No you haven't ! Where is the pillow ?
Honestly it looks more like toe up. Unless you got that mushroom to work amazingly against physics?
Measured the unequalized device here.
How much in dB is low enough?
No you haven't ! Where is the pillow ?
Honestly it looks more like toe up. Unless you got that mushroom to work amazingly against physics?
I suspect you got too much of direct sound there.
Did you measure the level of direct sound and wall reflections ?
- Elias
I think the "mushroom" is what I thought had a nice frontal radial pattern (..and vertical as it turns out as well).
Basically he has 2 180 deg frontals side-to-side, and aimed at their respective walls (or ipsalateral). A radial rather cardiod horizontal pattern I'd imagine.
The front wall has an absorber.
Because of the positioning in the center of the room, the loudspeakers are more distant from their respective side wall and as a result presents less reflective pressure than a speaker closer to it - particularly below 2 kHz for an average design with traditional stereo triangle.
In other words its pretty much an ideal.
Again, for the design and it's implementation: thumbs-up!
Many things will be better if you first remove the stereo cross talk. Ever tried Ambiophonics yet? There is a running thread here on that too.
- Elias
- Elias
I have to admit that when I listen to any audio system, I am always listening for specific properties, image localization over frequency, etc. It's hard for me to just listen and not think about specifics. When I have a non-educated listener come in and listen, they only know how to listen in a general way. I think there is still value in getting their opinions, as long as I know how to interpret what they say.
Toole, by the way, would probably be the first to admit that his experiments are riddled with variables, and not God's word. We are all on a path of learning. Nobody knows it all.
Toole, by the way, would probably be the first to admit that his experiments are riddled with variables, and not God's word. We are all on a path of learning. Nobody knows it all.
There's a button for that too. Never tried the exact ambiophonics yet. What are the recording requirements for it? I know it's different than standard stereo.
Dan
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-13.html#post2692605
In my case 5-10 dB lower direct sound compared to side wall reflections was enough to hide the speaker itself to enable imaging. That was above about 700 Hz. I was only able to achieve that with the pillow fast and easily for the test purposes.
- Elias
..that's weird.
When I tried these sorts of experiments I always had imaging. Side barrier, frontal barrier, even in a different room to some extent.
Assuming your experiment with the sofa pillow was with directive drivers at higher freq.s.. I wonder what it would be like for you if you put absorbers on the walls? Perhaps adjusting output above 700 Hz and listening to the differences.
I sort of did that, at least removed all the side wall reflections:
http://www.diyaudio.com/forums/multi-way/10962-stereolith-loudspeakers-question-10.html#post2529717 post #472
After this there was sound coming only from the box, no imaging of any kind when panned signal left to right.
- Elias
Nonetheless, voting for anechoic says to me someone has spent too much time in highly treated mixing studios or mixing with headphones.
Like I said, I don't like / understand the design of this experiment. And if the finding was that *anything* other than anechoic or maximum HF dispersion was rejected, then I am really suspicious about either the experimental design or my comprehension.
Markus speaks the truth. It is bizarre that anyone thinks multichannel forces the listener to be surrounded with the instruments. Perhaps a naive multichaneller?
The point is that if your speakers are pointing straight down the room there will be a lot of near wall reflection and a little far side reflection, but if you toe them in the loss of near wall reflection will be greater than the gain in far wall reflection, so sum of L and R side wall reflection decreases, i.e. toe-ing in causes a reduction in side wall reflections energy. Am I wrong?
Three times repeating the point now?Like I said on page 2, post #62.There isn't one, stereo reproduction is flawed.
OK that's the fourth time
Last edited:
Cool, thanks for the link. A lot to read so give me a while before I give'r a go. Seems a lot like what Princeton was doing. Brings me back to what I love about binaural recording/playback. Nothing more chilling than hearing a large dog growl 2 meters behind you from a binaural recording.
Dan
It's been said before and it will be said again. It is a truism that needs to be restated from time to time.
Last edited:
If imaging in the lower midrange is sensed primarily by timing comparisons rather than amplitude comparisons, AND those timing comparison imaging cues get corrupted by the additional set of timing cues that result from inter-aural crosstalk in a playback scenario, then it would make sense that latteral room reflections would stand a chance of re-creating a sense of distinct timing cue information in the lower midrange frequencies that somehow dominates the more two dimensional cues that are coming out of the speakers.
My system makes the most of embedded timing cue info by having substantial cancellation at the sides below about 500HZ, since they are an open baffle vertical line array (from 100HZ to 1.4kHZ). Open baffle/dipole speakers generate fake ambience on the Z axis but have rolled off projection to the side walls of the room below about 500HZ. I add in my own brew of the Carver Holographic generator circuit to substantially reduce the inter-aural crosstalk, and I've got embedded imaging cues and fake recording studio imaging cues putting images all over the room. Of coarse the hologram only works if you sit within a few inches of the right place, but I usually sit in a specific place when doing serious listening anyway.
Perhaps most people would virtually never sit in the exact right place, or they want it to be at its best for other listeners in the room, in which case I'd say forget the hologram and if the room does it well, go for those latteral side wall reflections. Maybe even toe in the speakers more to vary the ratio of direct to reflected. If inter-aural crosstalk and room acoustics are going to do a lot of damage to a reproduction anyway (and they usually will), I'd recommend to strive for technical accuracy as we know it, as a default (a high resolution digital EQ, set with pink noise and a calibrated mic, for example), and then adjust tone controls or whatever for the most enjoyable sound.
If the room has really bad acoustics, I'd look at either controlled directivity speakers that aim most of the sound at the listeners who are substantially acoustically absorbative, or lots of smaller speakers placed all over the room. That's what Don Davis (Syn-Aud-Con seminar) taught us they do in churches and other large highly reverberant rooms. I don't have a better simple idea for highly reverberant rooms.
My system makes the most of embedded timing cue info by having substantial cancellation at the sides below about 500HZ, since they are an open baffle vertical line array (from 100HZ to 1.4kHZ). Open baffle/dipole speakers generate fake ambience on the Z axis but have rolled off projection to the side walls of the room below about 500HZ. I add in my own brew of the Carver Holographic generator circuit to substantially reduce the inter-aural crosstalk, and I've got embedded imaging cues and fake recording studio imaging cues putting images all over the room. Of coarse the hologram only works if you sit within a few inches of the right place, but I usually sit in a specific place when doing serious listening anyway.
Perhaps most people would virtually never sit in the exact right place, or they want it to be at its best for other listeners in the room, in which case I'd say forget the hologram and if the room does it well, go for those latteral side wall reflections. Maybe even toe in the speakers more to vary the ratio of direct to reflected. If inter-aural crosstalk and room acoustics are going to do a lot of damage to a reproduction anyway (and they usually will), I'd recommend to strive for technical accuracy as we know it, as a default (a high resolution digital EQ, set with pink noise and a calibrated mic, for example), and then adjust tone controls or whatever for the most enjoyable sound.
If the room has really bad acoustics, I'd look at either controlled directivity speakers that aim most of the sound at the listeners who are substantially acoustically absorbative, or lots of smaller speakers placed all over the room. That's what Don Davis (Syn-Aud-Con seminar) taught us they do in churches and other large highly reverberant rooms. I don't have a better simple idea for highly reverberant rooms.
The way I would look at the results is that when asked to vote on which sound presentation they "preferred", there was a tendency to vote for what they were used to and most exposed to.
"Naive listeners" had almost certainly never heard anechoic or highly damped / highly directional speaker configurations with very high direct/reflected ratio before. Because it sounded "alien" to them, they all voted for the response with lots of reflections because that is what they're familiar with and used to listening to at home.
Of those professionals that had been exposed to both scenarios in their work, the results were almost 50/50, which suggests that its more a matter of taste and expectation rather than one being better than the other.
To dig deeper into this 50/50 mix it would be necessary to see whether their personal music tastes (or professional mixing history) affected this result - for example some types of music benefit from minimal added room signature (generally those with a lot of reverberant field contained in the recording) and when trying to master such a recording added reflections in the listening room are a hindrance.
Make no mistake - added reflections and reverb in a typical home living room mask many small details - its easy to hear small details in ambience of some recordings on headphones that simply get lost in the reverb tail in typical speaker listening due to room reflections.
On the other hand recordings that are fairly dry and consist mostly of artificially mixed sound sources tend to benefit from a bit of room reverb to liven them up.
It's not really a matter of how "live" the reverberation of the sound should be, but rather should the recording be the major contributor or the room. Nobody likes the sound of a recording with no reverb played in an anechoic chamber, and I don't think anyone here is suggesting that as a good idea. Even those of us who like directional speakers know full well that there are still room reflections and reverb in the room, albeit a bit more controlled. (And directional speakers are only ever directional above a certain frequency, which is not the same as testing a wide dispersion speaker in an anechoic chamber, where there are no reflections at any frequency)
The real question in this thread is high direct/reflected ratio versus low direct/reflected ratio, not no reflections at all versus lots of reflections. Nobody is going to turn their living room into an anechoic chamber, so the fully anechoic response is irrelevant to the discussion.
Last edited:
I'm reading through the Flindell paper to try and make a little more sense of it. As previously mentioned the added reflections could be rolled off from a range of corner frequencies to, in effect, make the simulated loudspeaker more or less directional at high frequencies. There were two trials, one with a 6dB per Octave slope for all rolloffs and a second with 3dB per Octave rolloffs. In addition there were 4 types of music and a 5th selection of solo speech.
Although I mentioned that there was a bimodal preference for the two directivity extremes amongst the professional listeners, it isn't bimodal amongst the individual choices. For example solo speech is strongly preferred with the fully rolled off settings (most anechoic). With the slower rolloff trials the professional group leaned towards wider dispersion, at least with the non speech selections. With the steeper rolloff cases there was no preference for wide simulated dispersion.
The naive group preferred wide dispersion fairly uniformly across the two rolloff rates and for all types of music.
Regarding the rating scale portion of the test, the listeners were asked to rank the various options on a 1 to 10 scale for "image quality", "spaciousness", and "overall impression". The results for all listeners and all options are pretty well random and in the 5 to 6 range. Note that not only does this not show that stereo imaging was decreased with increased reflections, neither does it show that spaciousness was increased. (Should I put that in the largest font available?) It also doesn't give a higher numerical ranking for wide dispersion which the naive group clearly preferred in the previous forced choice test.
Careful what you conclude from that.
David S.
Although I mentioned that there was a bimodal preference for the two directivity extremes amongst the professional listeners, it isn't bimodal amongst the individual choices. For example solo speech is strongly preferred with the fully rolled off settings (most anechoic). With the slower rolloff trials the professional group leaned towards wider dispersion, at least with the non speech selections. With the steeper rolloff cases there was no preference for wide simulated dispersion.
The naive group preferred wide dispersion fairly uniformly across the two rolloff rates and for all types of music.
Regarding the rating scale portion of the test, the listeners were asked to rank the various options on a 1 to 10 scale for "image quality", "spaciousness", and "overall impression". The results for all listeners and all options are pretty well random and in the 5 to 6 range. Note that not only does this not show that stereo imaging was decreased with increased reflections, neither does it show that spaciousness was increased. (Should I put that in the largest font available?) It also doesn't give a higher numerical ranking for wide dispersion which the naive group clearly preferred in the previous forced choice test.
Careful what you conclude from that.
David S.
Your questions deal with very basic stuff. If you want to join the discussion, you might want to first educate yourself a little. You should buy Floyd Toole's Sound Reproduction. Parts of it can be found here:
Sound reproduction: loudspeakers and ... - Google Boeken
Here's a good summary of the issues we're discussing: http://www.aes.org/e-lib/download.cfm?ID=13686&name=harman
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.