Recording Impulse Responses for Speaker Virtualization
Jun 24, 2023 at 3:57 AM Post #1,531 of 1,817
Jun 24, 2023 at 6:55 PM Post #1,532 of 1,817
Jun 24, 2023 at 8:25 PM Post #1,533 of 1,817
Thank you @castleofargh and @bigshot for your answers.

@castleofargh It's nice that the HRTF doesn't change with the sound level. So we can measure our speakers at the ideal sound level which minimize background noise interferences and distortions and then we can adjust the level on headphones without altering the results.
The sources you shared are very interesting, especially the two last. It seems that the cues for the distance are not the same below and above the critical distance. ILD seems to be predominant below while the direct to reverberant ratio is more important above. My idea initially was to do measurements really close to the speaker in-order to minimize the reverberations and room modes and then adjust the reverberations using Impulcifer decay time parameter or an another software. Adding reverberations accordingly to the distance may allow to mask ILD cues while increasing direct to reverberant ratio cues. But I'm not sure that doing this at playback will be as consistent as in real-life.
 
Jun 25, 2023 at 1:34 AM Post #1,534 of 1,817
Votre français est impeccable M. Casteofargh. Vous habitez en France, si ma mémoire m’est fidèle ?
I am French(born in Corsica), but try to minimize non-English on the forum because when in Rome... And also because the posting guidelines say so. ^_^'

Thank you @castleofargh and @bigshot for your answers.

@castleofargh It's nice that the HRTF doesn't change with the sound level. So we can measure our speakers at the ideal sound level which minimize background noise interferences and distortions and then we can adjust the level on headphones without altering the results.
The sources you shared are very interesting, especially the two last. It seems that the cues for the distance are not the same below and above the critical distance. ILD seems to be predominant below while the direct to reverberant ratio is more important above. My idea initially was to do measurements really close to the speaker in-order to minimize the reverberations and room modes and then adjust the reverberations using Impulcifer decay time parameter or an another software. Adding reverberations accordingly to the distance may allow to mask ILD cues while increasing direct to reverberant ratio cues. But I'm not sure that doing this at playback will be as consistent as in real-life.
I don't remember exactly, but when I tried fooling around with reverb(long before I had recorded anything at my own ears, so take it with a grain of salt), I felt that regardless of spatial cues, I didn't have a big tolerance for reverb or delays. A small increase would be fine and helped with the general feeling of space(not necessarily distance for me, but a more enveloping sound was easy to get. I simply didn't like large reverb or long delays between direct and secondary cues for music listening (for some movies it was still fine up to a point). IDK why exactly, but I tend to stick with small amounts of reverb.
Anyway, once you have binaural mics, you can experiment and find out how you in particular react to this and the actual impact on the PRIR for distance between you and the speakers, or between the speakers and the walls. And of course, how much your field of view alters the sound interpretation.
You'll rapidly get fed up with testing, most people do, but so long as you have the desire for answers, speaker simulation is an interesting and IMO gratifying personal journey.
 
Jun 25, 2023 at 8:50 AM Post #1,535 of 1,817
Do you know any software that allows to play with reverb, maybe with settings for early reflections too ?
I think I will try soon. I have to see what to buy exactly and I'm thinking right now of how to have consistent measurements when rotating myself since I plan to use only one speaker. Maybe I will first measure front left and right with cheap and probably bad speakers I already have, just to see how close to them I can go and to train myself a little.
 
Last edited:
Jun 26, 2023 at 12:17 PM Post #1,538 of 1,817
It works when you change hesuvi from automatic to stereo
fact is that i want it to do automatically, in fact hesuvi checks if the app is using stereo channel so it works automatically when i'm either in games and in VLC, but when it comes with chrome and youtube or such different apps it doesn't work and i have to "force" using the content format to stereo

if there's a way to automatically call hesuvi in stereo mode when watching youtube would be awesome, this is the only thing i'm asking
 
Last edited:
Jun 27, 2023 at 6:17 PM Post #1,539 of 1,817
Hello everyone,

I think I will soon buy the hardware needed for my measurements. I'm a little nervous about the microphones, I don't want to ruin them when removing the wings or the casing. I tried some foam in my ear and it seems to go out of the ear canal a little, so it will be difficult to try deep or shallow insertion. I'm thinking of using juste a little amount of foam to maintain the microphones at the ear canal opening, or earbud tips. Something like these exemples
https://www.head-fi.org/threads/rec...r-speaker-virtualization.890719/post-16016609
https://www.head-fi.org/threads/rec...r-speaker-virtualization.890719/post-16502879
https://www.head-fi.org/threads/rec...r-speaker-virtualization.890719/post-16653902
https://www.head-fi.org/threads/rec...r-speaker-virtualization.890719/post-16934860
Do you think it's a good idea ? I would like to not glue them in order to be able to change my mind if one solution is not as good as what I thought, do you thinks it's possible ?

It seems that the Primo EM258 have more noise than the Sound Professionals, I'm hesitating between the SP-TFB-2, the MFS-TFB-2 and the MS-TFB-2-MKII. Did some of you be able to compare them ? Does the XLR version have better performances than the jack versions ? If I go to jack, is the Mogami cable really better than the cable labeled as premium ?
 
Jun 27, 2023 at 9:15 PM Post #1,540 of 1,817
Not sure about the cables.

But regarding the mics. The second link you posted is mine and I got better results not using those but using in ear foam plugs. I used a hot glue gun to stick mine to the foam and also cut the foam to a much smaller length so the mics were almost flush with my ear canals. The hot glue can easily be peeled off the mics if not much is used.
The foam didn’t slip as much and mostly stayed in place which I think was important

Hope you get some good measurements and don’t give up after a few tries. I’m still wowed by mine everyday and I did mine around a year ago.

Are you doing it for the surround sound?
 
Last edited:
Jun 27, 2023 at 10:00 PM Post #1,541 of 1,817
A question for those using this for surround sound.

What media player are you using? And any important settings that need to be implemented?

Do you use lav filter for audio, and should I bitstream if I’m connected via DAC usb?

Is there any windows sound settings to change/untick for surround. (I’m finding unticking exclusive mode is better for me)
 
Jun 27, 2023 at 11:12 PM Post #1,542 of 1,817
should I bitstream if I’m connected via DAC usb?
No, definitely not. Bitstreamed surround formats must be decoded to multichannel PCM before they can be processed by HeSuVi or other speaker virtualization software.

I used VLC player.
I had quite some trouble before I had it working for surround sound formats. I followed some guide, I don't remember now where I found that, maybe I can try to find it later (if nobody else helped you before I have time).

Edit: Actually one problem I had at some point was that I couldn't get the decoding to multichannel PCM going. (And I don't even now who's doing that decoding: Windows or VLC player?). Instead it decoded and downmixed to stereo PCM, which is not what we want.
 
Last edited:
Jun 28, 2023 at 12:40 AM Post #1,543 of 1,817
Jun 28, 2023 at 5:31 AM Post #1,544 of 1,817
No, definitely not. Bitstreamed surround formats must be decoded to multichannel PCM before they can be processed by HeSuVi or other speaker virtualization software.

I used VLC player.
I had quite some trouble before I had it working for surround sound formats. I followed some guide, I don't remember now where I found that, maybe I can try to find it later (if nobody else helped you before I have time).

Edit: Actually one problem I had at some point was that I couldn't get the decoding to multichannel PCM going. (And I don't even now who's doing that decoding: Windows or VLC player?). Instead it decoded and downmixed to stereo PCM, which is not what we want.

Thanks for the reply. Just to be clear my surround works amazingly well. There’s a scene in ford vs Ferrari where the main guy is test driving a car and some blokes are talking. As the camera is focused on the guys talking you can hear the car making its way around the circuit. The distance and actual ability to be able to tell the car is turning a bend after a stretch is unbelievably lifelike.

I’m asking incase there’s little things/settings that I may have overlooked or maybe I’m just so deep in this rabbit hole of perfection that I keep looking for things to improve.

I use mpc hc with lav filters
 
Jun 28, 2023 at 7:32 AM Post #1,545 of 1,817
Thanks for the reply. Just to be clear my surround works amazingly well.
Ahaa, from your questions I got the impression you didn't have it working at all other than in stereo (which would have been surprising knowing how busy you have been with all this).
 

Users who are viewing this thread

Back
Top