It doesn't. As you already know, I definitely am not the expert on the topic. IR to me is a concept that is core in nature in the capturing of an acoustical signature. Basic.
(Sorry, I edited my post #186 after you replied; please measure HF extension if possible.)
Perhaps your added early front-wall "diffusion"/"liveliness" broadened HF dispersion hence your "sweet spot". Late side-wall reflections however would be partially absorbed/diffused/randomized so "phase shifts" would be even less directionally useful....
The OP's claim seems reasonable to me. The larger midrange (or horn) will have higher sensitivity (dB/Vrms) and more directivity than a smaller driver. To produce a given SPL (100dB) the smaller driver will be working alot harder (THD, cone movement) with it's lower directivity in the midrange freqs. The bigger driver (or horn) should have a cleaner and more directive pattern, so likely a better image.
Hi, in my case I did not treat the front wall (actually a large window pane so impossible to treat), but i treated the back wall and the late side-wall reflections with some mf/hf absorption and diffuser panels, in order to create uncorrelated late reflections coming from the front wall and front part of the room (after having hit the back wall and side walls). I only had the front corners treated with basstraps and partial diffusion in the front corners. Also partially treated the ceiling. So I can relate to your statement 👍
I no longer live there and am in the process of moving to a new home. I have some measurements from my system and pictures of the room with treatments somewhere on this forum that show an even and short waterfall decay and GD but how it is perceived is another story. I'll measure again when I am done with my new listening room.
"Well I've stated many times "my theory" (something really mundane) that "great imaging" came from >10khz very high frequencies in phase (better, time-aligned as well) with lower frequencies forming a coherent bundle of fundamental/harmonics. Same is required for great details and transcient response. As we all know very high frequency soundwave is extremely directional." (wchang)
^This is true I think, but only for the direct sound; as soon as the bouncing of the walls and ceiling starts, the phase relationship between fundamentals and harmonics blur. To start with a coherent speaker is a good idea for imaging, however. I am tempted to say that the build of a good system starts with the acoustics (preferably via measurements), and then choosing the right speaker for the room. Then proper imaging will not be a problem in general.
Last edited:
Stereo → Binaural (Front Imaging)
- HRTF Processing:
- Use personalized or generic HRTFs (e.g., Smyth Realiser A16, Waves Nx) to spatialize stereo into binaural.
- Tools: Impulcifer (free), DearVR Pro ($299).
- Crosstalk Cancellation (CTC):
- Apply CTC algorithms (e.g., BACCH SP) to simulate frontal "speaker-like" imaging by canceling interaural crosstalk.
- Early Reflections:
- Add 10-20ms room reflections (e.g., Convology XT) to externalize sound beyond the head.
- Mid/Side Processing:
- Isolate center channel (mid) for frontal focus, apply HRTF to sides for width.
Binaural → Stereo/Multi-Channel
- Transaural Playback:
- Use CTC (e.g., BACCH SP) to cancel speaker crosstalk, preserving binaural cues over loudspeakers.
- Ambisonic Decoding:
- Convert binaural to Ambisonics (B-format), then decode to 5.1/7.1 (e.g., Reaper + IEM Plugins).
- Downmixing:
- Extract direct sound (dry) and diffuse field (wet) via Dolby Atmos Renderer, map to speaker arrays.
Key Science:
- Frontal imaging requires ITD/ILD + pinna cues (HRTF) and early reflections (≤20ms).
- Binaural-to-speaker CTC demands head tracking (e.g., Waves Nx Head Tracker) for accuracy.
- Record stereo with Impulcifer (HRTF measurement).
- Process via Reaper + ReaVerb (convolution).
- For playback, use BACCH4Mac (CTC).
The more directivity, the lesser expectation of room treatment. That is so alluring to me.
Been there, done that. It’s OK, not sure 85%. I was hoping for better, but it’s a step in the right direction. I do know that neat perfection can be achieved.
Listening to anything in the absence of environmental feedback mechanisms is uncomfortable, or outside of the sensory routines we have developed over years of development and practiced learning........it may be fun to go down the rabbit hole, i just don't think you'll get many folks to join you for the sake of it.
I'll lay something out there if you wanna experience something along the imaging front lines........listen in a space with a dome ceiling and a stereo pair of speakers placed 180 degrees apart and upfiring...........you'll be spoiled for life.
While I hear this often, many now have experience with noise cancelling headphones which demonstrate this point fairly well. While it's a little unusual at first, it's hardly uncomfortable and we can clearly hear how the signal to noise ratio improves the listening experience.
With headphones, maybe not. We have closed shell headphones for many years that keep noise out reasonably well. But the BBC document from post #191 clearly states that the trend towards "dead"control rooms with speakers made people feel uncomfortable. It can be a personal preference, though.
And: "One anticipated problem arose from the knowledge
that the stereophonic illusion did not work so well in
an anechoic environment unless the listener was very
close to the centre-line between the loudspeakers." (quote)
Quoting an earlier post..
One advantage of the approach I use is that the power is tonally matched to the direct, which is one of the factors important for making the room not call attention to itself.
This is how I address the issue while improving imaging. Whereas the paper talks about using the room to impose directivity, it is now often done using directive speakers.
One advantage of the approach I use is that the power is tonally matched to the direct, which is one of the factors important for making the room not call attention to itself.
I think most people are more bothered by the small sweet spot that can accompany very high DI. I don't but some obviously do. When you look at it simply, there is a Direct Sound level and an Indirect sound level. I see people displaying it as a ratio sometimes... 60/40 Direct/Indirect I think someone commented above talking about the same thing....
In regards to what I am saying about directivity. You can use directivity to create a 60/40 Direct:Indirect sound ratio at the listening point. Without ever touching room treatment, you will still have an increase in Clarity comparable to the 60/40 situation created with room treatment. Upon applying room treatment to the high directivity situation, since there is less energy to attenuate to start with, you will need less acoustic treatment as a result if you even feel the need room treatment at this point... At least that looks good on paper.
I think it really is a matter of RT60. I think having short enough RT60 times is more potent to increasing accuracy of the signal perception than getting the reflections to be even toned. I recall vaguely, discussion that the brain is able to separate the room from Direct signal with a few conditions being met, which involve attenuating early reflections. The high DI situation, may start off in a position of high clarity due to this making any room treatment, more focused on the area where the system loses DI or becomes modal according to the room.
With high DI speakers (which I prefer -like camplo), the problem I experienced is due to fewer, but louder first reflections from backwall and sidewall near backwall sitting plm 3 ft/1m from backwall; thus causing some kind of comb filtering effect (head in vice, or image suffers). I was able to solve this however, so with proper room treatment you can have the clarity and imaging, and the room out of the way.
Last edited:
this does not make sense to me; why would there be fewer, but a select few that are louder? Isn't the whole idea of high DI to reduce room reflections as uniformly over as wide a frequency range as possible? How would a high DI speaker have prominent backwall reflections?
A speaker with high Directivity Index project it's sound energy in a more or less narrow beam, so it creates less side reflections. This is indeed frequency dependent and you want an even dispersion over as much of the frequency band as possible, whether it is a narrow beam or a wide beam with constant DI. With high DI speaker I meant that the sound energy is more aimed at the listening spot and close around it, maybe I am not describing well or mixing things up, I'm sorry for that... semantics and all. My speakers do not have a constant DI but are more and more beaming towards higher frequencies. In this case you get fewer reflections to the side in the treble range, but more from the back wall and surrounding walls. It works out fine for the on axis frequency as the horn start losing efficiency in the higher frequency and this beaming matches the gain from the horn in lower frequencies. But the rise in DI poses problems that I fixed with diffusion and doing so, the sweetspot widened and imaging improved. A constant DI is preferable whether it is a narrow or wide dispersion, but not always attainable depending on the horn design. I hope I made myself understandable.
(source: https://www.audioholics.com/loudspeaker-design/understanding-loudspeaker-measurements)
Since this is a DIY speaker forum, I’ve always accepted the reality in that the performance is a combination of the speaker and the room. And as this bond can never be broken, the best chances for success are to optimize each to work with the other to create an environment. I mentioned Polk SDA tech earlier…….a GREAT example of a system that does work really well in rooms that meet some basic criteria not requiring extensive treatment……space and symmetry. Smaller spaces that remain symmetric can simply scale down the system. Non symmetric spaces would require a more robust, active approach that does not yet exist other than in the massive and super expensive B&O monstrosity….and this tech is beyond what most here would be willing to tackle to the point of practicality. Barefoot does their thing…..but it’s too clinical for the audiophile elite and they’ve failed at their crossover attempt.
Modified or personalized SDA…..that’s the answer.
Modified or personalized SDA…..that’s the answer.
Horneydude, with the typical 90 degree or so directivity you should be able to illuminate the entire area of the five room boundary surfaces behind you, naturally leading to a diversity of timing and angles.
I find that designing broadband for much higher DI than that starts to become exceedingly difficult to achieve in the given space using conventional methods.
I find that designing broadband for much higher DI than that starts to become exceedingly difficult to achieve in the given space using conventional methods.
Mayhem, are u still working on extracting carusos voice? Just want to hear that, since u been doing it for years.