Can I do more for my 2.1 system?

[Calypso]

Working to improve my 2.1 SQ through use of a MiniDSP FLEX and REW which has been both overwhelming and interesting. After spending many hours reading, watching YT videos, re-watching YT videos, I have managed to setup and collect measurements from REW, generate PEQ filters for all 3 speakers, establish cross over filter points and slopes, and gains. I also optimized the MiniDSP by using the PEQs on input and then again later as part of the output chain; quasi doubling the number of PEQs.

At each step, I spend quite a bit of time listening to the new configuration and comparing it, as best I could, to the unchanged configurations stored as one of the MiniDSP profiles. I do have some room modes visible in the waterfall that can only be addressed with bass traps but that will happen later.

I was just wondering if there are any other steps that I should still consider, for example, accounting for the different arrival times? If I understood correctly, use of FIRs with RePhase is more for speaker builds and not for systems like a 2.1.

Please correct me if I misunderstood and open to suggestions for my next step. Thanks.

 

[sm52]

accounting for the different arrival times

This is the main task, the first step. This should have been done before creating filters. Because by coordinating the time of the three speakers now, you will have to redo the filters again.

 

[Calypso]

This is the main task, the first step. This should have been done before creating filters. Because by coordinating the time of the three speakers now, you will have to redo the filters again.

Is there a video or guide that explains the process of determining the arrival times and what is entered into the MiniDSP?

 

[sm52]

There are many videos on Youtube and discussion threads on avnirvana. Write in the search time alignment REW miniDSP.

 

[Calypso]

@John Mulcahy - Hi John, if you have a moment can you provide me with your guidance? Thanks.

 

[JStewart]

Hey Calypso, I know you reached out to John, but may I offer this link to time alignment on MiniDSP’s website?

https://www.minidsp.com/applications/rew/speaker-time-alignment#tabulate

This guide will work well for setting alignment between the mains (or individual drivers in a speaker). Subwoofer to mains alignment requires a different strategy of aligning the phase of the mains and sub around the crossover to get the best SPL support. REW’s alignment tool can be used for this. You will find some information on the alignment tool on this page of the REW help

All SPL Graph

So, my suggestion is work through the mains alignment and if you have some questions or when you’re ready to move on to the subwoofer let us know.

 

[Calypso]

Hey Calypso, I know you reached out to John, but may I offer this link to time alignment on MiniDSP’s website?

https://www.minidsp.com/applications/rew/speaker-time-alignment#tabulate

This guide will work well for setting alignment between the mains (or individual drivers in a speaker). Subwoofer to mains alignment requires a different strategy of aligning the phase of the mains and sub around the crossover to get the best SPL support. REW’s alignment tool can be used for this. You will find some information on the alignment tool on this page of the REW help

All SPL Graph

So, my suggestion is work through the mains alignment and if you have some questions or when you’re ready to move on to the subwoofer let us know.

Appreciate the tips. I having been doing my research and landed at the link you provided and have since, time aligned the two mains and one subwoofer. Then I went back and regenerated all the PEQs and having been listening and comparing my changes to other profiles stored in the MiniDSP. Going now to listen for a week or so before making any changes. Ideas for a next step?

 

[JStewart]

Appreciate the tips. I having been doing my research and landed at the link you provided and have since, time aligned the two mains and one subwoofer. Then I went back and regenerated all the PEQs and having been listening and comparing my changes to other profiles stored in the MiniDSP. Going now to listen for a week or so before making any changes. Ideas for a next step?

Ummm, enjoy the music?
Glad you got it sorted!

 

[sm52]

time aligned the two mains and one subwoofer.

If you post an mdat file with measurements of three speakers before correction and measurements with the speakers corrected, we can evaluate them and give some advice.

 

[Calypso]

If you post an mdat file with measurements of three speakers before correction and measurements with the speakers corrected, we can evaluate them and give some advice.

Will do!

 

[Calypso]

@sm52 - Attached are the measurements for all 3 speakers with only the corrected time delay. Included are the 4 PEQs, 2 for left speaker and 2 for the right speaker. Two are loaded at the input stage of MiniDSP FLEX and the other 2 are loaded and applied during the matrix processing. The higher frequency PEQ (5K-20K Hz) which are minimal in effect, are applied first since both the left and right channels are routed to the sub for combining by the MiniDSP. I do have some room modes that I cannot correct with speaker placement or PEQs, so I am happy to just optimise these and enjoy the outcome. Hope this is not too confusing.

 

[sm52]

measurements for all 3 speakers

I looked at all the measurements. There is no point in making an assessment yet. Each measurement involves two speakers. I can't understand why this is so. You select output R, for example, but there are two impulses in the measurement. The same happens if output L is selected. One speaker must participate in one measurement. Second. You are not using time synchronization. Without this, it is impossible to coordinate several speakers. You should select one speaker for the reference, for example L. And the second one at the same time for the measurement, for example R. Once the reference is selected, it should be kept the same for the measurement of all the speakers you want to match. It’s even better to use a separate speaker, the fourth one, for the reference. Twitter, or having a tweeter. Further. For best results, make the full frequency range 0-24 kHz. Including for the sub.
First measurement. Output L is a reference, and it is also for measurement, you need to disconnect the input wires from the subwoofer. Second measurement. Output L is reference, output R is for measurement, the input wires must be disconnected from the subwoofer. Third measurement. Output L is reference, output R is for measurement, disconnect the wires from the right speaker, connect only the right channel to the subwoofer.

 

[Calypso]

@sm52 - Here are measurements (individual L/R/Sub) using the tweeter on the left speaker as a timing reference but the subwoofer sweep was just to 300Hz, so will re-run later. Not sure what to do to correct the IR issue. I do get a chirp sound at the beginning and end of the sweep which is contrary to the description in the MiniDSP document.

 

[sm52]

Now the measurements are correct. The screenshots show impulses from the left and right speakers. When measuring the left one, after 0.88 ms a second impulse appears, which should not exist. When measuring the right one, after 1.33 ms a second impulse appears, which should not exist. With such a small difference between the impulses, you may not notice the sound of the two speakers by ear. But it seems like there are always two speakers. Somewhere in the REW or miniDSP settings, some setting sends a measurement sweep to not one speaker, but two.
As for the relationship between the main speakers and the sub, it is previously necessary to make a subwoofer 2.9 m closer than now, or 8.7 ms earlier than now.

 

[Calypso]

Ok, I will see if I can track down the reason for the second impulse. What were the steps you used to calculate the time delay?

 

[sm52]

'Overlays' button, 'Impuls' tab. I have highlighted the time distance approximately. Because due to the double impulse, it is not clear which time point should be considered the beginning.

 

[Calypso]

@sm52 I ran the measurements but was unable to find out how to correct the second Impulse. I did note that the chirp is present prior to and after the sweep, which is not documented as such in the MiniDSP document. I also calculated the time delay, entered it into MiniDSP and re-ran the measurements.

'Overlays' button, 'Impuls' tab. I have highlighted the time distance approximately. Because due to the double impulse, it is not clear which time point should be considered the beginning.

I tried but could not find any way to correct this, so will have to ask in the forum. Does this mean no further work can be done? PEQs?

 

[Calypso]

'Overlays' button, 'Impuls' tab. I have highlighted the time distance approximately. Because due to the double impulse, it is not clear which time point should be considered the beginning.

Can I use the method described in the MiniDSP document (referenced above) to calculate the respective time delays?

 

[JStewart]

Can I use the method described in the MiniDSP document (referenced above) to calculate the respective time delays?

Either way gets the same answer.

A REW chirp after the sine sweep was added after the miniDSP doc was written. No worry there.
You might check the XOs or routing in the MiniDSP for the source of the 2nd impulse if not done already. You can post screenshots if you like.

 

[sm52]

Does this mean no further work can be done? PEQs?

Yes. First you need to align the time.

Can I use the method described in the MiniDSP document (referenced above) to calculate the respective time delays?

You can do this for all speakers except the subwoofer. System delay for the sub indicates the beginning of the sub impulse. For proper matching, you must use the first large peak of the subwoofer impulse. But this is approximate, close to the ideal time. Next you need to look at the phase matching of the main speakers and subwoofer. As a result of the adjustment, the matching point will move further. I indicated the approximate time because there is no certainty about the starting point.

To get even better matching, the microphone should be placed at the same distance from the main speakers. Then the sub will be matched to the central point of the listening position. Therefore, any slight movement of the listener from this point will not greatly affect the matching of all speakers.

 

[JStewart]

@sm52 , I downloaded the mdats from post 11 and 17 and I didn't see any measurements with a timing reference which is weird because the OP said he heard the chirps. Here's an example:

Are you seeing something different?

 

[sm52]

Are you seeing something different?

The measurements at post 17 use Acoustic reference.

 

[Calypso]

Here are the settings on MiniDSP and the .mdat for the recorded sweep of just the left speaker.

 

[sm52]

Try disabling the 'Play dither before sweep' option in the measurement window.
After starting the measurement for the left speaker (output L, ref output L), move close to the right speaker to hear if it play the sweep while measuring the left speaker.

 

[Calypso]

Try disabling the 'Play dither before sweep' option in the measurement window.
After starting the measurement for the left speaker (output L, ref output L), move close to the right speaker to hear if it play the sweep while measuring the left speaker.

No joy, 2nd impulse is still visible and the right speaker is dead silent during the sweep. @sm52 I attached the REW preferences settings hoping for a hint.