Why should you time align
Short answer: you'll get the most accurate imaging
Detailed answer: see my post on the reason time alignment is needed
What you will need
Behringer DCX
Measurement mic that will interface with DCX (ECM8000 or similar)
XLR cable
Step 1: equipment setup
1. Place your mic in your listening position. For this test you don't require a mic tripod, since you can probably simply place it in your couch, where your ears are when listening.
2. Connect the mic to input C on Behringer DCX
Step 2: run Auto EQ
1. Press the setup button
2. With the page button navigate to page 2 - auto align
3. Press OK
All channels will be muted. Un-mute all the channels by pressing all of them. A series of pulses will play through each channel.
4. Set gain
Rotate the dial to set the gain on DCX. If you have a preamp upstream, it will not work since DCX is acting as a source. You will need to set the volume quite high or the routine will not work.
5. Click OK
DCX will play a series of pulses and measure the time delay. It will then assign settings to align.
Step 3: review settings
1. Press channel 1 then via the page button navigate to page 7 - polarity.
You may notice the polarity of some channels has been inverted. Click through all the channels to see what has been done.
2. Review delay settings on page 8.
If you have a subwoofer included, don't be surprised if the other channels have been delayed by a large amount like 8m or 20 ms. DCX is correcting for phase shift as well, not simply the physical location of the drivers.
Why does DCX do it differently each time?
If you run the routine again, with nothing altered, you will get different settings. At first, they may appear radically different, but on closer inspection you will find the different settings are in fact doing the same thing in slightly different ways. It can correct differences with both polarity reversal and digital delay. If in a subsequent run it handles polarity differently, then it will compensate with different delay.
Confirming with measurements
You may wish to confirm with measurements. Investigate with frequency response where phase is also plotted. If you see a common 180 degrees phase shift at the crossover, DCX will simply invert polarity. If the shift isn't a neat fit, then digital delay may be used. If you know the phase shift at the crossover, you can work out the required digital delay.
First work out the time for one full cycle
Time = 1/frequency
eg. at 3k, time = 1/3000 = 0.3 ms
Delay = phase shift/180 x time for one cycle
So the delay = phase shift / 180 x 1/frequency
In this case, reversing the polarity at 3k is equivalent to 0.15 milliseconds.
Time alignment series
Why you should time align >
Is it really necessary?
Time alignment overview >
A bird's eye view of various ways that are used to time align speakers. Does physical offset really work?
Time alignment with Behringer DCX >
This is about as easy as it gets. You need a mic and a few minutes to run the auto routine.
Digital time alignment >
My preferred method. No need to build bizarre baffles that can introduce their own issues.
Subwoofer phase alignment
Yes, even subwoofers benefit, but not for the same reason
Paul...thanks for puting this up. I have read through that DCX manual many times...it requires a bit of editing and you have done a fine job!
ReplyDeleteIn my system ,the volume had to be so loud before a "reading" would occur, that I feared some damage would evenutate(to my ears, if nothing else).
You have given me confidence to tackle it again.
The only drawback for me at the moment, is that I prefer to use the DCX to manage the sub drivers only, in a range below that which my "Golds/cabinets" do not go anyway. In order to delay the Golds, they would have to go through the DCX, which I had been hoping to avoid.
I'm glad you found it helpful. You might like to check out the recent listening test thread on StereoNET and my post here. My impression has been that DCX is reasonably transparent in a compatible system. There are some that won't be satisfied with it's sonics - small differences can be picked in an instant switching test if the conditions are right. It's worth trying in your system - try to keep an open mind. You might also try for yourself a digital conversion test. Simply switch a loop cable vs DCX working with null settings so that it does analogue > digital > analogue conversions. You will find out exactly what the digital conversions do to the sound. I'd contend that most would find it difficult to pick a difference and would have to listen carefully.
ReplyDeleteHi Paul,
ReplyDeleteGreat column and great site. I just bought a DCX2496 and am preparing to install it in my home theater to crossover my two diy 3-way front L&R speakers.
I gather I just install it as I would a regular crossover (I have a Rane AC23 in place at the moment) then follow your instructions.
Does it test and align all the drivers in both L&R speakers at the same time?
Or does it just align the drivers in, say the left speaker with each other, and then align the drivers in the right speaker with each other, leaving the alignment between the left and right speakers to the surround processor?
Thank you for your consideration.
Best wishes,
Dave
(I don't mean to be anonymous, I'm just not sure how to do the other choices)
Auto align will cover all channels. It depends how you have it set up. If you have stereo link for a 3 way then it seems to choose the same settings for left and right. In other words, it will not adjust for the fact that the left speakers is closer to the mic, but it will delay the mid and tweeter on the left to align with the woofer. I haven't tried it with link off, but I would guess it would change how the channels are aligned.
ReplyDeleteHi Paul,
ReplyDeleteThanks to your great column, I was able to time align my two front speakers using the DCX. After reading your columns on the minidsp units, I'm thinking of getting a couple to use as crossovers and EQ's . It seems like for determining the delay times I could use the DCX as before, just write down the delay times and whether or not polarity is reversed, and use that information to set the minidsps. Do you see any problem with that?
Thanks for the great blog.
Best wishes,
Dave
Dave, I find that for subwoofers you can require more delay to the mains than MiniDSP will allow - that is a problem! Also, the times calculated by DCX are not always accurate. You need to check. One way to do this is with excess group delay on REW - check the online help files for that one.
ReplyDeleteHi Paul,
ReplyDeleteNice column and great job. I encountered a problem though which i didn't manage to pass through though when using the auto align . Here is what i did, i hooked up everything including the ecm, switched on the dcx and amps, ( i was running a 2way setup with sub). setup the crossover 2 way and proceeded to the auto align page. pressed ok and unmuted 4 channels then went ahead to crank up the gain on dcx outputs. I managed to hear the short sound bursts but they continued for sometime and seem to take forever to fnish so I opted to go manual.The environment was noisy so im wondering what I did wrong......which gain should be cranked up?
Thanks
Ibrahim
Noise can be a real problem and even in a very quiet room you have to make it ridiculously loud to work. The solution is to run it when and where background noise is low or you may not get it to work. DCX sends its signal direct to the power amps and there is only one gain control you can use. I should mention that it appears you can't trust the DCX measurement of time alignment entirely.
ReplyDeleteI have tried the auto test with the DEX ultra curve and found the test volume needed to be set at unreasonably high levels to even get it to run. I can't imagine what they are thinking when they set up the test program? While I was running my first attempt, with what sounded like a tornado in my living room, I could smell paper burning? I noticed the RTA was showing the 6 right most bands were at max and I still wasn't able to see any bars on the screen at the highest setting. I smelled the burning and shut it down. I thought maybe it was 1 of my 3 amps and couldn't find anything wrong. But then I thought about the settings and sure enough the system required the volume to be so loud that it burned up a coil in my electrostatic mid/tweeters crossover section. The smell was the paper cover and the former the coil was wrapped on. I removed the coil since I'm not using that crossover and it was fine but I question what good is it to set the sound level so high that it's setting up my system to sound "good" at levels I'll never ever play it? This makes no sense?
ReplyDeleteWe all know that the sound level will change the overall perceived sound when it quiet and when it's loud. I'm not using any loudness controls or anything and my system sounded great when I had analog stuff doing this.
I actually bought the DEX by accident while looking for the DCX. Luckily I was able to exchange it once they got some on stock. Now the only reason I wanted to get the DCX is for the time alignment settings since I made 2 sets of DIY 3 ways tri-amped speakers both using electostat mid/tweeters. I really wanted to use a Rane 23 as another poster mentioned but I missed my chance on 1 locally for $60! All others I saw were over $250! So the DCX offered the sane plus replace my analog EQ that has RTA pink noise tone sweep and pulses plus mic input for testing. Setting up that system took less than 30 minutes but I had polarity and time issues.
So now I'm trying to get things sounding close to before and finding the same issues as DEX. I must add that I have also tried miniDSP and found it to be simply useless. It altered the sound even when in bypass mode. There was no way I could even get it to be tolerable and after 3 days of struggling with a piece of garbage it went back. I have also tried a few others with even worse results, defective units seem to be a common factor with these and as I've found out interface has not improved since before 1999 when I saw Rockford fosgste had a digital DSP for cars and I knew right away that it was never going to work with that interface and multiple buttons pushes just to change a single setting.
Now here I am wishing I had the Rane crossover to do this. If I have to set my volume so loud that I see the mic input level hitting the red led and I'm still unable to run the test what do I do?
I'm baffled once again at the reasoning behind the setting. I live in a super quiet area and my ambient sound is in the low 20dbs? My breathing reads a gain on a meter so it's not loud noise floor. 1 of my amps has a fan but it only raises the sound to low 30s. They want me to set my volume in the upper 90s to low 100dbs and sure I'll never play it that loud. When I do these tests manually with a sound meter and an RTA sweep I'm able to get things set but I still end up with 1 speaker out of phase. If I manually change it then I have a other and so on in a circle. It sounds impossible but that's why I bought this. Any ideas how this always ends with 2 driver out of phase with the others? Like I said it's like a continuos loop of chasing the phase issue.
It's been a long time since I've used the auto align feature but yes I do recall that the SPL level required for it to work is insane. If you are running ESL panels I certainly wouldn't use the auto align.
DeleteYour experience with MiniDSP sounds quite odd. I've had very good results with many different MiniDSP units, starting with their basic 2x4 MiniDSP in a box ranging through to their more recent SHD units which are excellent.