 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
196.2.124.251
Hi all. Two quick questions about up/resampling.
#1. if the original files are 44.1kHz, and the hardware you use (DAC, USB interface etc.) only has 48kHz or 96kHz upsampling, is it better to upsample to one of those or keep the original 44.1kHz. The reason I ask is that I read a number of times that it's best to upsample in whole multiples, e.g. 88.2kHz etc. (still output 24-bit, so changing bit-depth but not sample-rate)
#2. if keeping 44.1kHz, in a program such as foobar, would it be better to leave out any sampling plugins in the DSP loop or is there still an advantage to using something like PPHS to resample 44.1kHz ultra-linear mode; or the secret-rabbit resampler at 44.1kHz etc.Cheers,
X
Follow Ups:
Xenithon:Good question you pose.
I would definately go to 96khz if you're going to go through the trouble of upsampling.
But I also would "invest" in a $30 Chaintech AV710 and see what true 44.1 sounds like with *no* additional resampling. The AV710 does not "force" you to 48 or 96khz but passes through whatever sample rate the source is (as I understand it).
I myself am leaning towards the "keep the signal chain bitperfect" camp these days from my experiments with different hardware and DSP based PC solutions.
Cheers,
Presto
Hi and thanks. I already use an Edirol UA-1EX which gets data via USB (ASIO) and can either send out in 44, 48 or 96kHz.The onboard soundcard is disabled altogether
nt
I have tried SRC at 88.2kHz and 96kHz starting with native 44.1 files. I like 96kHz the best.
Steve:Do you have a link to SRC? I can't find the straight SRC.
Thanks!
> Do you have a link to SRC? I can't find the straight SRC.There are (were) two SRC's floating around. The first one
(the one that became "famous") was for Foobar 8. This
was apparently taken (from source code freely available
under the GPL license) and made into a dll by a helpful
programmer and published on a number of Web sites.
But then the author of SRC, Erik de Castro Lopo,
(who was apparently -- this is my guess -- motivated by the fact
that some audio manufacturers had become interested in SRC for
commercial purposes) claimed that publishing the Foobar dll binary
plug-in was a violation of SRC's GPL license, and demanded that it
be removed from the Web. All the Western sites complied right
away, though SRC for Foobar 8 was available long after it
"shouldn't" have been on some Asian Web sites (from which somebody
like me has to download from without being able to read
in detail). That version is going to be very difficult to
locate, unless you can persuade somebody to e-mail it
to you privately (in violation of applicable laws, etc.,
etc.)However, now that the author of SRC has made the terms of
commercial use quite clear, he seems to have relented on
the private use of SRC with Foobar. So, from "Mega-Nerd"s
own Web site http://www.mega-nerd.com/SRC/fb2k.html you can now
download an SRC dll once again. However, this new SRC seems
to be for Foobar 9 only (I haven't tried it under 8, but I
wouldn't expect it to work). So to use it, you'd have to
upgrade to Foobar 9, if you haven't already.
Here is my experience regarding your question #1.I use 44.1/16 output from the PC via USB/spdif adapter into Behringer Ultramatch SRC 2496. Behringer is capable of upsampling to 88.2 and 96Khz in 16, 20 or 24bits.
44.1/16 or upsampled signal goes to Panasonic SA-XR 45 and Pana is capable to accept anything from 441.1/16 to 96/24.
To my ears 44.1/16 and upsampled (all combinations up to 96/24) sound different. I can tell them apart, but often I can not tell which one I like better. Also, I have found it is dependent on kind of music I am listening to and the quality of the original recording. Enough to drive you nuts and ruin enjoyable listening session and turn it into test lab chore.
It is possible that either Behringer or Panasonic work differently on different sampling rates, or both of them. Perhaps there is some other artifact, but for me, in my setup, 44.1 upsampled to 88.2 does sound different from 96. The difference is minuscule.
Before I embarked on PC journey, I used transport + dac combo and had DSP upsampler in the digital chain. At that time, I was also able to hear differences between 44.1 and 88.2 or 96KHz and it was also dependent on music material.
I have never had an opportunity to hear heavyweights of dCs class to determine how they sound on various upsampled frequnecies.
> #1. if the original files are 44.1kHz, and the hardware
> you use (DAC, USB interface etc.) only has 48kHz or 96kHz
> upsampling, is it better to upsample to one of those or
> keep the original 44.1kHz.There's no sense in going from 44.1 to 48 just for sound
(unless, of course, **you** think it makes an improvement).44.1 to 96 is considered (by those who "believe" in upsampling
at all) to be a "legitimate" path.> The reason I ask is that I read a number of times that it's
> best to upsample in whole multiples, e.g. 88.2kHz etc.There's a theoretical basis for this, but whether it translates
to a difference in sound is up to you. I don't take it very
seriously. Back in '99, when upsampling became a "hot"
issue thanks to dCS (using their 974 digital-to-digital converter),
and Bascom King and Jonathan Scull (in _Stereophile_) and
others raved about it, the upsampling conversions initially
reported on were 44.1 -> 96 or 44.1 -> 192 (non-integral
multiples).> (still output 24-bit, so changing bit-depth but not sample-rate)
? 44.1 -> 88.2 is certainly a change in sample rate.
> if keeping 44.1kHz, in a program such as foobar, would it
> be better to leave out any sampling plugins in the DSP loopYes.
> or is there still an advantage to using something like PPHS
> to resample 44.1kHz ultra-linear mode; or the secret-rabbit
> resampler at 44.1kHz etc.I can't imagine why there would be an advantage, again unless
it happens to sound better to you.If you're a DAC manufacturer (like, say, Bel Canto with their
DAC 1 and DAC 2, or PS Audio with their Link III) there's
a possible reason to use an ASRC (asynchronous sample-rate
converter) chip (a Crystal CS8420 in the case of the old
Bel Cantos; a TI/Burr-Brown SRC4192 in the case of the
new Link III) -- namely, to reclock the data going into
the DAC chip and eliminate jitter. In those cases, the
upsampling (to 96 in the case of the DAC 1; to 192 in the
case of the DAC 2 or Link III) isn't defeatable, so
if you happen to be using a 96KHz source (in the case
of the DAC 1) or a 192 kHz source (in the case of the
Link III) you'll be going through the ASRC with a 1:1 clock
ratio. But this is for jitter rejection, which you won't
get by using an upsampling DSP in Foobar.
I will be experimenting between the current setting (48kHz) and 44.1kHz with no sampling plugin. I tried 96kHz (various plugins such as PPHS and SRC) but did not like any. Sounded more metallic, more forced, unnatural. I listen to a lot of female vocal so it was very apparent.
> I tried 96kHz. . . Sounded more metallic, more forced, unnatural.That's odd, because it seems that most folks who don't like
upsampling (apart from the theoretical argument that it
couldn't possibly make any difference ;-> ) complain about
precisely the opposite effect -- that it softens the sound
too much. Sort of like the old objection that some folks
had against bitstream (or sigma-delta) DACs (as opposed to
multibit). That, and that upsampling adds an unnatural
ambience (sort of like fake reverb). I like both "effects"
myself, so I'm a fan of upsampling.
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2024, Audio Asylum, all rights reserved ]
