DiyMobileAudio.com Car Stereo Forum banner

1 - 20 of 27 Posts

·
Registered
2004 Rav4 JL900/5 Rss265-44HO Rockford DSR-1 Fountek FR88 RS180-4
Joined
·
61 Posts
Discussion Starter #1 (Edited)
Maybe some wizards will clear this up in a few key strokes, that would be nice.



278689
278690
This is a hypothetical situation involving the Dayton rss265ho and the Dayton rs75 I chose the Daytons because I own the ho and Dayton provides the phase info easily accessible.
If I chose to cross the ho at 200hz, the phase is ~20* advanced in relation. And if I choose to cross the rs75 at around 300hz the phase is lagging ~30*. I’ve been thought many times to use 24db LR filters to leave the phase unchanged overall. So, with the phase unchanged by the filter the sub is still advanced ~30* and the rs75 is lagging ~25*. Combined the phase appears to be about 50* mismatched. Now, phase audibility through different frequency ranges is of interest here, however my prime motivation behind this post is to discuss methods of getting as close to absolute coherency as is currently possible. So far the methods I’ve tossed around are.

1: Using time alignment, mathematically considering the general length of the waves involved and then calculating the distance to that point of phase.
I’m just guestamating here but around 200hz the wavelengths are a couple feet long. So changing the delay, say to add a distance equivalent to 30* either way seems detrimental to the overall sound arrival timing between individual drivers.

2: Different crossover slopes, if there are slopes that alter the phase ~45* that could possibly work. My concern would be the wider frequency overlap caused by the shallow slopes introducing a myriad of other complications.

3: Allpass filters, I’ve read there could be some complications involved, I don’t remember though.

4: DSP with phase rotation. This interests me probably the most. This doesn’t seem like a common feature anong DSPs though, perhaps because there are other simpler ways.

One of the reasons I like fullrange speakers is because of the phase coherency over such a broad range. If I can wrap my head around the topic, I can open up my options and employ more suitable choices for my specific design ambitions.
 

·
Registered
Joined
·
1,718 Posts
So I assume the data plots your showing are for the electrical phase measured by DATS? Once the sound waves leave the speaker and bounce around the interior of the car like a pin ball before reaching your ears - There's no chance they will representative of what's in those plots.
 

·
Registered
2004 Rav4 JL900/5 Rss265-44HO Rockford DSR-1 Fountek FR88 RS180-4
Joined
·
61 Posts
Discussion Starter #4 (Edited)
Hmm, good point I completely overlooked. Are you suggesting that even though I’m focused on the direct sound, the secondary chaos renders my concerns a moot point? And even my beloved fullrange coherency obsession lacks relevance in the automotive environment? If so, I can move forward in my life and install purpose built drivers for their relevant passbands. So in your opinion even with a tweeter mounted very close to the mid (electrical) phase doesn’t ultimately matter? Is acoustic phase generally flat across the passband?
 

·
Premium Member
Joined
·
4,981 Posts
Any filter slopes, or box group delay, will change the acoustic phase. The inductance will also change it to a small extent.
 

·
Registered
Joined
·
405 Posts
Forget about electrical phase. We use 24db LR to avoid acoustic phase issues. 6db = 90 degrees out, 12db = 180 out, 18db = 270 out, 24db = 0 out. The closer the tweeter is to the mid the better.
 

·
Registered
2004 Rav4 JL900/5 Rss265-44HO Rockford DSR-1 Fountek FR88 RS180-4
Joined
·
61 Posts
Discussion Starter #7
Are we saying that acoustic phase is generally linear though a speakers passband? More so, should I be concerned at all, having dsp in my arsenal? If there’s any benefit at all in correcting acoustic phase in a multistage car arrangement, I’m interested in figuring it out. If not, there are plenty of other things for me to obsess over.
 

·
Registered
2004 Rav4 JL900/5 Rss265-44HO Rockford DSR-1 Fountek FR88 RS180-4
Joined
·
61 Posts
Discussion Starter #9
Cool, now I feel like I’m getting somewhere. I’ve read many opinions about how detrimental crossing in the vocal 300-3000 range is, even outside where it’s not “as bad” is still bad in their mind. Like it’s some kind of mystic phenomenon that separates the pros from the Walmart parking lot car clubs. In the second video he mentioned how, for most enthusiasts his methods are sufficient. I’m interested in achieving the last ten percent, so to say. Pretty much everyone who’s replied to my questions knows a lot, do you guys concern yourselves with minute phase nonlinearity around crossover points?
 

·
Registered
Joined
·
405 Posts
I primarily concern myself with the summation at the crossover point. Ideally you want a 6db gain there. Keep beaming in mind too when considering crossover points.

278696
 

·
Registered
Joined
·
1,171 Posts
While nicks video is handy, it’s a very simple overview, you can swap phase and adjust time alignment incorrectly and get it to sum correctly also, a tape should get you something like

If you are after perfect phase, I am not far off with my current car, I play my sub to 55 (70 acoustic) midbass from around 70 to 165 acoustic, and mids from 165 to 3.5k acoustic, then 3.5k on up

I then use smaart to setup time delays and get the phase to overlay throughout the drivers passband using all the tricks in my arsenal, due to very careful choice of drivers and mounting locations to get the phase as good as I could get below the magic Schroeder frequency so I get as smooth and perfect a summation as possible with excellent, almost perfect summation below that freq around 300-350hz
 

·
Registered
2004 Rav4 JL900/5 Rss265-44HO Rockford DSR-1 Fountek FR88 RS180-4
Joined
·
61 Posts
Discussion Starter #13 (Edited)
While nicks video is handy, it’s a very simple overview, you can swap phase and adjust time alignment incorrectly and get it to sum correctly also, a tape should get you something like

If you are after perfect phase, I am not far off with my current car, I play my sub to 55 (70 acoustic) midbass from around 70 to 165 acoustic, and mids from 165 to 3.5k acoustic, then 3.5k on up

I then use smaart to setup time delays and get the phase to overlay throughout the drivers passband using all the tricks in my arsenal, due to very careful choice of drivers and mounting locations to get the phase as good as I could get below the magic Schroeder frequency so I get as smooth and perfect a summation as possible with excellent, almost perfect summation below that freq around 300-350hz
While nicks video is handy, it’s a very simple overview, you can swap phase and adjust time alignment incorrectly and get it to sum correctly also, a tape should get you something like

If you are after perfect phase, I am not far off with my current car, I play my sub to 55 (70 acoustic) midbass from around 70 to 165 acoustic, and mids from 165 to 3.5k acoustic, then 3.5k on up

I then use smaart to setup time delays and get the phase to overlay throughout the drivers passband using all the tricks in my arsenal, due to very careful choice of drivers and mounting locations to get the phase as good as I could get below the magic Schroeder frequency so I get as smooth and perfect a summation as possible with excellent, almost perfect summation below that freq around 300-350hz
Is phase more difficult/important below the schroder freq?

Do we need to be choosing our x-over points based on avoiding crossing in the vocal range, or does range not matter if phase is coherent? I mean can you crossover in the ears most phase sensitive frequencies, if those points are chosen for their phase linearities?

In my graphs used for example the electrical phase is shown, I mistook it for acoustic. Is acoustic phase linear in a drivers passband?

I’ve been kinda reluctant to accept the easy answer, apparently the ear can be trained to detect phase problems in crossover points, so it seems possible that even with proper summation “something“ can still be off.

Do you feel the comprimises I make using fullranges to cover my mids and highs are justified by my interest in phase coherency?
 

·
Registered
Joined
·
405 Posts
I’ve never used widebands although I do have some GS25’s coming for rear fill so not sure firsthand. It does eliminate one crossover point so there is that. In a perfect world we would use only one speaker for the whole spectrum. What widebands are you planning on using and where will they be mounted?
 

·
Registered
2004 Rav4 JL900/5 Rss265-44HO Rockford DSR-1 Fountek FR88 RS180-4
Joined
·
61 Posts
Discussion Starter #15
I’ve never used widebands although I do have some GS25’s coming for rear fill so not sure firsthand. It does eliminate one crossover point so there is that. In a perfect world we would use only one speaker for the whole spectrum. What widebands are you planning on using and where will they be mounted?
fountek fr88ex in .5ltr sealed sail enclosures dead on axis. They have their ups and downs. They do play plenty high and zaph measurements show good distortion overall, a decade ago...
I have a lively thread active, it’s the one about crossing a 4” directly to my subwoofer, you might enjoy.
 

·
Registered
Joined
·
1,171 Posts
Is phase more difficult/important below the schroder freq?

Do we need to be choosing our x-over points based on avoiding crossing in the vocal range, or does range not matter if phase is coherent? I mean can you crossover in the ears most phase sensitive frequencies, if those points are chosen for their phase linearities?

In my graphs used for example the electrical phase is shown, I mistook it for acoustic. Is acoustic phase linear in a drivers passband?

I’ve been kinda reluctant to accept the easy answer, apparently the ear can be trained to detect phase problems in crossover points, so it seems possible that even with proper summation “something“ can still be off.

Do you feel the comprimises I make using fullranges to cover my mids and highs are justified by my interest in phase coherency?
If you can measure phase and see issues with a few measurements around the ears using something like smaart you can then aim to get crossovers in the cleanest locations and allow you to get better summation in crossovers

if home audio can put crossovers in the 2.5-3k range and pull it off with passive crossovers then we can with active and more control, so it wouldn’t be something I am overly concerned about, widebands/full ranges do have a place for a very simple effective front stage, however I will also say I like the sparkle and air a good tweeter brings to a front stage, so for me I’d rather have that extra top end detail of a tweeter, but each to there own

Last year I tuned a car that won every round of Emma sq competition it went to... it used hybrid unity 6x9 and unity 2” as its front stage

the lower in frequency you get the more important phase is, so subs to midbass is very phase dependant... 80% phase/20% eq, levels... tweeters are more 50/50 for me

Acoustic phase is generally whatever the car dictates and as good as your speaker locations make it... electrical phase is a paper thing, you don’t need to be concerned with it as like I say the car messes with it and makes it whatever it ends up as

The hardest part is getting crossovers free of phase issues so you get good summation and coherence, then eqing to a setting that allows the drivers to shine 👍🏼

i hope this helps somewhat 😎
 

·
Registered
2004 Rav4 JL900/5 Rss265-44HO Rockford DSR-1 Fountek FR88 RS180-4
Joined
·
61 Posts
Discussion Starter #17
If you can measure phase and see issues with a few measurements around the ears using something like smaart you can then aim to get crossovers in the cleanest locations and allow you to get better summation in crossovers

if home audio can put crossovers in the 2.5-3k range and pull it off with passive crossovers then we can with active and more control, so it wouldn’t be something I am overly concerned about, widebands/full ranges do have a place for a very simple effective front stage, however I will also say I like the sparkle and air a good tweeter brings to a front stage, so for me I’d rather have that extra top end detail of a tweeter, but each to there own

Last year I tuned a car that won every round of Emma sq competition it went to... it used hybrid unity 6x9 and unity 2” as its front stage

the lower in frequency you get the more important phase is, so subs to midbass is very phase dependant... 80% phase/20% eq, levels... tweeters are more 50/50 for me

Acoustic phase is generally whatever the car dictates and as good as your speaker locations make it... electrical phase is a paper thing, you don’t need to be concerned with it as like I say the car messes with it and makes it whatever it ends up as

The hardest part is getting crossovers free of phase issues so you get good summation and coherence, then eqing to a setting that allows the drivers to shine 👍🏼

i hope this helps somewhat 😎
Extremely helpful my friend!!! I can’t give my thanks enough to the people helping me and everyone else on here!!!

I’ve seen a DSP that can rotate phase in 15* intervals, I forget which one. Would that be a very helpful feature?

So, there is nothing magical about phase, it just affects things like comb filters and summation? Like I learned in car audio 101? I need to be testing mounting locations for combining and nulls for the smoothest overall response? I’m telling you, I was convinced phase was some strange thing only audible to SQ judges and superheroes. See, I read audiophiles talking about hearing phase and $150,000 speakers perfectly phase coherent, but they might be blowing smoke up each others bum...

How about the new mini dsp C-dsp with DIRAC? Seems to take a lot of work out of tuning, leaving only the finishing touches to someone with capable ears.

I’m really enjoying growing in the hobby, just so many variables to consider. The fact that people win competitions with widebanders really stokes my fire.

I’m so grateful for the info, and answering so many questions. I’d love if you answered everything I just wrote but if I had to sum it up.

• Choose speaker locations based on creating a wide and deep soundstage.

• Choose speaker location based on overall in car phase and most importantly coherency at crossover.
 

·
Registered
Joined
·
1,171 Posts
Extremely helpful my friend!!! I can’t give my thanks enough to the people helping me and everyone else on here!!!

I’ve seen a DSP that can rotate phase in 15* intervals, I forget which one. Would that be a very helpful feature?

So, there is nothing magical about phase, it just affects things like comb filters and summation? Like I learned in car audio 101? I need to be testing mounting locations for combining and nulls for the smoothest overall response? I’m telling you, I was convinced phase was some strange thing only audible to SQ judges and superheroes. See, I read audiophiles talking about hearing phase and $150,000 speakers perfectly phase coherent, but they might be blowing smoke up each others bum...

How about the new mini dsp C-dsp with DIRAC? Seems to take a lot of work out of tuning, leaving only the finishing touches to someone with capable ears.

I’m really enjoying growing in the hobby, just so many variables to consider. The fact that people win competitions with widebanders really stokes my fire.

I’m so grateful for the info, and answering so many questions. I’d love if you answered everything I just wrote but if I had to sum it up.

• Choose speaker locations based on creating a wide and deep soundstage.

• Choose speaker location based on overall in car phase and most importantly coherency at crossover.
some people are very good at hearing phase issues, it’s like everything, once you’ve heard something and know what to listen for it’s very hard to unhear it so then every time you listen you then hear it if that makes sense, so when they don’t hear it it will be very obvious when it’s not there, some people also probably talk out of there arris 😂

Helix can do steps of 8 degrees or something similar, it can be handy but only if you can measure it to be able to set it with that resolution 🙃

The issue you have is being able to measure it as with time alignment and phase adjustment there are 100’s of combinations, and anyone saying they can guess it by ear is not as accurate as seeing it as you adjust it live with so emerging like smaart

speaker locations for deep stages are often different from wide, just pick the one that gives the best presentation of the width to the ears, you can’t measure this, it’s a case of try speakers in different locations and aiming (on axis or close to it is generally better for me for detail, not always best for reflections, often widebands are better aimed at the respective headrest so the reflection is kept minimal from the opposite side window (for the passenger side normally)) playing a single speaker at a time and work out where the audio sounds as focused as possible, this may not be at the driver due to reflections and the way your brain works out where the source of sound is coming from... it’s amazing how reflections can make a speaker sound like it’s coming from a location halfway up the windscreen at certain freqs

This location will also likely have the least phase issues due to the lack of reflections

The last thing you said isn’t quite accurate, you pick locations that are cleanest with regard to phase, otherwise your imaging will split at some freqs and be coherent at others (think when you have speakers out of phase on purpose and the two sources sound like they are in two different locations instead of being in the centre... but in this case the image splits at certain frequencys only)

then you would pick crossovers where both speakers have minimal phase issues, so if the midbass has a phase issue at 210 and the mids have an issue at 120... then you’d crossover at 165 as an example, this way you will have the most coherence through the crossover region from 125-200 👍🏼
Dirac is a different beast using for filters to adapt phase and make it more linear at the listening position, I am very curious about using it in my car as it’s a massive advantage if done correctly 👍🏼 You still have to set time alignment and crossovers first, you can’t put tune a bad install to some degree in every case, I would always choose to get the install as good as possible and then use electronics as the icing on the cake after it’s as good as it can be au naturel so to speak 👍🏼
 

·
Registered
2004 Rav4 JL900/5 Rss265-44HO Rockford DSR-1 Fountek FR88 RS180-4
Joined
·
61 Posts
Discussion Starter #19
Dumdum you’re a f~{%:ing gentleman and a scholar 🙏
 
1 - 20 of 27 Posts
Top