DSP tuning for below dummy level

[ThxOne Premium Member]

So all of this time I've been flattening out the hip hop and EDM music that I listen to? Sounds about right.

The V boost bass and highs and cuts mids... in reality, in your ears the graph is more flat than anything else at that point.

 

[Original wis]

The V boost bass and highs and cuts mids... in reality, in your ears the graph is more flat than anything else at that point.

Gotcha. And now instead of a head full of questions I have a head full of ideas. Tomorrow after work I'm going to tape measure the heck outta the car cabin.

Especially because my tweeters are in the sail panels, the mid-ranges are in the kick panels, and the mid bass are in the bottom of the door. From what I'm understanding from this discussion I'm going to need to do a lot of tuning so that the mid-range frequencies are brought up in the car without being muddied by the mid bass.

 

[ThxOne Premium Member]

Gotcha. And now instead of a head full of questions I have a head full of ideas. Tomorrow after work I'm going to tape measure the heck outta the car cabin.

Tape, string and a balloon. Hang the balloon where your head settles in the car normally while you drive. measure to the balloon.

ALSO!!! Measure the sound while you are in the car. The EQ will change as your body will reflect and absorb frequencies.

 

[Original wis]

Tape, string and a balloon. Hang the balloon where your head settles in the car normally while you drive. measure to the balloon.

ALSO!!! Measure the sound while you are in the car. The EQ will change as your body will reflect and absorb frequencies.

Thanks again once I get it figured out I'll update my build log and let you guys know how it turned out. You guys are pretty awesome.

 

[El Wapo]

I took a couple screen shots from the app and I will attempt to decribe each shot in detail and explain what they mean. This first shot is the first part of the setup. Each speaker is assigned a channel. This channel will be used to adjust that speaker and that one only.
View attachment 41804

Crossover adjustment is going to take a range of freq. that you can use to adj. where the speaker will play in the frequency spectrum between 20 hz- and 20khz. This is the human baseline for hearing. If it sas HPF (high pass filter) it means that anything lower than that will not play through the driver. If it is LPF (low pass filter) it is the exact opposite. Nothing higher than that freq. will be allowed to play.
If you use a bandpass crossover it will take a small frequency range and it will have a filter on both ends.
The "filters" we use are dictated by slope. We can adjust them in 6, 12, 18, and 24 db per octave So let's take the two octaves of 20 and 50 hz. If our crossover is using a 6db slope the roll off or volume drop playing between the two is very gradual. If it is 24 db the dropoff is much sharper.
View attachment 41805

Equalization. This can really make a system shine. It mentions that the dsp has the ability to adjust 31 bands which means that we can adjust each octave in the freq. spectrum. 31. Human hearing is centered around 2-3khz. One of the links I showed you has a list of where instruments play to give you a better idea. When we adjust an eq we always want to cut frequencies and never add.

View attachment 41806

Delay/Time alignment. We use this to set when the sound of the speaker arrives to us. This is useful when you hav one midrange that is closer to you than the other one. We used to solve this problem back in the day by putting midrange and tweeters in the kicks.
Take a ruler and measure the distance between the drivers and passenger footwell. It may be less than 6 inches. Now measure from your drivers side A pillar and the passenger A pillar. See the difference. This is why we use time alignment. With that even with the wild difference in distance we can have the sound arrive at our ears at the same time.

Thank you for explaining this for us!
I was lost on all this and now have at least a starting place of knowledge. I was thinking I might not use a dsp in my next build due to difficulty of tuning.

 

[winkychevelle 10+ year member]

I haven't read all the post here but here is my process.

I use a measurement mic to see the response curve of each individual speaker. And adjust the individual eq for each speaker 1 at a time to get as flat of a response as possible from each individual driver.

I then use the mic to adjust each speakers amplitude or volume level so they appear the same volume at the the point of which you are centering the sound stage, drivers seat or center of vehicle.

Then I go through the process of adjusting the time alignment so that the sounds both tweeter and mid of each front door reach my ear at the same time. This helps adjust the height of your sound stage when you have mids in the doors and tweets on the sails, dash, or a pillars. After that I adjust for left and right. You can set your sound stage position several ways I prefer the drivers seat since I'm most commonly the only person in my truck. But you could also set it to be centered for a more even listening experience between driver and passenger side.

After all is said and done i set my eq for the whole system to my ears and what I like.

Naturally your ears pick up the mid range better than the bass and tweeter range so it's not uncommon for your eq curve to look like a smiley face.

I often use the decks coarser eq to do the final to the ear adjustment.

Your bass will be low during testing so adjust it afterwards to your level of liking.

Double check you aren't clipping when it's all said and done.

You may lose a few db of overall loudness due to having to boost certain frequencies. This in turn means a lower gain position in order to not clip the signal. Most of the time it's not an issue since most people only care about db of the subs which are largely unaffected. A few db on highs isn't a big deal. You could always have a surplus of power to make up for this as well.

 

[SlugButter]

I have a DSP that basically turned into a fancy pants electronic crossover. I still haven’t went through the process of setting it up the way its supposed to be.

 

[Jimi77 5,000+ posts]

I got the deadner thing so I'm okay with that I've already deadened my entire car. I guess this is what I'm getting at:

I listen to a lot of EDM and rap it's so I don't want everything in the front of me at the center. I want to be sitting in the middle of it because sounds move left and right in DJ mixes and beats so I want to be where I'm like, in it. This is like if I were watching a symphony or an orchestra or something and I was sitting in the crowd. I want to sound like I'm at a rave in the middle of a club or something. So how do I do that?

Or like for instance if I'm listening to Korn, I don't want to be in the crowd in front of them, I want to be on stage next to Jonathan. Do you get what I mean?

Being in a car helps. If you want to be "on-stage" then move the speakers next to you instead of in front of you.

 

[mat3833 10+ year member]

Oof, welcome to the beginning of the end my friend. I read most of the information here and I agree with the advice given. I'll lay out the process I went through(and went through, and went through, and through, and will go through again) and give some suggestions.

Before you do any of this, take a bit of time and set your time alignment using whatever DSP method. You will have to change this multiple times, but it's better to do it now and later. One suggestion is to add a baseline 20ms delay to everything. So if your dsp calculates 9ms for a speaker, manually set it to 29ms.

Step 1 is to download REW on a laptop.

Step 2 is to get a calibrated Mic(I use the Dayton iMM-6)

Step 3 is getting your baseline measurements. One thing I would suggest is getting a baseline of individual speaker pairs as well as the entire system as a whole.
For me that's midbass, midrange, tweeters, and subs. You will be tuning the EQ for each speaker pair.

Step 4 is picking a "house curve" to tune too. Everybody likes different curved but a basic curve would be the "JBL house curve" you can find info on house curves here: https://www.diymobileaudio.com/threads/rew-house-curve-files.259337/page-2

Once you have your curve picked and your baselines this is where the time-consuming part comes in.

Step 5 is making your first big adjustments. Focus on the large peaks and dips. Try to pick one peak and one dip and bring them back to your house curve or within 1-2db. I like to work in pairs, one peak and one valley per change. This will take a bit of time, but should be pretty simple once you get the hang of it. Go through all your speaker pairs until you have things mostly in line with your chosen curve.

Step 6 is taking another measurement for each speaker pair and your total system. Look at how close these graphs match your chosen house curve. And listen to your stereo. This is a great time to use your Headunit EQ and mess around a bit. Cut a specific band and not how it sounds. I wouldn't recommend boosting anything on the HU, but cutting is perfectly fine and can give you an idea of what sound you like.

II'll answer any questions on this part and then continue on with level-matching, time alignment, and final tweaks next post.

Matt

Edit: I forgot. When taking measurements, sit in your car and move the mic in a slow figure 8 at ear level in front of your face. You also want to keep a consistent volume level throughout all measurements.

 

[Original wis]

Gotcha. Incidentally today while I was at lunch on work I was playing some old school metal that I like to listen to every once in awhile and I realized that if you actually stopped and listen for a second you can figure out where every single person is as well as where every single instrument is pretty much. It actually shifted my perspective on listening to music actually sat there like an idiot for about 30 minutes listening to different songs and picking out where everything in the song was in relation to like I would actually be. So I kind of get the wisdom of setting the front stage is if you're at a concert. It actually makes watching music videos make more sense cuz everything is actually playing where it is in relation to the music video as well.

That also means that there really isn't a such thing as SQL. There's a certain level of volume where it's going to be impossible to be discerning enough to appreciate it.

 

[Original wis]

What's a house curve again? Is curve the scientifically correct word for like EQ settings?

Okay so for sags let's say that I bypass the laptop the Rew software the pink noise and I just set all of my sound equipment to flat and use the access DSP app. And my own ears. What's the worst case scenario if I just do it that way?

Don't get me wrong. From from a scientific standpoint I completely understand the value of using the pink noise that r e w software and the microphone.

I'm trying to establish a baseline so that I can evaluate and what the value is of using the mic and r e w software as opposed to just doing it by ear. Because I think either way I'm going to be able to tune the sound to something that I personally like.

 

[mat3833 10+ year member]

What's a house curve again? Is curve the scientifically correct word for like EQ settings?

Okay so for sags let's say that I bypass the laptop the Rew software the pink noise and I just set all of my sound equipment to flat and use the access DSP app. And my own ears. What's the worst case scenario if I just do it that way?

Don't get me wrong. From from a scientific standpoint I completely understand the value of using the pink noise that r e w software and the microphone.

I'm trying to establish a baseline so that I can evaluate and what the value is of using the mic and r e w software as opposed to just doing it by ear. Because I think either way I'm going to be able to tune the sound to something that I personally like.

You can tune by ear. Nothing really wrong with it honestly. I just recently got a mic and started using a DSP, and it's a world of difference. Everything is just more balanced and it just sounds better.

A house curve is just an EQ target. Generally you aim for an in-car response with a "boost" of 6db from 20hz tapering down to zero around 150-200hz, flat through the midrange, and either a slight shelf of 1-2db above 8kish or a slight dip in the opposite direction.

We are much more sensitive to "midrange" frequencies, so the 500hz-4500hz band is what needs the most attention to sound right.

Matt

 

[Original wis]

You can tune by ear. Nothing really wrong with it honestly. I just recently got a mic and started using a DSP, and it's a world of difference. Everything is just more balanced and it just sounds better.

A house curve is just an EQ target. Generally you aim for an in-car response with a "boost" of 6db from 20hz tapering down to zero around 150-200hz, flat through the midrange, and either a slight shelf of 1-2db above 8kish or a slight dip in the opposite direction.

We are much more sensitive to "midrange" frequencies, so the 500hz-4500hz band is what needs the most attention to sound right.

Matt

Pretty much my thoughts thanks for confirming

 

[Original wis]

You can tune by ear. Nothing really wrong with it honestly. I just recently got a mic and started using a DSP, and it's a world of difference. Everything is just more balanced and it just sounds better.

A house curve is just an EQ target. Generally you aim for an in-car response with a "boost" of 6db from 20hz tapering down to zero around 150-200hz, flat through the midrange, and either a slight shelf of 1-2db above 8kish or a slight dip in the opposite direction.

We are much more sensitive to "midrange" frequencies, so the 500hz-4500hz band is what needs the most attention to sound right.

Matt

Are there any rules for using 1 crossover and 2 DSPs? One for front stage, one for rear fill? Crossover for assisting the overall levels while you're driving?

 

[mat3833 10+ year member]

Are there any rules for using 1 crossover and 2 DSPs? One for front stage, one for rear fill? Crossover for assisting the overall levels while you're driving?

I'm a little confused. You shouldnt need a dedicated crossover(passive crossover) if you have a dsp. I'm running an active 3 way setup with the DSP handling the crossover stuff.

If you are running passive crossovers your tuning process will be a little different since you have less adjustability on individual drivers, but your goal is the same.

Matt