27 replies to this topic

[Winno]

My HX-D2 lets me adjust balance and has full time alignment features.

I've been setting up the ta and tried balance adjust first (so that I get a homologous sound across the entire stage) and then ta for focus.

I then tried it the other way around.

Let's forget (if we can) driver placement for a bit and just talk about the above two factors. If we have to take it into consideration, let's asume that it's symetrical on each side and say a 4 way set up with mids and tweeters on dash, midbases in doors and subs in the hatch.

How did you set yours up and why?

Cheers

Side comment: I am noticing that image solidity and placement is alterable through adjustment of tweeter phase when I thought that treble was influenced more through intensity rather than phase.

Edited by Winno, 02 November 2009 - 09:13 PM.

[~Samuel~]

Did the chicken come before the egg?

Sorry i cant be of more help... Back to SPL i go

[BMWTurbo]

How are you adjusting the phase?

In a single seat system, I personally tune in the following order TA/Phase/Polartiy - Xover - Levels - EQ - Levels - TA - Repeat as required.

I use Impulse Response measured at a single abitrary point about mid-way between the ears and set the 'phase' (via polarity) so that all drivers are in phase and that the Impulse response is within the smallest gradation on the unit. Each driver is tested by itself.

Once this is set I look at selectiing crossover points (testing above TA setting will have been done with each driver playing an extended range, whilst taknig care not to damage the driver), via knowledge of the drivers and plotted response curves.

Next I look at level matching each driver so that the ouput is roughly 'similar' at the listening position and then set about EQ'ing down some of the larger peaks in response.

Once you've done this you may need to revisit the level's as they will change depending on 'cut' amounts and you may wish to tweak further the level and TA in order to solidify the 'stage'.

[Pulse-R]

To be honest, and give the easy answer... there is no 'right way' to do it.

The longer answer says to set the TA first, based on relative distances as measured. then set the crossovers to suit the drivers' response, then the phase if required (depends on placement and crossovers).
Now adjust the EQ and levels to give good sound, then tweak the TA to firm up the image, and a bit more EQ or TA to reduce steering.

[muzzy66]

It's something that everyone (I imagine) would do in their own way.

I always set my crossovers first, because that's something I can set objectively based on driver data, and pretty much get right first go. After this, I usually set up the gain levels (objectively, based on a combination of 1w/1m sensitivity and input power), and then set the time alignment (based on a combination of distance, and simulated group delay).

I use objective data to set all of these initial values which usually gets me around 95% on the mark, then do some fine-tuning by ear to get that last 5%.

It's an unorthodox method for sure, but but it's worked very well for me so far - and no matter how much 'by ear' tuning I do, I always end up somewhere within 10% of those 'initial' values I set at the start.

As stated above there is no 'right or wrong' method really - this is just the way I like to do it

[syd-monster]

similarly (looking at responce, not staging, as staging comes in between 5 & 6)
1 - study equipment to get a basic x-over point set up
2 - set all gains on minimum
3 - start with everything in phase on all items, terminals & settings
4 - set gains with a meter or ol' school ear
5 - adjust cross x-over points, slopes
6 - compare to my reference
7 - repeat, from step 3 until close
8 - borrow, steal, beg for RTA sweep
8.5 - now set level, re-adjust gains, and xover points & slopes with RTA in hand
9 - compromise between what I like and what is realtively linear and then to a know curve, such as fletcher muson,
10 - go back to 3 and perhaps introduce some EQ. or T/A if I have too.

[Pulse-R]

I find that the T/A is interactivew with the crossover slopes and levels... due to cancellations/etc.. so some fine tuning is needed in there.

[Winno]

Thanks guys.
I have spent a bit of time studying and thinking about this topic and Muzzy and Syd, the order that you mention is that which is used by most of the well known and regarded tuners.

Most agree that levels on all drivers need to be set first.

I have a question further in relation to this though;

When they say levels need to be set to identical levels from all drivers, does this mean:
1. The actual output in dB from each driver is the same
or,
2. The perceived level at the main listening position is the same (which means that the relative output from the left side is set higher than the right)

I must admit that in the past I have just winged it and been very lucky. Now I'm really getting into following a documented process and it's fun. My system is my dream system and I have waited for years to get this far and I naturally want to get the best out of it now.

Thanks so far.
Winno

[Gonadman2]

I must be the worst Sq'er ever, I just whack a good CD in and tweak until it starts to sound good. Muzzy, using objective data is fine, but sometimes the installation will dictate different crossovers/slopes than you might otherwise think. Point in case was my tangs, they allegedly play to 20khz, and I had them set to cross to the tweeter at 10khz. Unfortunately the firing angle of the midrange dictated a lower x-over than what the 'objective' data said topping out at 6.3khz and 12db slope. Using objective data as a starting point is fine, but I think that much tweaking needs to be done before you can truly say that you got x-over points 'right.'

I suppose I do go through some of the steps here, I've never thought about it really.

[muzzy66]

Gonadman2, on Nov 3 2009, 10:10 PM, said:

I must be the worst Sq'er ever, I just whack a good CD in and tweak until it starts to sound good. Muzzy, using objective data is fine, but sometimes the installation will dictate different crossovers/slopes than you might otherwise think. Point in case was my tangs, they allegedly play to 20khz, and I had them set to cross to the tweeter at 10khz. Unfortunately the firing angle of the midrange dictated a lower x-over than what the 'objective' data said topping out at 6.3khz and 12db slope. Using objective data as a starting point is fine, but I think that much tweaking needs to be done before you can truly say that you got x-over points 'right.'

I suppose I do go through some of the steps here, I've never thought about it really.

Definately - I always look at what angles I'll be using, and then take that into consideration when setting the crossovers.

For example, if the speakers are going to be installed 60 deg off axis, then I'll check the 60-deg off axis response of the drivers in order to determine which crossover points may or may not potentially work well.

Sometimes neither crossover point is perfect on paper. The perfect example was with my current system - the tweeters play cleanest when crossed at 2.5khz, but the mid only played flat to around ~2.3khz before beginning to roll off. It was a decision between crossing at 2khz (potentially higher distortion on tweeters) or crossing at 2.5khz (small dip in midrange). It all came down to determining how noticable that little 1dB (or so) dip would be, and the only way to know for sure was to try both settings and determine it by ear.

A big reason why an on-axis response chart on it's own is practically useless for a car system - I don't think i've seen a single car stereo in which the speakers are truly 100% on axis ... they tend to be at 15deg off axis at the very least, and even this is enough to limit top end response.

But anyhoo, back to the story...!

Edited by muzzy66, 04 November 2009 - 12:02 AM.

[muzzy66]

Edit: Accidental double post - internet playing up!

Edited by muzzy66, 03 November 2009 - 11:57 PM.

[Pulse-R]

two-seat (Real Men's SQ) tuning means getting the levels the same (or close) at two listening positions

[br85]

Pulse-R, on Nov 4 2009, 05:18 PM, said:

two-seat (Real Men's SQ) tuning means getting the levels the same (or close) at two listening positions

A challenge for the sake of it if you ask me. When I can walk into a mastering studio and sit beside the engineer and get the same levels as him from my seat, I might try it in a car. Until then, I might just keep playing with the little boys who can only get good levels from one seat.

[Winno]

Who cares about the passengers anyway, especially when they don't care themselves!

I must admit that after sitting in alot of cars now here and overseas, I am yet to find a car that images as solidly from both seats as it does from the driver's seat.
There will always be a compromise to take some of the driver's side goodness and give it to the passenger as well.

[Gonadman2]

Winno, on Nov 5 2009, 04:55 AM, said:

Who cares about the passengers anyway, especially when they don't care themselves!

I must admit that after sitting in alot of cars now here and overseas, I am yet to find a car that images as solidly from both seats as it does from the driver's seat.
There will always be a compromise to take some of the driver's side goodness and give it to the passenger as well.

I agree with this 100%. I have never heard a 2 seat car and come away thinking "man I gotta try that!"