021 GMC Sierra AT4
Joined
·
1,072 Posts
Discussion Starter · #1 ·
Thank you in advance for any help here!

Set up will be:

2021 GMC Sierra Sierra AT4
Helix P6 Amp for GB60/15 Active (Front), GS62 (Rear)
JL HD750.1 for 2 x JL 12TW3 (Stealthbox)
UMIK-1 Mic (Also have a IMM-6)

Started to prep for my upcoming build and first time tuning. I had some questions specific to setting up channel gain and output levels before I use the both the Helix RTA/Auto EQ (after reading on CAJ) and/or REW. Iin preparation of the tuning process, I have been playing with the PC Tool in demo mode for a week or so and absorbing what I can from Youtube tutorials for Helix and REW.

I plan to use the AF House Curve and will set my sub amp gain with an O Scope at 40hz (then double check with a multimeter vs JL's recommended voltage, they have usually been damn close in the past) before I tune.

Question #1 - Left / Right Gain Balance
Can/Should I do this quick and dirty using an IMM-6 (or REW w/ UMIK-1) to get SPL within a couple db of each other? Or set them to equal gain/output and let the tuning process (whether AutoEQ or REW) vette this out?

Question #2 - Subwoofer vs Door Speakers/Tweeter Gain
I generally see that most house curves have sub 100 Hz +10db (and tapering off) for the most part and wondering how folks adjust this before tuning? Or if they adjust this at all?

In my brain I am thinking that my door speakers and tweeters should all be set -10db (or close) in the Helix software before I tune but looking for advice. Is this pointless before tuning?

Question #3 - Tuning Order
I have seen a few different strategies using the HelixRTA / AutoEQ feature. My strategy right now is:
  1. Right Front (GB60 & GB15), No Sub
  2. Right Rear (GS62), No Sub
  3. Right Front and Rear, No Sub
  4. Left Front (GB60 & GB15), No Sub
  5. Left Rear (GS62), No Sub
  6. Left Front and Rear, No Sub
  7. Sub Only
  8. All Channels
How does this look? Any advantages to tuning the left side vs right side first? I know most folks tune each driver individually but I have been reading very promising things about the AutoEQ in the Helix RTA and most people tend to group drivers there. I am planning to tweak specifically in REW after the fact but I know I will be pressed for time when I first get my rig set up and won't have the hours to spare to tune fully in REW as a noob.

Hopefully @SkizeR takes a look at this :) Seems to the local Helix guru here on DIYMA and CAJ.

Thanks to anyone that can help!
 

· Registered
Joined
·
4,486 Posts
Don’t forget to drop the sub 6db when tuning vs left or right, so when it’s tuned with left and right (which should sum +6db at lower freqs ideally) the level will stay relatively correct and balanced still to the house curve

as to left vs right levels, normally the drivers side ends up somewhere between 3-6db down on the passenger side depending on pathlength difference between the two

the reason for tuning separately is so that you can make sure the crossovers are at correct freqs and move them if need be so you can find good clean smooth response which indicates good phase also and means the crossovers will work as they should
 

· Banned
Joined
·
24,246 Posts
Question 1) Don't waste your time doing it quick and dirty just to "clean up" on top of it. This will just cause you to use more filters for no reason and waste time

Question 2a) I'm not sure I fully understand this question, but there's nothing to adjust yet.
2b) No, don't do this. Just leave them and take measurements and see what is going on first.

Question 3) If using Helix Auto EQ, this looks good. If using REW, do individual speakers at a time. But again, just tune one way or the other. Otherwise, you are adding unnecessary steps and clutter to your tune and its process.
 

· Super Moderator
2021 GMC Sierra AT4
Joined
·
1,072 Posts
Discussion Starter · #5 ·
Question 1) Don't waste your time doing it quick and dirty just to "clean up" on top of it. This will just cause you to use more filters for no reason and waste time

Question 2a) I'm not sure I fully understand this question, but there's nothing to adjust yet.
2b) No, don't do this. Just leave them and take measurements and see what is going on first.

Question 3) If using Helix Auto EQ, this looks good. If using REW, do individual speakers at a time. But again, just tune one way or the other. Otherwise, you are adding unnecessary steps and clutter to your tune and its process.
Yup, this all makes sense to me, very much appreciate the response!
 

· Registered
Joined
·
1,022 Posts
1. Get levels as close to your house curve as posible using only channel gain & output level on main screen.
2. Same answer as #1
3. for auto EQ tune entire left side then right side. If only one sub then do as dumdum said raise sub 6db
 
  • Like
Reactions: RickWilson

· Super Moderator
2021 GMC Sierra AT4
Joined
·
1,072 Posts
Discussion Starter · #7 ·
Found a reply from Andy (on the Audiofrog FB Group) that helped clear up some advice you gave above. Posting because hopefully it helps someone else and it's a quick reference for me.

Regarding gains vs. noise:

"Setting gains is primarily a matter of maximizing system output balanced against noise and distortion.
So, we want the system to be as loud as is necessary without any noise and without unnecessary distortion.

In an old analog system with a head unit, an EQ, several crossovers and amps, this was a matter of making sure that the inputs and outputs of each of those processors in the chain were set correctly. So, we wanted to make sure that no inputs were overdriven and no outputs were set so high that they'd make noise.

In a system with a DSP, we don't have a chain of inputs and outputs to manage, really. We have one device that will do all of the processing and it will drive the amplifiers.

The notion that we would use amplifier gains to prevent the speakers from being overpowered came later, but it's mostly a ruse--and that should have been exposed in reading that link about tweeters and power.

In the DSP, there is an analog preamp which is used to set the analog input level so that it doesn't provide too much voltage to the A/D convertor. That maximum input voltage may be 1V, may be 2V or 2.8V or whatever. No matter how much voltage exists on the RCA leads, that will have to be attenuated in order not to overdrive the A/D because if there's distortion there, it will be converted in to the digital signal and then it cannot be removed using the level controls inside the DSP.

So, for this part, we aren't as concerned about maximizing **** as we should be about NEVER allowing this to clip. So, we want to set the input sensitivity in the DSP to prevent clipping the A/D.

If you're using an aftermarket radio with 4V preamp outs, for example, this is pretty simple. No need for measuring and sine waves and all that. Just set the input sensitivity for 4V or a little higher--maybe 4.5V. There. Done. No clipping.

If you're using the speaker level outputs of a factory radio and it doesn't have some kind of boost rail for more output than can be made with a standard amp IC, then the maximum output voltage is about 8.5V. So, you can set this at about 9V and you'll be safe.

If you need to be more precise, then you need to use pink noise played by the radio. You can use the input RTA or some other RTA to measure the frequency response of the head unit. Then, to set the input gain appropriately, you'd generate a sine wave at the frequency with the highest output and use that to set the gain. You could do that by watching the clip light in the software or by measuring that output with a VOM and setting the gain using the numbers.
Using pink noise and a clip light is not the right process, no matter what anyone else says. With pink noise from 20-20k, 20Hz will be the loudest (highest voltage) frequency and the voltage will be gradually reduced as frequency increases. Do not set the gain using a VOM and pink noise.

So, once we set the input sensitivity to prevent clipping, then we have to move on to the rest of the system.
Once we choose a target curve, there will be one combination of digital and analog level that will hit the target and provide the lowest noise and distortion. The bass will be the loudest, followed by the midbass, then the midrange and then the tweeters. Chances are that you have about 10 times the amplifier power you'll need for a midrange in a 3-way and about 100 times as much power as you need for tweeters. You may be able to use most of the amp for the midbass.

So this process of making sure our tweeter amplifier can make 150 watts without distortion is a waste of time because we'll never ever need it to do that. If we set the gain so the amplifier can make that much power and our DSP crashes or we input an incorrect setting, then we've allowed that condition to destroy the speaker for no good reason. We could have used the amplifier to prevent some of that.

So, since there's ONE solution set here that hits the target and minimizes noise and distortion, some of this ridiculousness can be avoided.

Now, we've set the input sensitivity of the DSP. Next, we can reduce the output level of the DSP to give us some room to work. We don't want to clip the digital signal in the DSP either. If we start with all of the channel levels at 0dB (maximum) then we don't have any room to boost at low frequencies, but, depending on the program material, we might have a lot of room at higher frequencies.

But, if we clip the low frequencies in the digital signal inside the DSP, we'll hear that clipping as nasty sounding midrange when there's a bass note that causes the clipping.

So, we reduce the level of the digital signal to give us room to work. Start with -6dB or -12dB. The system won't be loud, but for setting the EQ, that doesn't matter.


Next, put in the pink noise and set the amplifier gains to get close to the target. Now, we've turned down the power that we don't need and we've balanced the amplifier gains with one another. Turning down the tweeter amp reduces hiss, so we've addressed noise at the point most likely to generate that noise.

Then, we EQ. This is just to set the frequency response so it hits the target. Once we're there, then we set the level of the whole system.

We can set that level first by increasing the output levels of the DSP channels by turning them all up by the same amount while listening for distortion and noise. We need the system to play loudly enough to satisfy us and no louder. The more level we add, the more noise we add.

So, turn those levels up until it's loud enough and there's no annoying distortion and noise. If you get to all the way up and it still isn't loud enough, then you can move to the amplifier gains. Watch the RTA and increase the amplifier gains so that the curve stays the same but the level of the whole system is increased. At the point at which you hear distortion, that's it. No more. At the point at which the noise is annoying, that's it. No more.

Under some conditions and EQ usage, you may max out the digital signal but have more volume left if you stop at -3 dB, for example, on the DSP levels. Then, with no distortion there, you can increase the amplifier gains.

This takes practice, but ultimately it's a much better process of maximizing signal and reducing noise than some arbitrary process of using sine waves to do this in a system in which we have up to 100x the power we need.
Much of this can be further complicated by user interfaces that don't provide the documentation and the tools that are required or by processes suggested by people who are primarily concerned with maximizing bass for SPL competitions.

The point of an audio system isn't to maximize power delivery to every set of speakers in a multi-way system. It's to maximize system output once the "sound" of that system has been optimized."

Credit: Andy Wehmeyer

Felt like this had some really valuable logic when it comes to tuning, noob or not.
 
1 - 7 of 7 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top