Treble needs also good dispersion, in my stereo Mls measurements it shows a cut of at 15kHz when the dispersion changes above 10kHz (dome tweeter). It effects the brightens and HF details. So I think it good to have good 30 degree dispersion above 15kHz to have HF on your listening position and have a wider sweetspot
My horns do better although limited dispersion it evolves over the hole bandwidth like often shown here.
They have a smaller sweet spot on listening position due limited dispersion. That is why better dispersion is required IMO.
Hi Earl
Hey to be clear, I was not suggesting you would have a hard time imagining this, rather if you were like some in the thread, if you had never heard it, the description might sound a bit implausible.
The first time I ever heard the apparent lack of connection between the physical depth of the source and the image the brain derives was with some Quad ESL-63’s I was repairing (removing the annoying protection spark gaps) for my old boss at Intersonics.
There were times when a voice really seemed like it was located somewhere behind the speaker. I had already built plenty of electrostatic speakers at that time (why he allowed me to tinker with his) and it was fixing them that I think made him interested when I said I had built a subwoofer using a servomotor from a local junk store.
Prototype 3 sounded good enough to demonstrate at work and after hearing it, Roy said I could start a small speaker division as long as it only cost a little space and lighting etc.
It was doing an AES paper on the Servodrive Subwoofers in the way old days that I met you for the first time too . I recall you had just done a couple papers on band pass systems with absolutely mind numbing (for me) math so when you got up to ask a question in the Q&A after, my heart resumed beating when it was something I could answer haha.
Anyway, when I heard that effect from Roy’s Quad’s , I though huh, that is weird but then thinking about how it worked (assembly of a simple partial sphere wave front by progressively delayed concentric rings) that made sense. At the time, I was not allowed to pursue anything at work using normal drivers so it was a while before I could do anything but bass.
Fast forward 15 odd years to working on full range rotary speakers at Quantum Sound and a bunch more to the full range horns and the same thing is much more audible. To me the name “source identity” seemed to fit the ease with which one can identify the depth location with one’s eyes closed.
Speaking of ones eyes being closed leading to the connection between what we see and what our brain hears. This was a great documentary in general but the parts about hearing process should be a wake up call for those who scoff unsighted (without knowledge) testing of all things ears.
Notice, even when you know consciously, your hearing process is so well trained that your eyes over ride your ears. You only hear reality when your eyes are closed / diverted IE: that knowledge removed.
Try The McGurk Effect! - Horizon: Is Seeing Believing? - BBC Two - YouTube
Best,
Tom
Hey to be clear, I was not suggesting you would have a hard time imagining this, rather if you were like some in the thread, if you had never heard it, the description might sound a bit implausible.
The first time I ever heard the apparent lack of connection between the physical depth of the source and the image the brain derives was with some Quad ESL-63’s I was repairing (removing the annoying protection spark gaps) for my old boss at Intersonics.
There were times when a voice really seemed like it was located somewhere behind the speaker. I had already built plenty of electrostatic speakers at that time (why he allowed me to tinker with his) and it was fixing them that I think made him interested when I said I had built a subwoofer using a servomotor from a local junk store.
Prototype 3 sounded good enough to demonstrate at work and after hearing it, Roy said I could start a small speaker division as long as it only cost a little space and lighting etc.
It was doing an AES paper on the Servodrive Subwoofers in the way old days that I met you for the first time too . I recall you had just done a couple papers on band pass systems with absolutely mind numbing (for me) math so when you got up to ask a question in the Q&A after, my heart resumed beating when it was something I could answer haha.
Anyway, when I heard that effect from Roy’s Quad’s , I though huh, that is weird but then thinking about how it worked (assembly of a simple partial sphere wave front by progressively delayed concentric rings) that made sense. At the time, I was not allowed to pursue anything at work using normal drivers so it was a while before I could do anything but bass.
Fast forward 15 odd years to working on full range rotary speakers at Quantum Sound and a bunch more to the full range horns and the same thing is much more audible. To me the name “source identity” seemed to fit the ease with which one can identify the depth location with one’s eyes closed.
Speaking of ones eyes being closed leading to the connection between what we see and what our brain hears. This was a great documentary in general but the parts about hearing process should be a wake up call for those who scoff unsighted (without knowledge) testing of all things ears.
Notice, even when you know consciously, your hearing process is so well trained that your eyes over ride your ears. You only hear reality when your eyes are closed / diverted IE: that knowledge removed.
Try The McGurk Effect! - Horizon: Is Seeing Believing? - BBC Two - YouTube
Best,
Tom
All this discusion of speakers being apparent as the source got me thinking. I believe the recording is somewhat to blame if material localizes at one speaker or the other and that we might be able to simulate it.
I found a web site that had anechoic recordings of various insturments.
Index of /Public/Anecoic
This includes cuts from an older Denon recording of a full orchestra surrounded by acoustic absorption, as well as solo instruments in an anechoic chamber. I've picked one of a piano in the left channel and soprano in the right. It sounds like electric piano, probably easier to get into an anechoic chamber! (mozartdry)
Cool Edit Pro has a very elaborate reverberator. I took the same file and first reduced the separation until the span between the performers seemed about right and then added an appropriate reverb. (mozartwet)
Just to make the comparison more obvious I added a clip that switches between the two mixes every few seconds. (mix of wet and dry).
Note that these are not .zip files, but are .mp3 files. Copy them to your desk top and change the .zip ending to .mp3.
Have a listen and see if the reverberated version "leaves" your loudspeakers more effectively than the dry version. I've only been listening on headphones but there difference is dramatic.
Finally, to see if even simple changes can put the image in a more natural space, I used a function called "multitap" that makes a simple multitap reverberator. Here the original signal is recirculated via 2 delays, one at about 25ms and one at about 35 ms. These are set to delay and mix to the opposite channel, much like two discrete hard reflections. They also recirculate, much as the sound between two parallel hard surfaces would. That is what it sounds like, two soloists between two hard walls, but much more convincing than the anechoic sources. I didn't reduce the seperation between channels on this version. (mozart multitap)
Have a listen and give us your comments.
David S.
I found a web site that had anechoic recordings of various insturments.
Index of /Public/Anecoic
This includes cuts from an older Denon recording of a full orchestra surrounded by acoustic absorption, as well as solo instruments in an anechoic chamber. I've picked one of a piano in the left channel and soprano in the right. It sounds like electric piano, probably easier to get into an anechoic chamber! (mozartdry)
Cool Edit Pro has a very elaborate reverberator. I took the same file and first reduced the separation until the span between the performers seemed about right and then added an appropriate reverb. (mozartwet)
Just to make the comparison more obvious I added a clip that switches between the two mixes every few seconds. (mix of wet and dry).
Note that these are not .zip files, but are .mp3 files. Copy them to your desk top and change the .zip ending to .mp3.
Have a listen and see if the reverberated version "leaves" your loudspeakers more effectively than the dry version. I've only been listening on headphones but there difference is dramatic.
Finally, to see if even simple changes can put the image in a more natural space, I used a function called "multitap" that makes a simple multitap reverberator. Here the original signal is recirculated via 2 delays, one at about 25ms and one at about 35 ms. These are set to delay and mix to the opposite channel, much like two discrete hard reflections. They also recirculate, much as the sound between two parallel hard surfaces would. That is what it sounds like, two soloists between two hard walls, but much more convincing than the anechoic sources. I didn't reduce the seperation between channels on this version. (mozart multitap)
Have a listen and give us your comments.
David S.
The problem was, apart from the folder, the dispersion from the dome mid.
Either way I use Tannoy DCs which are pretty much 90deg conical from 1.2k to 16k. Works for me but I wouldn't want more than 90deg. Too much grief from the room and mine are quite versatile when it comes to positioning.
In the textbook on Spatial Hearing that i am reading, they describe experiemnents with a single source in front of the listener. As the frequency of the source increases the percieved height also increases.
dave
This test is a good initiative.
Listened on head phones or checked in Audition, the differences are as you say dramatic. I will say that only the wet can make illusion, the dry is catastrophic, the multitap is psychedelic.
But with my main system that is a side firing in a big room, with XTC, rear speakers and other features, then the dry shows a well placed voice (25 %) with a good timbre and articulation, very close. The absence of reverberation is not a problem, just an unusual characteristic, like if I was not any more in my room but in a highly stuffed space.
So this sample is perfectly listenable (except for the crappy piano).
The wet sounds artificial, the reverberation is too simple, concerning more the voice than the piano, doesn't result in depth too. The localisations are not affected. Not my favorite.
The multitap is a problem for me : terrible on head phones but honestly I can't say that there's such a difference when I listen on the speakers. Absolutely...annoying.
Listened on head phones or checked in Audition, the differences are as you say dramatic. I will say that only the wet can make illusion, the dry is catastrophic, the multitap is psychedelic.
But with my main system that is a side firing in a big room, with XTC, rear speakers and other features, then the dry shows a well placed voice (25 %) with a good timbre and articulation, very close. The absence of reverberation is not a problem, just an unusual characteristic, like if I was not any more in my room but in a highly stuffed space.
So this sample is perfectly listenable (except for the crappy piano).
The wet sounds artificial, the reverberation is too simple, concerning more the voice than the piano, doesn't result in depth too. The localisations are not affected. Not my favorite.
The multitap is a problem for me : terrible on head phones but honestly I can't say that there's such a difference when I listen on the speakers. Absolutely...annoying.
interesting, so the acoustics of the room is not at all superimposed on the recording
again sth must be wrong with Your hearing as Your experience clearly goes against the prevailing and sound theory You really have to post a picture and diagram of it for the people here to understand at least a bit of what You are talking about
I would't say the 4th should sound natural, unless you like your performers outside between 2 brick walls! It does make a difference, on headphones at least, in that they have been put into some type of space rather than stuck in one ear or the other.
I'll burn a CD and try it on my speakers. Cool edit has a great reverberator with lots of presets, but each is the same program with a large number of acoutics variables individually set (room volume, decay, diffusion, etc). Maybe I need to pick one that is overblown on headphones and less subtle on speakers?
The dry orchestra recordings are interesting too. What facinates me is that you usually aren't aware of a concert hall's acoustics, but then you can't turn it off and on, can you? With a dry recording through the reverberator you can turn it off and on, and with reverberation added you do loose clarity (you can't "count the violins") but pick up naturalness and a much more musical quality.
Very cool.
Thanks for trying it Casull, Radugazon.
David S.
For me your mp3s work "like advertised"
. "mozartdry" is firmly glued to each speaker (or to the outermost positions of my headphone panorama).
"mozartwet" moves the piano and the voice slighty inward (nearer to the center), but also markedly towards the frontwall. The speakers disappear completely.
"multitap" makes the sources a bit more diffuse than "mozartdry", but they stay fixed to the speakers.
I believe that you would get an equivalent disappearing act as in "mozartwet", if you add a small amount of the left channel signal to the right channel of "mozartdry" and vice versa - effectively moving the sources just a little bit towards the center of the stereo base.
Rudolf
For me your mp3s work "like advertised"
I wonder how your system is different than the others?
I'll try a few more options tonight.
Thanks!
In other thread came up a journal article by Bennett, Barker and Edeko which is relevant for speaker disappearance point of view too if simultaneous phantom imaging is desired:
http://www.google.fi/url?sa=t&rct=j&q=a%20new%20approach%20to%20the%20assessment%20of%20stereophonic%20sound%20system%20performance&source=web&cd=8&sqi=2&ved=0CF4QFjAH&url=http%3A%2F%2Fdecoy.iki.fi%2Fdsound%2Fambisonic%2Fmotherlode%2Fsource%2FA%2520new%2520approach%2520to%2520the%2520assessment%2520of%2520stereophonic_Bennett%2520et%2520al_1985.pdf&ei=xAbQTv2HNqXg4QTUz5A6&usg=AFQjCNGq_H_meWCl-u3DRNMOSknDFctKRQ&cad=rja
They proved by theoretical considerations and confirmed that by listening tests that at high frequencies speakers can be localised as the image shift into the speakers.
"... the central image disappears, and a source appears located at each loudspeaker."
"A further increase in frequency ... finally all components cluster around the location of the loudspeakers."
Of course this only applies for a conventional stereo triangle
- Elias
http://www.google.fi/url?sa=t&rct=j&q=a%20new%20approach%20to%20the%20assessment%20of%20stereophonic%20sound%20system%20performance&source=web&cd=8&sqi=2&ved=0CF4QFjAH&url=http%3A%2F%2Fdecoy.iki.fi%2Fdsound%2Fambisonic%2Fmotherlode%2Fsource%2FA%2520new%2520approach%2520to%2520the%2520assessment%2520of%2520stereophonic_Bennett%2520et%2520al_1985.pdf&ei=xAbQTv2HNqXg4QTUz5A6&usg=AFQjCNGq_H_meWCl-u3DRNMOSknDFctKRQ&cad=rja
They proved by theoretical considerations and confirmed that by listening tests that at high frequencies speakers can be localised as the image shift into the speakers.
"... the central image disappears, and a source appears located at each loudspeaker."
"A further increase in frequency ... finally all components cluster around the location of the loudspeakers."
Of course this only applies for a conventional stereo triangle
- Elias
"True" figure 8 dipole without baffle for mids and highs.
I don't listen to signals at 6 kHz with just 1/3 oct bandwith, as they did. The fundamentals of any reasonable musical signal will completely dominate the directional impression.
Attachments
My experience is completely opposite with any conventional stereo triangle, the high freqs dominate spatial perception.
- Elias
... and i feel this is why stereo may usually work quite well.
Nevertheless experiments using high frequency and narrowband signals
may be useful to show weaknesses of stereo.
The widely observed functionality of stereo
- and the practical experience in the production process -
IMO relies on applying it to "common" genres of music.
Stereo does not seem to be a universal mapping system allowing
proper lateralisation of arbitrary phantom sources.
"True" figure 8 dipole without baffle for mids and highs.
Maybe Toole has it wrong. Maybe the dipole is preferable and only works poorly when the stereo signal is too simplistic and confined unaturally to a single channel?
David S.
I subscribe to that. It is a crutch for sure, but for me it works beautifully - from bone whistles to synthesisers.
Maybe Toole had the wrong dipoles? Modern dipoles with dynamic drivers and minimal baffle size can achieve far smoother frequency response, less distortion and better controlled directivity than the ESLs tested by Toole and Olive.
Rudolf,
I see you've made a few changes. What tweeter are you using now?
Dave,
My set up is a pair of B2031P in equilateral triangle, tweeter slightly above ear height. Distance to rear wall is approximately 28" from wall to drivers, and distance to side walls is >3 feet.
Tweeters are the Visaton FRSX5. They are very basic and bottom budget cone drivers, but with an unrivaled forward directivity (> 2kHz):
More information on the FRS5X.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.