Time alignment/delay questions.

Discussion in 'Archived Threads 2001-2004' started by Jeremy Anderson, Nov 1, 2001.

  1. Jeremy Anderson

    Jeremy Anderson Screenwriter

    Joined:
    Nov 23, 1999
    Messages:
    1,049
    Likes Received:
    0
    Trophy Points:
    0
    After a series of quick brownouts reset all the settings on my Onkyo 595, I was going through and resetting all my speaker distances. My system consists of Polk R10 mains and surrounds, CS175i center, PSW350 sub in a 12'x14'x8' room. The distance from listening position to both mains is 13' (measuring to the tweeter). The distance to the center is 11' and to surrounds is 8' and 10' (due to placement issues on the right surround). I have a few questions that I hope someone can answer:
    1) Can room acoustics or placement throw off the time alignment of the center speaker? I was trying different settings for the center just out of curiosity, and it seemed like setting it to 10' (3ms delay) instead of 11' made the front three speakers sound like one cohesive unit instead of separate speakers.
    2) While I know this is a cardinal sin here, the center speaker is unfortunately in an entertainment center's nook, hanging about 4 inches out of the little cubbyhole and angled down toward the listening position (and when a woman insists that you can't move it somewhere else, you just CAN'T). Could that be causing time alignment problems?
    3) If the room itself is affecting time alignment, is there a way to objectively measure what the center delay should be to counteract it? Or can someone tell me what quality exactly I should be listening for when it is properly adjusted?
    4) In the case of my surrounds, I have one at 8' and one at 10' distance. Do I split the difference and set the receiver to 9' or do I set it to whichever speaker is closest (in this case, 8')?
    I know this is mostly me being picky, but that's just me. My mains and surrounds are currently wall-mounted (via keyhole slot), but I'm gonna put them on B-tech mounts soon to both get them off the wall and allow for better aiming. Any help anyone can offer would be nice, and I can put up a quickie page of pictures of this room if it would help.
     
  2. Guy Kuo

    Guy Kuo Supporting Actor

    Joined:
    Mar 6, 1999
    Messages:
    581
    Likes Received:
    0
    Trophy Points:
    0
    I can think of two ways to objectively look at the delay issue.
    a. Use the ETF program with its output routed to both your center and a main speaker. Look at the impulse response and you'll see the impulse break up if timing is off. A little expensive to do if you don't already have the program for other acoustical tuning.
    b. Use the analog RS SPL meter and AVIA's speaker phase testing signals. Basically, you take advantage of those test generating the same signal on two speakers but alternating in phase. With the SPL meter in fast mode, you can watch the constructive and destructive interference. Play the left/center phase test and SLOWLY move the SPL meter left/right at your listening distance and find the point at which the needle undergoes the largest swing between in-phase and out-of-phase. That is the place there is maximal reinforcement and cancellation. Which means that position is on the acoustical midpoint plane of those speakers. If the center speaker is too close (or not delayed enough) that position will be pushed to the left of your prime listening position. Visualize it mentally as a triangle between the speakers and the meter. As you swing the base of the triangle (the line between the two speakers) the meter must also move to stay on the bisecting perpendicular. The triangle moves as a unit.
    This phasing method is extremly precise. A 1/2 inch positioning error will be detectable. I've previously described this in the forums. I don't believe anyone has ever reported they tried it.
    ------------------
    Guy Kuo
    www.ovationsw.com
    Ovation Software, the Home of AVIA DVD
     
  3. Bob McElfresh

    Bob McElfresh Producer

    Joined:
    May 22, 1999
    Messages:
    5,182
    Likes Received:
    0
    Trophy Points:
    0
    Well the rule is that sound will travel 1.1 feet per miliSecond (unless you are in Denver).
    This is to compensate for when the same sound erupts from all 3 front speakers at the same time.
    Room acoustics is normally not an issue because this calculation is based on direct-sound from the drivers to your ears.
    What your receiver is doing is taking the distance to one of your fronts (F) and the distance to the center (C) and calculating:
    Difference = F - C = 13-11 = 2 feet
    So your receiver is putting a ~ 2 Msec delay on the center sound.
    When you tell the receiver the center is 10 feet, it is calculating:
    Difference = F - C = 13 - 10 = 3 feet
    So your receiver is applying a ~ 3 mSec delay to the center.
    If setting your distance to 10' instead of 11' (3 ms instead of 2 ms) sounds cleaner, go for it.
    But what you are really doing in INCREASING the delay. This has the same effect as shoving the center back 1 foot with each miliSecond delay. This has the fun effect of making the sound-stage more deep.
    And for the rears..go with the 9' setting. Just make sure you are using a SPL meter to level-adjust to compensate for the different distances in terms of VOLUME.
    Good luck.
     
  4. Jeremy Anderson

    Jeremy Anderson Screenwriter

    Joined:
    Nov 23, 1999
    Messages:
    1,049
    Likes Received:
    0
    Trophy Points:
    0
    Bob, I've got the whole shebang calibrated with an SPL (and the right surround is +1db up from the left because of the difference in distance). I just wasn't sure what delay setting to use. I'll try 9 and see how that works out. If I can't find a happy medium on that, I'll likely rearrange so that I'm equidistant from each surround (since it's more of a seating problem than anything).
     
  5. Guy Kuo

    Guy Kuo Supporting Actor

    Joined:
    Mar 6, 1999
    Messages:
    581
    Likes Received:
    0
    Trophy Points:
    0
    If the center speaker is too close (or insufficiently delayed), its sound arrives earlier than it should. Since I'm having you phase check it against the signal from the left main speaker you have to position the meter closer to the left speaker to make the signals time align. That's why I say it would shift the point to the left (nearer the left speaker).
    You can move the meter in either an arc or line parallel to the back wall. It doesn't make a difference because it is actually a plane acoustically equidistant from both speakers which the meter is crossing. Either way, you cross that plane. I have you move the meter to find the max swing point because it is a lot easier to locate the current position and then alter it by moving the speakers slightly or changing the delay.
    Give it a try. It should make a lot more sense once you see how the meter behaves when it hits the magical plane. Just imagine a rigid triangle with the SPL meter at the apex and the two speakers forming the base. Once you know the direction you need to move the vertex, you know how to move the base.
    ------------------
    Guy Kuo
    www.ovationsw.com
    Ovation Software, the Home of AVIA DVD
     
  6. Jeremy Anderson

    Jeremy Anderson Screenwriter

    Joined:
    Nov 23, 1999
    Messages:
    1,049
    Likes Received:
    0
    Trophy Points:
    0
    Wow Guy... you weren't kidding! The point where constructive/destructive made the needle swing the most WAS shifted toward the left speaker. Changing the delay to 3ms (13' mains, 10' center) on the center channel brought that point pretty much dead between the speakers. Once again, you (and the Avia disc) continue to astound...
    Now, could this possibly help with tweaking my surround delay, since Avia has the left main/left surround phase tones as well?
     
  7. Guy Kuo

    Guy Kuo Supporting Actor

    Joined:
    Mar 6, 1999
    Messages:
    581
    Likes Received:
    0
    Trophy Points:
    0
    I wouldn't be confident in using it on the surrounds unless they are direct radiators. If the surrounds are dipolars, you have out of phase components coming at all times. The results would be unpredictable. This works for the main front speakers quite well. I'd invite you to do it for the left and right mains as well as the center to bring all three into exact alignment. Then sit with your head there and note how imaging is with both stereo music and soundtracks. I won't further influence your before you try that little exercise.
    ------------------
    Guy Kuo
    www.ovationsw.com
    Ovation Software, the Home of AVIA DVD
     
  8. Jeremy Anderson

    Jeremy Anderson Screenwriter

    Joined:
    Nov 23, 1999
    Messages:
    1,049
    Likes Received:
    0
    Trophy Points:
    0
    Well, my mains and surrounds are identical (Polk R10's) so it shouldn't be a problem to try it. The worst that can happen is I don't like how it sounds and change it back.
    I can't do anything about the positioning of the mains themselves right now (because they're attached to the wall via keyhole slots), but when I get the B-tech speaker mounts, I'll be sure to use all of this knowledge to properly place them. Thanks again for the advice!
     
  9. Jeremy Anderson

    Jeremy Anderson Screenwriter

    Joined:
    Nov 23, 1999
    Messages:
    1,049
    Likes Received:
    0
    Trophy Points:
    0
    Just wanted to follow up:
    I spent a good three hours listening to music last night in DPL-II after readjusting my speaker delays, and the difference is absolutely night and day. Changing the center delay to 3ms made the front soundstage... "seamless" is the best word I guess.
    I also did some experimenting with the phase tests between my left main and left surround (since both are direct radiators) and modified the surround delay accordingly (to 3ms instead of the previous 5). WOW! Talk about a difference! With Radiohead's OK COMPUTER in DPL-II, sounds that used to seem like they were coming from the speakers themselves were suddenly hovering in space between the speakers.
    Musically, it's like having a whole different system. Whereas before, mid-bass sounded muddy and flat from my center with music, suddenly I'm able to hear every note. It's amazing how much of a difference it made. I haven't really done much movie watching with it yet, but I'll throw SHREK at it tonight and see how that sounds. Thanks again Guy, and keep kickin' ass!
     
  10. Guy Kuo

    Guy Kuo Supporting Actor

    Joined:
    Mar 6, 1999
    Messages:
    581
    Likes Received:
    0
    Trophy Points:
    0
    Good job, Jeremy. I'm pleased to hear you found the technique helpful. It was particulary interesting to read how you did with the surrounds. I've not tried that here because I have dipolars. Like you, I was quite pleased with what getting the speakers into phase at the same place did for imaging. You can set things by distance or ms on the processor, but you aren't guaranteed those settings actually generate the exact amount of delay needed. Actually measuring the effects of phasing lets you compensate for the entire playback chain so you actually do get things aligned. This lets you get those speakers into position and delay for very tight timing. Thanks for trying it.
    ------------------
    Guy Kuo
    www.ovationsw.com
    Ovation Software, the Home of AVIA DVD
     
  11. Will

    Will Guest

    I've heard that time delays in the different speakers
    should not be done if an external analog crossover (like
    the Outlaw ICBM) is ALSO used.
    This isn't a problem when the digital crossover processing is
    done in the same unit as the digital time delay processing.
    But apparently it is a problem when analog crossover
    processing is performed afterwards (in an external unit).
    [Edited last by Will on November 02, 2001 at 03:50 PM]
     
  12. Guy Kuo

    Guy Kuo Supporting Actor

    Joined:
    Mar 6, 1999
    Messages:
    581
    Likes Received:
    0
    Trophy Points:
    0
    Yes, Will that is true. The channel delays must be applied after the crossover. Let's look at that with just two channels to make it easier to understand. Let's assume that channel A normally needs a 3 ms delay to arrive in phase with channel B at the prime listening position. Another way of saying this is delaying channel A by 3 ms causes signals which appear simultaneously in both channels to both arrive at the listener at the same time. Bass information which is extracted from both the channels is done at the machine (before any delay) so that bass info is also derived from bass info from the same instant in time and therefore in phase relative to each other. If that bass info is also delayed appropriately to arrive at the listener simultaneously (subwoofer delay), then all the info from the two channels which come from the same instant in the track arrive simultaneously at the listener. Perfect and ideal.
    Now we change the situation and add an external crossover network which lacks the ability to compensate for unequal speaker distances. If the various channels of info going into the crossover are not from the same instant in time (i.e., one or more delays are unequal from others) then bass will be added together from different instances in time. That would create a phase difference in bass being summed from the various channels. The result is cancellation or reinforcent of bass depending on its frequency - uneven bass response. The upper frequencies which are passed by the crossover would still be more or less delayed appropriately to arrive simultaneously at the listener, but the bass would already have been decimated before it even got to the subwoofer. For this reason, you get to choose between the following options.
    a. Preserve flatness of bass response by setting all delays to zero. This properly sums the bass together, but sacrifices the ability to time align the main speakers via delay settings. If you want to achieve both time alignment and bass response flatness, you need to physically place all your speakers at the same distance from the sweet spot.
    b. Forego bass response flatness by allowing out of phase bass mixing in the crossover device. This preserves main speaker time alignment, but defeats the reason you got the external crossover in the first place - better bass.
    Neither scenario is ideal. If you had the ability to place delays AFTER the external crossover, then you would have the best of both worlds. Lacking that you get to prioritize or move your speakers. One approach mentioned in the ICBM manual is to simply give up on letting any position in the room achieve perfect time alignment. After all, that only can be true for ONE position. I tend to disagree with the notion that if only one spot can be "perfect" there is no reason to create it. I may share the room with others, but there are times when I WANT to be in that spot for maximal enjoyment.
    ------------------
    Guy Kuo
    www.ovationsw.com
    Ovation Software, the Home of AVIA DVD
     
  13. John Kotches

    John Kotches Cinematographer

    Joined:
    Mar 14, 2000
    Messages:
    2,635
    Likes Received:
    0
    Trophy Points:
    0
    Guy,
    This is one big reason why neither SACD nor DVD-A will truly blossom as formats until a digital output standard exists.
    DVD-A is a series of compromises unless you have the $$$ to go for the Meridian setup which is fabulously expensive, and sounds pretty damned amazing!
    I'm fortunate enough in my latest arrangement of the room to have the front trio of speakers and the subwoofer equidistant. The surrounds are only 1' offset, which is a vast improvement over earlier configurations in my listening room. So timing issues will they exist, are fairly minimal.
    Regards,
    ------------------
    John Kotches
    Contributing Writer
    Secrets of Home Theater and High Fidelity
     
  14. scott_tinari

    scott_tinari Stunt Coordinator

    Joined:
    Feb 10, 2001
    Messages:
    121
    Likes Received:
    0
    Trophy Points:
    0
    This message is for GUY or anyone else that knows. I have a pioneer elite 35 reciever and I have the energy take 5 speakers. Now my question is on the reciever should I just plug in the distance from my listneing position from the front and back speakers into the reciever? Also my reciver doesnt have the (ms) just (ft). But someone on here said you can make the time delay tighter or something like that. My current delay is 8 feet for the 2 front speaker 8.5 for the center and 6.5 for the surrounds, can I adjust the delay to make it sound better or just keep it the way it is. Thanks
     
  15. Clinton McClure

    Clinton McClure Casual Enthusiast
    Supporter

    Joined:
    Jun 28, 1999
    Messages:
    4,412
    Likes Received:
    413
    Trophy Points:
    4,110
    Location:
    Central Arkansas
    Real Name:
    Clint
    Guy - Great information! This morning while following your steps to set my center delay, I discovered I also have a surround delay. (Yamaha is wonderful at not telling you things like this in the user's manual.)
    Is there a formula which can be used to set the surround delay? They are dipoles mounted on 32" tall stands about 4' from either side of the listening position. I am guessing from the factory the delay was set at 20ms because I have not adjusted it, but that is what it is set on.
    Bob - I just re-read your post and will follow your formula to reset my surround delay. I wonder if this will change any since they are dipoles...
    Thanks. [​IMG]
    ------------------
    My DVD Collection Casa del Clint
    [Edited last by Clinton on November 08, 2001 at 02:01 PM]
     
  16. Guy Kuo

    Guy Kuo Supporting Actor

    Joined:
    Mar 6, 1999
    Messages:
    581
    Likes Received:
    0
    Trophy Points:
    0
    With THX spec'ed processors, the speaker delays are set up by entering distances from listener to speaker in feet or meters. The processor then takes that information to calculate the appropriate number of ms delay which should be applied for each surround format. On non-THX systems, the delays are sometimes set in milliseconds. Approximately one ms delay is needed to "move" a speaker a foot farther away.
    Older Dolby Prologic only systems are a bit more complex to set up because the rear surround channel signal needs to be intentionally delayed so sound from the surrounds arrive 15 to 30 ms AFTER main channel sounds. This was needed because the extracted non-discrete surround channel could be contaminated by front channel sound. By adding the delay, the Haas effect could be used to mask the earlier arriving, undesired front channel sound which came out of the surround speakers. (BTW, If you are still stuck with a a receiver with only Prologic, I recommend you fine adjust the delay by applying an axe to the front plate. That way you can upgrade to a receiver with DD and DTS decoding capability.)
    So with Prologic you'd set up the delays such that the front three channels all arrive simultaneously and the surrounds will arrive about 20 ms later. Take into account the difference in speaker distances to make this happen.
    On a modern receiver, DD and DTS formats carry discrete surround channels. There is no longer a need for Haas effect to hide anything unless you play back something in Prologic. In the cases of DD and DTS, the delays are used to ensure that sounds from all channels arrive simultaneously at the user regardless of their actual physical distance. It is for very fine tuning and testing of this "in phase" state that I have described the SPL meter/AVIA phase signal method in this thread.
    Dipoles are trickier to test with this method because the speaker produces both in and out of phase sound components. Summing such a mix doesn't give as clean cancellation/reinforcement effects. It is possible to temporarily replace a dipole with a direct radiating speaker to test the delays for proper phasing. Once the delays are set, the dipole can be once more brought into use. The disadvantage of swapping out the speaker is that you will like loose that 1/2" precision which is obtainable with the phasing method.
    Thanks the for words of encouragement, Clinton. This is a topic which takes far too long to explain given the very small audience that takes advantage of the information.
    ------------------
    Guy Kuo
    www.ovationsw.com
    Ovation Software, the Home of AVIA DVD
     
  17. Clinton McClure

    Clinton McClure Casual Enthusiast
    Supporter

    Joined:
    Jun 28, 1999
    Messages:
    4,412
    Likes Received:
    413
    Trophy Points:
    4,110
    Location:
    Central Arkansas
    Real Name:
    Clint
    Here's what I've discovered:
    My mains are 8' from listening position, and my center is 7' away, and surrounds 4' away.
    I was able to decrease my overall center delay from 5ms to 1ms and the sound stage is unbelievable!
    It turns out my receiver has a seperate surround delay for DD/dts, prologic, & each DSP mode has it's own delay.
    Since the surrounds are only 4' away from the listening position, I used the same basic method {F - C = CD (only altered it to) F - S = SD}. I have my surrounds set at 4ms for DD/dts. The prologic was defaulted for 20ms, with a range of 15 - 30ms, like you (Guy) have stated previously. Using the same formula, I lowered the prologic surround delay to 19ms with good results.
    Guy and Bob, thanks again! [​IMG]
    ------------------
    My DVD Collection Casa del Clint
     
  18. Jeremy Anderson

    Jeremy Anderson Screenwriter

    Joined:
    Nov 23, 1999
    Messages:
    1,049
    Likes Received:
    0
    Trophy Points:
    0
    The reason I brought this topic up in the first place is that I've never seen it covered much here on the HTF (or any other forum, for that matter). I think it's an issue that too many people ignore, and I have to admit that I was guilty of that for my 2 year tenure with my old JVC DD receiver. But since then, I've upgraded to an Onkyo 595, gotten all new speakers, a new DVD player, rearranged the furniture, put concrete slabs under and on top of my subwoofer, bought thick curtains to dampen reflections off my window -- in short, I'm being obsessive. But none of those little tweaks/upgrades made as massive a difference as getting my speaker delays lined up right. Thanks again, Guy!
    Since we're on the topic, will this issue be covered in more detail on Ovation's upcoming disc for S&V? I already have Avia, but if the S&V disc will have more comprehensive testing of any sort, I'll definitely buy it.
     
  19. Will

    Will Guest

    This is great information!
     
  20. Guy Kuo

    Guy Kuo Supporting Actor

    Joined:
    Mar 6, 1999
    Messages:
    581
    Likes Received:
    0
    Trophy Points:
    0
    The Haas effect is the propensity of our hearing to register only the earlier occurance of a sound if the duplicate occurs about 20 ms later. Basically, if the second occurance happens during that window, we tend not to perceive the 2nd occurance. Longer than that of a delay we hear it as an echo. This was important for Prologic because front center channel sound leaked into the rear channels. By taking advantage of the Haas effect (by delaying the surrounds) the leaked sound could be psychoacoustically masked.
    Time delay and alignment does make a big difference in imaging. It is a pity that so few people really pay attention to this easily optimized detail.
    ------------------
    Guy Kuo
    www.ovationsw.com
    Ovation Software, the Home of AVIA DVD
     

Share This Page