L/R speaker measurement and PEQ

[Emiyanez]

Hi all,

I have been playing around with REW and a calibrated microphone. So far, I have measured both speakers simultaneously and the sub separately from my listening position, adjusted crossover frequency and delay, measured the whole system together, fitted the curve to Harman in REW and exported the filters from 20-200 Hz into the Wiim App PEQ. I got rid of room resonances and my system never sounded that good.

Now, I would like to understand whether I could benefit from the independent L/R PEQ in the Wiim Ultra and I wonder how I could make the best out of it. Based on what I have read, I was planning to make near field on-axis individual measurements of each speaker (simulating anechoic conditions) and try to match their curves as much as possible by applying L/R PEQ from 200 to 20000 Hz in order to compensate for speaker differences and have a better sound presentation (Erin also talked about it in his WiiM Ultra video). I will then re-do the subwoofer integration, the global measurements from listening position and the correction for room resonances.

However, I came across this comment from John Mulcahy from REW in the AVNirvana forum:

"As an aside, applying EQ above a couple of hundred Hz is a risky business that can easily do more harm than good, responses can vary a lot over short distances so filters you apply may have a very different effect than you expect just a short distance away from the measurement point. If you do apply EQ at higher frequencies best keeping the filter Q values low and typically best applying the same EQ to both channels (somewhat counter to what you are attempting) as our brains are pretty good at distinguishing the direct sound from the speaker from the contribution of the rooms' reflections (which are included in your measurement) so what you hear is often not what you see on the measurement".

Does this mean that L/R individual adjustment above hundreds Hz are not recommended? Is this true even for on-axis measurements at short distance to correct for speakers inconsistencies?

 

[TurtlePaul]

Yes to both questions. EQ above ~300-500 hz is hard and often counterproductive.

 

[NTK]

John's comment is with regard to the in-room steady state response, not the free field on-axis or listening window response.

If you have reliable anechoic (free field) on-axis or listening windows frequency response curves for your speakers, by all means use the information and EQ your speakers to flat frequency response all the way to 20 kHz.

 

[TurtlePaul]

If you have reliable anechoic (free field) on-axis or listening windows frequency response curves for your speakers, by all means use the information and EQ your speakers to flat frequency response all the way to 20 kHz.

I don’t agree. Plenty of speakers EQing them to on-axis flat makes them worse. That is the whole point of measuring directivity.

 

[NTK]

I don’t agree. Plenty of speakers EQing them to on-axis flat makes them worse. That is the whole point of measuring directivity.

It is true that you cannot correct directivity error with EQ. The only solution is better speakers.

However, if you already have the speakers, EQ'ing their on-axis frequency response will let you maximize the quality of the sound you can get out them. The on-axis FR is the FR of the direct sound, which precedes all reflected sound, and is the reference to which all subsequent sound will be compared against. It is therefore the most important sound from the speakers. The problem with directivity errors (which EQ cannot solve) is that they cause the reflected sound to have a timbre dissimilar to that of the direct sound, which our hearing often finds unnatural.

Quoting Dr Toole:

What is your favorite house curve

Just implemented this. Works great for music but may need some tuning for HT (This is the B&K curve in case you are wondering). While it looks steeper than it is, it is only about a 5dB drop from 50Hz to 20K Hz./ Depending on the high end of your speakers, that portion of the target may need...

www.audiosciencereview.com

​

...​

Third, a loudspeaker that is not flat on axis but which has well-behaved directivity as a function of frequency can benefit from in-room equalization, but in order to know that one needs comprehensive anechoic data. If one had such data, the optimum equalization (above the transition frequency) would be based on the anechoic data, not a room curve. Also, if one had the anechoic data, one should not have purchased the loudspeaker to begin with.​

...​

​

 

[kemmler3D]

It is true that you cannot correct directivity error with EQ. The only solution is better speakers.

However, if you already have the speakers, EQ'ing their on-axis frequency response will let you maximize the quality of the sound you can get out them. The on-axis FR is the FR of the direct sound, which precedes all reflected sound, and is the reference to which all subsequent sound will be compared against. It is therefore the most important sound from the speakers. The problem with directivity errors (which EQ cannot solve) is that they cause the reflected sound to have a timbre dissimilar to that of the direct sound, which our hearing often finds unnatural.

Quoting Dr Toole:

What is your favorite house curve

Just implemented this. Works great for music but may need some tuning for HT (This is the B&K curve in case you are wondering). While it looks steeper than it is, it is only about a 5dB drop from 50Hz to 20K Hz./ Depending on the high end of your speakers, that portion of the target may need...

www.audiosciencereview.com

​

...​

Third, a loudspeaker that is not flat on axis but which has well-behaved directivity as a function of frequency can benefit from in-room equalization, but in order to know that one needs comprehensive anechoic data. If one had such data, the optimum equalization (above the transition frequency) would be based on the anechoic data, not a room curve. Also, if one had the anechoic data, one should not have purchased the loudspeaker to begin with.​

...​

​

Tend to agree with @TurtlePaul that trying to EQ the speaker is much trickier. I agree it's not always 100% a bad idea, but it's much easier to cause problems inadvertently this way.

I would definitely not try to EQ the on-axis sound based on simple REW sweeps. Like you said, only use anechoic measurements for this, if you can find them or produce them yourself.

 

[TurtlePaul]

It is true that you cannot correct directivity error with EQ. The only solution is better speakers.

However, if you already have the speakers, EQ'ing their on-axis frequency response will let you maximize the quality of the sound you can get out them.

I don’t agree with that. Quite a few speakers that have directivity error with wide tweeter dispersion exhibit a dip in the low tweeter range (“BBC dip”) and/or choose to run the tweeter at an on-axis level below the woofer level. EQing those speakers up to flat only makes them sound brighter and worse.

The ‘auto-EQ’ which is posted in every speaker review focuses on optimizing the score and not solely optimizing on-axis. There are several well reviewed speakers where on-axis is clearly sacrificed to improve first reflection or power response: Revel F208 and Philharmonic Ceramic Mini, Ascend Sierra-1 v2 are three examples.

 

[Hayabusa]

Hi all,

I have been playing around with REW and a calibrated microphone. So far, I have measured both speakers simultaneously and the sub separately from my listening position, adjusted crossover frequency and delay, measured the whole system together, fitted the curve to Harman in REW and exported the filters from 20-200 Hz into the Wiim App PEQ. I got rid of room resonances and my system never sounded that good.

Now, I would like to understand whether I could benefit from the independent L/R PEQ in the Wiim Ultra and I wonder how I could make the best out of it. Based on what I have read, I was planning to make near field on-axis individual measurements of each speaker (simulating anechoic conditions) and try to match their curves as much as possible by applying L/R PEQ from 200 to 20000 Hz in order to compensate for speaker differences and have a better sound presentation (Erin also talked about it in his WiiM Ultra video). I will then re-do the subwoofer integration, the global measurements from listening position and the correction for room resonances.

However, I came across this comment from John Mulcahy from REW in the AVNirvana forum:

"As an aside, applying EQ above a couple of hundred Hz is a risky business that can easily do more harm than good, responses can vary a lot over short distances so filters you apply may have a very different effect than you expect just a short distance away from the measurement point. If you do apply EQ at higher frequencies best keeping the filter Q values low and typically best applying the same EQ to both channels (somewhat counter to what you are attempting) as our brains are pretty good at distinguishing the direct sound from the speaker from the contribution of the rooms' reflections (which are included in your measurement) so what you hear is often not what you see on the measurement".

Does this mean that L/R individual adjustment above hundreds Hz are not recommended? Is this true even for on-axis measurements at short distance to correct for speakers inconsistencies?

Playing mono white noise should be in dead center.
If not there could be a reason to have L/R EQ correcting this

 

[ernestcarl]

Hi all,

I have been playing around with REW and a calibrated microphone. So far, I have measured both speakers simultaneously and the sub separately from my listening position, adjusted crossover frequency and delay, measured the whole system together, fitted the curve to Harman in REW and exported the filters from 20-200 Hz into the Wiim App PEQ. I got rid of room resonances and my system never sounded that good.

Now, I would like to understand whether I could benefit from the independent L/R PEQ in the Wiim Ultra and I wonder how I could make the best out of it. Based on what I have read, I was planning to make near field on-axis individual measurements of each speaker (simulating anechoic conditions) and try to match their curves as much as possible by applying L/R PEQ from 200 to 20000 Hz in order to compensate for speaker differences and have a better sound presentation (Erin also talked about it in his WiiM Ultra video). I will then re-do the subwoofer integration, the global measurements from listening position and the correction for room resonances.

However, I came across this comment from John Mulcahy from REW in the AVNirvana forum:

"As an aside, applying EQ above a couple of hundred Hz is a risky business that can easily do more harm than good, responses can vary a lot over short distances so filters you apply may have a very different effect than you expect just a short distance away from the measurement point. If you do apply EQ at higher frequencies best keeping the filter Q values low and typically best applying the same EQ to both channels (somewhat counter to what you are attempting) as our brains are pretty good at distinguishing the direct sound from the speaker from the contribution of the rooms' reflections (which are included in your measurement) so what you hear is often not what you see on the measurement".

Does this mean that L/R individual adjustment above hundreds Hz are not recommended? Is this true even for on-axis measurements at short distance to correct for speakers inconsistencies?

Try taking moving microphone measurements (MMM).