[musicreo]

so, how about putting the same speaker data on the left and the right?

Can you explain what you mean by "same speaker data"?

 

[kalstein]

Can you explain what you mean by "same speaker data"?

left and right speaker.
Of course, I know that the data is bound to be different due to room acoustics.

 

[sander99]

left and right speaker.
Of course, I know that the data is bound to be different due to room acoustics.

Speaker and room information and personal hrtf information are intertangled and not seperable (at least not easy) in the HRIR measurements. And as someone mentioned your ears won't be identical (and your head won't be 100% symmetrical) so your hrtf won't be left-right symmetrical. So you can not simply use the same speaker data for left and right without messing up something regarding hrtf.

 

[musicreo]

left and right speaker.
Of course, I know that the data is bound to be different due to room acoustics.

You mean using the inverted left speaker measurement for the right ear?
I have tested it but the virtual center sounds much more "in your head".

 

[kalstein]

You mean using the inverted left speaker measurement for the right ear?
I have tested it but the virtual center sounds much more "in your head".

yes, i meant that. thanks for sharing your experience.

 

[johnn29]

Speaker characteristics are separated with room measurement though. So I believe a form of that already happens with room correction that'll then preserve HRTF.

 

[kalstein]

I changed my speaker from bose c20 to psb xb.
and recording again.
normal case is ok. but room correction case has error.
(It is a message that I have seen before.)

(venv) C:\Users\JHJ\Impulcifer>python impulcifer.py --test_signal="data/sweep-6.15s-48000Hz-32bit-2.93Hz-24000Hz.pkl" --dir_path="data/my_hrir" --specific_limit=5000 --generic_limit=2000 --room_mic_calibration="data/7063899_90deg.txt" --room_target="data/harman-in-room-loudspeaker-target.csv" --decay=300
Creating impulse response estimator...
Running room correction...
Traceback (most recent call last):
  File "impulcifer.py", line 556, in <module>
    main(**create_cli())
  File "impulcifer.py", line 60, in main
    plot=plot
  File "C:\Users\JHJ\Impulcifer\room_correction.py", line 83, in room_correction
    fr = ir.frequency_response()
  File "C:\Users\JHJ\Impulcifer\impulse_response.py", line 352, in frequency_response
    fr = FrequencyResponse(name='Frequency response', frequency=f[1::step], raw=m[1::step])
ValueError: slice step cannot be zero

and here is my record files.
https://drive.google.com/file/d/19nTS_r7eIBKulgWZ91GkHYPB_0i6AYqs/view?usp=sharing

@jaakkopasanen check it again please.

 

[jaakkopasanen]

My idea of being able to correct rooms up to 20 kHz really does seem silly now...

 

[johnn29]

Saw that video a while back - it's why I gave up on room correction above 500hz with my AVR! If you've not read it - his book is excellent too.

It's made my life a lot easier with making measurements - just a single room measurement is good and you cut if off at the same range.

Been mostly on Headphones during a heatwave here - my Marantz AVR pumps out way too much heat. In going back to real speakers though, the one thing that strikes me is your brain can really localize sound because of head movement. It's probably a real pain to do - but is head tracking on your radar? Waves has a blue tooth headtracker.

 

[jaakkopasanen]

Saw that video a while back - it's why I gave up on room correction above 500hz with my AVR! If you've not read it - his book is excellent too.

It's made my life a lot easier with making measurements - just a single room measurement is good and you cut if off at the same range.

Been mostly on Headphones during a heatwave here - my Marantz AVR pumps out way too much heat. In going back to real speakers though, the one thing that strikes me is your brain can really localize sound because of head movement. It's probably a real pain to do - but is head tracking on your radar? Waves has a blue tooth headtracker.

I would love to implement head tracking but that needs to be in the real time processor, not in Impulcifer. Impulcifer of corse would have support measurements with different angles. I'm certainly not sure I'm I'm guessing that head tracking with EqualizerApo would require me to fork it and implement a version of the convolver which supports real time updates. And that sounds like a lot of work.

 

[johnn29]

Ah yes, that would be a huge project. I remember you mentioned that a while back. In real life viewing you never really move your head that much. But when you get the freedom back it makes a difference. I still don't think it's essential.

 

[sander99]

I think headtracking is very important because the brain can initiate unvoluntary head movements to solve ambiguities in sound localisation.
Having said that, some people don't seem to need it, others do.
For me personally: Horizontal head tracking helps a lot. I haven't experienced vertical headtracking yet but think I don't need that so much: with my borrowed A16 I could look up and down and didn't get a sense that the virtual speakers were moving up and down with my head movements (except in extreme cases like looking straight upwards).
A funny thing is that many people who had the A16 demo thought that the horizontal and vertical headtracking were fully functional, while in fact only horizontal was (and only horizontal lookangles were measured).
With vertical head movements you don't change the strong ITD and ILD clues, so I suppose just seeing the loudspeakers overrides the absence of change of the more subtle hrtf filtering cues in that case.

 

[johnn29]

Interesting experience from A16 users.

i've read the research and I've been more mindful of how I localise sound. I guess it depends on how you listen - but for movies and music I'm not trying to isolate locations. Most movies want you glued to the screen, so not looking around and there's way too much going on to be able to localize individual sounds without getting out of the movie. Even with object based - I find it's all too loud and confusing for the brain - so you get the bubble instead and might notice a few louder sounds. It's cool to be able to do it on the Atmos/DTS:X demo clips, but I never do it in movies. The only exception is whenever I tweaked my system, I pay more attention to make sure everything is balanced and right.

Now when I'm in the real world, like in nature. I notice I localize sounds easily with head movement and tilting. But that's very different for me.

The one area I do find head tracking to work wonders is with my Mobius/Waves headphones. Because it's a generic HRTF, with coarse customization options for your head circumference, the center channel is much more in your head. The head movement there fools you into thinking the in head center is slightly in front of you - maybe 3 feet away. Which makes sense for a PC gaming headset. I'm excited to try out Apple's implementation of it.

Once I'm days into listening to my BRIR I do sometimes turn it off and on again, just to quickly "demonstrate" to my brain that the sound is really out there. I don't need to, but I think it helps - same with listening to speaker call out ID tests.

 

[Àedhàn Cassiel]

Hey guys... I'm already at my personal endgame with headphones, and I have the EQ settings I want to stick with forever (or until aging drastically changes my hearing). Meanwhile, I don't know the first thing about speakers or the speaker market. I don't own any speakers except for a passable bluetooth speaker to connect to my phone. I don't have any treated rooms or anything like that. Could anyone give me an idea what it would cost and how much trouble it would be for me to try to do the measurements for Impulcifer on my own? Would there be enough people like me to justify some kind of speaker loaner tour? Are people finding any noticeable difference between the different Sound Professionals microphones; is there good enough reason to go above the $20 model? Or would anyone happen to be near USA - SC? I'd gladly pay a few bucks to use someone else's setup.

I'm also wondering whether this can be used to make sense of how and why peoples' preferences in headphone FR differ from the Harman Curve. I've dialed my midrange preference in very clearly: it begins rising gradually in the lower mids, it doesn't begin the steep rise seen on the Harman Curve until past 2kHz, and it looks like Harman with a few dB less peak from 3kHz onwards. Might measurements like these help clarify whether I have different preferences from people that like the Harman Curve, or need a different headphone FR to have the same FR delivered to my ear drum that fans of the standard Harman Curve are getting delivered to theirs? Might binaural measurements together with Harman Curve FR calibrations be used to quickly dial in any given person's most preferred signature?

 

[sander99]

to have the same FR delivered to my ear drum that fans of the standard Harman Curve are getting delivered to theirs?

Subtle difference: Most natural would not be to get the same FR on your ear drums as on other people's ear drums, but the FR on your ear drums that your brain expects based on how sound coming from a distance outside your head will be filtered before it reaches your ear drums.