USB Strikes Back! Watch out AOIP - USB/Ethernet Chain beats All (at least for me)

Dec 31, 2016 at 4:59 PM Post #151 of 573
  Hi rb2013 i got my F-1 today from shenzhenaudio.com pretty fast finally and the original black alu box looks really good
bigsmile_face.gif
, im just listening to "Será Una Noche from marecordings", little bit strange music
bigsmile_face.gif
but the sound is awesome, no background noise anymore and very natural and relaxed sounding,.... unbelievable for that price i will not change back to the 5x more expensive mutec ..
 
Thanks and happy new year !


Did you try running the F1 into the Mutec?
 
I ask as my F1 arrived today and so I have simply replaced my Audiobreeze with the F1 while it is burning in .....and so I have the spdif >> Mutec3+USB at the moment.
 
I'll try without in due course.
 
M
 
Jan 1, 2017 at 10:43 AM Post #152 of 573
I have slotted he F1 into my system in place of my Audio Breeze DU-U8 and it now has about 20 hours use.
 
Initially I thought it was somewhat bright, although with a lot more high frequency detail. The bass was well resolved but some lower register information was not available on some tracks.
 
Now I would say the brightness is settling down, but the extra detail remains. The lower registers are developing.
 
This all sounds as though I have reservations, actually I think there is no doubt that this is a step up on my Audio Breeze.
 
I will try this sans Mutec in due course .....but, as it breaks in I want to be able to compare it to my last setup using my 'review' playlist, which means I only want to change one variable at a time.
 
M
 
Jan 1, 2017 at 11:39 AM Post #153 of 573
While this post doesn't perfectly relate to the topic, it does in part.
 
I've been using a Windows based software player called Infinity Blade (another player by the same author is called Bughead Emperor . . . these are the author's names, not mine).
 
Putting aside that it's the best sounding player I've ever heard and I have JRiver, HQPlayer, WTFPlay, and have used others, it has a feature that does relate to this thread.
 
The author, Hiroyuki Yokota, has suggested using a Panasonic Micro SD SLC card (8GB or larger, ideally) with his player as the player will automatically copy files to the card once the files are selected and it will clear the files off the card once they've been cleared from the player.
 
The player is not nearly as convenient to use as players which have their own custom remote app but, for me and many others, the improvement in sound quality is absolutely worth the trouble.
 
The relevant point for this thread is that the feature I'm mentioning means that no one has to figure out how to use multiple high capacity flash drives to implement Rob's very innovative solution. Using Infinity Blade, an 8GB SLC Micro SD card (the SLC type is important from what I understand) will suffice for a library of virtually any size.
 
For anyone who is interested, the links to download the free software are located below. The first link is for Skylake processors, the second link for all other processors.
 
Joel
 
 
http://oryaaaaa.world.coocan.jp/_userdata/Affeine_2_86_V5_x64_AVX2.zip
 
https://drive.google.com/folderview?id=0B-HZf3I4Da03Nng0Zk9KTGNlT2c&usp=sharing
 
Jan 1, 2017 at 12:07 PM Post #154 of 573
  While this post doesn't perfectly relate to the topic, it does in part.
 
I've been using a Windows based software player called Infinity Blade (another player by the same author is called Bughead Emperor . . . these are the author's names, not mine).
 
Putting aside that it's the best sounding player I've ever heard and I have JRiver, HQPlayer, WTFPlay, and have used others, it has a feature that does relate to this thread.
 
The author, Hiroyuki Yokota, has suggested using a Panasonic Micro SD SLC card (8GB or larger, ideally) with his player as the player will automatically copy files to the card once the files are selected and it will clear the files off the card once they've been cleared from the player.
 
The player is not nearly as convenient to use as players which have their own custom remote app but, for me and many others, the improvement in sound quality is absolutely worth the trouble.
 
The relevant point for this thread is that the feature I'm mentioning means that no one has to figure out how to use multiple high capacity flash drives to implement Rob's very innovative solution. Using Infinity Blade, an 8GB SLC Micro SD card (the SLC type is important from what I understand) will suffice for a library of virtually any size.
 
For anyone who is interested, the links to download the free software are located below. The first link is for Skylake processors, the second link for all other processors.
 
Joel
 
 
http://oryaaaaa.world.coocan.jp/_userdata/Affeine_2_86_V5_x64_AVX2.zip
 
https://drive.google.com/folderview?id=0B-HZf3I4Da03Nng0Zk9KTGNlT2c&usp=sharing


Thanks Joel - interesting information.
 
I guess a large capacity HDD stays in the picture to feed the SD card 'buffer'
 
Note that some on that CA thread commented on the better SQ of a USB stick versus an SD card.  They really are very much different in construction and operation - although both use NAND memory circuits.
 
The newer high capacity USB sticks have advanced crystal oscillators built right on the board - additionally they have 3.3V regulators.  In fact they are micro-computers in their own right - hence can be set as bootable devices.  A few pages back I posted more details on the differences between SD cards and USB sticks.
 
Now the true advantage in my implementation - is completely GI for the USB storage and file conversion in a 'clean' environment - powered by a separate low noise LPS from the PC (that is a separate LPS not attached to the PC in anyway in fact my PC server and the Startech are on separate dedicated AC noise isolation and common and differential mode filters).
 

 

 
The new $16,000 Aurender W20 uses a SDD to cache from the internal HDDs. 
 
But with both the SD Card buffer and the SDD buffer - you have the noisy HDD spinners still in operation.  The internal HDD servo control read heads and motor to spin the HDD at 5600rpm all create modulation power supply noise in the file conversion environment.  I think best to just eliminate all motors, like fans, HDDs and the like from the PC or file conversion Render.
http://www.aurender.com/page/w20
 
As far as I'm aware my ultra USB chain is the first to use large capacity USB sticks in this fashion.
 
Jan 1, 2017 at 12:09 PM Post #155 of 573
 
Did you try running the F1 into the Mutec?
 
I ask as my F1 arrived today and so I have simply replaced my Audiobreeze with the F1 while it is burning in .....and so I have the spdif >> Mutec3+USB at the moment.
 
I'll try without in due course.
 
M

This will be very interesting - the F-1 with and without the Mutec as SPDIF reclocker.  I guess you preferred the Breeze DU-U8 over the Mutec as a USB DDC?
 
  I have slotted he F1 into my system in place of my Audio Breeze DU-U8 and it now has about 20 hours use.
 
Initially I thought it was somewhat bright, although with a lot more high frequency detail. The bass was well resolved but some lower register information was not available on some tracks.
 
Now I would say the brightness is settling down, but the extra detail remains. The lower registers are developing.
 
This all sounds as though I have reservations, actually I think there is no doubt that this is a step up on my Audio Breeze.
 
I will try this sans Mutec in due course .....but, as it breaks in I want to be able to compare it to my last setup using my 'review' playlist, which means I only want to change one variable at a time.
 
M

Yes the F-1 needs 200 hours to fully settle in.  How are you powering it?
 
Jan 1, 2017 at 12:12 PM Post #156 of 573
Great reply, Rob.
 
I wish I could argue with your conclusion . . . it would save me a lot of trouble if I could. But I can't.
 
Now I'm going to have to have to buy one of these flash drives. Can you only recommend the flash drive you're using or can you give a recommended set of parameters for a flash drive such that if any drive meets those parameters, it's a good purchase?
 
Thanks for the information. I understand the benefits of your solution even better now.
 
Joel
 
Jan 1, 2017 at 12:13 PM Post #157 of 573
  • 11-26-2012, 12:28 AM
    #80

    The Computer Audiophile
    user-online.png

    Site Founder
    Join Date​
    Nov 2007
    Location​
    Minneapolis, MN, USA
    Posts​
    14,885


    Hi Yuri - I have no doubt the bits are identical whether pulled from an HDD or SSD. That's a nonissue for me. Where I believe differences may be found is in the electrical noise generated by either HDD or SSD. Most USB receiving chips in DACs are powered from the USB bus. This direct electrical link from DAC to computer may impact performance although not change the bits.

    A green HDD may use 6 watts versus an SSD using 0.05 watts. Could this be one piece of the puzzle? I'm not sure.

    I've been experimenting with computer PSUs and should have some objective measurements that show audible differences soon.





    0​
    0​
    0​
     



  • 11-26-2012, 12:46 AM
    #81

    Yuri
    user-offline.png

    Newbie
    Join Date​
    Aug 2012
    Posts​
    13


    Well, I was thinking about DACs with their own power supply (as the one I'm using). DACs powered by USB may be a different point.

    If USB voltage is not stable then it *may* affect digital-to-analog performance - assume that ground level (GND) is drifting then output may not be as precise as it should.

    This might be possible when:
    1) Computer's PSU is not well designed. Good one should handle 0.05 watts vs 6 watts without any spikes. And please note SSD can also generate power spikes during some operations (erase is most power consuming, and there are background tasks to relocate data on flash)
    2) There is no power stabilization in USB-powered DAC

    So if there is a *real* difference in sound quality then there should be some sort of explanation why. But with my SSD knowledge I can't imagine difference yet.

    The only thing comes to my mind is performance. HDD may be late to provide required data causing audio dropouts or freeze. This is why advised to have big memory buffers for track preloading.





    0​
    0​
    0​
     



  • 11-26-2012, 01:32 AM
    #82

    sandyk
    user-offline.png

    Masters Level Member
    Join Date​
    Aug 2009
    Location​
    Sydney Australia
    Posts​
    8,339


    You're trying to explain digital audio system with analog means where indeed power supply quality or line noise may affect the signal




    Many people will tell you that a USB memory stick, which has an inbuilt 3.3V voltage regulator , sounds a little better than either SSD or HDD. It's very easy to demonstrate using a high resolution sound system, that the same USB memory stick when supplied by a very high quality +5V Linear PSU, with vBus +5V disconnected at the PC end of a USB cable sounds noticeably better than when using USB power, even when both .wav files are played from System Memory using cPlay.
    I had no problems demonstrating that, and an even more highly contentious issue, to a Sydney based qualifiied E.E. at a listening session several weekends ago. The E.E has wide industry experience, and is also a DIYAudio member.("Owdeo")
    Alex



    "If you can't hear the difference between an original CD and a copy of your CD,
    you might as well give up your career as a tester. The difference between a reconstituted FLAC and full size WAV is much less than that, but it does exist. - Cookie Marenco"


    0​
    0​
    0​
     



  • 11-26-2012, 01:57 AM



 
 

 
Jan 1, 2017 at 12:30 PM Post #158 of 573
  Great reply, Rob.
 
I wish I could argue with your conclusion . . . it would save me a lot of trouble if I could. But I can't.
 
Now I'm going to have to have to buy one of these flash drives. Can you only recommend the flash drive you're using or can you give a recommended set of parameters for a flash drive such that if any drive meets those parameters, it's a good purchase?
 
Thanks for the information. I understand the benefits of your solution even better now.
 
Joel


Hi Joel,
 
I have one of these Sandisk SD card reader to USB boxes around.  I will try and load some music files on a high capacity SD card and mount in one of the REX slots to compare to a USB stick.  I should note that high capacity (128GB or higher) SD cards are far more prone to be corrupted then an equivalent USB stick.  The PC CPU (or other ASIC) must perform some of the memory allocation functions away from the SD card, like wear management.  This can lead to the exFAT tables to become incompatible or corrupted.  And never do a 'disk format' in Windows - this will likely lead to a corrupted SD card.
 
These functions on a USB stick are done by it's dedicated microprocessor - so much less vulnerable to corruption.
 
Beware there are many fake USB flashdrives (low capacity USB drives hacked to appear larger) - so stay away from Ebay and Amazon (just read the long horrendous feedback chains) - such a shame there is not better policing there on these shady sellers (like banning them!).  Some of these counterfeits come in very realistic name brand packaging.
 
So I went to Best Buyer and bought a name brand - PNY 256GB USB stick.  They are cheap - like $49.  At the time Best buy had a buy two or more computer accessories and get another 20% off.  So I bought three - all for $120.
 
Now the real key to my chain is the USB/Ethernet Startech to run both the USB music data chain to the F-1 and to run the USB 256GB sticks for storage.  So it serves a dual roll - and leverages the advantages of full galvanic isolation from the PC, SPARTAN 6 FPGA USB/TCPIP packet translation and the MEIYAN LPS feeding the Startech REX box.
 
So just running a USB stick in an open PC USB port is not likely to provide the same SQ effects.  One could get a separate USB hub and power it with a small separate LPS - but you would still not have the galvanic isolation from the PC USB noisy ground plane.
 
Cheers!
 
Jan 1, 2017 at 12:53 PM Post #159 of 573
 
Hi Joel,
 
I have one of these Sandisk SD card reader to USB boxes around.  I will try and load some music files on a high capacity SD card and mount in one of the REX slots to compare to a USB stick.  I should note that high capacity (128GB or higher) SD cards are far more prone to be corrupted then an equivalent USB stick.  The PC CPU (or other ASIC) must perform some of the memory allocation functions away from the SD card, like wear management.  This can lead to the exFAT tables to become incompatible or corrupted.  And never do a 'disk format' in Windows - this will likely lead to a corrupted SD card.
 
These functions on a USB stick are done by it's dedicated microprocessor - so much less vulnerable to corruption.
 
Beware there are many fake USB flashdrives (low capacity USB drives hacked to appear larger) - so stay away from Ebay and Amazon (just read the long horrendous feedback chains) - such a shame there is not better policing there on these shady sellers (like banning them!).  Some of these counterfeits come in very realistic name brand packaging.
 
So I went to Best Buyer and bought a name brand - PNY 256GB USB stick.  They are cheap - like $49.  At the time Best buy had a buy two or more computer accessories and get another 20% off.  So I bought three - all for $120.
 
Now the real key to my chain is the USB/Ethernet Startech to run both the USB music data chain to the F-1 and to run the USB 256GB sticks for storage.  So it serves a dual roll - and leverages the advantages of full galvanic isolation from the PC, SPARTAN 6 FPGA USB/TCPIP packet translation and the MEIYAN LPS feeding the Startech REX box.
 
So just running a USB stick in an open PC USB port is not likely to provide the same SQ effects.  One could get a separate USB hub and power it with a small separate LPS - but you would still not have the galvanic isolation from the PC USB noisy ground plane.
 
Cheers!

 
Thanks for the quick replies, Rob.
 
I'm not a technical guy, but I don't know whether there was a reason the author of the software I'm using specifically recommended Micro SD SLC cards or not. I'm not sure if a standard SD card will offer the same benefit.
 
That point aside, I have both the Rednet D16 and the Adnaco optical boards, which I believe could make an interesting combination for employing a slightly different flavor of the solution your suggesting. When I've tested that out, I'll hope to let everyone know of my results.
 
One more question: Do you believe there's any preference in using a USB 2.0 vs. USB 3.0 flash drive?
 
Thanks again.
 
Joel
 
Jan 1, 2017 at 1:10 PM Post #160 of 573
   
Thanks for the quick replies, Rob.
 
I'm not a technical guy, but I don't know whether there was a reason the author of the software I'm using specifically recommended Micro SD SLC cards or not. I'm not sure if a standard SD card will offer the same benefit.
 
That point aside, I have both the Rednet D16 and the Adnaco optical boards, which I believe could make an interesting combination for employing a slightly different flavor of the solution your suggesting. When I've tested that out, I'll hope to let everyone know of my results.
 
One more question: Do you believe there's any preference in using a USB 2.0 vs. USB 3.0 flash drive?
 
Thanks again.
 
Joel


Well it's likely due to the higher speed.
 
As most are aware SD Cards come in different classes - I posted info on this bakc on page 6 - post #79
http://www.head-fi.org/t/829639/usb-strikes-back-watch-out-aoip-usb-ethernet-chain-beats-all-at-least-for-me/75#post_13112032
 
In order to get the equivalent data throughput speed to USB 2.0 High Speed (35MB/s) - you need class 10 to just get to 10MB/s - UHS class 3 to get to 30MB/s.
 
On the Rednet D16 - now it would be great if it had a USB port.  Anyway the Startech/ICRON only support USB 2.0, so I had no need for a USB 3.0 stick. 
 
I'm able to run 352K and 384K PCM native files and DSD 256 without a hiccup from the USB sticks in the REX.  So the massive data throughput of USB 3.0 does not look necessary.
 
Jan 1, 2017 at 1:20 PM Post #161 of 573
Another reason for a SLC or single level cell Micro SD card is as a buffer it has to constantly filled and erased over and over.  The SLC SD cards have a higher cycle life.
 
The write speeds are dramtically lower then read speeds for all solid state media.  So that presents another major challenge to these solid state buffering schemes - the complexity (and higher noise?) needed for a constant write then read operations on the fly.
 
In case the music written once and then read from that storage.  In fact it is highly recommended to write all music files sequentially vs multiple loads simultaneously.  This 'fills' the NAND circuits in a more uniform fashion.  And eliminates the need for disk defraging - a very wear intensive operation on a solid state device.
 
http://forums.sandisk.com/t5/Mobile-memory-professional-cards/Which-are-MLC-TLC-MicroSD-cards/td-p/332885
 
Jan 1, 2017 at 1:33 PM Post #162 of 573
Someone posted this cable over on my PC Music server thread: http://www.head-fi.org/t/754654/new-pc-music-server-build-project-all-ssd-no-fans
 
May give that a try - for the SSD running the OS - just move it's power to the small TeraDak LPS
 
https://www.amazon.com/dp/B01537TMVM/ref=cm_sw_r_cp_ep_dp_TOcAybF9MJRAC
 
Jan 1, 2017 at 3:50 PM Post #163 of 573
  This will be very interesting - the F-1 with and without the Mutec as SPDIF reclocker.  I guess you preferred the Breeze DU-U8 over the Mutec as a USB DDC?
 
Yes the F-1 needs 200 hours to fully settle in.  How are you powering it?


Yes, I preferred the DU-U8 to the Mutec direct, had more dynamics.
 
The F1 is powered via the microRendu, which has Regen tech built in.
 
M
 
Jan 1, 2017 at 4:31 PM Post #165 of 573
I'll keep posting this until it get through people's heads. All asynchronous USB chains regardless of the differences in sound are all bit perfect. The music data being delivered to the DAC is the same in all cases. Why the audible difference? It's not cable, it's not USB jitter, it's not software, operating system or server hardware performance.

For those that have a rationale mind to ponder why the many billion dollar IT industry dependent on USB working 100% doesn't give a crap about USB issues whereas we dwell incessantly on it?

Because when it comes to digital-to-digital, USB works fine.

What we hear with our ears is the small signal noise on the +5V pin on the USB cable...Which gets into the circuits that produce the analog output from the DAC.

I know this post will yet again fall on deaf ears because an entire industry is propped up on promulgating obfuscation and disinformation. Please wake up!!! Threads like this are such a waste of time.

You can have the best USB sound your source and DAC can give by inserting at LDO regulated battery in the USB path. I use a $135cdn Musical Paradise MP-U1 and you can throw away all other devices and higher end cables.

Dan
 

Users who are viewing this thread

Back
Top