I have no idea of what is going on above 20kHz, i'd have to consult an expert, i'll ask the cat when it wakes up again. I reckon this is the 21st century equivalent of playing an lp backwards to hear the devils messages. Is it relevant in any way known to mankind?
Why is everything so complicated, twas never thus with ye-olde reel to reel machines. If the pushers ever get a handle on digital, the snake oil thread will take over the forum.
Why is everything so complicated, twas never thus with ye-olde reel to reel machines. If the pushers ever get a handle on digital, the snake oil thread will take over the forum.
The only "correct" waveform from a 16/44.1 file is the waveform out of the reconstruction filter. And there's no way to make it "more correct," unless you're going to claim that 32 bit processing is prescient.
se
Good point - this, https://www.youtube.com/watch?v=hyetrAePLTA, was done nearly 100 years ago ... there have always been creative people around ...
There are myriad ways of getting analogue from digital - a serious concept is having 65,536 tiny speakers on a baffle, and flicking DC on and off with a toggle switch on each "bit" speaker - where's your reconstruction filter now ... ?
Last edited:
Frank,
Unless you are going to have 65,536 discrete channels driving all those speakers I'm not so sure that would be any type of improvement? I can think of married ways to use an array like that but not sure it is relevant to the discussion. On the more points on a graph giving a better fit to a curve I don't think that is correct either. As long as you have enough points to describe the curvature with the peak value and crossing point on the X axis and a few other points that should mathematically be enough to describe any pure sine wave I think. Yes if you are looking with your eyes at the graph it may look more like an analog waveform but it shouldn't be a necessity to accurately describe the curve. Only one curve is going to fit the curvature if you have the minimum number of points. Now the math to describe a multi-tone waveform with multiple frequencies happening simultaneously is beyond my math skills and how that is encoded and retrieved.
Unless you are going to have 65,536 discrete channels driving all those speakers I'm not so sure that would be any type of improvement? I can think of married ways to use an array like that but not sure it is relevant to the discussion. On the more points on a graph giving a better fit to a curve I don't think that is correct either. As long as you have enough points to describe the curvature with the peak value and crossing point on the X axis and a few other points that should mathematically be enough to describe any pure sine wave I think. Yes if you are looking with your eyes at the graph it may look more like an analog waveform but it shouldn't be a necessity to accurately describe the curve. Only one curve is going to fit the curvature if you have the minimum number of points. Now the math to describe a multi-tone waveform with multiple frequencies happening simultaneously is beyond my math skills and how that is encoded and retrieved.
Here you go then ... the big boys were playing with this idea ... https://books.google.com.au/books?id=NOPpwVvNu44C&pg=PA36&lpg=PA36&dq=philips+research+digital+loudspeaker&source=bl&ots=qC9rQNSpOW&sig=1GUaD-69oBNfFN51JIq_YYb6wVM&hl=en&sa=X&ei=lfP8VLaJA4XHmAWfp4C4BQ&ved=0CI4BEOgBMBQ#v=onepage&q=philips%20research%20digital%20loudspeaker&f=false
We are not making anything "more correct" - we're stopping it being "more incorrect" - that's the thrust of the effort ...
I'd like to see you deleted also as a charlatan intellectual.. And, that is not a personal attack 'cause I said it isnt an attack but a belief of fact.
SE..... I've been messed with by professionals. You dont qualify. Is that the best you can argue your view point when all else fails resort to name calling....? Sad. Where is the inventiveness? originality? Put some real effort into it and see what happens
THx-RNMarsh
Last edited:
I'm guessing they dropped it at some point in the 30 years since that article was written.
NXT was a blind alley too. I guess we're stuck with what actually works well eh
Then why the hell are you arguing to upsample again after downsampling? Upsampling after you've downsampled to 16/44.1/48 isn't going to make it any more CORRECT or INCORRECT. But you essentially argued that upsampling would make it MORE CORRECT when you made your 20kHz sine wave "argument."
Are you not even keeping track of what you have been saying?
se
As an interesting exercise that I'm doing right at this very moment is playing music on those PC speakers, from CD source. This is bottom of the pecking order audio playback, everything wrong about it as being something worthwhile ... yet, this is giving me playback at the moment which some people would call "analogue", it has a fluid, easy, rich, sparkling quality on classical recordings, which makes it a delight to listen to. And at any time I can damage that quality, and give it that throwaway, unpleasant, edgy "digital" tonality that audiophiles hate, just by doing a few simple things to upset the electrical environment that the system sees. I can go from decent "correctness" to "get me outta here!!" sound by doing seemingly trivial things - because I'm upsetting the "correct" operation of the playback chain ... this is why one has to focus very carefully on all these "small things" ...
Last edited:
People in this thread who are much more knowledgeable than I have put a mountain of "real effort into it." Yet you remain stubbornly, WILFULLY ignorant on even some of the most basic concepts. You are a huge waste of time and shouldn't be allowed anywhere near the word "professional."
se
Amusingly, I have a super el cheapo NXT bit o' kit right in front of me, on a shelf - I picked it up some time ago, as one does,I'm guessing they dropped it at some point in the 30 years since that article was written.
NXT was a blind alley too. I guess we're stuck with what actually works well eh
- something to play with down the track perhaps ... .- Status
- Not open for further replies.