The Fiio X5 Thread
Apr 25, 2015 at 8:54 AM Post #17,206 of 19,652
  What UI did you had installed?
 
What were you doing the day before?
 
Does you computer has any virus by chance?
 
Had you tried keeping the power button pressed for at least 20 seconds?
 
Have you tried puttin an original firmware on card1, then keeping pressed upper left and power button, to see if you can reflash firmware?
 
these methods are how most people reverted it back to life, it is a few pages ago in this thread. I hope that one of these methods will work for you.

 
I held the power button for at least 20 seconds and that didn't seem to do anything. So I did the same thing with the reset button and then hit the power button on afterward. It surprisingly worked. Now that I think of it, I might have had this issue with the X3 before.
 
I'm not sure why it happened. I am using the default UI. I was using the X5 in DAC mode the entire day. I couldn't have done anything with the firmware because it wasn't doing anything with a regular power-on. The computer wasn't reading the X5 either.
 
Apr 25, 2015 at 8:55 AM Post #17,207 of 19,652
Apr 25, 2015 at 9:31 AM Post #17,209 of 19,652
The argument is clear to me from a qualitative point of view.  I'd accept the argument is solid in theory.  The question is whether the ultrasonic frequency's modulation of lower audible frequencies have an audible impact.  JaZZ offered the graph as an example which serves the discussion well.  I think over there at sound science we could get a pretty good argument going (again) about whether these modulations are audible.  Obviously that depends on how big they are, how much audible frequency they cancel, and how that kind of loss or distortion is perceived.

I have developed a much more effective filter.  I can no longer hear high frequencies well enough to be too worried about what JaZZ demonstrates are at least physically plausible distortions of the original audio signal.

This is of course why bit perfect rendering of digital files is so stylish right now (and maybe for a long time?).


It's just a theory.
There is no proof offered.
There are no practical examples.
Really it just seems to be a form of unwanted distortion.
 
Apr 25, 2015 at 9:33 AM Post #17,210 of 19,652
The term «subharmonic» may be misleading. In the context of a possible impact of ultrasonic frequencies on the audible frequency band it means interferences between ultrasonics or between ultrasonics and audible frequencies which would become audible. Therefore the term «difference tones» is more adequate in my book. My above example with the amplitude modulated sine waves serves for demonstrating a possible scenario of such an interference. Imagine that the original signal had this very amplitude modulation (let's say at 20 kHz for the sake of clarity, as it's very pronounced with this frequency), a beat – the equivalent of an interference with a foreign frequency, in the case at hand 44.1 kHz. (Un)fortunately the reconstruction filter eliminates the amplitude modulation in favor of a continuous sine wave – as is its task in the interest of preventing aliasing. But logically it has removed an information. Although 44.1 kHz is considered inaudible, the beat/amplitude modulation caused by it may absolutely be audible. So from this example one could get an idea of how ultrasonics could have an audible impact, despite being inaudible themselves.


Difference tones are Intermodulation Distortion, which you really don't want to hear.
IM is considered worse than THD.
 
Apr 25, 2015 at 9:47 AM Post #17,211 of 19,652
Difference tones are Intermodulation Distortion, which you really don't want to hear.
IM is considered worse than THD.

 
Yes, it may fall under this category in a wider sense. Nevertheless, if it provides the impression of higher lifelikeness like (possibly) in the case of ultrasonics manifesting themselves in the audio range... (I don't get it why you're so reluctant to at least allow your imagination to play with the idea of said effects.)
 
Apr 25, 2015 at 10:08 AM Post #17,212 of 19,652
Yes, it may fall under this category in a wider sense. Nevertheless, if it provides the impression of higher lifelikeness like (possibly) in the case of ultrasonics manifesting themselves in the audio range... (I don't get it why you're so reluctant to at least allow your imagination to play with the idea of said effects.)


Youu seem to discussing an artifact of incorrect sampling and make it into a virtue.
Anti-aliasing filters are designed to NOT allow (or minimize) these types of distortions.
 
Apr 25, 2015 at 10:09 AM Post #17,213 of 19,652
   
I held the power button for at least 20 seconds and that didn't seem to do anything. So I did the same thing with the reset button and then hit the power button on afterward. It surprisingly worked. Now that I think of it, I might have had this issue with the X3 before.
 
I'm not sure why it happened. I am using the default UI. I was using the X5 in DAC mode the entire day. I couldn't have done anything with the firmware because it wasn't doing anything with a regular power-on. The computer wasn't reading the X5 either.

all is good then!
 
Apr 25, 2015 at 10:37 AM Post #17,214 of 19,652
Youu seem to discussing an artifact of incorrect sampling and make it into a virtue. Anti-aliasing filters are designed to NOT allow (or minimize) these types of distortions.

 
No, you're misinterpreting it:
 
  Imagine that the original signal had this very amplitude modulation (let's say at 20 kHz for the sake of clarity, as it's very pronounced with this frequency), a beat – the equivalent of an interference with a foreign frequency, in the case at hand 44.1 kHz. (Un)fortunately the reconstruction filter eliminates the amplitude modulation in favor of a continuous sine wave – as is its task in the interest of preventing aliasing. But logically it has removed an information. Although 44.1 kHz is considered inaudible, the beat/amplitude modulation caused by it may absolutely be audible. So from this example one could get an idea of how ultrasonics could have an audible impact, despite being inaudible themselves.

 
So – in the above scenario! – the «distortion», as you call it, is in fact a component of the original signal, not a bad reconstruction. The «bad reconstruction» is the elimination of the amplitude modulation.
 
Apr 25, 2015 at 10:48 AM Post #17,215 of 19,652
No, you're misinterpreting it:


So – in the above scenario! – the «distortion», as you call it, is in fact a component of the original signal, not a bad reconstruction. The «bad reconstruction» is the elimination of the amplitude modulation.


Then why demonstrate this by showing a set of pure tones?
Do you have a real world example?
 
Apr 25, 2015 at 11:00 AM Post #17,216 of 19,652
Then why demonstrate this by showing a set of pure tones? Do you have a real world example?

 
I don't have any real example – it's just a theoretical scenario, but from my perspective a realistic one. You have to rely on your fantasy and imagination (provided you don't have any ideological reservation).
wink.gif
 
 
Apr 25, 2015 at 11:18 AM Post #17,217 of 19,652
  My Sirius XM radio subscription recently expired for my auto sound system so I thought I'd try using the X5 into the auxiliary port in the mean time.  I must say that I was very impressed with the quality of the X5 Line Out signal thru my car stereo.  The clarity of the DAC was truly exceptional through JVC stereo and speakers. Now I'm left debating whether I should renew my Sirius XM radio. . .

I often use my X5 in the car, and my car has never sounded better. I've been taking lots of moderate to long trips lately, as I have a kid who's doing the college shopping tour, and fortunately he likes a lot of my music.
 
I've never got along well with Sirius XM. Compressed digital crap. The only argument for it is variety.
 
Apr 25, 2015 at 11:39 AM Post #17,218 of 19,652
   
I don't have any real example – it's just a theoretical scenario, but from my perspective a realistic one. You have to rely on your fantasy and imagination (provided you don't have any ideological reservation).
wink.gif
 

 
   
No, you're misinterpreting it:
 
 
So – in the above scenario! – the «distortion», as you call it, is in fact a component of the original signal, not a bad reconstruction. The «bad reconstruction» is the elimination of the amplitude modulation.

 
 
 
 
Actually, if this exists, it can be demonstrated provided you have the necessary equipment. Use x5->line out->mic in. Try to use high pitch tones, ~16khz, these should produce ultrasonic sub harmonics. 
 
You can generate pure tones using specific software, the effect can be prooven, if it exists. 
 
Having a higher sample rate affects DAC stage deconding in other ways, from my code knowledge, the interpolation algorithm can create a much more precise, and less stressed wave based on more samples, consider DAC process as a drawing of a graphic of a certain function. so if   f(x)=sinX, where x takes values from 0 to plus infinite. This would result in 1khz tone. 
 
The DAC has samples, points, and based off them it needs to re-draw the original graphic, which never looks like the original wave, because no DAC up to the date does not try to guess what f(x) but it tries to take into account a few samples, and decide how the road between them looks like, and draws it into analog. 
 
Considering that it has more samples, 1 second of 1KHZ contains roughly 1000 cycles. this means that every cycle is defined by 44 samples. if the algorithm would had had say double, 88 samples, or even better 385 samples to work with, the road between the samples is shorter and easier to aproximate it to the real sine 1khz wave.
 
Apr 25, 2015 at 12:24 PM Post #17,220 of 19,652
 
Actually, if this exists, it can be demonstrated provided you have the necessary equipment. Use x5->line out->mic in. Try to use high pitch tones, ~16khz, these should produce ultrasonic sub harmonics. 
 
You can generate pure tones using specific software, the effect can be prooven, if it exists. 
 
Having a higher sample rate affects DAC stage deconding in other ways, from my code knowledge, the interpolation algorithm can create a much more precise, and less stressed wave based on more samples, consider DAC process as a drawing of a graphic of a certain function. so if   f(x)=sinX, where x takes values from 0 to plus infinite. This would result in 1khz tone. 
 
The DAC has samples, points, and based off them it needs to re-draw the original graphic, which never looks like the original wave, because no DAC up to the date does not try to guess what f(x) but it tries to take into account a few samples, and decide how the road between them looks like, and draws it into analog. 
 
Considering that it has more samples, 1 second of 1KHZ contains roughly 1000 cycles. this means that every cycle is defined by 44 samples. if the algorithm would had had say double, 88 samples, or even better 385 samples to work with, the road between the samples is shorter and easier to aproximate it to the real sine 1khz wave.

 
George, if you like to do it, more power to you! I propose to mix a 14 and a 30 kHz tone plus a 20 and a 42 kHz tone in a 176.4 or 192 kHz wave file and make a screenshot of the resulting curve. Then downconvert it to 44.1 kHz and make a screenshots again. Additionally send it through the FiiO X5's line out and sample it with 176.4 or 192 kHz to see what the anti-aliasing filter does to it. You may get more striking results with other tone combinations, so feel free to try whatever comes to mind, just take care that at least one of the sine waves is ultrasonic.
 

Users who are viewing this thread

Back
Top