|
|
|
|
| Re: INTERESTING READ why we need super high freq. even tho we can't hear them [message #206990 is a reply to message #206874 ] |
Mon, 18 December 2006 19:27   |
maxdimario
Messages: 3811
Registered: December 2004 |
Platinum Member |
|
|
sorry guys, but it's baby-logic.
all you have to do is use your ears to check.. there is improvement at lower (mid to high) freq.. ultrasonics are perhaps part of the equasion, perhaps not..
neve desks, as well as other broadcast-era desks can do 50-100KHz, but they have built-in filters at 15 or 20 KHz.. this was to remove any radio freq. from the signal path etc..
rest assured that an amplifier which cant go beyond 20 kHz even if there are no lowpass filters involved WILL NOT sound as clean and dynamic as an amp which can do 100KHz but is limited by engineering a high-freq cutoff.
techs such as lavry, and the other mathematics-based digi-people, insist that if the filter is perfect you dont need a high sampling rate, because all of the audible freq. components are there..
a simple and more efficient WORKING logic tells us that NOTHING is perfect. NOTHING incl. reconstruction filters..
by increasing the sample rate you are making a real-world improvement to the converter because you have more resolution in the top end..
PERIOD.
the whole concept of 'knowing exactly' why something sounds one way or another is bullshit.
the art in audio is to try and make distortion impossible by stretching the limits of the amp to the point that you have headroom as well as every other kind of 'room' for mistakes.
|
|
|
| Re: INTERESTING READ why we need super high freq. even tho we can't hear them [message #207019 is a reply to message #206990 ] |
Mon, 18 December 2006 21:03 |
 |
 Tomas Danko
Messages: 4614
Registered: May 2004
Location: Stockholm, Sweden |
Platinum Member |
|
|
| maxdimario wrote on Tue, 19 December 2006 01:27 |
by increasing the sample rate you are making a real-world improvement to the converter because you have more resolution in the top end..
|
Speaking strictly about digital recording there will be no increased "resolution" as you call it, once you've reached well above 20 kHz (ie 96 kHz for example). You will not capture any more details by going to 192 kHz for instance. Reading up on reconstruction filtering explains why.
As far as bad implementation goes regarding filter curves messing up the 20 kHz area and below, this is exactly why people use upsampling and oversampling. This has got nothing to do with the rate of sampling data, and hence upsampling/oversampling will solve this problem without having to record redundant information.
Once you're outside of the computers and digital gadgets, there are some good points when it comes to designing, for instance, a power amplifier with a bandwidth way past 20 kHz.
But that's a whole different enchilada.
"T(Z)= (n1+n2*Z^-1+n2*Z^-2)/(1+d1*z^-1+d2*z^-2)" - Mr. Dan Lavry
"Shaw baa laa raaw, sidle' yaa doot in dee splaa" . Mr Shooby Taylor
|
|
|
| Re: INTERESTING READ why we need super high freq. even tho we can't hear them [message #207079 is a reply to message #206874 ] |
Tue, 19 December 2006 03:14 |
iCombs
Messages: 542
Registered: March 2006
Location: Minneapollis |
Gold Member |
|
|
The best justification I have heard yet for higher sample rates (and not just higher bandwidth gear, which, IMHO, is a total crock. I don't need everything I touch to be able to do 5-100k. 20-20k will do just fine for 95% of it) is that you end up with more accurate curves with smoother transitions sample-to-sample, which results (IIRC) in better phase coherence within each track (as the waves produced more closely resemble true curves). The arguement for higher sample rates seems to me to make more sense than the "more bandwidth than we need by 5 times just because EVERYTHING will be more linear" arguement. Since when was linear any fun? Since when was linear rock and roll?
My .02.
PS - Arguements against linearity null and void in the presence of orchestral instrumentation and/or measurement.
Ian Combs
Producer/Engineer
Lightspeed Group, Inc.
----------------------
"Mista apareeatah... can I have maar beass at all frequencies?"
|
|
|
| Re: INTERESTING READ why we need super high freq. even tho we can't hear them [message #207092 is a reply to message #207079 ] |
Tue, 19 December 2006 05:12 |
|
Tomas Danko
Messages: 4614
Registered: May 2004
Location: Stockholm, Sweden |
Platinum Member |
|
|
| iCombs wrote on Tue, 19 December 2006 09:14 |
The best justification I have heard yet for higher sample rates (and not just higher bandwidth gear, which, IMHO, is a total crock. I don't need everything I touch to be able to do 5-100k. 20-20k will do just fine for 95% of it) is that you end up with more accurate curves with smoother transitions sample-to-sample, which results (IIRC) in better phase coherence within each track (as the waves produced more closely resemble true curves). The arguement for higher sample rates seems to me to make more sense than the "more bandwidth than we need by 5 times just because EVERYTHING will be more linear" arguement. Since when was linear any fun? Since when was linear rock and roll?
|
After the reconstruction filter the sample-to-sample transition will be just as smooth at 24/96 as at 24/192. There is nothing to be gained with higher sample rates in this regard. If there is a difference in transition it's because one of the devices are broken.
I find the role of the reconstruction filters to be one of the least understood, most people still think the final output is a series of steps. And as the great guy we know as Nika Aldrich used to say, you only need two points to draw a circle.
The devil may be in the details, but I think he's messing with implementation and not sample-to-sample transitions.
Cheers,
Tomas Danko
"T(Z)= (n1+n2*Z^-1+n2*Z^-2)/(1+d1*z^-1+d2*z^-2)" - Mr. Dan Lavry
"Shaw baa laa raaw, sidle' yaa doot in dee splaa" . Mr Shooby Taylor
|
|
|
| Re: INTERESTING READ why we need super high freq. even tho we can't hear them [message #207110 is a reply to message #206874 ] |
Tue, 19 December 2006 07:00 |
maxdimario
Messages: 3811
Registered: December 2004 |
Platinum Member |
|
|
theoretically...yes..
electronically...no.
math is a representation of electronics..it is NOT electronics.
and the detail is lost not at 12 KHz but 3 or 4 KHz.
I believe (intuitively) that it has to do with micro-phase issues as stated above, because the blandness that you get on the top end and thiner sound is very similar to some phase distortion you get out of inferior analog circuits incl. cheap opamps.
a digital recorder at 44.1 with a mixer made of IC's is a sure recipe for a thin and bland-sounding record. blech!
there may be a converter out there which sounds great at 44.1, or perhaps I should say as good at 44.1 than at 88.2 or 96 etc.
unfortunately when played back AT HOME or other places where this super-converter (which may or may not exist) is not available.. the result is that it DOES sound different.
EVEN IN THE ANALOG world, every machine which has claimed to get perfect results using smaller tape, at lower speeds, with smaller parts and less expense...has never really achieved what it claimed.
the usual way to sell those boxes was to start talking about specs.. of course.
|
|
|
| Re: INTERESTING READ why we need super high freq. even tho we can't hear them [message #207113 is a reply to message #207092 ] |
Tue, 19 December 2006 07:02 |
maxdimario
Messages: 3811
Registered: December 2004 |
Platinum Member |
|
|
| Quote: |
you only need two points to draw a circle
|
maybe YOU can draw a circle with 2 points. I can't
give me 50 points and I can get it almost right.
|
|
|
|
| Re: INTERESTING READ why we need super high freq. even tho we can't hear them [message #207127 is a reply to message #207113 ] |
Tue, 19 December 2006 08:47 |
Bill Mueller
Messages: 4123
Registered: April 2004
Location: Joppatowne Maryland |
Platinum Member |
|
|
Pop Quiz.
Bobby can hear 20khz. Bobby can NOT hear 21khz. Can Bobby hear the difference between a 20khz sine wave and a 20khz square wave?
Answer yes or no. Explain.
Best Regards,
Bill
"Don't take it personally. But this shit is a science." J.J.Blair
“The Internet is only a means of communication,” he wrote. “It is not an amorphous extraterrestrial body with an entitlement to norms that run counter to the fundamental principles of human rights. There is nothing in the criminal or civil law which legalizes that which is otherwise illegal simply because the transaction takes place over the Internet.” Irish judge, Peter Charleton
|
|
|
| Re: INTERESTING READ why we need super high freq. even tho we can't hear them [message #207158 is a reply to message #207127 ] |
Tue, 19 December 2006 10:10 |
|
no.
A 20khz square wave is a 20khz sine with overtones, that bobby cannot hear.
Bobby will hear a sine in both cases.
|
|
|
| Re: INTERESTING READ why we need super high freq. even tho we can't hear them [message #207167 is a reply to message #206874 ] |
Tue, 19 December 2006 10:36 |
maxdimario
Messages: 3811
Registered: December 2004 |
Platinum Member |
|
|
the point is not that a 20 KHz sine or square wave sound different, it's whether a converter can reproduce 3-15 KHz convincingly or not..
a 20KHZ square wave with a brickwall filter BECOMES a sine wave.
two points give enough information to make a circle, but this is assuming that the METHOD to extract the circle from the two points is perfect.. and it isn't. (I can't draw a circle perfectly BY HAND with two points)
the proof is in the pudding..
remember the PT bass test...
|
|
|
| Re: INTERESTING READ why we need super high freq. even tho we can't hear them [message #207195 is a reply to message #207167 ] |
Tue, 19 December 2006 11:59 |
Bill Mueller
Messages: 4123
Registered: April 2004
Location: Joppatowne Maryland |
Platinum Member |
|
|
Max,
Any audio system (including digital encoding system, analog electronic, magnetic and mechanical) with bandwidth equal or greater to the testing system (the listener) will encode inter-bandwidth information with resolution (bandwidth) equal to or greater than the testing system. That is the point. The listener is not able to perceive the detail in a 3khz square wave beyond their 20khz limit. Therefore there is no need to exceed the bandwidth of the testing system, by any great amount.
Simultaneously, and you know exactly well, that increased bandwidth also brings with it, increased noise, with no discernible improvement in fidelity. This is why every well designed audio circuit is bandwidth limited, to keep out thermal noise, radio frequency noise and non linear distortion.
Best Regards,
Bill
"Don't take it personally. But this shit is a science." J.J.Blair
“The Internet is only a means of communication,” he wrote. “It is not an amorphous extraterrestrial body with an entitlement to norms that run counter to the fundamental principles of human rights. There is nothing in the criminal or civil law which legalizes that which is otherwise illegal simply because the transaction takes place over the Internet.” Irish judge, Peter Charleton
|
|
|
|
Members
Search
FAQ
Register
Login
Home
