Room Correction

[ejspain]

This question is probably out there but I can't find anything.

Talking mostly 2 channel listening...

Is room correction (Audyssey, Dirac, ARC Genesis, MultEQ, and others) really necessary for electrostats?

Trying to think super simple here...from the older stats to current, ML has been crossing over the panels between 300 - 450hz since the 90s - where you probably don't want correction in that region anyways???

Depending on your model and listening taste, adding a sub, or 6 was "almost" necessary. A lot of the room correction softwares are great BUT since the stats are so freakin accurate; even the included sofware; for the Masterpiece Series) doesn't touch the panels, should we bother allowing other RC software to attempt making adjustments? Maybe exclude your LCR (if its all electrostats) and just allow the SW to set distance/levels for the subs, surrounds/atmos?

Thoughts?

 

[Robert D]

This question is probably out there but I can't find anything.

Talking mostly 2 channel listening...

Is room correction (Audyssey, Dirac, ARC Genesis, MultEQ, and others) really necessary for electrostats?

Trying to think super simple here...from the older stats to current, ML has been crossing over the panels between 300 - 450hz since the 90s - where you probably don't want correction in that region anyways???

Depending on your model and listening taste, adding a sub, or 6 was "almost" necessary. A lot of the room correction softwares are great BUT since the stats are so freakin accurate; even the included sofware doesn't touch the panels, should we bother allowing other RC software to attempt making adjustments? Maybe exclude your LCR (if its all electrostats) and just allow the SW to set distance/levels for the subs, surrounds/atmos?

Thoughts?

Personally, I run Audyssey on everything. All 11 speakers including my electrostatic fronts. When I listen to 2 channel for music, I don't notice a very big difference between having Audyssey on or off. I do feel that the "stage" is more realistic with Audyssey on. The music sounds more live. I like that, so I leave Audyssey on. Most receivers and preamps allow you to run the music "pure", with no DSP applied. So you can watch movies with it on and then off for 2 channel music. I see no reason not to run it, especially if you use the system for more than stereo music.

 

[ttocs]

Remember, it's Room Correction.

Room Correction apps attempt to alleviate issues caused by room/speaker interactions. If our speakers were just in anechoic chambers then we could just use them as is and be happy with the +-3dB spec.

I was never happy with Dirac correcting above about 500Hz until sometime last year when Dirac changed their app for the better with respect to how it measured stats.

In my case, Dirac helps to tame the upper frequencies which otherwise are just a bit higher than I like, and these days does a very good job.

The question of what to allow to be corrected, it's something that's easy to do with Dirac. I don't know about other apps.

 

[Robert D]

Remember, it's Room Correction.

Room Correction apps attempt to alleviate issues caused by room/speaker interactions. If our speakers were just in anechoic chambers then we could just use them as is and be happy with the +-3dB spec.

I was never happy with Dirac correcting above about 500Hz until sometime last year when Dirac changed their app for the better with respect to how it measured stats.

In my case, Dirac helps to tame the upper frequencies which otherwise are just a bit higher than I like, and these days does a very good job.

The question of what to allow to be corrected, it's something that's easy to do with Dirac. I don't know about other apps.

And, I'm sure you can turn off DIRAC with the push of a button too? It's nice, allows you to compare music with and without quickly.

 

[Brandon Hartwick]

To answer the original question, no room correction isn't necessary, it's just another tool at your disposal. Another way to go about things is to get a mic and run some sweeps with Room EQ Wizard so you can actually see what's happening.

This is a sweep I ran in my room last weekend with zero digital room correction or DSP. It needs some work judging by the graph, but it was better than I expected.

 

[ttocs]

This is a sweep I ran in my room last weekend with zero digital room correction or DSP. It needs some work judging by the graph, but it was better than I expected.
View attachment 23911

I'd love to know how you're getting over 105dB from a single speaker! And what speaker is it?

When my system is at a volume setting that makes movies peak at 107dB the volume is set at -18dB, and as shown below, the Left Front 13A (no subs for this measurement) measures a little over 80dB by itself. I upped the volume to -12dB for a comparison and to show something else, the woofers start to sag below about 40Hz at this volume setting which, if the entire system were playing a movie, would be peaking at 113dB SPL. My room is around 6000 cubic feet, so this is why I use a bunch of subwoofers, to get some good output without each one straining to do so. It's also why I've been using multiple subs with the 13A's for music during the last couple years.

I made the scale approximately the same as your example as a comparison. No Dirac.

This next plot is a comparison of the same traces like in the plot above, but with a different scale so it shows more variation, and the traces are now dashed lines. (I use this scale specifically because my noise floor is right at about 45dB around 25Hz.) Plus, I added the third trace to show what Dirac does. That range from just under 2000Hz to just over 4000Hz is very noticeable when listening to music, and I enjoy the sound much better when it's reduced as can be seen by the solid trace.

If the entire front wall of the house, the right wall in this room, wasn't all glass windows from floor to ceiling I'd be able to use more treatment so things would be better before adding room correction to the equation.

I've shown on another forum great looking response curves that sound awful. But the Dirac curve above sounds great. I point this out because it should be more common knowledge, but I don't think it is. There's more that goes into great sound than good "looking" response graphs. Just like some improvements cannot be measured, great measurements don't necessarily sound good.

 

[Brandon Hartwick]

I'd love to know how you're getting over 105dB from a single speaker! And what speaker is it?

This is the response curve from both speakers using REW at the listening position, the speakers are Aeries I's with active crossovers.

 

[ejspain]

I'd love to know how you're getting over 105dB from a single speaker! And what speaker is it?

When my system is at a volume setting that makes movies peak at 107dB the volume is set at -18dB, and as shown below, the Left Front 13A (no subs for this measurement) measures a little over 80dB by itself. I upped the volume to -12dB for a comparison and to show something else, the woofers start to sag below about 40Hz at this volume setting which, if the entire system were playing a movie, would be peaking at 113dB SPL. My room is around 6000 cubic feet, so this is why I use a bunch of subwoofers, to get some good output without each one straining to do so. It's also why I've been using multiple subs with the 13A's for music during the last couple years.

I made the scale approximately the same as your example as a comparison. No Dirac.
View attachment 23912


This next plot is a comparison of the same traces like in the plot above, but with a different scale so it shows more variation, and the traces are now dashed lines. (I use this scale specifically because my noise floor is right at about 45dB around 25Hz.) Plus, I added the third trace to show what Dirac does. That range from just under 2000Hz to just over 4000Hz is very noticeable when listening to music, and I enjoy the sound much better when it's reduced as can be seen by the solid trace.
View attachment 23913

If the entire front wall of the house, the right wall in this room, wasn't all glass windows from floor to ceiling I'd be able to use more treatment so things would be better before adding room correction to the equation.

I've shown on another forum great looking response curves that sound awful. But the Dirac curve above sounds great. I point this out because it should be more common knowledge, but I don't think it is. There's more that goes into great sound than good "looking" response graphs. Just like some improvements cannot be measured, great measurements don't necessarily sound good.

This is exactly my point. I've been beating my head against the wall assuming this very analysis. Why doesn't my music/movies sound phenomenal if I have good measurements? My ears also tend to like the opposite of what the measurements say. Glad I'm not crazy. Thanks @ttocs

 

[JonFo]

The question of what to allow to be corrected, it's something that's easy to do with Dirac. I don't know about other apps.

The latest Audyssey MultEQ-X pro-level editor (a Windows PC app) allows windowing (they call it 'curtains') to be set at either end of the scale. Such that I can limit corrections to 2K and below or from 80Hz on up on a speaker set to 'small'.

It also allows manual parametric EQ settings and even importing PEQ sets generated by REW. It seems pretty flexible.

 

[JonFo]

Why doesn't my music/movies sound phenomenal if I have good measurements?

Because frequency response is not the only measure we should be looking at, the temporal aspects are as, if not more, critical. And dipole panel speakers engage many temporal modes in a room. This is why I keep harping on reflection management so much.

If you look at your REW measurements using the Waterfall plot, you will see some ragged peaks and valleys up and down the frequency range, with the lower-end having longer decays (the mountain goes further forward), but notice the mess in the high-end as well.

Minimizing those decays is generally a good thing. The most effective is absorption, but it can also be improved with some placement tweaks.

The other metric to look at is the group-delay graph. This one is harder to use effectively, but is a great way to see if speakers are phase-aligned correctly. As you know, even an active crossover will skew phase (timing) at the crossover points, so getting a good in-room blend is important. Same for subs aligning to the mains, the GD graph will help you iterate down to the ms for the smoothest GD curves. I spent an hour on that iteration getting my MBM;s aligned to the primary sub. It paid off.

The bottom line, one can have a rough-looking frequency response curve but have great in-room performance if the temporal behaviors are mitigated.

 

[ttocs]

This is exactly my point. I've been beating my head against the wall assuming this very analysis. Why doesn't my music/movies sound phenomenal if I have good measurements? My ears also tend to like the opposite of what the measurements say. Glad I'm not crazy. Thanks @ttocs

This is another example.

I have tried to use MSO (Multi Sub Optimizer) to see how it works out. It's kinda a steep learning curve, but mainly because it's difficult to retain the liveliness while smoothing the response. I found that if I severely limited what MSO could do, it sounded better, but it wasn't as smooth.

The Blue trace is only MSO, the Red trace is only using a REW filter. The REW version sounds great! The beautiful looking MSO version sounds absolutely terrible and lifeless! No smoothing. LFE channel in REW.

As I've posted in another forum, be careful what you wish for.

 

[JonFo]

No smoothing. LFE channel in REW.

Scott, would you mind posting the Watrwefall plots for each of those measurements?

That would show the variance in decays.

 

[ejspain]

So who's in the Dallas, TX area or flying in to help this schmo get it right. I have a very basic understanding of the tools and how to read them but I don't know what I don't know. Really would help if someone who knows exactly what to do, could sit in my room and school me. I'll supply all the beer, whiskey, wine, grub...even room and board. Any takers??

 

[Robert D]

This is exactly my point. I've been beating my head against the wall assuming this very analysis. Why doesn't my music/movies sound phenomenal if I have good measurements? My ears also tend to like the opposite of what the measurements say. Glad I'm not crazy. Thanks @ttocs

After running Audyssey, I always end up turning up the volume knob of my sub a little bit and raise the crossover point a bit on my left/right mains from 40hz up to 60 or 80hz. The bass sounds better. I think on my system 60 Hz is the setting.

 

[ejspain]

After running Audyssey, I always end up turning up the volume knob of my sub a little bit and raise the crossover point a bit on my left/right mains from 40hz up to 60 or 80hz. The bass sounds better. I think on my system 60 Hz is the setting.

In my experience...which isn't saying much...Audyssey is great at presenting pretty curves but the worse at correcting bass response...or is it operator error? As a matter of fact, is Audyssey suppose to correct bass response or just flatten as much as possible based on room anomalies? IDK.

Not exactly sure how Audyssey determines crossovers but it'd be nice if it could detect speaker impedance from a tone and create a more accurate crossover or am I asking for too much? Anyhoo not a deal breaker...it's pretty simple to change. Can't make this stuff too easy or no one would buy it. Funny how the more you spend in this hobby, the more difficult it gets but I'm having a ball...I think.

 

[ttocs]

Scott, would you mind posting the Watrwefall plots for each of those measurements?

That would show the variance in decays.

'Ya mean, like this?

Below 45dB is the noise floor of my non-dedicated room. For the Waterfall plot I used REW default settings.

The MSO Waterfall looks better to me. However, Group Delay looks better with the REW EQ filter. Spectrum is a mixed bag, with the Peak Energy line being much better with the REW EQ filter.

 

[JonFo]

The MSO Waterfall looks better to me. However, Group Delay looks better with the REW EQ filter. Spectrum is a mixed bag, with the Peak Energy line being much better with the REW EQ filter.

Thanks, these are the ones.

Eyeballing these, the MSO looks better in most respects; the only issue with MSO seems it allowed a good bit of very low-end energy to remain. And your room clearly has a room mode in the 50Hz to 70Hz region, and MSO seems to tame it as best it can, but that one needs room treatments to mitigate. Large bass traps (such as RealTraps MondoTraps) would be my suggestion.

I am impressed with the MSO results, very smooth FR and time decay overall, that one should be the most accurate filter set. However, I can possibly see why you have a preference for the REW filter one, as it has that bump at 70 -80Hz, with much more retained energy (the 'mountain' extends past 300ms). The Spectrum plot for the REW filter shows that bump much clearer; it has a significant time shift, which is also reflected in the GD plot.
In general, we all prefer a bit more energy in the 50 to 80Hz range, as that's where the 'chest thump' is, and my guess is that's why you prefer the REW results. However, that time skew and significant retained energy (slow decay) means the bass is not well resolved in that region.

I'd look at the MSO filter sets and see what it's applying around that region and copy that into your REW filter and check it out and see if the retained energy is mitigated. But it will not sound as punchy.

If you want punch, consider redeploying a vented sub to a nearfield position and set up a bandpass filter from 45 to 120Hz for it. Then you can tune the amount of punch by adjusting the gain on that one sub. See the spectrogram of my MBM setup (back when I was running pretty hot, it's only up 2dB now, and the delay spike at 60Hz is the built-in fridge upstairs running full tilt)

More later, got to have breakfast with the spousal unit ...

 

[ejspain]

Thanks, these are the ones.

Eyeballing these, the MSO looks better in most respects; the only issue with MSO seems it allowed a good bit of very low-end energy to remain. And your room clearly has a room mode in the 50Hz to 70Hz region, and MSO seems to tame it as best it can, but that one needs room treatments to mitigate. Large bass traps (such as RealTraps MondoTraps) would be my suggestion.

I am impressed with the MSO results, very smooth FR and time decay overall, that one should be the most accurate filter set. However, I can possibly see why you have a preference for the REW filter one, as it has that bump at 70 -80Hz, with much more retained energy (the 'mountain' extends past 300ms). The Spectrum plot for the REW filter shows that bump much clearer; it has a significant time shift, which is also reflected in the GD plot.
In general, we all prefer a bit more energy in the 50 to 80Hz range, as that's where the 'chest thump' is, and my guess is that's why you prefer the REW results. However, that time skew and significant retained energy (slow decay) means the bass is not well resolved in that region.

I'd look at the MSO filter sets and see what it's applying around that region and copy that into your REW filter and check it out and see if the retained energy is mitigated. But it will not sound as punchy.

If you want punch, consider redeploying a vented sub to a nearfield position and set up a bandpass filter from 45 to 120Hz for it. Then you can tune the amount of punch by adjusting the gain on that one sub. See the spectrogram of my MBM setup (back when I was running pretty hot, it's only up 2dB now, and the delay spike at 60Hz is the built-in fridge upstairs running full tilt)

More later, got to have breakfast with the spousal unit ...

Man I wish you could come to Dallas and help me apply all that knowledge in my room.

 

[Fidji99]

Man I wish you could come to Dallas and help me apply all that knowledge in my room.

It is beyond Audyssey capabilities. And this is one of the reasons, why it more often than not sounds bad after DRC. Audyssey is notorioulsy bad above 300Hz, you need either Dirac or even something even more sophisticated like Trinnov to make it sound really good.

If you would use PC as a source, there is big choice of SW tools available, e.g. Focus FIdelity.

 

[ttocs]

Eyeballing these, the MSO looks better in most respects; the only issue with MSO seems it allowed a good bit of very low-end energy to remain. And your room clearly has a room mode in the 50Hz to 70Hz region, and MSO seems to tame it as best it can, but that one needs room treatments to mitigate. Large bass traps (such as RealTraps MondoTraps) would be my suggestion.

I am impressed with the MSO results, very smooth FR and time decay overall, that one should be the most accurate filter set. However, I can possibly see why you have a preference for the REW filter one, as it has that bump at 70 -80Hz, with much more retained energy (the 'mountain' extends past 300ms). The Spectrum plot for the REW filter shows that bump much clearer; it has a significant time shift, which is also reflected in the GD plot.

The big issue with the MSO filter is the bass has no dynamics left. The bass is blurry, non-distinct, lifeless, and not just in one small range, all over.

My subs are setup differently now, so I'll do more comparisons soon.