astr0b0y
Active Member
Nice write up, thanks. I’d really like to hear binaural through speakers without resorting to building a wall between them.
Too bad they don’t offer student pricing...
Too bad they don’t offer student pricing...
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
BACCH4Mac Pro Edition: a report
- Thread starter Dialectic
- Start date
OP
- Thread Starter
- #22
Yeah, it is pricey.
I don't think the experience can be characterized merely as binaural listening through speakers; that characterization implies that listening through headphones is ideal. In fact, listening to a binaural recording through headphones by most accounts does not reproduce even the spatial cues that one gets from listening to non-BACCH speakers. I think this is attributable to the fact that headphones circumvent the HRTFs of our ears, which enable us to hear sounds in three-dimensional space.
Last edited:
@soundArgument,
Thanks much for the thorough write up, and hittin' that barbed wire. It seems one of these will be in my future, with reasonable assurance that I'm not your avatar
It seems you bought the pro version? That version apparently has the ability to mix surround recordings into binaural playback - have you tried that?
Thanks much for the thorough write up, and hittin' that barbed wire. It seems one of these will be in my future, with reasonable assurance that I'm not your avatar
It seems you bought the pro version? That version apparently has the ability to mix surround recordings into binaural playback - have you tried that?
I thought, "I like when a someone lives up to the image presented in their avatar."
Then this!
What a letdown.
Then this!
What a letdown.
I may have to change it...
Well, I went ahead and ordered the BACCH package. I guess the worst that could happen is I don't like it - which seems unlikely given all the positive things I've read - and I'd have to pay a restocking fee.
I have on hand two pairs of ESL speakers (Sanders and JansZen), both of which are narrow directivity speakers preferred for BACCH. I'm excited to hear the results - ideally this could be the "envelopment" breakthrough I'm looking for.
I have on hand two pairs of ESL speakers (Sanders and JansZen), both of which are narrow directivity speakers preferred for BACCH. I'm excited to hear the results - ideally this could be the "envelopment" breakthrough I'm looking for.
OP
- Thread Starter
- #29
You will have excellent results. The BACCH SP-assisted Sanders electrostatic speakers that I heard in Mr. W.S. Lam's system in Hong Kong effected the most realistic simulation of music performed in real acoustic spaces that I have yet heard. The effect on some recordings was genuinely startling.
OP
- Thread Starter
- #30
I spoke to the Theoretica team last night and must issue a correction regarding my earlier statement regarding VST plugins. VST plugin functionality is available through the BACCH4Mac app, so convolution absolutely can be applied after the BACCH filters on a Mac.
I realize this is not full Roon integration, but it would simplify setup with a system like that of @dallasjustice.
My name is William Guiracoche and I am Development Engineer at Theoretica Applied Physics. On behalf of Theoretica I thank you all, and in particular the thread host soundArgument, for your lively and informative discussion. I offer below replies and clarifications to some of the questions and points raised in thread so far. I apologize for the length of this reply, which is due to the accumulated points to respond to. I will be happy to try to answer future questions in this thread. If you wish to ask questions offline, please do not hesitate to send them to [email protected].
We have never encountered or even heard of speakers that are incompatible with BACCH 3D Sound. Since any speaker, by definition, will yield an acoustical response to an electrical input, the impulse response (IR) of any speakers can be measured by the BACCH-SP's (or BACCH-dSP’s) IR measurement system, which is based on exciting the speaker with am exponential sine sweep and obtaining the impulse response by deconvolution from the binaurally recorded response at the entrance of the ear canals of the listener. This can be done for any pair of speakers and the resulting IR will always yield a BACCH filter that will significantly reduce crosstalk, unless the early reflections in the room have as much energy as the direct sound (which is seldom, if ever, the case for audiophile listening). Therefore, we suspect that the rumor that the BACCH-SP "would not calibrate with expensive digital active speakers made by a certain Swiss manufacturer of luxury audio products” is a false rumor.
As described in our FAQ, https://www.theoretica.us/faq.html, (Question 3: Does BACCH 3D Sound require special speakers?), the main concern for good imaging (with or without BACCH) is decreasing the ratio of reflected-to-direct sound energy. As mentioned by Mr. Borduin, in his second posting above, speakers like D&D or Kii do indeed give excellent results with BACCH because of their ability (controlled directivity) to minimize room reflections.
Regarding Mr. Borduin’s note about the angle between the speakers, I quote from our FAQ: While previous non-optimized crosstalk cancellation (XTC) methods required the speakers to be placed very close to each other (the so-called “dipole” configuration), this is not at all the case with BACCH® 3D Sound. To a zeroth-order, speakers placement (excluding the effects of room reflections) does not matter with BACCH 3D Sound. The 3D image you get through a BACCH filter created for a given speakers configuration will be essentially the same as the image obtained for a completely different (even asymmetric) speakers configuration as long as the BACCH filter corresponding for that configuration is used for listening. With BACCH 3D Sound the speakers are only the conduit of the sound waves to the ears of the listener, who then perceives the location of the original sound sources in the recording and not the location of the speakers. This is often a startling fact for audiophiles as they are used to a phantom image that is anchored in the speakers and whose dimensions (essentially width) is strongly dependent on the speakers span. This is why the recommended speakers span for regular stereo listening is the 60 degrees of the equilateral triangle. With BACCH 3D Sound we recommend the same stereo triangle only because this is what audiophiles are used to and because it allows them to compare the sound through the BACCH filter to that without it (by hitting the bypass button) and hear the significant enhancement to the imaging compared to a case they are well familiar with.
In response to dallasjustice’s question about integration with Roon: Many of our customers (in fact most of our European customers) use Roon as their source of audio with their BACCH-SP or BACCH4Mac systems. The integration is seamless and very straightforward. In the case of the BACCH-SP, all three models (Grand, adio and dio) can be configured as Roon Core (with Roon running inside the BACCH-SP’s CPU) or Roon Bridge (Roon audio coming from a core on the LAN) and the Roon engine (including Tidal) is controlled through Roon Remote running on the same iPad that controls the BACCH-SP. The audio output of Roon becomes simply one of the available inputs to the BACCH Purifier. The same is the case for BACCH4Mac, where the audio from Roon running (as core or Bridge) on the Mac is routed to the BACCH-dSP application for processing.
As to applying the BACCH filter in Roon: While Roon DSP Engine has a Convolution option that allows it to apply a filter in IR form to the audio, to our knowledge it cannot do the so-called “True Stereo” (2x2) convolution required for BACCH filters. Moreover, unlike some other audiophile file players like Audirvana and Pure Music, Roon at present does not allow loading a VST or AU plugin. Theoretica offers owners of BACCH4Mac a free license for its VST/AU plugin, the BACCHp, which can be loaded in any player (or DAW) that accepts VST/AU plugins and applies any BACCH filter generated by the BACCH-dSP application, without the need to launch that application. (Of course, then, the head tracking capability of BACCH-dSP will not be available and the listener is constrained to a fixed sweet spot.)
The reason why BACCH-dSP is a standalone application is because many of its features (e.g. head tracking) would be difficult to implement in the form of audio plugins (which are not as well suited for video processing) and because it is more difficult. with such plugins, to optimize the processing over multiple cores/processors as it is presently done in BACCH-dSP.
It is relevant in this context to mention a feature, called “Output Plugin” that was recently added to BACCH-dSP, which is its ability to load any VST or AU plugin at its output stage. This was done in response to some of our customers who wish to apply room correction IR filters (generated by many RC packages) inside BACCH-dSP.
In response to Mr. Borduin’s comment regarding the CPU requirement for extra convolution, we should clarify that adding a convolution (even a True Stereo one) at the output stage inside BACCH-dSP using a VST/AU convolution plugin is easily handled by the CPU of the recommended Macs (recent, 2012 or later, Mac Pro, Mac Mini, iMac, or MacBook Pro with a minimum of 16 GB of RAM). In fact, the BACCH 3D mixer module of the BACCH4Mac Pro package, on which soundArgument reported on 2/27, can use up to 20 such convolutions, simultaneously, at sampling rates up to 96 kHz with no problems, thanks to the parallel processing architecture of the BACCH-dSP application.
We have never encountered or even heard of speakers that are incompatible with BACCH 3D Sound. Since any speaker, by definition, will yield an acoustical response to an electrical input, the impulse response (IR) of any speakers can be measured by the BACCH-SP's (or BACCH-dSP’s) IR measurement system, which is based on exciting the speaker with am exponential sine sweep and obtaining the impulse response by deconvolution from the binaurally recorded response at the entrance of the ear canals of the listener. This can be done for any pair of speakers and the resulting IR will always yield a BACCH filter that will significantly reduce crosstalk, unless the early reflections in the room have as much energy as the direct sound (which is seldom, if ever, the case for audiophile listening). Therefore, we suspect that the rumor that the BACCH-SP "would not calibrate with expensive digital active speakers made by a certain Swiss manufacturer of luxury audio products” is a false rumor.
As described in our FAQ, https://www.theoretica.us/faq.html, (Question 3: Does BACCH 3D Sound require special speakers?), the main concern for good imaging (with or without BACCH) is decreasing the ratio of reflected-to-direct sound energy. As mentioned by Mr. Borduin, in his second posting above, speakers like D&D or Kii do indeed give excellent results with BACCH because of their ability (controlled directivity) to minimize room reflections.
Regarding Mr. Borduin’s note about the angle between the speakers, I quote from our FAQ: While previous non-optimized crosstalk cancellation (XTC) methods required the speakers to be placed very close to each other (the so-called “dipole” configuration), this is not at all the case with BACCH® 3D Sound. To a zeroth-order, speakers placement (excluding the effects of room reflections) does not matter with BACCH 3D Sound. The 3D image you get through a BACCH filter created for a given speakers configuration will be essentially the same as the image obtained for a completely different (even asymmetric) speakers configuration as long as the BACCH filter corresponding for that configuration is used for listening. With BACCH 3D Sound the speakers are only the conduit of the sound waves to the ears of the listener, who then perceives the location of the original sound sources in the recording and not the location of the speakers. This is often a startling fact for audiophiles as they are used to a phantom image that is anchored in the speakers and whose dimensions (essentially width) is strongly dependent on the speakers span. This is why the recommended speakers span for regular stereo listening is the 60 degrees of the equilateral triangle. With BACCH 3D Sound we recommend the same stereo triangle only because this is what audiophiles are used to and because it allows them to compare the sound through the BACCH filter to that without it (by hitting the bypass button) and hear the significant enhancement to the imaging compared to a case they are well familiar with.
In response to dallasjustice’s question about integration with Roon: Many of our customers (in fact most of our European customers) use Roon as their source of audio with their BACCH-SP or BACCH4Mac systems. The integration is seamless and very straightforward. In the case of the BACCH-SP, all three models (Grand, adio and dio) can be configured as Roon Core (with Roon running inside the BACCH-SP’s CPU) or Roon Bridge (Roon audio coming from a core on the LAN) and the Roon engine (including Tidal) is controlled through Roon Remote running on the same iPad that controls the BACCH-SP. The audio output of Roon becomes simply one of the available inputs to the BACCH Purifier. The same is the case for BACCH4Mac, where the audio from Roon running (as core or Bridge) on the Mac is routed to the BACCH-dSP application for processing.
As to applying the BACCH filter in Roon: While Roon DSP Engine has a Convolution option that allows it to apply a filter in IR form to the audio, to our knowledge it cannot do the so-called “True Stereo” (2x2) convolution required for BACCH filters. Moreover, unlike some other audiophile file players like Audirvana and Pure Music, Roon at present does not allow loading a VST or AU plugin. Theoretica offers owners of BACCH4Mac a free license for its VST/AU plugin, the BACCHp, which can be loaded in any player (or DAW) that accepts VST/AU plugins and applies any BACCH filter generated by the BACCH-dSP application, without the need to launch that application. (Of course, then, the head tracking capability of BACCH-dSP will not be available and the listener is constrained to a fixed sweet spot.)
The reason why BACCH-dSP is a standalone application is because many of its features (e.g. head tracking) would be difficult to implement in the form of audio plugins (which are not as well suited for video processing) and because it is more difficult. with such plugins, to optimize the processing over multiple cores/processors as it is presently done in BACCH-dSP.
It is relevant in this context to mention a feature, called “Output Plugin” that was recently added to BACCH-dSP, which is its ability to load any VST or AU plugin at its output stage. This was done in response to some of our customers who wish to apply room correction IR filters (generated by many RC packages) inside BACCH-dSP.
In response to Mr. Borduin’s comment regarding the CPU requirement for extra convolution, we should clarify that adding a convolution (even a True Stereo one) at the output stage inside BACCH-dSP using a VST/AU convolution plugin is easily handled by the CPU of the recommended Macs (recent, 2012 or later, Mac Pro, Mac Mini, iMac, or MacBook Pro with a minimum of 16 GB of RAM). In fact, the BACCH 3D mixer module of the BACCH4Mac Pro package, on which soundArgument reported on 2/27, can use up to 20 such convolutions, simultaneously, at sampling rates up to 96 kHz with no problems, thanks to the parallel processing architecture of the BACCH-dSP application.
dallasjustice
Major Contributor
This sounds like the best way to use convolution with Baccch. One could load a VST convolver as the “output plugin” in the BACCH-DSP. One could still use Roon and output to the BACCH-DSP. Am I understanding this right? Would the output plugin be able to handle channel routing/mixing as well?
Your understanding is correct.
BACCH-dSP's Output Plugin (OPI) is applied to the stereo output of BACCH-dSP. At present both the inputs and outputs of the OPI are stereo. The stereo output can be routed to any of the channels of the output device using BACCH-dSP's "I/O Mappings" capabilities. We could consider implementing multi-channel capabilities for the OPI to accommodate plugins with multiple outs (e.g. a crossover plugin) if this is deemed very useful for our customers.
William Guiracoche
Development Engineer
Theoretica Applied Physics
BACCH-dSP's Output Plugin (OPI) is applied to the stereo output of BACCH-dSP. At present both the inputs and outputs of the OPI are stereo. The stereo output can be routed to any of the channels of the output device using BACCH-dSP's "I/O Mappings" capabilities. We could consider implementing multi-channel capabilities for the OPI to accommodate plugins with multiple outs (e.g. a crossover plugin) if this is deemed very useful for our customers.
William Guiracoche
Development Engineer
Theoretica Applied Physics
dallasjustice
Major Contributor
Thanks William. I have a 4way setup with FIR crossovers.
I am happy BACCH has moved into software. I do think it would be best if BACCH built its own convolver to be used inside the BACCH-DSP. That would open the door to a wide range of setups, room correction softwares and routing. A convolver that accepts .wav impulses and .cfg files (e.g. Jriver or Roon Convolver) would work best, IMO.
Thank you.
Michael.
As I mentioned above BACCH-dSP can apply any VST/AU plugin at its output stage, but the present implementation is for a stereo output only. We are thinking of adding a crossover module at the output stage of BACCH-dSP that allows applying convolution-based crossover, mixing and routing. Would you please consider contacting us offline at [email protected], as someone who uses multiple-channel room correction and crossover, to share with us ideas on how best to configure such a module for audiophiles who have such set-ups?
William Guiracoche
Development Engineer
Theoretica Applied Physics
As I think about this, one basic challenge with incorporating additional convolution into BACCH might be head tracking. Simply adding post-BACCH FIR convolution in the form of VST plugin or similar would add unknown delays to the signal processing chain, and thereby delay reaction to head movements. I don't know the psychoacoustics well enough to say whether this would be a problem, but I suspect so based upon what I've read of small head movements being important to sound localization. If my speculation is true, then BACCH would need to have tight coupling to any additional convolution to account for signal processing delays.
EDIT: Never mind. Thinking about it just a bit more, any convolution delay would be constant and could easily be accounted for as part of the measurement/calibration process. In fact there might be some processing delay in SoundArgument's 8C speakers ...
Last edited:
Similar threads
