Recording Impulse Responses for Speaker Virtualization

[Kosorm]

Hi, im in the process of obtaining all of the gear needed for impulcifer. As my main use case for this is music i only plan on doing stereo for the time being, maybe doing 7.1 at another time. When reading the documentation I understand you can use a single speaker at different angles to get 7.1, but the section above that discusses stereo seems to assume a 2 speaker setup. Do i/should i have to bite the bullet and get 2 speakers in order to do the stereo setup?

Secondly (but not as important), im planning on buying the JBL 305p MKII for this, based on whether or not i end up needing 1 or 2 of them, is that the best use of my money for this (especially considering i will have no acoustic treatment in my apartment) or are there more highly recommended speaker setups for this around the same range?

Thanks in advance, this thread has been super helpful already!

 

[musicreo]

When reading the documentation I understand you can use a single speaker at different angles to get 7.1, but the section above that discusses stereo seems to assume a 2 speaker setup. Do i/should i have to bite the bullet and get 2 speakers in order to do the stereo setup?

If you do the rotation method you can use a single speaker for Stereo.

Secondly (but not as important), im planning on buying the JBL 305p MKII for this, based on whether or not i end up needing 1 or 2 of them, is that the best use of my money for this (especially considering i will have no acoustic treatment in my apartment) or are there more highly recommended speaker setups for this around the same range?

It is a decent speaker. A few years ago the speaker was highly recommend in its price range but I don't knqw if it still the case.

 

[Kosorm]

If you do the rotation method you can use a single speaker for Stereo.

It is a decent speaker. A few years ago the speaker was highly recommend in its price range but I don't knqw if it still the case.

Thanks so much for the response! Glad to know i only need to buy one speaker haha

 

[castleofargh]

If there is no plan to actually use a speaker system, I know it's not great, but I would absolutely get something cool on Amazon, and return it once I'm done with the measurements.
I say Amazon because of the long return policy. Most people need time to figure out everything, and often get puzzled by the amount of possible variation between 2 measurement attempts. I know I did.

 

[Kosorm]

If there is no plan to actually use a speaker system, I know it's not great, but I would absolutely get something cool on Amazon, and return it once I'm done with the measurements.
I say Amazon because of the long return policy. Most people need time to figure out everything, and often get puzzled by the amount of possible variation between 2 measurement attempts. I know I did.

Ill definitely look into that as an option, as that would probably be a lot more convenient. You wouldnt by chance have any recommendations for a speaker that fits that criteria on amazon within a few hundred bucks of that would you? If not thanks anyways for the idea

 

[castleofargh]

I don't. I've been out of the speaker market for a long time(It's the kind of stuff where you get a pair, and it might just outlive you).

 

[Lion3]

Hello everyone, it's been a while.

I’ve been contemplating the topic of upmixing APOs for quite some time at the request of a few Korean users, and I’d like to share a small result of that effort. The reason I haven’t enjoyed upmixing until now is that none of the upmixers I’ve listened to (for example, Auro3D) have come close to what I expected. The channel separation wasn't correct, and overall, they just weren’t satisfying. This includes everything from GitHub projects, custom upmixing solutions, and resources I found through Google searches.


https://drive.google.com/drive/folders/1m0gFQkdPJEIiJ0syRilghdVJ_2cX9vpl

You just need to download the text file and the convolution files in the link. Then, simply edit the text file to update the paths of the two downloaded convolution files.
Of course, if you plan to use 5-channel or 7-channel audio, you will definitely need to use VB-Cable.

And then, you just need to place the downloaded upmix.txt file on top of the HeSuVi files like this.

Copy: VSL=0.8165*L+-0.5774*R
Copy: VSR=-0.5774*L+0.8165*R
Copy: SL=-1.0*VSL
Copy: SR=VSR

I applied the appropriate coefficients to L and R, and the left and right surround channels have opposite polarity (separated).

Copy: VRL=VSL+-0.488*VSR
Copy: VRR=-0.488*VSL+VSR
Copy: RL=-1.0*VRL
Copy: RR=VRR

Additionally, most of the scripts found online only mixed the L and R channels in the same way, but I had to extract the surround back channels from the separated surrounds. This was necessary for proper separation.
With this setup, I was able to confirm that most tracks played correctly, with the front center, surround, and surround back producing distinct sounds without HRTF smearing.
The delay for each channel will vary depending on the room and measurement impulse, but in my case, a delay of around 10 ms for the sides and 20 ms for the rear worked well. As long as it's within the 5–30 ms range, which doesn't exceed the Haas threshold, you can adjust it according to your preference while listening.

If you're an Impulcifer + HeSuVi user with 7.1 channel BRIRs, it's worth trying out at least once.

 

[morgin]

Hello everyone, it's been a while.

I’ve been contemplating the topic of upmixing APOs for quite some time at the request of a few Korean users, and I’d like to share a small result of that effort. The reason I haven’t enjoyed upmixing until now is that none of the upmixers I’ve listened to (for example, Auro3D) have come close to what I expected. The channel separation wasn't correct, and overall, they just weren’t satisfying. This includes everything from GitHub projects, custom upmixing solutions, and resources I found through Google searches.


https://drive.google.com/drive/folders/1m0gFQkdPJEIiJ0syRilghdVJ_2cX9vpl

You just need to download the text file and the convolution files in the link. Then, simply edit the text file to update the paths of the two downloaded convolution files.
Of course, if you plan to use 5-channel or 7-channel audio, you will definitely need to use VB-Cable.



And then, you just need to place the downloaded upmix.txt file on top of the HeSuVi files like this.

Copy: VSL=0.8165*L+-0.5774*R
Copy: VSR=-0.5774*L+0.8165*R
Copy: SL=-1.0*VSL
Copy: SR=VSR

I applied the appropriate coefficients to L and R, and the left and right surround channels have opposite polarity (separated).

Copy: VRL=VSL+-0.488*VSR
Copy: VRR=-0.488*VSL+VSR
Copy: RL=-1.0*VRL
Copy: RR=VRR

Additionally, most of the scripts found online only mixed the L and R channels in the same way, but I had to extract the surround back channels from the separated surrounds. This was necessary for proper separation.
With this setup, I was able to confirm that most tracks played correctly, with the front center, surround, and surround back producing distinct sounds without HRTF smearing.
The delay for each channel will vary depending on the room and measurement impulse, but in my case, a delay of around 10 ms for the sides and 20 ms for the rear worked well. As long as it's within the 5–30 ms range, which doesn't exceed the Haas threshold, you can adjust it according to your preference while listening.

If you're an Impulcifer + HeSuVi user with 7.1 channel BRIRs, it's worth trying out at least once.

I’ll definitely will be trying this if I can workout how to edit the text file.

Have you tried Dolby atmos app on the Microsoft store. I still get goosebumps when watching Dolby atmos movies

 

[morgin]

Do I add both “upmix 4ch.txt” and “upmix.txt” into eqapo or just one?

And do I need to add

Copy: VSL=0.8165*L+-0.5774*R
Copy: VSR=-0.5774*L+0.8165*R
Copy: SL=-1.0*VSL
Copy: SR=VSR

And

Copy: VRL=VSL+-0.488*VSR
Copy: VRR=-0.488*VSL+VSR
Copy: RL=-1.0*VRL
Copy: RR=VRR

To the txt files. I can see it’s already in the upmix.txt

 

[Lion3]

Do I add both “upmix 4ch.txt” and “upmix.txt” into eqapo or just one?

And do I need to add

Copy: VSL=0.8165*L+-0.5774*R
Copy: VSR=-0.5774*L+0.8165*R
Copy: SL=-1.0*VSL
Copy: SR=VSR

And

Copy: VRL=VSL+-0.488*VSR
Copy: VRR=-0.488*VSL+VSR
Copy: RL=-1.0*VRL
Copy: RR=VRR

To the txt files. I can see it’s already in the upmix.txt

Upmix only . not 4ch.(If it's 7 channels)
7.1, 5.0, 4.0, 6.0 all seemed to work well when I heard it myself or asked some users to listen to it, but some users said symptoms that seemed to have some composition (4ch users), so I made a separate document about it.
First of all, I'm going to take a closer look at the 4ch side.

And no. I remember using Dolby Atmos before, but I don't use it now.

 

[Lion3]

To the txt files. I can see it’s already in the upmix.txt

also check my link again. I uploaded upmix-edit. that's right

 

[Kosorm]

Hey might be a bit of a dumb question and i might be a bit over cautious, but i finally got everything necessary to do this and i just want to be absolutely sure that once i put my iebms into the vxlr+'s, its then safe for me to turn on 48v power on my audio interface? Its only beforehand that its unsafe to do so right?

 

[MrHaelscheir]

Hey might be a bit of a dumb question and i might be a bit over cautious, but i finally got everything necessary to do this and i just want to be absolutely sure that once i put my iebms into the vxlr+'s, its then safe for me to turn on 48v power on my audio interface? Its only beforehand that its unsafe to do so right?

If your IEBMs are designed for plug-in-power, then yes, it is only safe to turn on the 48V phantom power when they are connected via the VXLR+ adapters. I'm not sure if some folks would go as far as to first turn on the phantom power, turn the interface off, plug in the IEBMs and adapters, and then turn the interface on.

 

[sander99]

its then safe for me to turn on 48v power on my audio interface?

Actually, I never really thought about this before but are we talking about DIY in-ear-microphones (bare capsels with bare wires soldered to them, glued on a piece of foam or something like that) on 48V phantom power and the safety of the person putting them in his or her ears?

If so my advice is:
STOP! HALT! ABORT!
STOP! HALT! ABORT!
STOP! HALT! ABORT!
STOP! HALT! ABORT!
STOP! HALT! ABORT!
STOP! HALT! ABORT!

(But maybe this is about something else and you all know what you are doing.)

 

[Kosorm]

Actually, I never really thought about this before but are we talking about DIY in-ear-microphones (bare capsels with bare wires soldered to them, glued on a piece of foam or something like that) on 48V phantom power and the safety of the person putting them in his or her ears?

If so my advice is:
STOP! HALT! ABORT!
STOP! HALT! ABORT!
STOP! HALT! ABORT!
STOP! HALT! ABORT!
STOP! HALT! ABORT!
STOP! HALT! ABORT!

(But maybe this is about something else and you all know what you are doing.)

Just using the cheaper sound professionals iebms recommended on the impulcifer github with the hooks cut off