advertisement


Cable to super tweeters

Supertweeters do make a big subjective difference to the sound quality of most speakers, surprisingly almost always making the bass sound better defined and more taut. I would guess it probably has to do with enhancing the portrayal of the cues around the performed note that tell us more about what it actually sounds like (ie. does the bass guitar resonate the snare on the snare drum at a certain note?). As I have said above a traditional supertweeter implementation is clearly operational at audible frequencies of 10kHz and above - I have mine set with an 18kHz crossover point with a 27Ohm resistor in parallel and it makes a big difference in the audible range - albeit with full range speakers that naturally roll-off above 12-15KHZ.

If I do the hearing tests I can hear up to about 17-18kHz (29 years old), although in every test instance you are only hearing the aggregate of how much the frequency response of the speaker and of your ears has deteriorated at this point.
 
sybil said:
Supertweeters do make a big subjective difference to the sound quality of most speakers, surprisingly almost always making the bass sound better defined and more taut. I would guess it probably has to do with enhancing the portrayal of the cues around the performed note that tell us more about what it actually sounds like (ie. does the bass guitar resonate the snare on the snare drum at a certain note?). As I have said above a traditional supertweeter implementation is clearly operational at audible frequencies of 10kHz and above - I have mine set with an 18kHz crossover point with a 27Ohm resistor in parallel and it makes a big difference in the audible range - albeit with full range speakers that naturally roll-off above 12-15KHZ.

If I do the hearing tests I can hear up to about 17-18kHz (29 years old), although in every test instance you are only hearing the aggregate of how much the frequency response of the speaker and of your ears has deteriorated at this point.
Click to expand...

Interesting. Snares are quite easy to hear with conventional (22k) tweeters, however, and all this is predicated on vinyl replay, I'm guessing.
 
It's worth bearing in mind that many tympani produce harmonics way above 20khz. It seems that most people feel that extended treble response makes sounds such as cymbals and bells sound more authentic and identifiable.

So it's easy to hear a snare with Tannoys limited to 15khz but they might sound more realistic if supertweeters are added.
 
Merlin, that is exactly what I was getting at - any instrument is going to set up resonances in adjacent objects that themselves may well produce harmonics way beyond 20KHz. The bass guitar causing the snare to rattle was the most obvious example I could think of of a recorded bass instrument that could only be reproduced in full using an extended range tweeter. Not to say it won't sound great without a supertweeter though, especially if you have well balanced speakers with an extended (flat to 20KHz) frequency response.
 
it would be cool to find a microphone that is in widespread use that's flat n' clean upto 40khz.... i think you will find it's tough to get one that can withstand high sound pressure levels....and mic pre's.... i know some neve gear went upto 60khz fairly flat but no vintage or modern mics really do that....plus instrument levels themselves are way down in level at those frequencies.

i think what simon and tenson are saying is that the power response and dispersion may be helped but not a great deal else.

remember cd is pretty rolled off around 19khz anyway.

a lot of a-harmonic mush lives up there too.

the 20hz to 20 khz recommendation is pretty good a guide.
 
Darryl,

I'd tend to agree but would suggest that it's what is happening in the system between 20khz and 40khz that has an affect on what we hear.

Brickwall filters and break up modes result in phase errors that we are particularly sensitive too. The impulse response of a resonance may well have an affect despite being above the frequency that we can hear a steady state test tone.

I agree that mics suffer the same problems. Directional mics suffer phase issues I'm told and few reach 40khz even today. But it's not so much the information that they capture that I am thinking of as much as their behavior once ultrasonic. A smooth roll off above 15khz is preferable to and break up mode or brickwall filter above 20khz.

It was Peter Lyngdorf who many years ago was mentioning the EU/Kef/B&O tests where they were able to demonstrate that people were able to identify transient information at higher frequencies than the standard 20-20 range. If that was true, then it would stand to reason that we could hear the effects of breakups and brickwall filters in the time domain.

Certainly the most authentic systems I've listened to have had extended responses with no obvious break up, and I can also identify high resolution recordings over red book blind. It's an interesting area - and for once, not one that seems to be the exclusive preserve of the audiophool extremists.
 
merlin said:
Brickwall filters ... result in phase errors
Click to expand...

They don't have to. And thus most in existence simply don't.


that we are particularly sensitive too.
Click to expand...

Please show me accepted studies that indicate a particular sensitivity to phase errors. I have a pile that claim the opposite.

It was Peter Lyngdorf who many years ago was mentioning the EU/Kef/B&O tests where they were able to demonstrate that people were able to identify transient information at higher frequencies than the standard 20-20 range.
Click to expand...

Back then I got the final reports of the Archimedes project. It did not investigate such things, and thus did not claim the above. What it did investigate was the perception of reproduced sound in rooms, and that is a context in which typically inter-aural timing accuity gets discussed. This goes down to nanoseconds, as is widely known. Misunderstanding or misrepresenting this fact may lead to false claims that human auditory perception is equivalent to several 100 kHz of bandwidth.


I can also identify high resolution recordings over red book blind. It's an interesting area -
Click to expand...

Excellent. I take it you want to volunteer in the simple test I am hoping to conduct early next year?
 
I knew I shouldn't have posted something controversial on an internet forum :)

I retract everything.

I am however intrigued by your choice of Quads if phase coherence is a non issue.
 
I really don't see what my Quads have to do with your claims on the audibility of effects in the 20-40kHz range, including signal path non-linear phase distortion.

Do you volunteer for the test? I really need discerning listeners.
 
Werner I'm afraid I really don't have the time or the inclination to participate in any form of bake off/test/trial these days. My interest in audio theory is really not that strong.

Once upon a time maybe but not now.

All I will say is I will not make frivolous claims for much in the audio world but I have repeatedly heard the differences I mention under completely blind conditions. I ripped vinyl to AIFF at 24/96 then used Pure Vinyl to make a Red Book copy for portable devices.

When selected for playback via iPad's remote app, there is absolutely nothing to identify which is which - a situation I faced with all of these albums. Everytime I correctly identified the high res original. It is not easy or obvious but there is a readily identifiable difference and the files were level matched. Now you can claim all manner of reasons for this but there is no power of suggestion at work.

It is also remarkably easy to introduce a filter above people's hearing threshold and for them to notice the difference as an increase in bass. Again. Blind. Again no power of suggestion. I would be interested in your findings. You clearly have the time and inclination to explore these phenomenon - hopefully with an open mind.
 
merlin said:
I ripped vinyl to AIFF at 24/96 then used Pure Vinyl to make a Red Book copy for portable devices.....Everytime I correctly identified the high res original.
Click to expand...

I would be far more inclined to accept that what you were identifying was the sonic 'footprint' of the software, not inherent limitations of the format. It is quite simple to make 16/44 digital copies of vinyl that are entirely indistinguishable from the original, even using a fairly cheap analogue to digital converter.
 
Plutox said:
I would be far more inclined to accept that what you were identifying was the sonic 'footprint' of the software, not inherent limitations of the format. It is quite simple to make 16/44 digital copies of vinyl that are entirely indistinguishable from the original, even using a fairly cheap analogue to digital converter.
Click to expand...

Anything is possible in life I guess.

People will often point to objective science and claim that the digital decimation must be to blame in this case, yet when objective science shows no evidence of any "footprint" left by the decimation process they will usually ignore that.
 
Given that I claim to be able to convert any sensibly modulated hi-res recording to 16/44 and that you will be unable to distinguish the two by listening alone (at least, when played on a half-decent DAC), my confidence is boosted by the fact that, even at its grubbiest, 16/44 is capable of performance an order of magnitude beyond that of vinyl.

In objective terms, I'm sceptical that relatively low rent software like Pure Vinyl will make that good a job of the sample rate conversion but I could be wrong.
 
I'd recommend checking the specifications on their website and maybe trying it out FOC. See if you can produce two files with a "footprint" that you can bring back to the forum and show where the degradation lies.

http://www.channld.com/pure-vinyl_src.html

Would you care to explain how the "low rent software" approach of Channel D (again have a look at the link) is inferior to say studio processing of the last decade or the SRC performed in your "half decent DAC"?

I'd also be interested in the signal chain you are using, from source component to monitor.
 
I would try it out, were I a McUser. I've tried a great many SRC programs in my quest and it's apparent (to me at any rate) that few cut the mustard, particularly when the target sampling rate is 44.1kHz*. If, at under $200, they are offering a conversion module that competes well with the top-end of the industry I'm sure there will be plaudits. When evaluating SRC systems, I tend to use material that is 40-50dB lower than it would typically be, and apply gain following the processing. That way, processing artefacts are far more noticeable.

The curve you link-to does look good, but the others cited for comparison don't exactly stack up in terms of an industry-wide beauty pageant of SRC software. FWIW, in my view, the more revealing curve (on the Infinitewave web site) that correlates with audible artefacts is not the sweep test but the gunk that accompanies the 1kHz pure tone test.


* I assume this is because the impact of the necessary filtering comes so perilously close to the audible region. Even 48kHz seems to be an easier target sampling rate.
 
You must log in or register to reply here.


advertisement


Back
Top