[XingYu]

I always use headphone for hrir(brir) system on PC.
But I purchased Apple's earpods wire and tried with it. The results were better than I thought.

I first listened to the files made for the 800s, and except for the ultra-high notes, the feeling from 500 to 6000 Hz, which corresponds to the important HRTF, was pretty much the same.
Because the bass is calibrated for the 800s, the bass in the apple earpods was a bit jarring.

So I asked the developer about how to use autoeq, and I tried it. and not bad. EQ.csv from AutoEQ (HD800s to APple Earpods)
However, I felt HRTF's response was broken, and I downloaded messages of 800s and airpods from Rtings and other places.
I also overlapped the headphone files that I had measured with my ears and carefully adjusted the base part while looking at its similarities.

After doing a little twist from about 50 to 120hz (digital virtual loopback), I ran impulcifer with this, and it sounds almost like the headphones I originally listened to.

i tested it with a combination of pure direct sound and octa subwoofer responses, and the earpods work very well.

And I always did that in the speaker system, and I always listened to my headphones only when I made up my mind. I never listened to it at any other time.
And I put my earphones in my phone today and sat outside on the bench listening to music.
It's so emotional.
The singer is singing for me in front of the bench. It's really a small performance in everyday life.

I searched this and that for my phone and found out about Rootless Jamesdsp.
I originally thought this application would only support two-channel stereo. (half brir only)
However, this has confirmed support for 4-channel True Stereo and works very well. I didn't have to listen to it on my PC anymore.
Only when creating a wav file for this, it automatically detects 4 channels within the application if you only create it in the order of LL, LR, RL, and RR.

The weather is cold because it's winter, but I still feel warm listening to the song.




I recommend a winter song by a Korean female singer for this time.

 

[morgin]

Nice song!

I’m hoping there’s a way to make this work on iPhone too. But I have a android device so will be testing this Jamesdsp

 

[Brandon7s]

I searched this and that for my phone and found out about Rootless Jamesdsp.
I originally thought this application would only support two-channel stereo. (half brir only)
However, this has confirmed support for 4-channel True Stereo and works very well. I didn't have to listen to it on my PC anymore.
Only when creating a wav file for this, it automatically detects 4 channels within the application if you only create it in the order of LL, LR, RL, and RR.

Wonderful! I've been searching everywhere for a way to use HRIRs on my Android phone so I will give this a shot. I already have all of my HRIRs in TrueStereo format (as well as HeSuVi) because my DAW convolution plugin only supports TrueStereo, so this should be easy for me to try.

By the way, the new Master Series The Sound Professional MKII mic came in and I've made a few great HRIRs with those over the long holiday. I'll take some pictures when I have the time, as they are much larger than I thought they would be. I assumed that they'd be the same size as the MKI, but they are not. They BARELY fit into my ears, but they DO fit. I don't think these would work for a lot of people though. But if they do fit then I can definitely recommend them. Their sensitivity and noise floor is miles better than any of the other in-ear mics I've used, including the MKI.

When I normally capture the IRs from my speakers with Impulcifer, I have to raise the mic preamp gain to +35db, and I also have to set the speaker levels to louder than I would normally listen in casual use. With these new mics, I can set the preamp gain to +12db and I lowered the overall speaker volume down to my normal listening levels and I get the same input levels that I was getting with the prior mics, but with a significantly reduced noise floor. It's hard to say if the new mic has made a noticeable difference in the quality of the HRIRs since I've only taken a couple measurements with them, but I will note that the bass frequencies are better balanced between the left and right channels with the few that I have taken so far. I suspect that the greatly reduced noise floor is helping to keep the low-volume measurements more aligned with each other, particularly between 20 and 70hz, which is well into the falloff my speakers in this room due to a room node notching a lot of bass content below 90hz. The speakers themselves don't roll off until 50hz, but the room makes the rolloff look much sooner.

I've not had much free time recently but I will post a picture comparing the new mic's size to the original ones as soon as I have the opportunity. I've not forgotten about trying the subwoofer IRs that you sent me, XingYu! I will try that out and respond with the results once I have more time as well.

 

[jaakkopasanen]

Someone here should start making open ear canal microphones and selling them. Using as small capsule as possible with a some kind of fitting or clamp to keep the capsule in place at the center of the ear canal opening. Currently the biggest problem with Impulcifer is the headphone compensation because close ear canal microphones prevent ear drum from loading the headphones and therefore the headphones won't have the same frequency response as when worn normally.

 

[XingYu]

I've not had much free time recently but I will post a picture comparing the new mic's size to the original ones as soon as I have the opportunity. I've not forgotten about trying the subwoofer IRs that you sent me, XingYu! I will try that out and respond with the results once I have more time as well.

Good! First of all, it was difficult to check what file you usually listen to, and it may be a little different from the way I used to do it because I put it on my own.
So if when you hear it, it's similar to what you originally hear, but it feels like it's definitely filled with bass, it's working well. Then tell me again then. I'll align you with a more clear and luxurious octa subwoofer room response.

Someone here should start making open ear canal microphones and selling them

Take my money

 

[XingYu]

Wonderful! I've been searching everywhere for a way to use HRIRs on my Android phone so I will give this a shot. I already have all of my HRIRs in TrueStereo format (as well as HeSuVi) because my DAW convolution plugin only supports TrueStereo, so this should be easy for me to try.

Yes. Remember that. LL, LR, RL, RR in order.

 

[musicreo]

By the way, the new Master Series The Sound Professional MKII mic came in and I've made a few great HRIRs with those over the long holiday. I'll take some pictures when I have the time, as they are much larger than I thought they would be. I assumed that they'd be the same size as the MKI, but they are not. They BARELY fit into my ears, but they DO fit. I don't think these would work for a lot of people though. But if they do fit then I can definitely recommend them. Their sensitivity and noise floor is miles better than any of the other in-ear mics I've used, including the MKI.

The bigger capsules always have a better SNR and are more sensitive. But it not possible to place them into the ear canal entry. For me the placement into ear canal entry compared to the placement directly before the entry made a difference for the accurate perception of directions. However, recording was much easier with my bigger capsules.

 

[conql]

Someone here should start making open ear canal microphones and selling them. Using as small capsule as possible with a some kind of fitting or clamp to keep the capsule in place at the center of the ear canal opening. Currently the biggest problem with Impulcifer is the headphone compensation because close ear canal microphones prevent ear drum from loading the headphones and therefore the headphones won't have the same frequency response as when worn normally.

I recall that @musicreo once shared a method for creating open ear canal microphones quite some time ago. I followed those instructions and crafted my own using 4mm microphones with a 68dB SNR. The audio quality is indeed superior to that of closed ear canal microphones, though the improvement is not drastic. I still find myself making significant adjustments to the bass and treble balance, and the high frequencies (above 8kHz) don't quite match the sound of my loudspeakers. While they sound very **plausible**, the difference is clearly there when A/B testing. The microphones are quite consistent and minimally affect my hearing, allowing me to test the results without removal. When apply HPEQ only, the frequency response look very flat. I've also experimented with various insertion depths but didn't observe any major differences in quality.

Reflecting on this, I'm still uncertain about the exact cause of these issues, but I suspect it might be due to:
1. The 4mm diameter of the microphones might still be too large.
2. The microphones are not centered in the ear canal but making contact with it.
3. The lack of acoustic treatment in my highly reverberant room.
4. Impulcifer's alteration of ITD.
5. My headphones (Sony MDR V6) might not be good enough.

Not sure how significant each of these problems is, nor can I rule out the possibility of other factors influencing the results. Just some food for thought.

 

[XingYu]

3. The lack of acoustic treatment in my highly reverberant room.
4. Impulcifer's alteration of ITD.

ITD chack can possible in Read.me in folder. (Btw, If ITD is not same, it can be fix to same. ITD is just delay when recorded.)
Reverberant ... Can you messure in REW?

 

[XingYu]

I recorded the screen for some Korean HRIR/BRIR users.
It is a loopback using VB cable and Audacity.
As you can see
Use an EQ txt file made by REW, or apply an FIR filter, or use a pre-amp to make speaker deviation or headphone deviation gain matching, etc
Hang what you're going to do on the equalizer apo VB cable. (If you're going to do left and right, of course, the channels are left and right.)
Then, import the FL,FR.wav file or Headphones.wav to Audacity.
And you just have to record it. For the filter to be applied, of course it has to be included in the recording, so we're going to re-record it in loopback
These are the things that you can do before you use the impulses
After creating, undermining, mischiefing, applying VST, or doing something crazy in REW
You can loop it back like this again.

This is a basic process and concept.
It is very good to manually correct the gain matching and deviation.
And if I apply this, I can even do what I've done before by virtually creating and aligning responses. (Of course it has to decompose the Hesuvi file to force the delay again.)

 

[musicreo]

The things you show in the video could also be done with the benchmark of EQ-APO.

 

[XingYu]

The things you show in the video could also be done with the benchmark of EQ-APO.

Benchmark can record make 4ch true stereo or 14ch All each track?
As i know, To do that, you would have to record each track in mono.
Also, Audacity was more appropriate for my purpose because if the Impulse makes an error, only certain tracks need to be forced to play 20us to 40us fast or slow.

 

[XingYu]

I found something strange while making seven-channel files and correcting other people's.
I thought the impulcifer was time-aligned to help with this because the impulse peaks of the recordings sometimes don't match exactly.

So I copied the same FL,FR file and tested it after renaming it SL,SR/BL,BR/FC,X. As it was, all the peaks should be exactly the same because they are the same file.

And I finally found it correctly.
front channel is always zero.
The side channel is always 5 samples fast based on 48000 Samplate (about 100us)
Center channel is 5 samples slow (100us)
The rear channel is fine.
And what's wrong with this is that the delays in the crosstalk channel are also changed at will. (Such as FL-R , SL-R things.)
Because this is the same file, it must be located in the same place...

I think this is going to be a bug.
And there's one more suggestion I'd like to make.
I force zero alignment via digital loopback in rew and place as many exact crosstalk channel delays as I want or should apply.
I want you to add it so that I can turn off the function for impulse alignment during the impulcifer calibration and processing time align.
Sometimes the impulcifer over-detects that the impulse peaks are almost identical but slightly different in shape and changes the delay. Therefore, such files often have to manually re-adjust the peak after processing.

 

[jaakkopasanen]

I found something strange while making seven-channel files and correcting other people's.
I thought the impulcifer was time-aligned to help with this because the impulse peaks of the recordings sometimes don't match exactly.

So I copied the same FL,FR file and tested it after renaming it SL,SR/BL,BR/FC,X. As it was, all the peaks should be exactly the same because they are the same file.

And I finally found it correctly.
front channel is always zero.
The side channel is always 5 samples fast based on 48000 Samplate (about 100us)
Center channel is 5 samples slow (100us)
The rear channel is fine.
And what's wrong with this is that the delays in the crosstalk channel are also changed at will. (Such as FL-R , SL-R things.)
Because this is the same file, it must be located in the same place...

I think this is going to be a bug.
And there's one more suggestion I'd like to make.
I force zero alignment via digital loopback in rew and place as many exact crosstalk channel delays as I want or should apply.
I want you to add it so that I can turn off the function for impulse alignment during the impulcifer calibration and processing time align.
Sometimes the impulcifer over-detects that the impulse peaks are almost identical but slightly different in shape and changes the delay. Therefore, such files often have to manually re-adjust the peak after processing.

If all speakers are equal distance from the center of the head, the delay from different angles arrive at the ears ever so slightly different times. Impulcifer adds these delays intentionally.

 

[XingYu]

If all speakers are equal distance from the center of the head, the delay from different angles arrive at the ears ever so slightly different times. Impulcifer adds these delays intentionally.

Can I or we also disable it? I or some certain people would love it if I could deactivate it.
I've been calibrating the delay by forcing it in the file that's already been made, and this has been pretty inefficient.
It's not a huge difference, but at least I felt the smoothest channel switch when operating with each crosstalk delay at the right timing (main response is 0ms).
However, when I heard what was done by the impulcifer while calibrating some of the users' files, I heard each channel buried on the fast channel side in the order of the delay because their rooms were not as perfect as my measurements. If the response was perfect, even if the delay was 100us different, it didn't lead to a big problem, but at least in the environment of the in-room, the delay that started was correct only when it was 0.
If it is cumbersome for you to update and the user can modify and deactivate something in the python file, I would appreciate it if you could let me know.