Re: [Sursound] Distance perception
On 07/26/2011 02:18 AM, Marc Lavallée wrote: Imagine two rooms with proper acoustic characteristics and treatments for ambisonics reproduction: the first is 3mX4m and the other is four times larger in surface (9mX12m). In both rooms there's a horizontal hexagon of speakers, and 5 speakers are against a wall. regardless of room size, they will require a bit of equalisation. if the speakers are designed to be close to a boundary surface, the one that's not against a wall needs (gentle) bass boost. vice versa, if your speakers are designed to be free-standing, the five speakers need some attenuation at LF. if you're still shopping for speakers, i found that the genelec 8030 have a nice built-in bass eq which can be used to deal with this issue. i was able to even out the bass response of a rig where most speakers are next to two boundary surfaces and a few only next to one. of course, you could also do this in software. When NFC is applied in both rooms, do they sound the same in terms of distance perception when playing the same recording? NFC is not a constant. the amount of NFC depends on the distance to the speaker. Or is the same sound object appear to be twice as far in the largest room? actually, if you hope to get distance perception so good that the notion of twice as far begins to make sense, then you're in for some heartache. that's why i said distance cues are gimmickry earlier. the actual curvature of the soundfield (which is all that NFC does for you) is not a very robust distance cue. the delay of the (reproduced) floor reflection is a lot more helpful, as is the ratio of direct to reverberated sound (but the latter doesn't help soundman john with his spitfires). so why get gung-ho about a cue of secondary importance, for a perception apparatus that doesn't care much anyways... the problem is that your listening room floor reflection will always be different from and stronger than the recorded floor reflection, which pulls the image towards the speaker circle. if you close your eyes and find yourself able to suspend your disbelief long enough to actually imagine yourself in a cathedral listening to an organ, then rejoice and be happy. don't spoil the magic by gauging the distance. it's not going to happen. the sad and simple fact is that _no_ surround rig can get the distance unambiguously right in any but anechoic conditions. and before you run off to shop for styrofoam, be warned that most recordings would sound utter crap in anechoic conditions, because nobody mixes for that. moreover, the phasing problems of our beloved ambisonic technique would become very obnoxious indeed. the deader you make your room, the more hope you have to get precise distance information. at the same time, the rig will sound less pleasant and artefacts will become more obvious. since humans suck at absolute distance perception anyways, your best bet is to be content with some degree of distance discriminination. that is, you want to hear the woodwinds _somewhere_behind_ the strings. you wouldn't normally care how many metres. this usually works well if the recording is ok. Apart from widening the listening sweet spot, are larger rooms better at reproducing distance cues when using the same speaker configuration? Is distance perception directly related to speaker distances? as mentioned before, the floor reflection is a very strong distance cue at close range under semi-anechoic conditions (i.e. if you want to gauge the distance of that sabre-toothed tiger or the potential mating candidate). if you're right next to the sound source, the floor reflection will have the longest delay. far away, the delay will be negligible. the general case is dly = 340 / (sqrt(ear_height^2 + half_distance^2) * 2) when a listening room first reflection is strong and early, it will dominate your sense of distance. in that sense, larger rigs have the potential to be less intrusive wrt distance perception. -- Jörn Nettingsmeier Lortzingstr. 11, 45128 Essen, Tel. +49 177 7937487 Meister für Veranstaltungstechnik (Bühne/Studio) Tonmeister VDT http://stackingdwarves.net ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Distance perception
sorry, itchy trigger finger... On 07/26/2011 10:14 AM, Jörn Nettingsmeier wrote: as mentioned before, the floor reflection is a very strong distance cue at close range under semi-anechoic conditions (i.e. if you want to gauge the distance of that sabre-toothed tiger or the potential mating candidate). uhm, i realise that the latter example is a bit dated - who meets significant others in the great outdoors, these days. for clubbing, the dominant cue should be direct-to-reverb ratio, unless you have to fall back to olfactory and visual cues entirely because of the extremely loud music. if you're right next to the sound source, the floor reflection will have the longest delay. far away, the delay will be negligible. the general case is dly = 340 / (sqrt(ear_height^2 + half_distance^2) * 2) minus the straight-path delay of course: dly = 340 / (sqrt(ear_height^2 + (distance/2)^2) * 2 - distance) -- Jörn Nettingsmeier Lortzingstr. 11, 45128 Essen, Tel. +49 177 7937487 Meister für Veranstaltungstechnik (Bühne/Studio) Tonmeister VDT http://stackingdwarves.net ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Distance perception
Tue, 26 Jul 2011 10:14:50 +0200, Jörn Nettingsmeier netti...@stackingdwarves.net wrote : regardless of room size, they will require a bit of equalisation. if the speakers are designed to be close to a boundary surface, the one that's not against a wall needs (gentle) bass boost. vice versa, if your speakers are designed to be free-standing, the five speakers need some attenuation at LF. The smallest KEF eggs should be fine against the walls, as you already wrote me. Two will be free-standing (front and back) and will need a bit more electronic correction. if you're still shopping for speakers, i found that the genelec 8030 have a nice built-in bass eq which can be used to deal with this issue. i was able to even out the bass response of a rig where most speakers are next to two boundary surfaces and a few only next to one. of course, you could also do this in software. I will do it in software. It's a domestic setup, so I don't need expensive active speakers and cabling; I prefer to use very small speakers with lamp cords. the problem is that your listening room floor reflection will always be different from and stronger than the recorded floor reflection, which pulls the image towards the speaker circle. Then less reflections means less localization of the speakers? the sad and simple fact is that _no_ surround rig can get the distance unambiguously right in any but anechoic conditions. and before you run off to shop for styrofoam, be warned that most recordings would sound utter crap in anechoic conditions, because nobody mixes for that. What follows is just my opinion. We are free to record and mix in any imaginable ways, so recordings sound imperfect in most situations. There's little a listening room can do to beautify recordings and reproduction systems, unless the room is considered as a musical instrument. Even in a small room with too much acoustic treatment, I may be pleasantly surprised by some very good recordings, and find some qualities in some very bad recordings; anything can happen in the middle, and low expectations is the key to happiness. What I expect from listening to ambisonic recordings is a better envelopment and a sense of realism not found in stereo recordings. I also expect some new experiences from field recordings and electroacoustic music for ambisonics. I also want to compare ambisonics to other reproduction methods; maybe stereo and 5.1 are not so bad... The other use for all those speakers is to add a bit of hall reverberation to some dry stereo recordings. moreover, the phasing problems of our beloved ambisonic technique would become very obnoxious indeed. Obnoxious phasing problems? Now I'm afraid! ;-) Maybe I spent decades listening to obnoxious problems I never noticed... I'll do my best to control phasing problems at the sweet spot. the deader you make your room, the more hope you have to get precise distance information. at the same time, the rig will sound less pleasant and artefacts will become more obvious. since humans suck at absolute distance perception anyways, your best bet is to be content with some degree of distance discriminination. that is, you want to hear the woodwinds _somewhere_behind_ the strings. you wouldn't normally care how many metres. this usually works well if the recording is ok. I found good acoustic panels, and I have to decide how much surface to cover. I once built large and thick panels to cover half of the walls and 2/3 of the ceiling. There was also a wool carpet with foam under it. The room was so dead that I was able to listen to my heart beat. I remember how sharp the stereo image was and how the speakers were not easy to localize with good recordings. Of course the room was a bit oppressive, almost like a recording booth... I hope to find a better compromise between analytic listening and listening for enjoyment. when a listening room first reflection is strong and early, it will dominate your sense of distance. in that sense, larger rigs have the potential to be less intrusive wrt distance perception. Right: first reflections should be better controlled in a small room. if you're right next to the sound source, the floor reflection will have the longest delay. far away, the delay will be negligible. the general case is dly = 340 / (sqrt(ear_height^2 + half_distance^2) * 2) minus the straight-path delay of course: dly = 340 / (sqrt(ear_height^2 + (distance/2)^2) * 2 - distance) So floor and ceiling reflections also need to be controlled, even more in a small room. The difficulty is how to leave some harmless and lively reflections. Maybe that adding a few small diffusors would be a good compromise. Thanks! -- Marc ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Distance perception
Some papers that may be of interest: Takahashi, A Novel View of Hearing in Reverberation, Neuron, Volume 62, Issue 1, 6-7, 16 April 2009 doi:10.1016/j.neuron.2009.04.004 Devore, et al., Accurate Sound Localization in Reverberant Environments Is Mediated by Robust Encoding of Spatial Cues in the Auditory Midbrain, Neuron, Volume 62, Issue 1, 123-134, 16 April 2009 doi:10.1016/j.neuron.2009.02.018 Antje Ihlefeld and Barbara G. Shinn-Cunningham, Effect of source spectrum on sound localization in an everyday reverberant room, J. Acoust. Soc. Am. Volume 130, Issue 1, pp. 324-333 (2011) http://dx.doi.org/10.1121/1.3596476 -- Aaron Heller (hel...@ai.sri.com) Menlo Park, CA US ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Distance perception
On Tue, Jul 26, 2011 at 04:35:39PM -0400, Marc Lavallée wrote: I will do it in software. It's a domestic setup, so I don't need expensive active speakers and cabling; I prefer to use very small speakers with lamp cords. Lamp cords ? Use at least 2.5 mm^2 ! the problem is that your listening room floor reflection will always be different from and stronger than the recorded floor reflection, which pulls the image towards the speaker circle. Then less reflections means less localization of the speakers? Yes, in general this is true, and it's quite logical - we use reflections to build up an 'acoustic picture' of a space, and in turn that is used to aid localisation. If the cues provided by room reflections dominate those reproduced from the recording you can't but identify the speakers as the source. the sad and simple fact is that _no_ surround rig can get the distance unambiguously right in any but anechoic conditions. and before you run off to shop for styrofoam, be warned that most recordings would sound utter crap in anechoic conditions, because nobody mixes for that. That is really a very valid observation. Almost all recordings rely on the listener's room acoustics to do part of the work. And studio control rooms usually have well controlled acoustics, but they are by no means anechoic. Which means that something similar is expected of the listening environment. What I expect from listening to ambisonic recordings is a better envelopment and a sense of realism not found in stereo recordings. I also expect some new experiences from field recordings and electroacoustic music for ambisonics. I also want to compare ambisonics to other reproduction methods; maybe stereo and 5.1 are not so bad... They are not. Very nice results can be achieved with either. The other use for all those speakers is to add a bit of hall reverberation to some dry stereo recordings. Depends a bit on the type of music you are listening to, but in general that is a good idea for any type of music that is normally played in concert hall like environments. There is another thing which I can't explain ATM. I've been working lately most of the time in a studio that has a regular octagon of speakers for Ambisonic monitoring. But half of the work done there is just stereo. The thing is that I very much prefer listening to stereo using two virtual speakers panned into 3rd order AMB rather than sending L,R directly to two of the speakers. But I can't ATM explain why. So floor and ceiling reflections also need to be controlled, even more in a small room. Yes. I recently moved home, and my new working environment is a rather small and boxy room. Its only redeeming feature is that the ceiling is not horizontal but inclined by 15 degress or so. The floor is hardwood, nice for recording but in this case it doesn't help for listening. When I first listened to some reference recordings in this place I was 'not amused' at all. But putting a thick carpet in front of the speakers changed the picture quite dramatically. The room is still a disaster for good LF response, but otherwise it has become acceptable by reducing a very strong floor reflection. The difficulty is how to leave some harmless and lively reflections. Maybe that adding a few small diffusors would be a good compromise. Diffusers are almost never a bad idea. Ciao, -- FA ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Distance perception
On 2011-07-26, Fons Adriaensen wrote: Lamp cords ? Use at least 2.5 mm^2 ! Where does this come from? I've never though cable geometry matters much at all, unless your pumping so much power through a cable over such a long distance that you have to worry about ohmic heating and the like. And even there, I've always thought changing resistance would mostly affect a tube end stage, which we've almost done away with already in favour of the A/B class solid state one. And at audio frequencies, shouldn't even feedback oscillation and its kin be well below perceptual thresholds? True, my cables are multistrand ones with approximately that cross-sectional area per polarity. But not because of some esoteric, audiophile reason. It's because that's what they sell the cheapest as speaker cable in my local shop. -- Sampo Syreeni, aka decoy - de...@iki.fi, http://decoy.iki.fi/front +358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2 ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Distance perception
On 07/27/2011 12:41 AM, Sampo Syreeni wrote: On 2011-07-26, Fons Adriaensen wrote: I certainly don't want you to waste your money on fancy speaker cables. Never thought otherwise. That's obviously never been what we do here. ;) But resistance does matter, so a good cross section such as 2.5 mm^2 puts you on the safe side. What I was trying to ask is, what's the real problem with resistance, especially with regard to a passive speaker and a modern, A/B class solid state end stage? I mean, I don't really see cable resistance shifting their operating point much, even with feedback, within the audible range. What is it that I'm missing? power transmission impedance matching. if you look at the spec sheet of a commercial p.a. amplifier, 9 times out of 10 you will see twice the power rating for 4 ohm loads than for 8 ohms. usually this means you connect two 8 ohm enclosures in parallel for an optimum load. but obviously any resistance of the wire will limit the power you can draw from the amp. say you're using the really cheap NYM 3G1.5 wire, which has about 14 ohms per km. for a practical speaker line length of 20m, that's 0.3 ohms. i won't make a fool of myself here by giving precise numbers after a day of mixing and three bottles of beer, but it's easy to see that 0.3 compared to 4 ohms is a significant fraction. -- Jörn Nettingsmeier Lortzingstr. 11, 45128 Essen, Tel. +49 177 7937487 Meister für Veranstaltungstechnik (Bühne/Studio) Tonmeister VDT http://stackingdwarves.net ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Distance perception
Tue, 26 Jul 2011 21:32:26 +, Fons Adriaensen f...@linuxaudio.org wrote : Lamp cords ? Use at least 2.5 mm^2 ! I'll use less than 10 meters of cabling to drive 10W max in each tiny 6 ohms speaker. So I'm not worried at all. Gauge 14 or 16 should be fine: http://www.roger-russell.com/wire/wire.htm The thing is that I very much prefer listening to stereo using two virtual speakers panned into 3rd order AMB rather than sending L,R directly to two of the speakers. It's very interesting! How large is the resulting stereo image? Is your technique documented somewhere? Can it work with a horizontal hexagon? With 2rd order AMB? -- Marc ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] speaker cable resistance [was Distance perception]
Am I right in thinking that the resistive component of the speaker's impedance is effectively in series with its inductance? Say, 5 ohms for an 8 ohm speaker? If so cable resistance is not so frightening for domestic runs. It's not so difficult to provide negative output impedance to counter the wire+speaker resistance. This can significantly increase damping and is sometimes used in powered 'active' speakers where all parameters can be controlled. Neil Adams At 06:41 7/27/2011, Sampo Syreeni wrote: On 2011-07-26, Fons Adriaensen wrote: I certainly don't want you to waste your money on fancy speaker cables. Never thought otherwise. That's obviously never been what we do here. ;) But resistance does matter, so a good cross section such as 2.5 mm^2 puts you on the safe side. What I was trying to ask is, what's the real problem with resistance, especially with regard to a passive speaker and a modern, A/B class solid state end stage? I mean, I don't really see cable resistance shifting their operating point much, even with feedback, within the audible range. What is it that I'm missing? -- Sampo Syreeni, aka decoy - de...@iki.fi, http://decoy.iki.fi/front +358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2 ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] speaker cable resistance [was Distance perception]
If you do not do something tricky with the amplifier-- and no commercial consumer audio amplifier intended for universal use does this trickiness, or none I am aware of-- then the cable impedance operates as part of the amplifier output impedance. This means that the amplifier will not be flat into the speaker load unless the speaker load is constant resistive impedance- which effectively never happens. It does not take a lot of resistance for this to matter. I suppose everyone knows how to compute this. In the unlikely event that anyone does not, it is explained(reasonably clearly I hope) here: http://www.regonaudio.com/Why%20Amplifiers%20Don't%20Always%20Sound%20Right.html This is mostly about amplifier output impedance --the article could have been subtitled Why you should forget about tube amps! But cable impedance just adds itself to amplifier output impedance (or in the case of cable capacitance, it is shunted over,in parallel not in series). So the calculations are similar. Do these things matter? Actually , in times past at least, lots of amplifiers even solid state ones were not in fact flat into actual speaker loads. Tom Holman had an article (:Manufacturer's Response)in The Absolute Sound early on--issue 26 as I recall-- measuring the actual response of a bunch of solid state amps into a real speaker. ALmost none of them was anywhere near flat to within the (small) threshold of audibility of frequency response shifts. (Holman's Apt design was the only one that was, in the group surveyed). There is quite a lot of reason to believe that this sort of thing is the main, maybe the only, reason that amplifiers sound different, to the extent that they do(except for amps with high distortion or clipping at the operating level). Robert On Wed, 27 Jul 2011, Neil Marcia Adams wrote: Am I right in thinking that the resistive component of the speaker's impedance is effectively in series with its inductance? Say, 5 ohms for an 8 ohm speaker? If so cable resistance is not so frightening for domestic runs. It's not so difficult to provide negative output impedance to counter the wire+speaker resistance. This can significantly increase damping and is sometimes used in powered 'active' speakers where all parameters can be controlled. Neil Adams At 06:41 7/27/2011, Sampo Syreeni wrote: On 2011-07-26, Fons Adriaensen wrote: I certainly don't want you to waste your money on fancy speaker cables. Never thought otherwise. That's obviously never been what we do here. ;) But resistance does matter, so a good cross section such as 2.5 mm^2 puts you on the safe side. What I was trying to ask is, what's the real problem with resistance, especially with regard to a passive speaker and a modern, A/B class solid state end stage? I mean, I don't really see cable resistance shifting their operating point much, even with feedback, within the audible range. What is it that I'm missing? -- Sampo Syreeni, aka decoy - de...@iki.fi, http://decoy.iki.fi/front +358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2 ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound
Re: [Sursound] Distance perception
On 26/07/11 3:41 p.m., Sampo Syreeni wrote: On 2011-07-26, Fons Adriaensen wrote: I certainly don't want you to waste your money on fancy speaker cables. Never thought otherwise. That's obviously never been what we do here. ;) But resistance does matter, so a good cross section such as 2.5 mm^2 puts you on the safe side. What I was trying to ask is, what's the real problem with resistance, especially with regard to a passive speaker and a modern, A/B class solid state end stage? I mean, I don't really see cable resistance shifting their operating point much, even with feedback, within the audible range. What is it that I'm missing? I swapped out some lamp cable on the speakers of a stereo setup some years back with some cheap stranded speaker cable I bought at Costco. Each core of the cable was about 3 times the cross sectional area of the lamp cord (each core of the speaker cable was about 3/16 in dia). The distances were not great, 5 or 6 feet. The improvement in stereo imaging was huge. Previously the image had wandered around between the speakers seemingly at random, now it was rock solid at the point wherever it was when I recorded it. ___ Sursound mailing list Sursound@music.vt.edu https://mail.music.vt.edu/mailman/listinfo/sursound