The Audiophile Thread

[progmatist]

Would you buy a color TV set before they started broadcasting in color?

When local stations were in the process of upgrading to HD, many people did in fact buy HD sets to watch their DVD collection. Before there was such a thing as HD DVD or Blu-Ray.

Is there any multi-channel material out there that isn't just a gimmick?

If they simply recreate the ambience of the hall, it does add to the experience. Resphi's Pines of Rome in particular MUST be heard in surround. In the closing march when additional brass starts playing up in the balconies, coming out of the rear channels is quite dramatic.

[Firth]

When local stations were in the process of upgrading to HD, many people did in fact buy HD sets to watch their DVD collection. Before there was such a thing as HD DVD or Blu-Ray.



If they simply recreate the ambience of the hall, it does add to the experience. Resphi's Pines of Rome in particular MUST be heard in surround. In the closing march when additional brass starts playing up in the balconies, coming out of the rear channels is quite dramatic.

I had a set box which produced HD TV channels and I had a Sony with s-video inputs. I set the set top to 480i on an HD channel and result was a significant improvement.

[rcarlberg]

Speaking of stereo, read this review of the Polk Legend model 800, with advanced SDA.

Okay, I read the review and watched the guy's video. The first thing I noticed is they never mention what this "Polk SDA" technology is, despite claiming how wonderful it is. Doing a little Googling, I found this explanation of the "Stereo Dimensional Array":

By positioning additional drivers to deliver a precisely derived cancellation signal to your ears, SDA enables you to hear the sonic image just as the recording captured it.

I noticed in the video the guy talked about connecting a cable between the speakers, that is not connected to the amplifier. He also noted that the speakers sound best when placed rather close together. The soundstage image extends beyond the speakers. And the SDA, when engaged, causes some intermodulation distortion with the non-SDA midrange and tweeter.

Yup, what you got there is 180° out-of-phase ("cancellation") signal fed to the duplicate high-end drivers ("additional drivers") which are aimed away from the listener. This simulates the effect you'd get from rear radiation on a planar driver.

I have played extensively with rear radiation in various speakers designs over the years, and I am very familiar with the effect. In fact I wrote about this in an article that was published in 1982. The effect can widen and deepen the apparent soundstage, especially if your room is too small and your speakers are too close together.

But it's a trick, a gimmick. It is adding something that isn't in the original recording. It works better on some recordings, as the reviewer notes, than it does on other recordings, because it is not a neutral effect. Depending on the mixing, adding out-of-phase signal to each side CAN widen the image, as anyone who has ever played with channel mixing on a digital audio workstation can tell you. Remove the center channel, make the wing speakers sound farther apart.

It's not magic. It's also not beneficial to all material.

[rcarlberg]

If they simply recreate the ambience of the hall, it does add to the experience. Resphi's Pines of Rome in particular MUST be heard in surround. In the closing march when additional brass starts playing up in the balconies, coming out of the rear channels is quite dramatic.

Yes, if the music was recorded "in a hall" and the recording didn't really capture the hall acoustics, then recreating the hall with echoed rear speakers could be considered "realistic."

What if it wasn't recorded in a hall though? Adding "hall acoustics" would not be a benefit, would not be a step toward realism.

[Firth]

Okay, I read the review and watched the guy's video. The first thing I noticed is they never mention what this "Polk SDA" technology is, despite claiming how wonderful it is. Doing a little Googling, I found this explanation of the "Stereo Dimensional Array":

I noticed in the video the guy talked about connecting a cable between the speakers, that is not connected to the amplifier. He also noted that the speakers sound best when placed rather close together. The soundstage image extends beyond the speakers. And the SDA, when engaged, causes some intermodulation distortion with the non-SDA midrange and tweeter.

Yup, what you got there is 180° out-of-phase ("cancellation") signal fed to the duplicate high-end drivers ("additional drivers") which are aimed away from the listener. This simulates the effect you'd get from rear radiation on a planar driver.

I have played extensively with rear radiation in various speakers designs over the years, and I am very familiar with the effect. In fact I wrote about this in an article that was published in 1982. The effect can widen and deepen the apparent soundstage, especially if your room is too small and your speakers are too close together.

But it's a trick, a gimmick. It is adding something that isn't in the original recording. It works better on some recordings, as the reviewer notes, than it does on other recordings, because it is not a neutral effect. Depending on the mixing, adding out-of-phase signal to each side CAN widen the image, as anyone who has ever played with channel mixing on a digital audio workstation can tell you. Remove the center channel, make the wing speakers sound farther apart.

It's not magic. It's also not beneficial to all material.

There was a link on that site for details. Regular stereo isn’t beneficial to most material, wrt imaging. But have you heard them? If you haven’t heard them, you are blowing much irrelevant smoke, and you ain’t persuading me that you really have heard the difference between these and what has been done.

[Quantum Cat]

Ultimately, it is all in the ear of the beholder. The favorite speakers I ever owned were Alison : One speakers purchased around 1980. They sounded amazing to me all who heard them without exception. They were designed by the late Roy Alison of Acoustic Research fame and featured "Stabilized Radiation Loading" which as far as I can understand it, compensates for the back wave that reflects off walls and back to the woofers thereby affecting their performance. Mr. Alison saw the room and the speaker as one system and his designed his products accordingly. I wish I had those Allison's now.

[rcarlberg]

Yes, the Allison One's angled cabinetry and multiple drivers build on the design of the AR-LST and the Bose 901. Like Magneplanars and Martin Logans the rear radiation couples with the room to create a huge soundstage seemingly coming from behind the speakers. Well, I guess it really is coming from behind the speakers, isn't it.

Every room is a component of a speaker system, unless you place them in the center of the room and sit between them (as a friend of mine with a $500k system does). How the room couples is determined by shape, reflectivity, angle, size, type of material, height, carpeting.... so many factors it's impossible to simplify. Certain rules are universal though: avoid glass (windows and sliding doors). Avoid low ceilings. Don't over-dampen the room either. With MOST speakers in MOST rooms the ideal geometry is an equilateral triangle, with the speakers and listener about equally spaced. Most speakers require a minimum of about 10-12 feet distance to develop the bass notes. Placing the listening couch/chair several feet in front of the rear wall prevents dominate reflections.

Beyond these simple guidelines, all sorts of "if-thens" and "yeah-buts" apply. It's an art as much as a science.

[rcarlberg]

Regular stereo isn’t beneficial to most material, wrt imaging.

I'm sorry, I have read and re-read this statement a dozen times, and I cannot figure out what you were trying to say. Stereo isn't beneficial with regard to imaging? Huh? Are saying only mono will image?

[Quantum Cat]

Yes indeed rcarlberg, the room is absolutely vital to the sound. My main system is currently in a pretty crappy basement with a very low ceiling. It would sound way better in a nice acoustically live room with a high ceiling. It sounds great when I sit in the middle, so I can relate to your wealthy friend. Putting a $500K system in that room would be like driving a Ferarri in rush hour traffic: it might stroke your ego, but it's a senseless thing to do.

I still mourn the loss of Roy's Allison Acoustics.

[Firth]

I'm sorry, I have read and re-read this statement a dozen times, and I cannot figure out what you were trying to say. Stereo isn't beneficial with regard to imaging? Huh? Are saying only mono will image?

Most stereo speakers are horizontally dispersive and as a result each ear hears both speakers, which is precisely what SDA solved. The current edition of the Legend 800, took this to a higher fidelity by cancelling this distortion in a wider bandwidth. And as that review points out, it does affect timbre. However, IMO a small timbre effect is easily adapted to by the ear/mind. Planar speakers like the Maggies have a narrow beam and a sweet spot where this distortion is much less. The rear reflections are decorrelated in time and don’t impact the direct path as speakers which have wider horizontal dispersion. Headphones are perfect, in that each ear hears a channel. However, because the entire ear isn’t involved in collecting sound, headphones lack good spatial characteristics. I actually don’t listen to stereo any more, I listen to stereo recording in extended stereo which uses 5.1 speakers with precise time alignment. In this case, if a signal is equal in L/R channels, each ear will receive only the center channel. The L-R falls into L and SL, of course with appropriate gains and delays, and the R-L falls into R and SR with appropriate gains and delays. Speaker timbres are matched. Because this is done in very precise DSP (32 bit 192KHZ) and doesn’t sound artificial like what I have heard in the past.

[rcarlberg]

Wow. A lot to unpack here.

I'm not sure I want SDA to "solve" the "problem" of my ears hearing both speakers. When we sit down in front of a performing ensemble, we use both ears. The cello on the right and the celesta on the left do not go into one ear each only. Why a speaker manufacturer would think this is a problem that needs solving is beyond me.

The guy in the video mentions that headphones have 100% separation, but the bass notes are not felt in the viscera which is a major part of the experience of bass. Even beyond that, though, 100% separation isn't always a good thing. Engineers generally don't mix on headphones for that reason. Headphones can be great, on material mixed for headphones, but on normal recordings -- and especially normal recordings played back in a normal listening situation -- total separation is actually a disadvantage.

"Headphones lack good spatial characteristics" because the whole ear isn't involved? Are you talking about the pinna? In humans, our pinna are not movable (unlike, say, cats) so we gain spatial cues by moving our heads around (like most dogs). The video points out that doing this with headphones on simply causes the aural image to move with our head, unless we have some kind of 3D software that holds the image still. Listening to music, unlike say hunting in the woods, the ability to "echolocate" the source of a sound is far less important. It might be a cool effect, I dunno. But it doesn't seem very essential.

[rcarlberg]

. I actually don’t listen to stereo any more, I listen to stereo recording in extended stereo which uses 5.1 speakers with precise time alignment. In this case, if a signal is equal in L/R channels, each ear will receive only the center channel. The L-R falls into L and SL, of course with appropriate gains and delays, and the R-L falls into R and SR with appropriate gains and delays. Speaker timbres are matched. Because this is done in very precise DSP (32 bit 192KHZ) and doesn’t sound artificial like what I have heard in the past.

So a STEREO signal -- which is two channels of information, right? -- is fed to a 5.1 rig, which is 3.1 more playback channels than you have source material. Somehow, those extra channels have to be derived or created or simulated, since you're not playing a 5.1 source.

Makes me question how this can be more accurate or more realistic than stereo.

[Firth]

If the sound seems to come from a speaker in a stereo setup, that is not imaging. Imaging refers to instruments seeming to be located in space and beyond the boundaries of speakers. That’s what the large Maggies did for me, amazing.

[Firth]

So a STEREO signal -- which is two channels of information, right? -- is fed to a 5.1 rig, which is 3.1 more playback channels than you have source material. Somehow, those extra channels have to be derived or created or simulated, since you're not playing a 5.1 source.

Makes me question how this can be more accurate or more realistic than stereo.

If two speakers are in a speaker cabinet and time aligned before combining, that is not simulated. Neither is this extended stereo. It provides the correct localization as heard at the mixing desk with headphones or with near field stereo monitors.

[rcarlberg]

That went right over my head. Either you are using terms in contradiction to their actual meanings, or you misunderstand basic concepts, or you're describing something completely foreign to me. Somehow, we're not communicating.

[Firth]

That went right over my head. Either you are using terms in contradiction to their actual meanings, or you misunderstand basic concepts, or you're describing something completely foreign to me. Somehow, we're not communicating.

You can say that extended creates derived channels, however what I attempted to relate is that even if a signal is split and goes to a different speaker, and the two signals are identical at the speakers ear, it will localize to somewhere in space, not the speaker locations. If the gain is adjusted so that the localization is in the left field mostly for L versus SL, the image will seem in the front but expanded.

[rcarlberg]

Oh okay, I didn't get that at all from what you wrote. Yes, if you have 2.0 in front of you, and 2.0 in back of you, and both pair are perfectly balanced with each othe, then the signal will seem to come (more-or-less) from the left and right, not front-and-rear. It's like my friend who sits between his speakers in the middle of the room.

But with more room effect.

And zero flexibility in where you sit, or move around.

[3LockBox]

I can add a bit of experience with Polk SDA 'technology', which one might describe as cross-talk technology.

I tried finding images of the speakers I had, to no avail. If you're familiar with Polk Audio 10A speakers that's what mine look like with regard to front bafflle driver placement. My speakers were 21in wide, 13in tall and 9.5in deep. Two 6.5" drivers side by each, soft dome tweeter between the two woofers but mounted above the woofers. A 10" passive radiator resided on the rear baffle. The speakers were placed on 30" tall stands.

To my understanding., a midrange signal from the left speaker's inner driver was sent, out of phase to the right speaker's inner driver, and vice-versa. This was done with a supplied two-prong cable.

When I bought these speakers I lived in an apartment, that was a converted motel, a typical "no tell motel" layout albeit with a kitchen. Cinder block walls, shag carpet (minimal padding) and popcorn texture ceiling - a real dive. I followed the suggested placement of the speakers as prescribed by Polk and they were adequately powered. They were a marvel, provided one was in the 'sweet spot'. Whether tape or CD, the imaging and soundstaging of these speakers were incredible. I moved three or four more times afterward, each time adhering to the proper setup as suggested by Polk. They were great sounding speakers. However, they never imaged as good as they did in that dive apartment.

[rcarlberg]

Yes, Lockbox, that sounds like my understanding of SDA. On the left speaker, you have one midrange+tweeter which is Left (L), and one which is Right but 180° out of phase (-R). On the right you have R plus -L.

What that does is remove almost all center channel high end, signal that would be common to both speakers.

This exaggerates the stereo separation.

If you had no stereo separation--say, with mono material--the high end would effectively cancel itself out. This would also be true of any material mixed to the center of a stereo field, as many vocals are. In fact, it is common practice in most multi-track recordings to record instruments and vocals monaurally, and mix them across the stereo field. Anything in the middle of such a mix would essentially be REMOVED by the speakers.

I'm guessing there's probably a delay circuit in there, of a few milliseconds, otherwise the cancellations would be too egregious. Maybe that's what the "advanced SDA" is?

Incidentally? This is the exact same methodology employed by Bob Carver in his so-called "Sonic Hologram Generator."

Which itself is a phase-inverted version of the Aphex Audio Exciter.

Which was responsible for the exaggerated sibilance of much of the music of the 1980s. Which was the cause of the complaints that early CDs sounded harsh.

[Firth]

That is an interpretation which excludes much detail, not presented by the inventors. Which is why you need to hear these.

[Firth]

This article describes the constraining of the sound stage between the speakers for normal stereo.
https://www.polkaudio.com/en-us/blog...sda-technology

[Firth]

Think about where locations of sounds should be perceived. If a binaural mix placed instruments at different places across the front of the listener using headphones, the difference in amplitude in each ear would be correlated to location. Equal amplitudes would be perceived directly in front, and if the amplitude was purely in the right side, the location of the instrument would be perceived way off to the right. In order to really replicate how the location of sound sources are perceived with speakers external to the ears, speakers would need to be located at every possible instrument location across the field of listening. Using only two speakers to represent sound from these locations is not accurate unless the instruments are located where the speakers are located.

[3LockBox]

Incidentally? This is the exact same methodology employed by Bob Carver in his so-called "Sonic Hologram Generator."

I remember hearing a demo of Carver's Sonic Holographer back when. I remember it seemed to smear the sound, not unlike much of the "stereo expansion" DSP found in modern playback apps. My SDA CRS did not (compact reference speaker). However, I did hear the flagship SDA SRS (series reference speaker) with 8 mid-woofers and 4 tweeters that I personally liked better without the SDA cable.

[rcarlberg]

That is an interpretation which excludes much detail, not presented by the inventors. Which is why you need to hear these.

I'd like to sometime. I won't pass up the opportunity should it ever arise.

[rcarlberg]

This article describes the constraining of the sound stage between the speakers for normal stereo.
https://www.polkaudio.com/en-us/blog...sda-technology

I have read that article before, and it's based on a fallacy.

Our ears rely on cues to help us locate sounds. When something—a musical instrument, for example—produces a sound, that sound arrives at our closest ear first. By the time the sound reaches our other ear, it’s a little bit delayed, and a little bit softer. When that performance is recorded and replicated on two stereo speakers, the left speaker reproduces left-channel information while the right channel produces right-channel information. The channels are reproduced separately, but the instant the sound enters the room, the channels are no longer separated. We hear the sound coming from the left speaker in our left ear, and the sound coming from the right speaker in our right ear—but we also hear some of the sound from the left speaker crossing over into our right ear, and vice versa. This natural distortion, called interaural crosstalk, makes it difficult to tell where sounds are coming from, like we can in the real world.

Here's the fallacy: a stereo recording of a stereo soundfield will already INCLUDE those aural clues as to direction, with the right speaker emitting information from the left channel just a hair delayed and just a hair softer, and vice versa. The audio clues to localization are built-in... that's why it's called a "stereo recording." Out in the wilderness our ears always hear with "interaural crosstalk" -- that's how we localize sounds. And that's why the stereo signal coming out of your speakers should be reproduced as neutrally and accurately as possible, so the clues embedded in the recording can be heard as clearly as possible.

That said, there are OTHER speaker design issues which can definitely smear the left/right localization cues, what's called the imaging of your speakers... we can launch into that whole discussion if you're interested?

That is an interpretation which excludes much detail, not presented by the inventors.

Actually, the article repeats almost all of the points I made, now that I have re-read it.