DiyMobileAudio.com Car Stereo Forum banner

941 - 960 of 983 Posts

·
Registered
Joined
·
829 Posts
I’m sure my reply won’t be satisfying, and it is just typed words unless you could hear my car, and you may come away from that experience still not satisfied. But I’m happy with my system - maybe not 100% satisfied, but nearly 90%.

At the present time, I have three main tunes on the car. 1) 3511 crossover point between tweets and mid range 2) 2000 crossover point between tweets and mid range 3) same as 1 but with a 2db dip in the curve around 2.5k for the rowdy stuff (Smashing Pumpkins, Flyleaf, Breaking Benjamin).

All three with the same method. Crossovers, Time Alignment (obtained by completing a 7 channel run), no EQ or levels in the Plug-In, 2 channel Dirac with a curve that I have developed based off a 1/2 Whitledge curve. All three sound very similar because of the curve (I believe). I perceive a lot of detail, staging/ imaging that is better than when I used other processors, and bass response that is accurate. A local friend has heard the car several times, and two trusted friends will hear it next weekend and will tell me if it is decent to their ears or if it needs to be set on fire and rolled down a hill.

But to answer your question - I get repeatable results that provides a sound I’m happy with.


Sent from my iPhone using Tapatalk Pro
Glad to hear it. I appreciate the response as well. Any special method for mic placement when running Dirac or straight from the manual?

My goal right now is to try out different speakers in the car without having to manually tune each setup. If Dirac can do what it does consistently, it would allow for the speakers themselves being the only variable and not the tune.

I know that with past auto-tune dsps, people would have a difficult time getting the same auto-tune result even when using the same exact settings (including the crossovers) with the same exact speakers and in the same locations.
 

·
Registered
Joined
·
1,732 Posts
Verify please.

Switching to off only changes the voltage input from 2-4 or 8-12 correct? Just making sure I'm reading this correctly.

View attachment 269895
That is correct, I recommend using a magnifying glass to verify they are all where they should be. I had one that was switched on, took about a month to figure out where the clipping was coming from.
 

·
Registered
Joined
·
2,615 Posts
But I’m happy with my system - maybe not 100% satisfied, but nearly 90%.
I've heard an earlier iteration of this car and I'd be 90% satisfied as well :ROFLMAO::LOL: In all seriousness, I've heard including mine 4 cars with DL now, all running various equipment in a variety of locations and they all sound great.

I'm completely satisfied with mine, I have zero desire to do anything with it other than enjoy listening to it. Someone with a more critical ear may point things out, but I'm happy. My process is repeatable, manually set crossovers, use Dirac to get levels, then run a 2 channel Dirac and done.

Any special method for mic placement when running Dirac or straight from the manual?
This will be the biggest thing to figure out as what works in my car, or simply to my preference, might not be correct for yours. I believe we all follow the manual, but other than the center initial microphone measurement, the other 8 can vary. Some make smaller size boxes with the mic positioning, others larger boxes. I find i like the larger box, but I believe @bertholomey and @Truthunter use a smaller box.
 

·
Premium Member
Joined
·
925 Posts
Looks like Dirac 3.0 is a go on miniDSP’s DDRC-24 and DDRC-22 as well as the multichannel DDRC-88A and DDRC88-D..

Version 3.0 incorporates Dirac Live Bass Control support for better integration of subwoofers:

Hopefully the C-DSP 8x12DL gets 2.0 or 3.0 support!
 

·
Registered
Joined
·
34 Posts
Installed the C-DSP 8x12DL yesterday in between my Alpine PXA-H800 and my amps. Alpine is processing stereo SPDIF input from my factory nav through Dolby Pro Logic II to 5.1 surround, then sent into the C-DSP and processed as 6 discrete Dirac Live channels. All crossovers and delays set on the C-DSP. Tuned to a modified Wisdom curve the first go-around.

I can't compare how this sounds to Dirac Live 2.0. I was fortunate enough to encounter the previous version of DL software not asking for microphone permission on my MacBook Pro...so I updated everything fresh to fix the problem.

So I'm running Dirac Live 3.0...but not actively splitting a 20Hz-20kHz signal into its constituents...the sub channel is already separated by the Alpine. All channels connected straight via 6 RCA cables.

Regardless...best audio I've heard in a car, if not in any context. Expansive, holographic soundstage. The cliche of hearing things in music I've never heard before this setup.

I'll certainly tweak the curve, but this is, in general, "it".
 

·
Registered
Joined
·
1,732 Posts
So, now that this thing is seeing widespread use, can anyone tell me how often the software needs to connect to the internet, beyond the initial license verification? Can a Dirac tune be run offline?
 

·
Registered
Joined
·
1,687 Posts
I've been using Dirac 3.0 for a few weeks now... first as a beta and now as the release. Only once had to unplug the USB & reconnect it to the processor for it to connect without giving me some weird error when I would launch D3 during beta testing. But never had D3 crash on me or lock up once while in use. I'm using a dell latitude with Windows 10 on it.

So, now that this thing is seeing widespread use, can anyone tell me how often the software needs to connect to the internet, beyond the initial license verification? Can a Dirac tune be run offline?
In 1.0 it seem to only need to connect when loading the filter correction file to the DSP.

But I think 3.0 is different. You need to actually sign into your Dirac account In order to save files and be able to load them later. Also, unlike 1.0, there is no way to load a saved file and adjust things in the filter design screen without being connected to the DSP.
 

·
Registered
Joined
·
40 Posts
So I've used Dirac the first time ever. Still learning.
I ran 5 channel dirac first (2way+sub), got the levels and delays of each speaker.
Then I ran 2 channel with a chair location. (left tw+left midbass+1 sub output and right tw+right midbass + 2nd sub output). I used Harman's +10db curve (low end boost).
I did just 1 center measurement.

Good things: the center of the stage is awesome. But it's too narrow at the center. (Kind of when most instruments are at the center).
What I don't like - is the synthetic sound, don't know how to call it properly. Maybe - overly compressed or something.
I'm sure there are a lot of mistakes and room for improvement.
 

·
Registered
Joined
·
2,615 Posts
You need to actually sign into your Dirac account In order to save files and be able to load them later.
I don't think you need to do this, I'll double check later, but am fairly certain I don't have a Dirac account, yet have been able to both save and load a file.

Good things: the center of the stage is awesome. But it's too narrow at the center. (Kind of when most instruments are at the center).
How far apart were your microphone measurements? I get both a solid center stage, but also plenty of width. Same with any other car running Dirac I've listened to, they don't sound narrow in the center at all.

I take my measurements about shoulder width apart and about that distance front to back as well. YMMV of course as everyone's car is different.
 

·
Registered
Joined
·
40 Posts
I take my measurements about shoulder width apart and about that distance front to back as well. YMMV of course as everyone's car is different.
I did just 1 center measurement, not 9. Wasn't moving it. But I will do measurements for all positions. I think this will play a big role.
 

·
Registered
Joined
·
34 Posts
I did just 1 center measurement, not 9. Wasn't moving it. But I will do measurements for all positions. I think this will play a big role.
Definitely take your 9 measurements and relisten.

That synthetic sound is likely partly a result of the tune that resulted from that single measurement. When I first got the PXA-H800 and ran Imprint (its autotune), I first used a single measurement point (out of a total of 6 possible IIRC for a single seat optimization). The resulting curve as well as the corrected impulse response would look amazing on the filter prediction (the process was similar to Dirac Live), but the nature of the sound was as you described.

Because of the myriad issues resulting from the volume of a car cabin, the reverberant nature of the chamber, the sheer amount of glass surface in any car leading to reflections galore...you've got to measure at multiple points to handle curve-fitting above the Schroeder frequency (~300Hz in a typical car) without overcorrecting. To a point, more data will result in a better tune, so long as the algorithm (or tuner) knows what to do with the information...what to address and what to ignore relative to how humans perceive sound. From my single tune and based on reading the mid-level papers published by the Dirac team...Dirac Live will do a good job of interpreting what you feed it, so long as you feed it.

There's a good chance you'll want to tweak the curve after the tune is complete and filters are applied, but I imagine the overall system response will sound more natural to you with more measurements to inform the algorithm's behavior.
 

·
Registered
Joined
·
40 Posts
Definitely take your 9 measurements and relisten.
Sure, I will re-set the tune today. Including the amp gains and will take 9 measurements.
As for the 5/7 channel run (when we want to get the delays & levels) - are 9 measurements also necessary?

Another noob question about the routing of the sub in the Minidsp itself. I'm confused because I've seen screenshots with different variations.
I have 2 subs running from 1 channel amp. I run 2 RCA's from the MiniDSP outputs (out 5 & 6) to the amp inputs - 2 RCA in.
I use SPDIF input (split into 7&8 channels in the Minidsp first tab). In the MIXER tab - I summed outputs 5&6 into 1 channel ( for example - Dirac 5).
In the Dirac/Routing tab - do I also have to sum the signal?

UPDATE:
Re-adjusted the gains and crossovers a little. Ran 9 measurements with 2 channel Dirac (left+sub output 1, right + sub output 2). And I'd have to say - I'm impressed. The tonality and stage are much better, everything blends nicely. I especially like the higher frequencies. But, as anyone is experiencing - there's just not enough bass. I tried to pull the target in the bass region a few times, it became a little bit better, but still, it's just not enough. I want it to be tactile (2x15 IB subs). Anyone found a solution to make the bass right?
 

·
Registered
Joined
·
49 Posts
So does it seem safe to update to the newest firmware? I've never liked the fact that full utilization of the DSP requires an internet connection, and this seems to have gotten even worse. It's a bit of an issue for me, since all work on the car is done in my garage at the back of my property. No WiFi signal from the house, and spotty cell phone connection to tether the laptop to. But if the upgrade is worthwhile, sound-wise, I guess I'll give it a go. Anyone know if the firmware can be reverted back if there are issues with the new?
 

·
Registered
Joined
·
34 Posts
So does it seem safe to update to the newest firmware?
Nothing to compare to prior here, but no problems to note - updated to solve the problem I had with the Dirac Live software not requesting mic access. No visibility on whether the firmware can be reverted back, unfortunately.

Re-adjusted the gains and crossovers a little. Ran 9 measurements with 2 channel Dirac (left+sub output 1, right + sub output 2). And I'd have to say - I'm impressed. The tonality and stage are much better, everything blends nicely. I especially like the higher frequencies. But, as anyone is experiencing - there's just not enough bass. I tried to pull the target in the bass region a few times, it became a little bit better, but still, it's just not enough. I want it to be tactile (2x15 IB subs). Anyone found a solution to make the bass right?
Glad to hear the first part! : )

From skimming your past posts, we've got similar taste in sub setups - I have a single AE SBP15, infinite baffle of course. And I also like tactile bass - think sub-30Hz. Why I went IB.

I tuned to this curve created by banshee28

I chose that curve as a starting point, not an assumed end point, knowing two things:

1) It would likely need steeper dropoff in the treble based on the proximity of most of my frequency coverage to glass and the listener (coaxes in top of doors play 280Hz and up are effectively aimed at the opposite doors' glass, center channel plays 100Hz and up firing into the windshield...not to mention a certain degree of horn loading from the intersection of the windshield and dash. (I made a custom 5/8" open cell foam microsuede-topped dashmat to trap some of the highest frequency effects, but can't do everything and has negligible effect until the highest frequencies based on physics.)

2) It would likely need a rise in bass in that lowest ~half octave down to 20Hz. I.e. a smooth rise to 20Hz rather than flat from 60 to 20. That's partly based on the kinds of music I listen to...everything from classical to hip hop, but enough hip hop with subterranean 808 lines that don't come out as ear-level for the whole melodic passage note to note without a steady rise in the amplitude of frequency response as you descend. It's also based on my typical listening volume, which is relatively loud, but not ridiculous (think 80, maybe 85db). So Fletcher Munson likely has some effect, even relative to the volume the studio engineer likely listened at while making the mix. (Meaning that while producers DO boost the lowest bass regions based on their own hearing to compensate, if I'm not listening quite as loud, and I'm also dealing with overcoming low-frequency exhaust noise in my car, I'll still need that slight slope in the lowest bass...if I'm listening to music with semi-meaningful frequency content that low...which I am. I can name several tracks with an 808 line that goes close to and even below 20Hz.)

So. I haven't tweaked the curve yet...but from my past tuning and general ear ability, Dirac Live tuned to what I fed it. Other responsibilities are taking precedent for me right now, but I'll measure with REW within the next week to confirm, as well as tweak to tonal taste with a 31 band EQ in the signal path (laptop), measure the final resulting response with pink noise, then tune to that. And I'll share what I tuned to, with the caveat that I only vouch for it in my car with my particular selection of speakers, placement, crossovers, deadening, listening volume...you get the picture. :D

Post a screenshot of the curve you're tuning to, or attach the export if you can. What kind of music are you listening to?

Perhaps an unnecessary additional note, but seems pertinent since you mentioned liking the higher frequencies (possibly the detail you hear)...depending on your cabin and front stage setup, you might also need more treble attenuation than your curve is dictating, which could solve a large portion of your bass conundrum. In the past, I'd put massive bass boosts in the custom curves I'd tune to in the lowest frequencies...and for the aforementioned reasons, I'll likely still need a smooth boost at the very bottom...but make sure subtly louder-than-tonally-accurate treble isn't preventing you from hearing (and also feeling) your lowest bass. All of these curves are just starting points. The nature of the install (especially speaker location and crossover choice, i.e. where you're actually sending parts of your frequency response) dictates how you tweak it IMO.

Hope you enjoyed the book : D and let us know what you tuned to and what you measure, etc. We will both get there!
 

·
Registered
Joined
·
40 Posts
Had to retune the crossovers manually and do manual EQ in minidsp. I have some big problem with midbass in the listening position due to the cabin. ( The measurement is perfect when measured nearside in the legs 60-300 very linear). But in the listening position I have a 13db peak from 180-300 Hz in the left midbass and 125-200 peak right side).

Dirac shows this problem, but it still leaves a dip between sub and midbass. So I have to solve this as much as I can first.

Anyway.
There's a new feature In Dirac live 3 called Bass Control. Anyone used it? I don't see it.
Maybe this feature will solve the lack of bass problem.
 

·
Registered
Joined
·
1,687 Posts
Unfortunately, from what some of us were told during beta testing, MiniDSP is not supporting the bass control add-on for multi-channel Dirac at this time.
 
941 - 960 of 983 Posts
Top