Home Theater Forum and Systems banner

Port Timing

Tags
port timing
3K views 28 replies 5 participants last post by  jtalden 
#1 ·
Both my mains and the sub are ported. I remember reading a while back that ported speakers disrupt relative timing between speakers and subs. Is this true and if so, can it be easily taken care of with external delays provided by a miniDSP or equivalent?
 
#2 ·
Ported speakers just have a different SPL and phase response (different rolloff rates).
Ported or not, the objective is just to create close phase tracking between the SW and mains.

If the best timing alignment results in the phase crossing at the acoustic XO rather than tracking closely throughout the entire XO range it is possible to fine tune it by changing the SPL/phase rolloff rate of one of drivers:
> Plugging the mains ports will change the rolloff rate from 24 to 12 dB.
> Changing the relative filter settings (as from LR24/LR24 to LR24/But18) for the LPF and HPF XO filters may be an option.
> Changing the relative XO filter frequencies (HPF vs LPF settings) also has some effect.
> Possibly some the Phase control dials on some SWs will shift phase rate of the rolloff through the XO range. Others may just be adjusting the overall SW delay. I don't have any SW's with those controls to test.

It all depends on the setup controls available to the hobbyist. Most people don't have many, if any, options. Fine tuning like that is very complicated and time consuming. It's the playground of dedicated tweakers. It is very unlikely to be detectable in sound quality.

Just pick the best phase tracking for the XO settings you choose. There are better way to spend one's time.
 
#3 ·
Yeah... you can spend a lifetime on this problem, and only hear the most minute of improvements. The big question is how dissatisfied you are with your current sound... if you're pretty happy at the moment, I wouldn't recommend going down that rabbit hole. Everything suggested above is spot-on, but also easier said than done in a lot of cases.
 
#4 ·
I read the following article and his explanation made a lot of sense about this very topic.

It seemed worth the effort to give it a try and ended up following his advice for my 'music only' system. I'm amazed at how much better/clearer and how the sound is 'louder' on lower volume setting. Well worth the effort and I highly recommend it.

All my ports are sealed and my mains are crossed over @ 80hz as directed. I am using sub drivers designed for a sealed cabinet. (NHT)

BTW... an important part of the set up was using the subwoofer location in figure Fig 7. 'An asymmetrical layout.'




This made a huge difference in my small 13' x 15' room. I can actually balance the bass at the listening position by using the volume control on the two subwoofers. Very cool idea. I run the sub closest to the sofa at a low crossover and slightly lower volume so it's not detectable in the least. Both subs are firing aimed away from the listening position at the wall to cut down any localization. The bass is very tight and it really does "fatten up the base". So now the system plays louder on less volume. Helped with nulls also.

"A ported sub for home use is even more wrong than ported mains. Now you would be attempting to acoustically add together in the room at least SIX low frequency sources with differing phase and frequency slope conditions: the LF drivers in your two mains, their ports, the sub driver, and its port."


http://www.soundoctor.com/whitepapers/subs.htm
 
#5 ·
"A ported sub for home use is even more wrong than ported mains. Now you would be attempting to acoustically add together in the room at least SIX low frequency sources with differing phase and frequency slope conditions: the LF drivers in your two mains, their ports, the sub driver, and its port."


http://www.soundoctor.com/whitepapers/subs.htm
This quote and some other comments in the linked article reflects a partial misunderstanding of the SW to mains XO setup issues in my opinion:
> I see no technical argument for a clear preference of a ported Vs sealed SW. They have slightly different advantages/disadvantages so for a given situation one may be a little better choice than the other. Pointing to 1 or 2 advantages/disadvantages without context to how it helps or hurts a particular situation is not helpful.
> We also cannot predict if ported Vs sealed mains will work more ideally for phase tracking in a given situation. It is the overall system acoustic XO characteristics that is main factor. This is a function of the room and SW placement as well as the electrical XO settings (including the important delay setting).
> The typical small listening room is often the primary actor in the success of the setup. Only careful measurements will determined the best setup from the various options available. Most often, due primarily to room acoustics of a given setup, it is a matter of choosing a compromise XO setting between several less than ideal solutions.
 
#11 ·
1. Yes
2. If I am doing the analysis just use the on-floor distances to the baffles. The distances are theoretically measured as line of sight LP to the baffle at driver height. The floor distances are plenty good enough as it's the relative distances to the mains that is more important than the absolute distances. The SW distance is the one that may change distance anyway so the starting distance is not a big concern. It's just good to know that it was approximately correct to start with. Otherwise a big needed distance change would make me think I made a calculation mistake.
3. If the SW LPF is the only change (the mains stay the same slope) then just measure the SW with each setting and if there is a significant difference we can chose the setting that provides the better results. The total measurements is then 4 rather than 3.
> Lmain
> Rmain
> SW12
> SW24

Please advise the XO settings and distance settings used. That can be done right in the REW notes for each measurement. Please set the sweep so it extends well past the XO range. I normally just sweep full range.
 
#13 · (Edited)
Well, maybe not so soon. :R
I was unable to acquire a full set of measurements with the acoustic timing reference. And I'm not so sure the ones I did take are valid.

REW Configured for ASIO / HDMI connection:
  • AVR is recognized and active
  • Only 2 channels appear in the OUTPUT selector instead of 6 as shown in Windows Manage Audio Devices
  • Scarlett 2i2 input (mic) is recognized but IDLE <---- SHOWSTOPPER

REW Configured for JAVA / RCA connections:
  • Both channels properly appear in the I/O selectors of the PREFERENCES window.
  • Either channel is able to be selected as an acoustic timing reference of the MEASUREMENTS window.
  • Short-burst timing signal is heard before full-range test signal
  • Test signal plays through L+R+SUB <---- SHOWSTOPPER

Now this is where I fell on my face... how do you measure ONLY the right channel with the left channel assigned as an acoustic reference? Output-1 (left) of my 2i2 soundcard is connected to both analog inputs of my AVR through a Y-splitter. So when the AVR is in stereo mode for this measurement, both channels (and the sub) play the full-range test signal. The only way I can see to silence the left channel is to unplug its RCA connection after the timing signal is generated. There must be a better way! Is the solution to use the same speaker for timing as measuring? How then does that work for the sub, which needs the L/R inputs connected to receive summed-to-mono bass below the xover?

Aside from manipulating AVR inputs as shown below (lifted from the REW101 Guide), additional speaker combinations can be created by disconnecting power amp inputs:
Text Font Document
 
#15 ·
I didn't read carefully all your posts regarding the 2i2 but be advised I never got ASIO4ALL to work consistently with that unit. It only was stable using the Focusrite ASIO driver. That doesn't let you use the HDMI though if I understand correctly. I just use a loopback cable for timing and send the analog signal to the driver using an analog cable. That may be a real pain if you cannot easily mute/unplug the other drivers so they don't play.


So hopefully ASIO4ALL will be stable for you.


I am not following closely as I have no power at home; 665,000 homes impacted. I'm working at a library now. They say it may be several days for some customers get back up. I'm hoping they get me some power sooner than that though.
 
#19 ·
I'm hoping they get me some power sooner than that though.
It was good to hear you were in the first half!

....
2. If I am doing the analysis just use the on-floor distances to the baffles. The distances are theoretically measured as line of sight LP to the baffle at driver height. The floor distances are plenty good enough as it's the relative distances to the mains that is more important than the absolute distances. The SW distance is the one that may change distance anyway so the starting distance is not a big concern. It's just good to know that it was approximately correct to start with. Otherwise a big needed distance change would make me think I made a calculation mistake.
3. If the SW LPF is the only change (the mains stay the same slope) then just measure the SW with each setting and if there is a significant difference we can chose the setting that provides the better results. The total measurements is then 4 rather than 3.
> Lmain
> Rmain
> SW12
> SW24

Please advise the XO settings and distance settings used. That can be done right in the REW notes for each measurement. Please set the sweep so it extends well past the XO range. I normally just sweep full range.
Here's my first stab at the data you asked for. The measurements are for the room after acoustic treatments. I have measurements for the baseline and intermediate steps if you'd like to see any of those. Please let me know if there's anything I should have done differently.
View attachment Room Layout 4A (12dBpo 5xTraps).mdat
View attachment Room Layout 4A (24dBpo 5xTraps).mdat
 
#20 ·
Lou,
There is a problem with this data. The IR positions are inconsistent. I am still trying to fully understand what may have gone wrong. There appears to be some problem with the loopback timing setup. We need to assure loopback timing is working properly before we try again.

Per the info panel in REW:
> The Mic is plugged into the left channel

Given that case:
> The TRS/TRS (or TS/TS) loopback cable should be connected to the 2i2 right output to right input.
> The Timing Channel should be set to 'Right' in REW.
> The measurement channel should be set to 'left' and the left output of the 2i2 is connected to the receivers left CD/Aux/DVD/etc (not multichannel) input for measuring the left main and to the right input for measuring the right main. [there are various ways to do this differently depending on the equipment being used. I didn't go back and try see what equipment you have or how it is configured.
> The music mode should be 'Stereo' for evaluation of the 2 front channels.

The info panel in REW shows:
> The timing channel set to 'left'. That is the mic position! Possibly that is the only problem?
> The output/measurement channel is set to 'both'. I am not sure if that would explain the issue. I would always use 'left' in this case.

Please consider what may be wrong and correct it. To provide more specific help I would need all your setup information.

You should remeasure the left and right mains alone several times and confirm to yourself that the IR location and shape is repeatable for each main channel and that the left and right channel IR locations are very close to each other when the mic is carefully centered and the distances/delays are for each channel are set identically.

Once we confirm the measuring setup then we can then trust the 4 measurements that I need for my analysis. You can compare my recommendation to the results you obtained using trial and error with the SW phase control. That method can probably work just fine, but I have no experience with it.
 
#21 ·
Lou,
There is a problem with this data. The IR positions are inconsistent. I am still trying to fully understand what may have gone wrong.
Ha-hah, just like software - my measurements always need debug! After measuring all FR, I first click on a single graph to the left and then the IMPULSE button in the upper right. Next I click on LIMITS and GENERATE MINIMUM PHASE (taking defaults). I can scroll around to find the trace, but the only way I can really come up with anything resembling what I've seen in the guides and here on the site is to enter a t=0 OFFSET equal to the delay in msec shown in that particular measurement's notes section. I may not have done that for every measurement that needed it.

There appears to be some problem with the loopback timing setup. We need to assure loopback timing is working properly before we try again. Per the info panel in REW:
> The Mic is plugged into the left channel

Given that case:
> The TRS/TRS (or TS/TS) loopback cable should be connected to the 2i2 right output to right input.
> The Timing Channel should be set to 'Right' in REW.
> The measurement channel should be set to 'left' and the left output of the 2i2 is connected to the receivers left CD/Aux/DVD/etc (not multichannel) input for measuring the left main and to the right input for measuring the right main. [there are various ways to do this differently depending on the equipment being used. I didn't go back and try see what equipment you have or how it is configured.
> The music mode should be 'Stereo' for evaluation of the 2 front channels.
  • Yes. CH01 IN = Mic.
  • Yes. Loopback cable connected between CH02 IN and CH02 OUT.
  • No. AFAIK, the timing Channel can only be selected when using HDMI. The MEASUREMENTS PANEL allowed me to specify whether no timing ref is used, a loopback timing ref is used, or an acoustic timing ref is used. But for JAVA with hardwired connections, REW does not offer a selection for left or right timing channel.
  • Yes. I set the measurement channel to LEFT using the INPUT selector in PREFERENCES.
    L ONLY: 2i2 CH01 OUT to Left Analog IN only / Mode=Stereo / Sub=Off.
    R ONLY: 2i2 CH01 OUT to Right Analog IN only / Mode=Stereo / Sub=Off.
    C ONLY: 2i2 CH01 OUT to Left Analog IN only / Disconnect AVR center out & right out / Connect AVR left out to amp center in / Mode=Stereo / Sub=Off.
    SUB ONLY: 2i2 CH01 OUT to Left & Right Analog IN / Turn off power amp / Mode=Stereo / Sub=On.
    C+SUB: Same as C ONLY, except turn sub on.
    L+R+SUB: 2i2 CH01 OUT to Left & Right Analog IN / Disconnect AVR center out / Mode=Stereo / Sub=On


The info panel in REW shows:
> The timing channel set to 'left'. That is the mic position! Possibly that is the only problem?
> The output/measurement channel is set to 'both'. I am not sure if that would explain the issue. I would always use 'left' in this case.
After taking new measurements exercising the new settings noted above, the info panel still shows the timing is being taken from LINE OUT = LEFT. I'm assuming that LEFT = CH01 for the 2i2. This happened even after changing OUTPUT in the MEASUREMENT PANEL from BOTH to LEFT. I will retry the measurements with OUTPUT set to RIGHT to see if that fixes the issue before I ask you to reanalyze.

Please consider what may be wrong and correct it. To provide more specific help I would need all your setup information.
Your help is much appreciated. You shouldn't have to hold my hand with every aspect of setup, but I may not even be aware that's where the problem lies. I'd like to try your suggestions and then revisit the issue. I should have another mdat file tomorrow.

You should remeasure the left and right mains alone several times and confirm to yourself that the IR location and shape is repeatable for each main channel and that the left and right channel IR locations are very close to each other when the mic is carefully centered and the distances/delays are for each channel are set identically.

Once we confirm the measuring setup then we can then trust the 4 measurements that I need for my analysis. You can compare my recommendation to the results you obtained using trial and error with the SW phase control. That method can probably work just fine, but I have no experience with it.
Boy do you ever have confidence in me! I really have no clue what I'm doing when generating IR data. Can you tell me if I should be offsetting t=0 per the delay specified in the notes section? That's the only way I've even been able to find the trace! I'll try the procedure you mention above. Who knows, I may make you proud!
 
#24 ·
I will do the SW12 analysis today.

Just continue to use the Java driver. I think you have it working okay. Hopefully using the loopback cable several measurements were made on the same speaker and it was confirmed all the impulses fell at the identical time on the impulse chart. As long as there is timing repeatability, and it's possible to route the signal to the measurement speaker, that is all that is needed.
 
#25 ·
Lou,
I reviewed the data you provided for SW12:
> The measurements there were not what I need for my normal phase tracking analysis so that method couldn't be used.
> The data there does follow a good trial and error method. One I mentioned in the 4th bullet of Post-2. That is; using the phase control on the SW to see which setting best supports and smooths the SPL response. From the data we find that the last measurement was the best setting. That setting is not labeled, but the sequence before it would suggest that it is with the phase control set to 180°. The setting of 135° (next to last measurement) is almost identical. Either of these 2 settings looks very good for SPL support.

We could get new measurement that fit my requirements for my normal phase tracking analysis, but that is not likely to result in any significant improvement to either of these 2 options.

I suggest you use the settings of the final measurement in that series. It works very well for the left and right channel. Try it for the center channel to assure it works well there also. If it does not, do not change the SW settings. Just adjust the distance of the center main ±1, 2, 3 ft and choose the distance that provides the best SPL response. It will probably look good at the current setting.
 
#28 ·
Lou,
I reviewed the data you provided for SW12:
> The measurements there were not what I need for my normal phase tracking analysis so that method couldn't be used.
> The data there does follow a good trial and error method. One I mentioned in the 4th bullet of Post-2. That is; using the phase control on the SW to see which setting best supports and smooths the SPL response. From the data we find that the last measurement was the best setting. That setting is not labeled, but the sequence before it would suggest that it is with the phase control set to 180°. The setting of 135° (next to last measurement) is almost identical. Either of these 2 settings looks very good for SPL support.

We could get new measurement that fit my requirements for my normal phase tracking analysis, but that is not likely to result in any significant improvement to either of these 2 options.
Thank you once again for your time and attention to detail. I'm disappointed in myself for not supplying what you needed. I'm also disappointed we couldn't take this exercise to completion. I was hoping to learn more about IR graphs, but I feel like I've failed to grasp the basics as demonstrated by others in similar threads. Part of the issue may be my lack of a clear-cut goal, which was initially vague but later morphed into timing adjustment between mains and sub. If we stop now, my take-away would be that timing adjustments are a tweaker's playing field which offer (for most) a relatively small subjective return-on-investment. I would also include that satisfactory results can be obtained through the phase tracking method, and/or by manipulating distance settings. To my credit, however, I did learn quite a bit about ASIO4ALL and some nuances of the measurement process. All valuable lessons to carry forward!

I suggest you use the settings of the final measurement in that series. It works very well for the left and right channel. Try it for the center channel to assure it works well there also. If it does not, do not change the SW settings. Just adjust the distance of the center main ±1, 2, 3 ft and choose the distance that provides the best SPL response. It will probably look good at the current setting.
Perfect - exactly the advice I was looking for! This information is golden to me, so thanks for sharing! I sincerely appreciate your effort surrounding these issues, and hope to learn more from you in the future.
 
#27 ·
I did cover the center channel issue in the last paragraph in Post-25. It requires you to do the same kind of test you did on left and right, but instead of the SW phase control (leave that at 180°), change the center channel distance to find the best setting for SPL.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top