[
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2008, Audio Asylum, all rights reserved ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]
|
Home
/ FAQ
/ News Classifieds / Events |
Audio Asylum Thread Printer |
Get a view of an entire thread on one page |
92.238.196.31
In Reply to: RE: My USB DAC research and some questions... posted by audioengr on December 21, 2008 at 22:11:35
How do you mean better than hardware? If you're talking about software playback with upsampling straight to a DAC compared to playback with a hardware upsampler in series how do you separate out the SRC quality from the playback process?
I've *processed* files with different SRCs including loopback recording the output of a hardware upsampler and then playing the files back with the same software/hardware; secret rabbit was noticeably worse than r8brain which was slightly worse again than the hardware SRC. 44.1 -> 96 or 192 was not as good as 44.1 -> 88.2 or 176.4.
To me Foobar 8.3 with secret rabbit 44.1 -> 96 smooths the high frequencies but is very woolly sounding, I've never really liked it.
Disregarding my opinion and looking at http://src.infinitewave.ca/ secret rabbit 1.2 doesnt actually plot very well against r8brain or, for example, the Weiss hardware SFC2.
How do you mean better than hardware? If you're talking about software playback with upsampling straight to a DAC compared to playback with a hardware upsampler in series how do you separate out the SRC quality from the playback process?
I have the same DAC with hardware upsampler and then one without, so I can hear the hardware upsampler and then the software SRC.
I've *processed* files with different SRCs including loopback recording the output of a hardware upsampler and then playing the files back with the same software/hardware; secret rabbit was noticeably worse than r8brain which was slightly worse again than the hardware SRC. 44.1 -> 96 or 192 was not as good as 44.1 -> 88.2 or 176.4.
To me Foobar 8.3 with secret rabbit 44.1 -> 96 smooths the high frequencies but is very woolly sounding, I've never really liked it.
What operating system and means of bypassing kmixer are you using? This is critical. With XP unmapped, software SRC is excellent. If you dont bypass kmixer, it will sound worse than hardware upsamplers.
steve N.
vista and wasapi output sounds better than anything i've tried on xp. 'Andy' has posted a really excellent minimal waspi player. I get the best playback quality by dragging wav files from my hard drive to a ram disk and then also defragging the ram (lots of useful defrag tools for this, some allow you to right click and defrag files). Once you get to this level running the laptop just off battery gives a noticeable improvement too.
I've tested wasapi and ks to be bit perfect with loop back recordings. Has to be that rather broadly defined and all-encompassing 'jitter' thing that gives these differences in real-time playback quality. But certainly with the same playback software and hardware there are clear differences between different sample rate converters.
"But certainly with the same playback software and hardware there are clear differences between different sample rate converters."
Differences can be verified objectively as well as subjectively. Sometimes the difference between two SRCs on identical input is only 30 db down.
(In the case of upsamplers, I only looked at differences below 20 kHz.)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
How do you mean better than hardware? If you're talking about software playback with upsampling straight to a DAC compared to playback with a hardware upsampler in series how do you separate out the SRC quality from the playback process?
I've *processed* files with different SRCs including loopback recording the output of a hardware upsampler and then playing the files back with the same software/hardware; secret rabbit was noticeably worse than r8brain which was slightly worse again than the hardware SRC. 44.1 -> 96 or 192 was not as good as 44.1 -> 88.2 or 176.4.
To me Foobar 8.3 with secret rabbit 44.1 -> 96 smooths the high frequencies but is very woolly sounding, I've never really liked it.
Disregarding my opinion and looking at http://src.infinitewave.ca/ secret rabbit 1.2 doesnt actually plot very well against r8brain or, for example, the Weiss hardware SFC2.
I agree with the above post. redbook upsampled x2,x4,x8 DOES sound better than the fractions you get with 96k or 192k.
BTW, did you try the Saracon software upsampler?
It should match or besten the quality of the Weiss hardware upsampler.
havent tried the weiss software or hardware src, i was comparing to dCS which isn't tested on that site. Realtime playback introduces jitter related variables so as i said difficult to compare sound quality even if it was the exact same algorithm running in windows. If the algorithm is the same though and you record to a file then it should actually negate jitter effects so it's quite possible in those circumstances that the weiss' hardware and software src may sound the same. If the hardware had fantastic power supply quality and near-perfect PLLs and the like it may actually sound substantially better but then your comparing jitter performance and not src performance :)
Post a Followup:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2008, Audio Asylum, all rights reserved ]