[DDDamian]

Get the feeling we're chasing ghosts here? 

 

[JaZZ]

Spoiler: Warning: Spoiler!

Quote:

Regarding DSD, this blog should be read by everyone who thinks it's a better format (relatively short without too much tech talk). If you are going to take the time and effort to seek out DSD or convert your library to DSD you should, at minimum, understand why and what benefit you will be getting, if any at all.

http://www.mojo-audio.com/blog/dsd-vs-pcm-myth-vs-truth/

DSD was essentially introduced as a cheap way for Sony to archive their music library. Marketing caught wind of it and upsold the 'benefits' to consumers and SACD was born - which there really are no benefits except for a cheaper and less expensive way to approximate true audio fidelity using cheap Delta-Sigma DAC chips. At the expense of quality we are forever caught up in a marketing game of high numbers that actually is hurting the fidelity of our music. Similar to what has been happening with the dynamic range loudness wars.

In the end DSD doesn't actually provide more detail than high res PCM recordings. From the blog............
The big issue I have with DSD is that to clean up the noise you need to add noise shaping algorithms and to me it just kills the life of the music. What I hear from DSD is digital, plastic, and lifeless. Not only that, but in direct tests I have done the audible difference is basically non-existent except for noted above. I dare say the differences that may exist are for the worse. I admit, I initially fell for the marketing trap but over time became aware of the 'how and why' of DSD because something just sounded off to me.
Everything about DSD to me screams compromise. PCM is fine for me and I'd rather not occupy my precious storage space with a format that, in my opinion, is inferior, and nothing more than marketing.

As always, this is my perspective on the formats, but the facts are there. They back up what I hear and if you read the link provided they may back up your experiences as well. BUT at the end of the day if you just love DSD and love what you hear from it who am I to say anyone is wrong. Just enjoy what floats your boat.

I don't want this thread turning in to a DSD vs PCM battle. We should keep on topic but there has been so much said about DSD quality with the X5ii I thought it appropriate to post this.

 
Just to add – relating to DSD64: A sampling rate of 2.822 MHz allows for at best 176 amplitude values per single wave at 16 kHz – the equivalent of 7,46 bit in PCM. At 20 Hz there are up to 141,120 amplitude values available, corresponding to 17,1 bit. Considering positive and negative amplitude, there are in fact less than half of it available, which means 60–70 ampli­tude values per single wave. But this only in the case of a pure 16 kHz sine wave. With multi-tone signals it gets even much worse. So the high-frequency resolution is effectively extremely poor – and no noise shaping can fix that. It can just remove the noise associated to the poor resolution in the amplitude domain.

 

[Joe Bloggs]

Just to add – relating to DSD64: A sampling rate of 2,822 MHz allows for at best 176 amplitude values per single wave at 16 kHz – the equivalent of 7,46 bit in PCM. At 20 Hz there are up to 141,120 amplitude values available, corresponding to 17,1 bit. Considering positive and negative amplitude, there are in fact less than half of it available, which means 60–70 ampli­tude values per single wave. But this only in the case of a pure 16 kHz sine wave. With multi-tone signals it gets even much worse. So the high-frequency resolution is effectively extremely poor – and no noise shaping can fix that. It can just remove the noise associated to the poor resolution in the amplitude domain.

That's not how delta-sigma works though--there's no encoding in the bitstream for how big a step up or down each step is. No, each step is actually all the way up or all the way down... it is more like a noise-shaped 1-bit PCM and the output is lowpass filtered to obtain the approximation of the original recording. The calculation of effective bit depth is much more complex than the above and inextricable from the noise shaping algorithm used. Certainly there's more than 17.1 bit available at 20Hz.

 

[JaZZ]

That's not how delta-sigma works though--there's no encoding in the bitstream for how big a step up or down each step is. No, each step is actually all the way up or all the way down... it is more like a noise-shaped 1-bit PCM and the output is lowpass filtered to obtain the approximation of the original recording. The calculation of effective bit depth is much more complex than the above and inextricable from the noise shaping algorithm used. Certainly there's more than 17.1 bit available at 20Hz.

 
I think you're misinterpreting my post and the displayed scenario. The 1-bit architecture is the basis of my criticism.
 
Take the 2,822,000 samples per second. Divide them by 16,000. You get ~176 samples per ¹/_16,000 second. So you have 88 samples at your disposal for one half wave. With a triangle wave you can use 44 samples saying «up» and 44 samples saying «down» in a row – representing 44 amplitude values. With a sine wave it's a bit less, because you have some deviating samples in between to achieve the sine shape. So that's the dynamic resolution disposable at 16,000 Hz with a pure sine wave. With a complex signal it may be much less.
 
You mentioned 20 Hz: 2,822,000 divided by 20 makes 141,100. That's 70,550 samples per (triangle!) half wave – corresponding to...
 
Hey Joe!

 

 

[Joe Bloggs]

That's not how delta-sigma works though--there's no encoding in the bitstream for how big a step up or down each step is. No, each step is actually all the way up or all the way down... it is more like a noise-shaped 1-bit PCM and the output is lowpass filtered to obtain the approximation of the original recording. The calculation of effective bit depth is much more complex than the above and inextricable from the noise shaping algorithm used. Certainly there's more than 17.1 bit available at 20Hz.

I think you're misinterpreting my post and the displayed scenario. The 1-bit architecture is the fundament of my criticism.

Take the 2,822,000 samples per second. Divide them by 16,000. You get ~176 samples per ¹/_16,000 second. So you have 88 samples at your disposal for one half wave. With a triangle wave you can use 44 samples saying «up» and 44 samples saying «down» in a row – representing 44 amplitude values. With a sine wave it's a bit less, because you have some deviating samples in between to achieve the sine shape. So that's the dynamic resolution disposable at 16,000 Hz with a pure sine wave. With a complex signal it may be much less.

You mentioned 20 Hz: 2,822,000 divided by 20 makes 141,100. That's 70,550 samples per single (triangle!) wave – corresponding to...

Hey Joe!

I didn't know we go back together that long

But remember I said that there's something fundamentally wrong with this analysis. theaudiohobby even somehow managed to quote a relevant snippet pointing out your mistake, even though he couldn't tell his hertz from his bits...

 

[Light - Man]

I didn't know we go back together that long

But remember I said that there's something fundamentally wrong with this analysis. theaudiohobby even somehow managed to quote a relevant snippet pointing out your mistake, even though he couldn't tell his hertz from his bits...

 
Joe, I know my Hertz from my Bits, one ia a car rental shop and the other is A Bit of Music and A Bit of Fun.
 
Light-Man would like to hear what type of music Joe and the Fiio team listen to - the fun bit is optional!

 
 

Here we go Guys a new thread called:
 
A Bit of Music and a Bit of Fun
 
http://www.head-fi.org/t/775356/a-bit-of-music-and-a-bit-of-fun
 
We are hoping for some good music and different music from all parts of the world in any language.
 
No rules - just music/ video links and a bit of fun (optional) and anything you want to share with us (legally of course).

 
Cheers, Light-Man

 
 

 

 

[JaZZ]

I didn't know we go back together that long

But remember I said that there's something fundamentally wrong with this analysis. theaudiohobby even somehow managed to quote a relevant snippet pointing out your mistake, even though he couldn't tell his hertz from his bits...

 
I didn't remember that, just realized it by checking back. What is that «something»?
 
Why not post it in the old thread to not disturb this FiiO X5 II thread any further!

(Since Head-Fi people tend to get impatient quite quickly.)

 

[BRCMRGN]

I have a dumb question. Does the X5ii support 200gb Micro SDs? Because I'm running out of space on my 128gb ones.

Works fine. I have 2 in mine. Format the cards in the X5ii to be safe.

 

[ASpencer]

  Works fine. I have 2 in mine. Format the cards in the X5ii to be safe.

 
Sound advice.  The new firmware actually improved my performance with those, namely updating the media library.

 

[Centrism]

Works fine. I have 2 in mine. Format the cards in the X5ii to be safe.

Sweet! Thank you. Now time to save my money.

 

[El Zilcho]

I haven't had a chance to try again to get the DAC driver working on my Windows 10 computer yet.  I have read a couple things about unsigned drivers on Windows 10, and some seem to say that even if you disable the verification process for install, it will turn back on and the drivers will fail to load after you reboot.  There are mixed responses about possible fixes.
 
The USB DAC function isn't the primary reason for getting the X5ii, so it won't totally ruin the experience for me if I can't use it, but I'd like to use it if possible.  I'll mess with it some more when I get a chance, but has anyone had any luck themselves with the DAC on Windows 10?  Any chance of actually getting signed drivers down the road, in case I can't make it work without too much difficulty?  

 

[Klonatans]

Just received this gadget. So far very positive impressions. Will make direct SQ comparisons with my AK100 Mk2 soon.
 
A non-audiophile question. Does an armband for X5 II exist?

 

[shamar75]

I have been very impressed so far with the sound quality of my X5. It has such good engineering put into its design, chip choice, quality of amplifiers, programming/firmware, etc.

If I were to pick one thing as a complaint, it would have to be the control wheel. I would describe it as horrible. It would have been better to put a 4 way control button in its place.

I scroll my wheel clockwise, it skips, cursor stops moving at times, even scroll direction changes and scrolls up and down the list without changing the direction of scroll.

Is it so hard to make an accurate navigation device?

I am thankful though that FiiO placed the forwards, backwards, and other 2 buttons. I have resorted to using the 4 buttons for navigation. I wish there was a way to disable/remove the wheel.

 

[Dobrescu George]

  I haven't had a chance to try again to get the DAC driver working on my Windows 10 computer yet.  I have read a couple things about unsigned drivers on Windows 10, and some seem to say that even if you disable the verification process for install, it will turn back on and the drivers will fail to load after you reboot.  There are mixed responses about possible fixes.
 
The USB DAC function isn't the primary reason for getting the X5ii, so it won't totally ruin the experience for me if I can't use it, but I'd like to use it if possible.  I'll mess with it some more when I get a chance, but has anyone had any luck themselves with the DAC on Windows 10?  Any chance of actually getting signed drivers down the road, in case I can't make it work without too much difficulty?  

win10 and unsigned drivers, it works without a problem. 

 

[xsk3l3t0rx]

  Just received this gadget. So far very positive impressions. Will make direct SQ comparisons with my AK100 Mk2 soon.
 
A non-audiophile question. Does an armband for X5 II exist?

no. i've been asking about this forever, and im getting no replies as to if theres even one in the pipeline. no one from fiio cares to answer, and everyone here is just going to tell you to cut your silicon case and put a strap thru it. i dont know if the X1 case fits, probably not cuz the dimensions arent the same. i really wish there was one for this, i'd love to use it at the gym...