Imaging issue leaning left

[Longrail]

Hey,

Have been trying to deal with a stereo imaging issue leaning left for just over a month now, countless hours and think I'm losing the plot, I am not getting anywhere, really hoping for some advice, full kit list below and pics but the issue is just the stereo pair. Have been extensively testing mono content, a series of sound effects and some music in Adobe Audition, I am using Dirac Live Bass Control but this issue happens with it off, the whole image leans left and higher frequencies around 10khz area can lean further left.

1. Swap L/R speakers
2. Toe in and out
3. Moving my head forwards and backwards
4. Increasing trim level for Front Right speaker, takes between1-5-3db to get dialogue content centered, this negatively affects other content.
5. I do hear a similar issue in headphones which suggests it is my hearing but it's not as bad with headphones.
6. Higher frequencies, around 7khz+ lean more noticeably left and particularly around 10khz even further left
7. EQ'ing wih DEQ2496 can help some of the content with higher frequencies, but then pulls other content too much to the right.
8. If I turn around 180 degrees it leans the opposite way
9. MDAT file link here if anyone can take a look, this is after Dirac (same issue with it off), fronts are limited to 280hz correction, I have tried matching the stereo pair frequency response very tightly with the DEQ2496 dsp but this does not fix the issue.
10. Speakers are placed equally between side walls, and my MLP is dead in the middle of both of them.

• KIT LIST:
• 7 x Adam A8X
• 4 x Adam A4V (heights)
• 2 x SVS SB3000
• Arcam AV41 Processor
• Behringer DEQ2496 DSP (for the front stereo pair issues)
• GIK Acoustics 244 4" broadband absorbers all around
• 2 x GIK Acoutics Turbo Trap
• Rockwool behind the front stage, about 1.1meteres of it behind Front Left, about 40cm behind Front Right, this is due to how the loft is built.

Front Left/Right REW response after running Dirac, fronts limited to 280hz correction

Time alignment:

ETC

Group Delay:

RT60:

 

[Keith_W]

Wow, your listening room reminds me of this:

Anyway your measurements seem to indicate that everything is in order, but there might be a couple of other things you haven't looked at. Post your MDAT, you will need to zip it first. If the measurements look good, the remaining possibility is that there is something wrong with your hearing. Sometimes wax build-up can cause this kind of problem.

 

[Longrail]

Haha, yeah does look a bit crazy when I look back at pictures of it. I did link the mdat file in the "here" word but I've zipped it and attached it to this post. I did actually book in a free hearing test today locally and they looked in my ears and said they are so full of wax they cannot see the ear drum. So before I can get a hearing test done I'm now looking to get a ear wax removal process done first hopefully in the next week or early in January.

 

[Keith_W]

I had a quick look. Your left/right measurements are absolutely identical.

In fact, your IACC's (Interaural Cross-Correlation) comparing left and right is >0.8 across the board for "early", it would be the envy of most ASR members, including me. That is pretty remarkable.

About the only flaw in your system is the very poorly aligned subwoofer (sub = blue, L speaker = red). It is about 30ms too late.

So yeah, I think the problem is your earwax. Good luck with it!

 

[gwing]

Ear wax does seem a favourite possibility. You might also want to listen to someone else's system, or get them to listen to yours, to check if they can hear the same leaning effect. Ask them if it leans, and if so to which side rather than suggest an answer though

 

[Longrail]

I had a quick look. Your left/right measurements are absolutely identical.

View attachment 497495

In fact, your IACC's (Interaural Cross-Correlation) comparing left and right is >0.8 across the board for "early", it would be the envy of most ASR members, including me. That is pretty remarkable.

View attachment 497494

About the only flaw in your system is the very poorly aligned subwoofer (sub = blue, L speaker = red). It is about 30ms too late.

So yeah, I think the problem is your earwax. Good luck with it!

Thank you for taking the time to look, I did not realise they were that well aligned, I only ever looked at the Impulse Response in REW. For the subwoofers I have 2 of them and Dirac always seems to set the delay of SUB1 around 0.29ms to 0.40ms, and keep SUB2 at 0.00. In a previous calibration with Dirac it did keep SUB1 at 0.00 and set SUB2 0.04ms.

With the poorly aligned sub, current sub delays is like below, would you advise to set the SUB1 back to 0.00 ?, Channel 13 = SUB1, Channel 14 = SUB2, I did try this the other day and it had little to no impact on frequency response and the way they sounded.

 

[Keith_W]

If a sub is delayed, all the other speakers need to be delayed to match the sub. This is because you can not move a subwoofer "forward in time" if you are using minimum-phase DSP. You did not indicate what DSP settings you were using when you took your measurement. IF it was taken with the subs already delayed, then it would be correct to return it to zero.

Anyway, I would use the zero crossing point after the initial energy build-up as the start of the subwoofer response. My quick eyeball in the previous post was wrong, it's closer to 24ms than 30ms You took measurements of L+sub, R+sub, and sub ... so I can't sim it for you. I would suggest aligning the start of the sub to the start of the main speakers, then shift the sub +/- 5ms until you see the flattest curve possible when the response of the sub and main speakers are summed. You can use REW's alignment tool. But since you are using Dirac ... well, I don't use Dirac so I can't tell you what to do there, sorry.

 

[kyuu]

To me, this response looks a bit strange:

For one, the bass keeps climbing down to ~15Hz. It's 10dB hotter than the response at 200Hz, seems like too much bass boost. Then, even more strangely, instead of naturally falling off a dB or two per decade, the response very gradually climbs from 200Hz to 10kHz and is 2-3dB hotter at 10kHz than at 200Hz. I don't think this explains the OP's problem, per se, it just doesn't look like a normal in-room measurement to me.

 

[Longrail]

If a sub is delayed, all the other speakers need to be delayed to match the sub. This is because you can not move a subwoofer "forward in time" if you are using minimum-phase DSP. You did not indicate what DSP settings you were using when you took your measurement. IF it was taken with the subs already delayed, then it would be correct to return it to zero.

View attachment 497514

Anyway, I would use the zero crossing point after the initial energy build-up as the start of the subwoofer response. My quick eyeball in the previous post was wrong, it's closer to 24ms than 30ms You took measurements of L+sub, R+sub, and sub ... so I can't sim it for you. I would suggest aligning the start of the sub to the start of the main speakers, then shift the sub +/- 5ms until you see the flattest curve possible when the response of the sub and main speakers are summed. You can use REW's alignment tool. But since you are using Dirac ... well, I don't use Dirac so I can't tell you what to do there, sorry.

I don't do anything special so I keep all speakers at defaults in the Arcam unit no adjustments at all (I do set the speaker types to LARGE so I can measure each speaker individually with REW), I then put the UMIK1 in MLP location, set the AV41 to Volume "75", play REW Pink Noise and get the speakers measuring the same, the subs are always 10db hotter, I then take a load of "PRE" measurements, then start the Dirac app and do a 7point measurement.

Here is the MDAT file I took before handing over to Dirac, it has all the separate measures, it's too big to attach here so I've linked it HERE

To me, this response looks a bit strange:

View attachment 497515

For one, the bass keeps climbing down to ~15Hz. It's 10dB hotter than the response at 200Hz, seems like too much bass boost. Then, even more strangely, instead of naturally falling off a dB or two per decade, the response very gradually climbs from 200Hz to 10kHz and is 2-3dB hotter at 10kHz than at 200Hz. I don't think this explains the OP's problem, per se, it just doesn't look like a normal in-room measurement to me.

So I do a 10db boost at the low end around 16.5hz at the moment, I like it for the most part but sometimes find it a bit much, below is the SUB curve and you can see that against the measured response, I don't mind a bit more treble, the Arcam unit has a global tone control for the Treble which tilts the response from around 2khz>, I didn't really like what it did when tilting down.

And the Fronts are like below:

 

[kyuu]

Well the bass boost is more of a subjective thing, of course. My Dirac curve has a 6dB boost and sometimes I think that may be too much, been considering generating a new filter using a 4 or 5dB boost instead.

The increasing response from 200Hz on is the more unusual bit. You shouldn't need a tone control or DSP for the response to fall off gradually, that should happen naturally. Your room is certainly very far from a normal room, but even given that it's quite unusual for the energy to increase as you go up in frequency. Although perhaps it's less that there's increasing energy and more that the somewhat messy response from ~300Hz to ~1.5kHz is just making it look that way and it's really just flat. I still would expect that to sound too bright.

 

[Longrail]

Well the bass boost is more of a subjective thing, of course. My Dirac curve has a 6dB boost and sometimes I think that may be too much, been considering generating a new filter using a 4 or 5dB boost instead.

The increasing response from 200Hz on is the more unusual bit. You shouldn't need a tone control or DSP for the response to fall off gradually, that should happen naturally. Your room is certainly very far from a normal room, but even given that it's quite unusual for the energy to increase as you go up in frequency. Although perhaps it's less that there's increasing energy and more that the somewhat messy response from ~300Hz to ~1.5kHz is just making it look that way and it's really just flat. I still would expect that to sound too bright.

I agree is it subjective, I also do agree 10db is a lot, I know it's a lot, for whatever reason most of the time in this room it doesn't sound overly powerful to me, except times like when watching the Blade Runner 2049 intro or scenes in Dune, then even +6db sounds sounds like a bit too much.

I do agree the 200hz+ sloping upwards is odd as it's normally the other way, I like the natural sound of the speakers in this room, whenever I have done full range correction it sounds like it takes the "life" out of the speakers and sounds quite flat/boring to me so I stop correction at 280hz currently. After my ears are sorted out I could give full range correction another go I guess.

 

[Cbdb2]

If you listen with headphones do you hear the same thing? If you swap headphone L and R does it switch sides (than its the gear) or stay on the same side (than its the ears)?

 

[Longrail]

If you listen with headphones do you hear the same thing? If you swap headphone L and R does it switch sides (than its the gear) or stay on the same side (than its the ears)?

Yep so I hear mostly the same thing with headphones using mono content leaning to the left, it's not as bad but it's pretty much the same, and if I turn the headphones around then the issue stays the same and still leans to the left.

 

[SteveC]

I've had a small hearing difference in my ears for many years. I have to set the balance control so that the right speaker is just a little louder. Needless to say a preamp or amp without a balance control is a non-starter for me. It's very easy for me to hear when it's set correctly (or as close as I can get it).

 

[dualazmak]

My post here would be of your reference and interest, I assume.
I wrote there "If you would be seriously interested in using the test tone signal tracks I prepared and applied in these my measurements and tunings, please simply PM me writing your wish."

 

[kyuu]

I do agree the 200hz+ sloping upwards is odd as it's normally the other way, I like the natural sound of the speakers in this room, whenever I have done full range correction it sounds like it takes the "life" out of the speakers and sounds quite flat/boring to me so I stop correction at 280hz currently. After my ears are sorted out I could give full range correction another go I guess.

To be clear I am absolutely not advocating for full-range correction. I'm quite the advocate for generally avoiding room-correction above Schroeder (280Hz is a good-enough guess).

 

[Longrail]

I played with the main speaker delays to the sub, changing Left from the current 4.85ms to 23.85ms and got them aligned like below (28.85 was the correct value but created a steep null at 100hz so backed off to 23.85), I only did a short song test and tried to A/B between the two millisecond changes but could not say I noticed much difference.

Teal = Front Left
Red = LFE

And Front Left before/after
Red = Before
Teal = After

 

[10PSPMemSticksForMiku2007]

Try thoroughly cleaning out your ears. I've experienced something similar. I have to clean mine regularly because I use foam earplugs so much.

 

[kawauso]

I played with the main speaker delays to the sub, changing Left from the current 4.85ms to 23.85ms and got them aligned like below (28.85 was the correct value but created a steep null at 100hz so backed off to 23.85), I only did a short song test and tried to A/B between the two millisecond changes but could not say I noticed much difference.

Teal = Front Left
Red = LFE

View attachment 497600

And Front Left before/after
Red = Before
Teal = After

View attachment 497601

If the subwoofers appear to be delayed after DLBC calibration, please don’t worry about it. That offset may be intentional. It’s either not audible at all, or, if it is, it’s likely a trade-off made to improve other parameters.

 

[Longrail]

Had my ears cleaned, good bunch of earwax removed, I still need to book in a hearing test separately, I believe it's helped with the 10khz area, however overall the image is still leaning left, I've run out of ideas on what to try with this.

1. Moved speakers about 1ft forward
2. Moved speakers further out left and right
3. Moved speakers as far back as they can go
4. Extreme toe-in and out
5. Tried speakers in a different room using an Audient iD14 interface
6. Tried a pair of small Genelec desktop speakers in the other room with the Audient interface
7. Tried speakers on their side, both with tweeters on the inside and outside
8. Tried intentionally delaying the Front Right speaker then running Dirac, Dirac just brings it back to the way it was before.
9. Tried moving an acoustic panel around me to see what that does, not much.

It seems certain sounds that have an "airy" ness to them seem to be centred while dialogue and various other content is not.