or Connect
Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › iFi Audio Nano iDSD discussion + impression
New Posts  All Forums:Forum Nav:

iFi Audio Nano iDSD discussion + impression - Page 39

post #571 of 1720
Quote:
Originally Posted by mink70 View Post
 

 

The reason is that many listeners find that it sounds better. The best sound, by far, that I've experienced from my MacBook is listening to native DSD, followed by PCM upsampled to DSD on the fly via JRiver 19. I don't have the audio design experience of either Gordon Rankin or Charlie Hansen (of Ayre), who seem to have passionately held opinions against DSD, but I don't see how any technical discussion could possibly trump listening. And yet these forums are full of people who just know something must sound bad because it contradicts their belief system. USB cables all sound the same, because they must, and people prefer tube amps because they like the sound of "colorations," and all digital data streams sound identical, because, after all, they're just ones and zeros. The truth is that we still know fairly little about how digital domain changes affect listening. How else to explain talented audio engineers disagreeing with each other about nearly everything pertaining to upsampling, software, DSD, and basic approaches to designing a DAC?


Hey Mink70...  can you tell me where the option is set in Jriver 19 to upsample PCM to DSD

 

thanks

post #572 of 1720
Quote:
Originally Posted by Dixter View Post
 


Hey Mink70...  can you tell me where the option is set in Jriver 19 to upsample PCM to DSD

 

thanks

Sure Dixter, though keep in mind that I'm using the OSX version. Go to Player-->Playback Options, click on DSP & Output Format, and then in the Output Encoding dropdown, select 2xDSD in DoP Format. I find that my music does sound better upsampled to 2xDSD, (probably because of superior noise behavior). Also, to play native DSD, go to Player-->Playback Options, click on Bitstreaming, and check DSD. It took me a while to figure this out, as JRiver does not have the most intuitive interface.


Edited by mink70 - 6/5/14 at 7:19am
post #573 of 1720

OK thanks...   I'll try that and see how it sounds... 

post #574 of 1720
Quote:
Originally Posted by Dixter View Post
 

OK thanks...   I'll try that and see how it sounds... 

Let us know. I'd be very curious to know what you think.

post #575 of 1720
Quote:
Originally Posted by kugino View Post

I'm on Gordon's side here. Just because something is DSD doesn't make it automatically sound better. Same like something in FLAC or mp3. If someone's creating a FLAC file from an mp3, I would say that's garbage. What matters is the mastering and the quality of the original source.

In the micro iDSD thread Thorstens from iFi/ARM gives a nice run-down on how hard it is to convert to/from PCM and DSD. Keep it in whatever the original format was...I don't get why you'd want to convert from PCM to DSD.


I can totally understand the logic here, because PCM does certain things well, and DSD does other things well, so when you convert the PCM to DSD you loose the stuff that PCM did well, and you dont get the advantage of what the DSD does well when its native. At least thats my non technical summary of what I have read. Totally makes sense. Thats why I have not played in DSD mode with my iDSD previously, because I dont have any native DSD, and upsampling to DSD made no sense.

 

Again, well crap, it does sound better. I will admit that the DSD over DoP on the old firmware and delivered upsampled with JRiver 19 the 'benefits' were a bit of a stretch. So, I went back to PCM and lived my life. Then I discovered ASIO worked better, then I thought to recheck bit-streaming up-sampled redbook to 2xDSD over ASIO. Sorry, its good. Oh and another thing... It was good enough for me to notice that my files on SD card sound better than those off my internal hard drive! ;) - peace.

post #576 of 1720
Quote:
Originally Posted by earwaxxer View Post
 

I will admit that the DSD over DoP on the old firmware and delivered upsampled with JRiver 19 the 'benefits' were a bit of a stretch.

 

Quick question: this new, firmware-enabled "direct DSD over ASIO" mode is only possible on WIndows, right? There's no similar option for OSX users? Or am I not understanding this right?

post #577 of 1720
Quote:
Originally Posted by mink70 View Post
 

 

Quick question: this new, firmware-enabled "direct DSD over ASIO" mode is only possible on WIndows, right? There's no similar option for OSX users? Or am I not understanding this right?

That I dont know, maybe the iFi folks can chime in on that. I know that WASAPI is windows, so I'm not sure if OSX uses ASIO by default etc.

post #578 of 1720

Did a quick experiment this afternoon using the iDSD Nano with JRiver 19 on a MacBook Pro running OS 10.6.8. Confirmed results on both the speaker setup (Shindo electronics and big vintage Tannoys) as well as headphones (AT W3000ANV--love these!). I played a redbook AIFF of the Miles Davis sextet playing "Circle," from "Miles Smiles," a track I love and know well, at the native sampling rate and upsampled to 24/96, 24/196, and 2xDSD over DoP. (I skipped 1xDSD because I found 2xDSD superior in every way). I also played the track in native DSD from an SACD rip (admittedly a different mix). Keep in mind that this one piece of software with one DAC over one OS, so the results are hardly global.

 

Here's what I found: native DSD is easily the best in every respect. For me, a revelation in terms of smoothness, color, clarity, delicacy and involvement. More like real music, a lot more. The best sound from digital I've heard. Among the upsampled formats, a mixed bag. (I tried to adjust for volume differences, but wasn't scientific about it). Native redbook was the least engaging. The music was rough sounding, like a kind of sonic sandpaper was applied. Maybe coarse is a better word. Grayish tonal colors. 24/96 sounded better: far more dynamic, with richer bass, more sound density, and more drive. Bigger images. A lot of fun. Tonal colors were resolutely gray, though, with a kind of grainy overlay. Miles's trumpet and Herbie Hancock's piano didn't sound all that much like a trumpet and a piano, which kind of ruined the punchy, rich sound. 24/192 cleaned up the timbres and made the music more colorful, but the expense of dynamics and drive and bass. Weird. Meh. 2xDSD: Here, the timbres of the instruments sounded MUCH better. The piano was now made of wood and string and the trumpet of brass. Best precision, clarity, and transparency of the bunch. On the flipside, the bass slam and scale of the music decreased (though I can't completely swear this wasn't partly due to volume differences). Not nearly as good as native DSD, but really quite nice and, most importantly, very engaging.

 

Take the above for what it's worth, which may not be much. But I wanted to share.

post #579 of 1720
Quote:
Originally Posted by analogsurviver View Post
 

I am struggling to get DSD128 playback with nano iDSD and HQPlayer - it keeps maxing out at sampling rate of 3072000, which is obviously too low to play the required 5644800 - despite setting 12288000 for the bit rate limit, etc. Does anyone know the system requirements for HQPlayer for DSD128 ?

 

It works OK with DSD64 .

The only thing that was published with HQPlayer was CPU support for SSE3 instruction set. You did not gave info on what is your setup is. On the window version of J.River you can download the trial version and run the benchmark under tools, A score of 3000 is better in converting PCM to DSD128 in real-time is what J.River recommends.

post #580 of 1720
Quote:
Originally Posted by john57 View Post
 

The only thing that was published with HQPlayer was CPU support for SSE3 instruction set. You did not gave info on what is your setup is. On the window version of J.River you can download the trial version and run the benchmark under tools, A score of 3000 is better in converting PCM to DSD128 in real-time is what J.River recommends.

Sorry for omitting the setup. It is Esprimo Q 9000, Intel i5 CPU  M520 @2.40 GHz    4GB RAM    64bit OS .

 

It is extremely nice as it is small ( think 4 DVD cases - a bit shorter and a bit wider, but that really IS it ) & QUIET, downsize is it can not be expanded/upgraded - no way (I am aware of ) to fit more RAM etc.

 

The latest version of Korg Audiogate, the V3.0.2, recommends CPU equal to or better than 2.66 GHz. .  As far as I can tell, it works as it should. - but I feel I might be on the brink of stretching it all. I also have trial versions of JRiver 19 and HQPlayer installed - and very curious thing happened yesterday. I had the same file opened in both JRiver and Audiogate - I naturally was interested in SQ difference(s) , if any . I spent last 24 hours or so digging out as much as I could on DSD players - also stumbled upon http://hi-fi-avenue.blogspot.com/2012/03/audiogate-gateway-to-good-sound.html . As a seasoned user of Audiogate, it naturally raised an eyebrow.

After finishing playing the same file in JRiver, Audiogate started to play - at double speed :eek:. Obviously, either the load on CPU was too much - or the two softs are similar and share more than something in common. Had to close down the JRiver and restart the Audiogate in order to listen to the file with Audiogate. Believe me, at that point the difference in SQ was the least of my concerns ...- but will update once the dust settles a bit.

 

I also cofigured foobar2000 for DSD ( over DoP ) and DXD playback. When playing DSD128, it generally outputs gibberish for a fraction of second(s) - and then plays fine. One has to be careful with volume at the start of the playback.  No problem with DXD or DSD64.

 

I could not get the JRiver to play DXD - it keeps going to 176/192 ? It must be me doing something wrong.

 

Thank you for bringing my attention to JRiver Benchmark. I just run the test and my score is 3068 - meaning it can run JRiver and nothing else for good performance.

 

I did also  look longingly at the XXHighEnd player - but after seeing the requirements had to back off.

 

 

Here the full JRiver Benchmark results for my setup:

 

=== Running Benchmarks (please do not interrupt) ===
 
Running 'Math' benchmark...
    Single-threaded integer math... 4,941 seconds
    Single-threaded floating point math... 2,790 seconds
    Multi-threaded integer math... 2,622 seconds
    Multi-threaded mixed math... 1,818 seconds
Score: 1561
 
Running 'Image' benchmark...
    Image creation / destruction... 0,226 seconds
    Flood filling... 1,240 seconds
    Direct copying... 1,232 seconds
    Small renders... 2,056 seconds
    Bilinear rendering... 1,966 seconds
    Bicubic rendering... 1,588 seconds
Score: 2648
 
Running 'Database' benchmark...
    Create database... 0,379 seconds
    Populate database... 1,882 seconds
    Save database... 0,463 seconds
    Reload database... 0,072 seconds
    Search database... 1,782 seconds
    Sort database... 1,498 seconds
    Group database... 0,931 seconds
Score: 3068
 
JRMark (version 19.0.128): 2426

Edited by analogsurviver - 6/6/14 at 1:18am
post #581 of 1720

Just got my iFi DSD going today. 

 

A suggestion for Nano owners: over on the NativeDSD.Com web site they have a new feature named "Just Listen" with free music downloads at DSD64, DSD128 and DSD256.

 

Most of the 8 music selections are recorded natively at DSD128 (2x SACD resolution) - but two of the 8 tracks are Classical Guitar recorded natively at DSD256 (4x SACD resolution)!  Very cool - and just the thing to test out the Nano.

Visit https://justlisten.nativedsd.com/albums/just-listen-1-compilation to download the files.

post #582 of 1720
Quote:
Originally Posted by bmoura View Post
 

Just got my iFi DSD going today. 

 

A suggestion for Nano owners: over on the NativeDSD.Com web site they have a new feature named "Just Listen" with free music downloads at DSD64, DSD128 and DSD256.

 

Most of the 8 music selections are recorded natively at DSD128 (2x SACD resolution) - but two of the 8 tracks are Classical Guitar recorded natively at DSD256 (4x SACD resolution)!  Very cool - and just the thing to test out the Nano.

Visit https://justlisten.nativedsd.com/albums/just-listen-1-compilation to download the files.

Thank you for sharing the link. I have been struggling to download DSD256 files from the link within this link 

http://www.computeraudiophile.com/f13-audiophile-downloads/originally-recorded-dsd256-piano-solo-sources-japanese-group-wechseldominate-19942/

for past few days - without sucess. The download starts, is extremely slow, and invariably quits at around 10% of the file. Never experienced such behaviour before from other sites - hope somebody else will have better luck with these.

post #583 of 1720
Quote:
Originally Posted by analogsurviver View Post
 

Thank you for sharing the link. I have been struggling to download DSD256 files from the link within this link 

http://www.computeraudiophile.com/f13-audiophile-downloads/originally-recorded-dsd256-piano-solo-sources-japanese-group-wechseldominate-19942/

for past few days - without sucess. The download starts, is extremely slow, and invariably quits at around 10% of the file. Never experienced such behaviour before from other sites - hope somebody else will have better luck with these.


Don't feel bad.  I've had problems trying to download the DSD256 files at Wechseldominate in Japan myself. 

 

Hopefully you'll have better luck with the new DSD256 files on NativeDSD.Com and their Just Listen section!

https://justlisten.nativedsd.com/albums/just-listen-1-compilation


Edited by bmoura - 6/6/14 at 2:08am
post #584 of 1720
Quote:
Originally Posted by bmoura View Post
 


Don't feel bad.  I've had problems trying to download the DSD256 files at Wechseldominate in Japan myself. 

 

Hopefully you'll have better luck with the new DSD256 files on NativeDSD.Com and their Just Listen section!

https://justlisten.nativedsd.com/albums/just-listen-1-compilation

Normally, I do not give up easily - but after failures to download over a weekend's time it was a bit of dissapointment. I am eager to hear how much better DSD256

sounds compared to DSD128. As DSD files, specially DSD256 files are large ( 1 GB = 5.5 minutes of audio ), I will wait with the downloads till the time I have unlimited data transfer from my IP - from midnight to 6 AM . Hopefully, my setup will be able to play them .

post #585 of 1720
Quote:
Originally Posted by analogsurviver View Post
 

Sorry for omitting the setup. It is Esprimo Q 9000, Intel i5 CPU  M520 @2.40 GHz    4GB RAM    64bit OS .

 

It is extremely nice as it is small ( think 4 DVD cases - a bit shorter and a bit wider, but that really IS it ) & QUIET, downsize is it can not be expanded/upgraded - no way (I am aware of ) to fit more RAM etc.

 

I also cofigured foobar2000 for DSD ( over DoP ) and DXD playback. When playing DSD128, it generally outputs gibberish for a fraction of second(s) - and then plays fine. One has to be careful with volume at the start of the playback.  No problem with DXD or DSD64.

 

I could not get the JRiver to play DXD - it keeps going to 176/192 ? It must be me doing something wrong.

 

Thank you for bringing my attention to JRiver Benchmark. I just run the test and my score is 3068 - meaning it can run JRiver and nothing else for good performance.

 

 

 

Okay, your CPU system speed is fine and that is not the bottleneck to play DXD or DSD. You should be able to convert PCM to DSD128 in real time just fine with J.River. What version of USB are you using USB 2 or USB 3?  Is your BIOS setting for USB set for Hi-speed? Is the iDSD on its own USB port? Normally you run just one audio playback software because of exclusive mode control setting of windows to the iDSD.  Make sure at this point you are not using DSP Studio in J.River just for the time being. J.River can show what is being send to the iDSD as well.

 

 

 

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Dedicated Source Components
Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › iFi Audio Nano iDSD discussion + impression