The problem with this is that 2 channel reproduction isn't a perfect approach. If all additions are undesirable then listening to two speakers in an anechoic chamber would have to be the answer. We know from trying that that it isn't very nice. Imaging falls within our head and there is no good sense of being in an appropriate space.
Although it seems to fly in the face of an ideal of never adding to what was recorded at the original venue, we can improve 2 channel reproduction with a multiplicity of room reflections. Accuracy may diminish but "plausibility" can be enhanced.
David S
It has been mentioned before:
John Watkinson demonstrates his speakers with recordings made on the spot of real musicians. So listeners can hear the real thing and the recording of it.
The demo's show no sign of spreading of sound sources at all and are very very much like the real event with only hints of harmonic distortion. If you use coincident microphone techniques his speakers can reveal every individual musician in the recording with no added artefacts from the listening room.
How is this possible? Humans have evolved in a reflective environment and we can distinguish direct sound from good reflections.
John Watkinson demonstrates his speakers with recordings made on the spot of real musicians. So listeners can hear the real thing and the recording of it.
The demo's show no sign of spreading of sound sources at all and are very very much like the real event with only hints of harmonic distortion. If you use coincident microphone techniques his speakers can reveal every individual musician in the recording with no added artefacts from the listening room.
How is this possible? Humans have evolved in a reflective environment and we can distinguish direct sound from good reflections.
yes, but more precisely it is not a function of location but of direction
so yes, one can get a sense that besides phantom source there are two split sound sources but still without floor reflection it is not possible to identify the speakers as those sources
on the other hand, the direction of those split soundsources is the direction of the speakers, what is not affected is the perception of depth and height of virtual soundstage and phantom position because the crosstalk comb filter provides no information as to the distance and height of the speakers as real sound sources, only the direction of them
but I think that when we add to the auditory direction cue from the crosstalk comb filter the visual cue then one can get an impression that one is able to localise the speakers as real sound sources
this is I believe what happened in Elias' experience with stereo flooder
Last edited:
..you weren't alone in spotting this.
even Markus agrees though strongly dissents:
Last edited:
Science in that paper?
Just to return briefly to the last part of the paper that started this thread.
The claims about Barkhausen limitations in loudspeakers are unsubstantiated and actually contradict the science as I understand it.
Did a search for prior threads but not much science there either.
Anyone here seen any substantiation or scientific support for the claim?
David
Just to return briefly to the last part of the paper that started this thread.
The claims about Barkhausen limitations in loudspeakers are unsubstantiated and actually contradict the science as I understand it.
Did a search for prior threads but not much science there either.
Anyone here seen any substantiation or scientific support for the claim?
David
this has been discussed here:
http://www.diyaudio.com/forums/multi-way/35936-best-sounding-magnet.html
I read an article in Electronics World by John Watkinson about loudspeaker magnets, as part of a series on loudspeakers. He replied to a letter criticising his comments on magnetic materials, saying;- "ceramic magnets are good.. for picking up swarfe when I'm machining the pole peices for my neodymium magnets".
He argues that ceramic is an insulator, while neodymium is a conductor, and so it prevents - if I remember the spelling - Barkhaussen noise. It's the movement of the magnetic domains apparently, and is also found in tape heads.
best summary:
John Watkinson is a F.A.E.S., so it's true. In this series, he was being deliberately provocative, calling loudspeaker design a subset of carpentry.
He was trying to inspire original thinking by pointing out the inadequacies of current technology, of which the magnetic circuit is a part. It was a very interesting series, but nothing to get too worried about.
Anechoic?
No, what we know is that
a) anechoic spaces can be disorienting and
b) (almost) no one listens to their stereo in an anechoic space.
That is not 'using an anechoic space is counterproductive to sound reproduction'
And the rest of your argument goes downhill from there.
It is exactly the early reflections which screw with your mind's attempt to visualise the space the music was recorded in.
I have no objection to adding reverb for euphonic effect but don't confuse that with improving accuracy. Especially if that euphonic effect is an attempt to cover blunders made in recording or compromised equipment design.
No, what we know is that
a) anechoic spaces can be disorienting and
b) (almost) no one listens to their stereo in an anechoic space.
That is not 'using an anechoic space is counterproductive to sound reproduction'
And the rest of your argument goes downhill from there.
It is exactly the early reflections which screw with your mind's attempt to visualise the space the music was recorded in.
I have no objection to adding reverb for euphonic effect but don't confuse that with improving accuracy. Especially if that euphonic effect is an attempt to cover blunders made in recording or compromised equipment design.
I very much agree thoglette. Stereo productions could rely on a specific reflection patterns in the consumer's room. But, this simply is not the case with current production techniques. That's why I earlier made the claim that (near) anechoic conditions are the goal in stereo reproduction (for highest accuracy).
Control rooms are generally leaning towards anechoic behavior or towards something that "doesn't interfere with the acoustics of the recorded space".
Even something like the Blackbird studio C room has reflections reduced by 30dB (!). This is more than most LEDE/RFZ rooms.
Control rooms are generally leaning towards anechoic behavior or towards something that "doesn't interfere with the acoustics of the recorded space".
Even something like the Blackbird studio C room has reflections reduced by 30dB (!). This is more than most LEDE/RFZ rooms.
Last edited:
Thanks for the reply.
My search found that discussion but there's not much science and no substantiation for the claim.
A shame that a paper that nominally "puts the science back into loudspeakers" ... doesn't.
Looks like yet more pseudo-scientific waste of time.
David
Isn't that exactly what I said?
I've spent enough time in anechoic chambers that I always found it a pleasant experience, peace and quiet! I have not set up stereo in one but there have been plenty of reports of the "imaging in your head" effect. I have listened to binaural by standing between, or nearly between, my speakers and the imaging is within the head.
I am happy to listen with headphones and consider them one end point of the continuum of higher directivity/reduced room reflections. The plus is that you have great seperation of as many sources as you can imagine from far left to far right. The down side is that the line of sources swings through your head.
I have quoted several times an experiment where a helper could lower two sheets of fiberglass to the sides of my head while I listened to normal stereo. Their position would block side wall reflections but not sound from the speaker. Although you were not previously aware of the wall reflections, their removal did flatten the image and make it a little less "real".
Its fine to look at waveforms and say "see, the signal is more accurate when I remove all room reflections", but this isn't realism. There is no musical experience (other than outdoor concerts?) where sound comes from a frontal quadrant only. Every study of concert hall acoustics from the 70's on reemphasizes that lateral reflections (from well outside the LR speaker positions) is crucial to the experience.
If you only have two channels then room reflections enhace the listening experience.
David S
Actually, on reflection, it is. I apologise.
It's been a long time since I was in an anechoic chamber but I recollect a similar impression of calm.
I did spent a fair bit of time trying to emulate one by setting up gear on an open lawn. Your comment on "flattening" of the sound (later on) is familiar.
Which is one of the reasons I bought a HeadRoom Airhead a long time ago.
That's a very interesting line of argument: that adding late reflections from way outside the front quadrant better simulates the original than nothing .
I accept the limitations of stereo's failure to reliably reproduce vectored sound but had not really considered the implications of head transfer function on the original vectored sound, particularly where the sound vector is well outside the frontal quadrant.
It points to a reason for the general preference for dead end/live end listening spaces.
It might be putting cheese and sauce on an imperfect schnitzel but if it tastes better....
Last edited:
Wouldn't the fiberglass introduce it's own reflection, just much closer?
I've been following this whole discussion with much interest. Admittedly, I'm not a scientist when it comes to this stuff, but I've always been of the mind that our listening environment should really strike a balance as far as reflected energy. Just enough so that a recording takes on a gentle flavor that identifies it as being in your room, but not so much that the intended impression of the recording frankensteins into something new.
It's like the analogy that pops up every once in a while. You wouldn't want to repaint a work of art to change the colors or add details you prefer, but if you hung it in your room, there's nothing wrong with making sure that the lights you hang around it cast the perfect amount of light to make it look good - and those lights may very well be different in someone else's room.
Depends on the material's flow resistivity. If it's not too high then the wave passes through the material and part of the energy is converted to heat.
I had great success reducing early reflections by placing 10cm thick Basotect boards around the speaker. If I find the time I'll build an absorptive "funnel" around my speakers. Currently they wear "hats":
Attachments
Last edited:
I'm happy that the discussion has come around to headphones. I was going to ask - has anybody here experimented with capturing their head related transfer functions (HRTFs) and used them with convolution? My route to loudspeakers actually came via headphones and the desire to have a nice measurement setup to emulate the Smyth Realiser. I need to finish my PhD thesis so for now I've only played around, capturing a few impulse responses with in-ear microphones. At the moment I EQ Dolby Headphone to my diffuse field HRTF which works pretty well.
I like the idea of virtualising speakers that defy physics and having control of cross talk etc. It seems to me that a lot of Realiser users miss the chance to have totally impossible speaker setups emulated.
I've also used headphones and speakers together a lot in the past. I live in an apartment so I don't like to make too much noise. If you run a virtualiser through your headphones and use L&R or just a center speaker you can pull the image out of your head. For this it's nice to use something very open like Koss Portapro so that the speaker's FR is not too much attenuated by the headphones. You need to be careful with EQ, of course. I've probably gotten the best immersion in movies without high SPL using this method.
At first I was concerned that the delayed sound from the speaker would mess things up, but in the end the precedence effect saves the day ala Linkwitz' Watson.
I like the idea of virtualising speakers that defy physics and having control of cross talk etc. It seems to me that a lot of Realiser users miss the chance to have totally impossible speaker setups emulated.
I've also used headphones and speakers together a lot in the past. I live in an apartment so I don't like to make too much noise. If you run a virtualiser through your headphones and use L&R or just a center speaker you can pull the image out of your head. For this it's nice to use something very open like Koss Portapro so that the speaker's FR is not too much attenuated by the headphones. You need to be careful with EQ, of course. I've probably gotten the best immersion in movies without high SPL using this method.
At first I was concerned that the delayed sound from the speaker would mess things up, but in the end the precedence effect saves the day ala Linkwitz' Watson.
I do own a Realiser and yes, I do miss that feature. Stephen did some enhancements though that actually allow one to use one single speaker for all directions.
So I have been watching this thread and wondering where it is going. There is some interesting psycho acoustical information going back and forth but here is my question. Disregarding the extremely limited audiophile users and looking at the common consumer where all the money is in the reproduction of music, what are the proposals here? While it is nice to talk about room treatments and live/dead rooms and anechoic chambers this is just not realistic in any sense in a normal home environment. So how do you take the information and use it to produce any usable application in the consumer market? I will give you that you do have an emerging market in the large scale acceptance of headphones but what changes to the physical design of headphones would be necessary or are all we are talking about are differences in recording technique? Two channel stereo production isn't going away anytime soon. Multichannel reproduction with the extremely cheap surround sound loudspeakers with limited frequency response and terrible directivity are not going to get you to where this discussion seems to be trying to drive the conversation. We are talking about extremely accurate sound producers to create any of the illusions that I think are being suggested here. So would someone please elucidate with a clear and concise few sentences what is the goal of all this? If it is only for a select few then it will never happen in a commercial sense, pure economics will determine if these ideas are viable.
Steven
Steven
even Markus agrees though strongly dissents:
Note: I was referring to the flaw in Tom's logic with stereo and the inability to improve it. Unfortunately I used Speaker Daves's entire quote - which I don't agree with. (..specifically I've NOT found that an anechoic reproduction produces an "in the head" sound. To be fair though I've never heard it done in a large chamber, though I have in a small one and have done it outside/free-field with ground absorption - which obviously couldn't entirely remove "ground" bounce at the lowest freq.s..)
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.