3D sound with my own XTC

[STC]

After 9 years of struggle with commercial XTCs, I decided to make my own that can play studio recordings as good as the full orchestra music. Took me about two months and here is the final product's sample. 

 

Recorded with 17cm spaced microphone from my listening seat. Only two speakers used. Recorded with Zoom H1 and Roland CS10EM microphone. Format of recording is 16/44.1. All tracks were played via JRiver through Reaper. All tracks were 16 bit 44.1kHz format except for Lux Aurumque which was MP3 downloaded from CDBaby. Just a short excerpt of each. 

 

1) Norah Jones -I got to see you again.

2)Famous Blue Rain Coat

3) Money - Pink Floyd

4) Lux Aurumque

5) Black Magic Woman

 

 

 

 

 

 

 

 

 

[STC]

Made further changes to make the bee circling the head. It also improved the vocal accuracy. Captured using phone.  The original file was 231Mb . The compressed version at 9.1Mb is here. 

 

 

[STC]

1 hour ago, AnotherSpin said:

Would you elaborate more about 17 cm microphone placement? 21 cm seem to be much more appropriate. Also, what kind of seat you used?

 

Just a standard 3 seaters from ikea. The 17 and 21cm is based on pinna to pinna measurement. There is no correct answer here. I think 17 is more correct as our actual hearing mechanism is deep inside. 

 

BTW, the last one was a direct iphone SE video. No spaced microphone. 

[STC]

1 hour ago, AnotherSpin said:

 

All Ikea related would not qualify as a high-end.

 

😂 so too my rear end. 

[STC]

37 minutes ago, sphinxsix said:

can understand you they add just about anything to it to make it cheaper.

 

I don’t understand. What is there to make them any cheaper when most of the product are either free or as low as $10. Perhaps, BACCH is the most expensive one at $50000 for their high end version which is now the gold standard. 

 

If you need the algorithm, please let me know.

[STC]

Oh.... sorry to disappoint you for not getting it.  

[STC]

XTC stands for crosstalk cancellation.  So far the gold standard is BACCH but their premium version is $50k. 

[STC]

Made further improvement to the crosstalk cancellation. I can now make the bee or the haircut sound at the ears. Also it looks like one setting can be applied for different genres. The tonal balance is correct.

 

Here is a video of 15 tracks excerpts which consist of tracks for vocals and some percussion. Also included 2L classical recordings. The Money and Window track should still able to produce the effect near the ears even on YouTube. I have been using the same tracks last two or three years for reference. Now time to review 100 over audio samples to find out which one sound the best. Lux Aurumque is also included as I find the Stereophile video of the same track is an interesting case study. That sounds almost original something which I never managed to record of the many high end stereo system that I could access. Waiting for this year's AV show and see if I can the same sound as in the Stereophile recording with stereo setup. 

 

The audio was captured using  omnidirectional microphones. This may not produce the exact drums attack but it can reveal much about the room acoustics. Sometimes, directional microphones may sound better with video but that not what you actually hear in the room. Hope to see more than 4 viewers for this video.  

 

 

 

[STC]

Now time to select the setting. These are my current 8 best settings among 100 over different implementation. Each of them has their own strength and weakness.  Out of these eight I have two which is accurate for my preference. Unfortunately, others in my family have their own preference among these 8 settings. Thankfully with Reaper it is just a click of the mouse to choose the preferred one. No need to change cables, amps or speakers to hear the difference in soundstage or details. This is not high end for audiophiles.   

[STC]

13 minutes ago, motberg said:

Is your algorithm for stereo files ? What applications will work ?

Similar to you, I have tried almost everything available and would consider the basic BACCH if it would not be mac based.

So far, my best results are using the Blackice (Jolida) Audio Foz SSX followed by passive vol. control with partial physical crosstalk barriers.. software solutions seem phasey, incoherent, unbalanced, etc. compared to this, but the separation effect  is not as pronounced as a full barrier or ambiophonics based software.

 

Crosstalk cancellation is always only for stereo. Either 2.0 or 4.0. In 4.0 like quadraphonic for the rear.

 

I am not familiar with Jolida. But if the cancellation is poorly executed than your observation is correct. Generally, from Carver Holographic and other XTC based on 60 degrees speakers separation which will not work correctly. 

 

Mine is based on Ralphs RACE and it can be converted to VST plugin or as a stand alone application. Currently, it is done through Reaper DAW. 

 

If you inly listen to music than BACCH is not necessary as you don’t need 20dB cancellation. I have experimented with Minidsp ambio plugin, Miller’s Ambiodsp and Soundpimp. IMHO, miller’s is the most accurate on and probably Minidsp but I have not used them for a long time. But preference wise for studio recordings Soundpimp sounds good.  Physical barrier will not work properly unless you use small pc speakers for experiment as the sound will leak to the other ear at concert level volume. 

 

Why not try Soundpimp? It is a stand alone application.  

 

Please feel free to PM if you need more info. I will look into Jolida as this is the first time I am hearing of it. I don’t think it is based on RACE. 

[STC]

@motberg what is your speakers angle when you used miniambio?

[STC]

18 minutes ago, motberg said:

iirc, 10 degrees per side... my miniambio is based on the old , lower res minidsp box.. and the new minidspHR version is a big step up in transparency. I use the HR for sub management ...

 

Mine too set at 20 degrees. The recording was at the arrangement. 

 

With Miniambio, you should set the delay to about 85 microseconds and attenuation about 2.5 db. Leave the centre at 0 first and listen to some classical. 

 

Choose any of the tracks in my above video and I will send you a preprocessed file so that you can play them without any cancellation like you play ordinary music. See if you find it different. That is a preprocessed file so play them like any other files and dont use Jolida or physical divider. Just dont forget to bring your speakers close to 20 degrees.  

 

The other thing you want to confirm first is the phase accuracy of your speakers amp and other in the chain. Some equipments are not phase accurate by design which can give some sense of spaciousness so in that case it wont work correctly. Apparently, some Mac pc too didn't work correctly despite disabling the internal sound card sound effect. If you are using multi driver speakers make sure they are phase coherent .  

[STC]

2 hours ago, motberg said:

Interesting you consider absolute phase an important attribute..

 

Not absolute phase. I apologize if I had use that term. I meant accurate phase between the two channels to the speakers. 

[STC]

The crosstalk cancellation development has been finalized and by comparisons to others such as AmbiophonicsDSP, MiniAmbio or Soundpimp this handles phase manipulated recordings like this and Amused to Death very well. Playing Money from Dark Side of The Moon. The playback was recorded using Sennheiser AMBEO headset Use headphones to hear the wide soundstage of Ambiophonics Interaural crosstalk cancellation system. More info at https://goo.gl/CF1C6B 

 

 

 

 

[STC]

Slight refinement. It is hard to adjust without actual measurements. Mad World - two years progress. 
 

 

[STC]

Without having another similar set of speakers placed at 60 degrees, it is extremely hard to work on the crosstalk to get the timbre correct. 
 

another try with Norah Jones
 

 

[STC]

Next refinement. Tried a new track. 
 

Alison Krauss- It wouldn’t have made a difference. 

 

 

 

[STC]

It took me 10 months to correctly measure the sound for tonal accuracy. Whatever difference you hear is down to the microphones, speakers coloration and room acoustics. Recorded with Sennheiser AMBEO. 
 

My system playing Norah Jones

 

 

[STC]

Redid the whole crosstalk again and discovered a minor error. The tonal balance improved further.  Ver16 R43 ( for my reference). 
 

Playing Jennifer Warnes - Famous Blue Raincoat. 

 

 

 

[STC]

I am now concentrating with the impulse response. Trying a new approach where I can increase the level much higher without hearing the echoes. Just smooth reverberation even at high level. 
 

( V17 R4)