[Bill R]

I've been hearing and reading a few things lately about how imaging has to do with content above 20k Hz. Being curious, I looked at the specs on my Panoramas, Essence, Sappires, and Emerald XLs, and I didn't see anything about what's going on above 20kHz. My amp is rated from 5 to 45kHz, and my CD source is rated from 4 to 44kHz. It seems odd that speakers would cut off at 20kHz. Wildly speculating, it seems more likely that there's rolloff above 20k, perhaps similar to how LF rolloff occurs... but again, that's just speculation.

Mike, can you share your insights here? Any technical data you could share would be welcome, and your own perceptions / experience re: what makes imaging work well would be interesting. About the only thing I can say for certain is that the Panoramas, Sapphires and Emerald XLs image very well.

[Mike Dzurko]

Bill:

Give me till later tonight or early tomorrow. Anyone else? Chime in!

[DR]

Bill,

I am no expert and am sure Mike will get back with a much more elegant answer. However, your CD drops off like a rock at about 20kHz. The 44kHz is the sampling rate.

I think imaging has a lot more to do with the mids than the highs. Most of the music you hear is in the mids with very little above 12kHz and very little below 100Hz. So a well centered voice, guitar, violin or whatever is what I consider good imaging. Maybe the voice is center and the guitarist slightly to the right. You get the picture. Depth also comes into play. Is the singer in front of the band (acoustic)?

My Sapphires excel at this.

Anyway that is my 2 cents worth (whoops, after today's market, 1 cent worth).

 

[stereocilia]

Bill, I'll second what DR says.

Not only does music live in the midrange, but most human hearing sensitivity rolls off before 20 KHz, often well before.  If you really dig, you can find debate about the importance of reproducing sound above or below the most-often quoted human hearing bandwidth of 20 Hz to 20 KHz.  But, as I understand it, 22.05 KHz is the highest theoretical frequency a CD can make.  I googled "CD sampling rate" to find:

http://www.cs.columbia.edu/~hgs/audio/44.1.html

I can tell you first-hand, the Sapphire XLs are stellar at imaging.

[DR]

One thing I have heard various people say is that extreme high frequencies can add a sense of "air". However, adding a super tweeter or something else to attain this would be futile for CD as Stereocilia says.

Here is one for you though, vinyl (with an all analogue chain) can go much higher in frequency!

[stereocilia]

Was it Stereophile this month that has an article showing an ultra-high frequency spectrograph of cd vs. analog?  Anyway, you don't have to strain to see the difference between those two graphs, that's for sure.

[Mike Dzurko]

I pretty much am in agreement with Stereocilia and DR. Extended highs (above 20K) at most, might add to some additional sense of air and space provided that info is actually on the recording (rare). And certainly not with Redbook CD . . . tape and LP are something different for sure.

To get back to one of Bill's other questions, yes, you are right. The speakers don't stop reproducing at 20K, rather, they are about 3db down from reference at this point with a gradual slope downward continuing on.

I think imaging is a great topic. There are many factors which contribute to great imaging. I'll add more when I get home tonight or tomorrow.

[Magnetar]

I've been hearing and reading a few things lately about how imaging has to do with content above 20k Hz. Being curious, I looked at the specs on my Panoramas, Essence, Sappires, and Emerald XLs, and I didn't see anything about what's going on above 20kHz. My amp is rated from 5 to 45kHz, and my CD source is rated from 4 to 44kHz. It seems odd that speakers would cut off at 20kHz. Wildly speculating, it seems more likely that there's rolloff above 20k, perhaps similar to how LF rolloff occurs... but again, that's just speculation.

Mike, can you share your insights here? Any technical data you could share would be welcome, and your own perceptions / experience re: what makes imaging work well would be interesting. About the only thing I can say for certain is that the Panoramas, Sapphires and Emerald XLs image very well.

Sorry to but in but CD's are limited to 22k - 44k is the sampling frequency

I have speakers that go out to 35k - With records, mostly 45s' there is ultra sonic information that MAY be positive to the sound - a good MC cartridge will track that high - CD's do not have frequencies this high, it was designed that way. SACD does, and so does DVD

James Boyk has some interesting thoughts and facts on high frequencies in playback to read - http://www.cco.caltech.edu/~boyk/spectra/spectra.htm A very interesting fellow.

[Mike Dzurko]

Thanks for the Jim Boyk link, a good read. In 05, John Potis wrote about his experiences with the $2000+ Murata supertweeters added to the Sapphire XLs:

http://www.6moons.com/audioreviews/murata/murata_2.html

[Joe_K]

Thanks for the Jim Boyk link, a good read. In 05, John Potis wrote about his experiences with the $2000+ Murata supertweeters added to the Sapphire XLs:

http://www.6moons.com/audioreviews/murata/murata_2.html

Nice writeup! Too bad they cost more than the Sapphires!
Other than wide bandwidth MC cartridges, there's also moving iron, (ie; variable reluctance) cartridges that go to 50khz. My old Sonus blue was one, although it has an annoying 6db peak at 45khz. Remember the CD-4 sub-carrier/decoder signal?
Cartridge Man currently makes two differant MI cartridges. (Not optimized for quad.)
I recall in the 80's people would disconnect the super tweeters, something about the extreme HF info could become irritating/fatiguing? Sounds like Murata has made serious advances with piezo technology.
If anyone does try these, please let us know your impressions.
JK

[bondmanp]

I defer to Mike and other engineering types, but I think there is an advantage to HF extension beyond 20kHz.  Unless I am mistaken, the behavior of dynamic drivers above 20kHz (their resonance, response patters, dispersion, etc.) have an affect on the audible frequencies in the top few octaves.  I think the concept is that extreme HF extension will minimize the adverse effects of sharp rool-offs and response peaks in the >20kHz range.  But, I could be completely wrong.

[richidoo]

Phase distortion (group delay) ruins imaging. It can happen anywhere in the recording / repro chain. Acoustic phase distortion is the biggest culprit, in the recording studio and in the listening room. Bass modes, comb-filtering, bad mic placement are clearly audible. But speaker crossovers are probably the second biggest source. Electronics are usually the lowest source, especially audiophile quality gear. Supertweeters are not required for perfect imaging, 10kHz sound has small enough wavelength to locate to high precision from long distance when there is no phase error. It is relatively easy to find a small bird singing in a tree 100+ feet away from you. But the new breed of high performance tweeters like RAAL and various berylliums make the phase error contained in recordings an increasing problem. Chesky seems to know how to get it right, they have some incredibly realistic recordings. Reference Recordings has great performances but not quite as good on the phase coloration, IMO. The recording sets the limit on how good the system can image. By "imaging" I mean locating sounds precisely within the stereo sound field all the way out to the speakers. A pop song with everything piled into the middle except vague spacey digital reverb is not good imaging, although even that can fail to center on cheap speakers.

The best imaging I ever heard was on full range (50-15k) single driver speakers with a great dac and all tube amps. Joe Henderson playing Mamacita on KennyDorham Trumpeta Tocatta. Speakers spread out 110 degrees, short triangle setup. To me it seemed like he was standing in the room, just to the left of the right speaker, I could not close my eyes and relax, I felt like he was a ghost standing right there playing for me. Intense and very emotional because I am afraid of ghosts. 

I heard Sapphire XL at RMAF 07 and 08. It is an amazing imager, even in a <8 foot listening triangle in a hotel room! Of course it was GIK treated, so acoustic phase distortion was minimized.

[DR]

What is GIK treated?

Thanks

[Joe_K]

DR,
It's a popular brand of room treatments.
http://www.gikacoustics.com/index.html
JK

[DR]

Thanks Joe!