Did you tried the theory, that flat power response actually is not that flat - it is rising?
I have curves of multiple measurement mics, and they are NOT the same. Somehow think, that it is exactly the case.
Also there may be the measurement error or very very unlucky baffle dimensions combination or similar: you have more SPL from tweeter off-axis on 2 separate regions: 3-9kHz, 11-20kHz on 10-20deg listening angle. I never observed such behavior. For my ears exactly the sweet spot is ~15-30 degrees.
Measurements done with Umik1 from MiniDSP. I would much rather have dedicated Mic with soundcard for initial measurements. Still convincing myself I need to buy it.
Baffle is far from optimal. But fixed constraint in this build. Grill cloth is mandatory. But measurements taken without cloth.
I tried reversing top bafle, so coax is closer to top. But that was not better.
Some ripple unavoidable.
With DSP it is possible to fix top octave to completely flat. I tried it and did not like it. Also I prefer the sound without to many corrections.
@PKAudio: Not my intention to hijack the thread. You measurements and review of sound convinced me to try the coax myself.
Measurements at 0 degree no EQ:
0 degree:
0 degree with EQ
20 degree with EQ:
Baffle is far from optimal. But fixed constraint in this build. Grill cloth is mandatory. But measurements taken without cloth.
I tried reversing top bafle, so coax is closer to top. But that was not better.
Some ripple unavoidable.
With DSP it is possible to fix top octave to completely flat. I tried it and did not like it. Also I prefer the sound without to many corrections.
@PKAudio: Not my intention to hijack the thread. You measurements and review of sound convinced me to try the coax myself.
Measurements at 0 degree no EQ:
0 degree:
0 degree with EQ
20 degree with EQ:
Ok,
From beginning:
I did not noticed, that your sim is done at #180 at 20 degrees, that explains some ripples. #182 looks even better.
The biggest influence for high freq ripples (uneven directivity) is the area very close to the tweeter. As this is coax - there are not much you can influence. At least physics say that. Everything looks more or less ok now.
I have 2 calibrated mics and a data from 3rd calibrated mic compared to other 2, so I can say that I have 3 calibrated mics. The measurements of them still differ by up to ~3dB in some narrow ranges, especially from 3 to 15kHz.
I want you to let the possibility, that your measurement microphone is not accurate enough and you get simulation that is textbook perfect, but may sound not good. I have measured just a few speakers, but doing everything to VituixCAD preference rating and as flat as possible is VERY close to what I like.
From beginning:
I did not noticed, that your sim is done at #180 at 20 degrees, that explains some ripples. #182 looks even better.
The biggest influence for high freq ripples (uneven directivity) is the area very close to the tweeter. As this is coax - there are not much you can influence. At least physics say that. Everything looks more or less ok now.
The more interesting part is here.
I have 2 calibrated mics and a data from 3rd calibrated mic compared to other 2, so I can say that I have 3 calibrated mics. The measurements of them still differ by up to ~3dB in some narrow ranges, especially from 3 to 15kHz.
I want you to let the possibility, that your measurement microphone is not accurate enough and you get simulation that is textbook perfect, but may sound not good. I have measured just a few speakers, but doing everything to VituixCAD preference rating and as flat as possible is VERY close to what I like.
@svp: Thanks. It definitely push me in direction, I need to up my measurement gear.
My filter is definitely on the soft side. It may be my Mic and room. Lately improved on my room acoustics,
so I have removed a lot of flutter echo. Over time this may change my preference.
But generally I would rather miss higher frequencies than too much. For example I find most speakers on Hifi shows to be too bright.
My litmus test is I want to be able to enjoy even worse/bad recordings.
I don’t have equipment to measure impedance.
My filter is definitely on the soft side. It may be my Mic and room. Lately improved on my room acoustics,
so I have removed a lot of flutter echo. Over time this may change my preference.
But generally I would rather miss higher frequencies than too much. For example I find most speakers on Hifi shows to be too bright.
My litmus test is I want to be able to enjoy even worse/bad recordings.
I don’t have equipment to measure impedance.
There is not much to up. Probably the only mistake you did is to buy usb mic. I invested into 2-channel audio interface, and now use it every day as external soundcard and headphone amp. My message is different: there MAY be incorrect response of the mic, it does not matter if it is analog or USB. You need more that one to compare results.
I doubt it... It still can be measurement/interpretation thing.
Your sim can be a little incorrect, because variable impedance does impact overall SPL curves.
For now you can try to use impedance from VituixCAD cabinet sim for woofer(s), There is export function. Use manufacturer data for tweeter as is. You probably will need to use another VituixCAD tool for that: SPL trace.
After this I think your sim will look somewhat different.
Not everywhere! But yeah, probably sibilance to 11 is a way to sell some stupid abominations like MBL's.
It mostly depends on the task of the crew who runs the equipment, but sometimes I really think, that 90% of all "audiophiles" are deaf and dumb.
Bad recordings sound worse on good equipment. Sorry...
Just saying hi in a couple of threads where I see my speakers have been mentioned. 👋🙂 I'm lurking here from time to time. I can't promise that I will (or even can) answer everything, but feel free to ask questions directly!
Thanks @sigbergaudio! Wondering if there's a reason you chose not to use the 6" coax in your designs?
@fb It was first used in our SBS.1 speaker where 5.5" was a good size for the design goals, it is also a little bit more tidy than the 6.5". Moving on the Manta it was practical to use the same driver. The 5.5" has impressive capacity, and in the Manta it is crossed high (600hz), so it worked well even with the 12" midbass. 🙂