Dont like the sound of room correction

[freedomgli]

At low to moderate listening levels at night when casually listening while others sleep, I generally prefer no room correction. But for almost all critical listening at moderate to higher volumes, I’m loving the Dirac Live Full Frequency and the default Dirac target curve handling the entire frequency range. I do have to raise the volume 6 to 10dB compared to no room correction. But it sounds superb. There is the occasional song that is poorly recorded that sounds better without Dirac but really I should just isolate those few songs and develop a target curve/ filter that works better for them. It’s taken me about a year and many system changes and measurements to get to this point. Early on my experience with Dirac was unsettling mostly due to inexperience and user error. But now I’m really happy.

 

[watchnerd]

Because we hear differently depending on sound pressure levels, the curve needs to change for it to sound the same. A lift in the bass and treble is needed at low volumes otherwise it will sound dull. A flattening out of the curve is needed at high volumes to prevent bass and treble from becoming out of balance and overwhelming.

You're right that it would be a hassle if you ride the volume a lot, which is why Dynamic EQ is such a killer feature with Audyssey. If Dirac were ever to implement such a feature for equal loudness it could cause me to change my setup. Not that I change the volume that much, but just so it always sounds good no matter what.

FYI I currently use room correction to 350 Hz in a symmetrical room with good speakers that don't need much help above that.

I understand the reasoning.

I just can't imagine being bothered enough to deal with the hassle.

 

[nick-v]

I've used Anthem ARC, Audyssey & Dirac Live in various products.

I've had success with all of them, but none of them sound "great" upon initially running the correction. They all need to be dialed in, and I think the vast majority of people that use these products (typically not on this forum) are likely not digging in to learn about how to properly tweak their correction.

My latest Denon AVR-X3700H did not sound good after initially running Audyssey. It made my large JBL Synthesis HDI-3800 towers sound like small bookshelf speakers. After purchasing the MultEQ Editor app and tweaking it now sounds almost as good as Dirac and Anthem ARC.

 

[farcurse]

I use a DDRC-22D to run Dirac room correction (Ascend Sierra 2-EX speakers with a single Rhythmik F12-SE subwoofer crossed over at 80 Hz using a MiniDSP 2x4 HD). When I first bought the device it was running Dirac Live v1. With this old version of the software I found good results correcting up to about 300 Hz to tame room effects. However, full range equalization left the top end feeling lifeless. It's hard to describe but vocals in particular sounded distant. I recently upgraded to Dirac Live v3 and the difference was dramatic. With full range equalization I'm getting better imaging than without. Vocals in particular sound more defined and easier to understand. I think this has to do with the FIR filters managing phase. I don't think the difference was due to errors in the measurement phase of the process: I re-ran measurements several times with identical results on both versions of Dirac Live and the results were consistent. Anyway, I'm really happy with the DDRC-22D running Dirac Live v3 full range. I maintain two active filters, one up to 300 Hz and the other full range, and almost always use the full range filter.

Just my 2 cents.

 

[jhaider]

Yes, it works similar to the implementation on the RME ADI-2. If you Google Audyssey Dynamic EQ, you can find the measured curves.

I don’t think they’ve changed since I measured them.

 

[Matt S]

Because we hear differently depending on sound pressure levels, the curve needs to change for it to sound the same. A lift in the bass and treble is needed at low volumes otherwise it will sound dull. A flattening out of the curve is needed at high volumes to prevent bass and treble from becoming out of balance and overwhelming.

You're right that it would be a hassle if you ride the volume a lot, which is why Dynamic EQ is such a killer feature with Audyssey. If Dirac were ever to implement such a feature for equal loudness it could cause me to change my setup. Not that I change the volume that much, but just so it always sounds good no matter what.

FYI I currently use room correction to 350 Hz in a symmetrical room with good speakers that don't need much help above that.

Such a killer feature, all done for you, automatically. So many users misinterpret its use, too.

Those complaining that Audyssey is bass light likely have the Dyamic EQ feature switched off.

 

[Newman]

I think you nailed it on the head. My brain expects the room to sound a certain way, and when it doesnt it seems unnatural to me

the word I would use would be “restricted”. Speakers sound “restricted” with the room correction on

I really don’t think that’s it. Just for starters, you are doing sighted listening, so any unconscious cognitive bias is having its way with you. (Simple example: “don’t mess with the finely honed sound that the speaker maker developed”. But that’s just an example: there could be any number in any form.) For mains, what exactly is being corrected? I.e. how well controlled is your room? How ideal the speaker placement and listener positioning? If you like the slightly bright K series, sound, then why correct above 500 Hz? For dessert, FIR is an ongoing issue for Audyssey and Dirac, but not for Anthem and MiniDSP.

So, although our brains “expect the room to sound a certain way”, that does not mean we expect a bad room to sound bad. e.g. what if you installed top-notch treatments to that room and completely changed its sound? How do you hold to your statement then?

cheers

 

[Chromatischism]

For dessert, FIR is an ongoing issue for Audyssey and Dirac, but not for Anthem and MiniDSP.

Issue? Care to elaborate?

 

[Newman]

Too much risk of the inevitable artefacts being audible. Unlike IIR, where the phase artefacts are generally proven with psychoacoustics and audio research listening experiments to be inaudible.

 

[Chromatischism]

Too much risk of the inevitable artefacts being audible. Unlike IIR, where the phase artefacts are generally proven with psychoacoustics and audio research listening experiments to be inaudible.

Do the artifacts show up in the FR?

 

[SpoOokY]

Such a killer feature, all done for you, automatically. So many users misinterpret its use, too.

Those complaining that Audyssey is bass light likely have the Dyamic EQ feature switched off.

That is not the issue with DynEQ. The main issue is that it is lifting surround speaker volumes too based on set volume. This completely destroys the balance between the fronts and rears. A dynamic bass increase is completely okay, but this is a no-go!

 

[Newman]

Do the artifacts show up in the FR?

Yes

 

[Matt S]

That is not the issue with DynEQ. The main issue is that it is lifting surround speaker volumes too based on set volume. This completely destroys the balance between the fronts and rears. A dynamic bass increase is completely okay, but this is a no-go!

Does it? Must admit to be using for 2.1 at the moment where its been generally very impressive.

 

[Chromatischism]

Yeah. Some of us anticipate it by reducing the surround levels, but it doesn't work for everyone.

 

[Senior NEET Engineer]

I don't think I can agree with that.

My friend has nice speakers that have a tremendous amount of sibilance with their stock sound.

Dirac with Harman curve applied from 20Hz to 20KHz cleaned that right up. And sibilance is in the many thousands of Hz.

By default his speakers sloped upwards from 1000 to 20,000 by about 5dB. After Dirac it sloped down by about 3dB from 1000 to 20,000 and that made a huge positive difference in the sound.

I almost couldn't listen to them before, and now they sound like some of the best I've heard.

"Never" was too extreme. I think it is a bad approach if you're trying to optimize sound quality, due to the way reflections are received in small rooms. But for home theater it can be good enough.

 

[SpoOokY]

Does it? Must admit to be using for 2.1 at the moment where its been generally very impressive.

It does pretty obviously. Best audible in games. You are facing a sound source, turn around and gradually this sound source, even though distance remains the same, becomes significantly louder. Logically, you cannot compensate this behaviour with a reduction of the surround speaker volume as then rears would be too quite when you turn the general volume down.
After I noticed that odd behaviour, I had to turn DynEQ off and fiddle around with MultiEQ app to get some kind of house curve. Without it Audyssey just sounds thin and lifeless.
Dirac does not have such a volume dependant bass compensation but actually I am not missing it.

 

[Matt S]

It does pretty obviously. Best audible in games. You are facing a sound source, turn around and gradually this sound source, even though distance remains the same, becomes significantly louder. Logically, you cannot compensate this behaviour with a reduction of the surround speaker volume as then rears would be too quite when you turn the general volume down.
After I noticed that odd behaviour, I had to turn DynEQ off and fiddle around with MultiEQ app to get some kind of house curve. Without it Audyssey just sounds thin and lifeless.
Dirac does not have such a volume dependant bass compensation but actually I am not missing it.

Clearly a design flaw. Surely they would have recieved this kind of feedback from users and addressed it? Seemingly not.

 

[Absolute]

I think every such topic deserve a bit of nuances to avoid inaccurate "truths". I was quick earlier to state that both Audiolense and Dirac degrade the sound quality in my setup above a few hundred hertz, but I did so without giving some context.

First of all, both Audiolense and Dirac have frequency dependent windows when they analyze the sound to avoid over-correction above the modal region. In Dirac's case this window depends on your choice of wide, semi-wide or narrow sweet-spot mode and how you measure/how many measurement points you use. One-point measurement in Dirac will result in a very nice graph but a real shit-show sound.
Audiolense is more advanced in the sense that it's entirely user-configurable how and where you want this frequency dependent window to work and results will vary based on that (and other settings).

Secondly, I have very flat speakers (JBL M2) with decent off-axis dispersion and acoustical treatment both on the ceiling and behind the listening position. The biggest bumps in the response above 800 hz comes from reflections from the couch and microphone stand, and that makes no sense to correct at the cost of direct sound accuracy.
If you have speakers with bigger direct-sound compromises you might end up in a situation where the room correction actually improves the direct sound, and then it's a different proposition entirely.

I think that the first win of any room correction software is that it requires users to buy a microphone and actually measure. The second win is that it forces users to fiddle a little with response curves to get it to sound ok. The third win is that it removes the focus from stuff that don't matter and moves it to stuff that do matter.

For this reason I truly believe that trying out room correction software will, at least in the long run, improve the sound of most users.

 

[SpoOokY]

Clearly a design flaw. Surely they would have recieved this kind of feedback from users and addressed it? Seemingly not.

Nope, it works as intended. Only the intention is a pretty bad one...

Edit: https://audyssey.zendesk.com/hc/en-...namic-eq-makes-surround-speakers-far-too-loud

 

[zzz2496]

@krizvi786 how exactly did you do the measurement? Is the measurement mic OK? What is your room/seating situation? I doubt it has anything to do with the DRC algorithm, but more to the room it self, or to the method of measurement...