DSPRelated.com
Forums

Sample rate converter comparison

Started by Erik de Castro Lopo January 27, 2007
Hi all,

I thought people might like to know about the sample rate converter
comparison which has been done here:

   http://src.infinitewav.ca/

My converter (Secret Rabbit Code) is not the best, but is definitely
not too bad.

Cheers,
Erik
-- 
+-----------------------------------------------------------+
  Erik de Castro Lopo
+-----------------------------------------------------------+
"Incompetence, like misery, seeks company." -- Erik Naggum
Erik de Castro Lopo wrote:
> Hi all, > > I thought people might like to know about the sample rate converter > comparison which has been done here: > > http://src.infinitewav.ca/ > > My converter (Secret Rabbit Code) is not the best, but is definitely > not too bad.
Erik, The link doesn't work for me. Jerry -- Engineering is the art of making what you want from things you can get. ¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
Jerry Avins  wrote:
> Erik de Castro Lopo wrote: > > > Hi all, > > > I thought people might like to know about the sample rate converter > > comparison which has been done here: >
[wrong link snipped]
> > > My converter (Secret Rabbit Code) is not the best, but is definitely > > not too bad.Erik, > > The link doesn't work for me.
Should be: http://src.infinitewave.ca/

Erik de Castro Lopo wrote:
> Hi all, > > I thought people might like to know about the sample rate converter > comparison which has been done here: > > http://src.infinitewav.ca/ > > My converter (Secret Rabbit Code) is not the best, but is definitely > not too bad. > > Cheers, > Erik
Interesting stuff. I found the phase behaviours of some of the resamplers very interesting. While they aren't too important for audio, they could cause major confusion in some other applications. It would have been nice to have a run time comparison, too - that would help to make it clear that SSRC in ZOH and linear) modes sells a lot of quality to get good speed. A scale or key would have been good, too. Cheers Marc
Marc Brooker wrote:

> It would have been nice to have a run time comparison, too - that would > help to make it clear that SSRC in ZOH and linear) modes sells a lot of > quality to get good speed.
I can assure you that those two converters are only there for completeness (so I can show people how bad linear interpolation is for audio) and for people wanting to mess up there sounds when trying to emulate 1980s sample playback synthesisers.
> A scale or key would have been good, too.
For the color scale? Its on the right hand side of the page. You may need to scroll over. Erik -- +-----------------------------------------------------------+ Erik de Castro Lopo +-----------------------------------------------------------+ "When you say "I wrote a program that crashed Windows", people just stare at you blankly and say "Hey, I got those with the system, *for free*." -- Linus Torvalds

Erik de Castro Lopo wrote:
> Marc Brooker wrote: > > It would have been nice to have a run time comparison, too - that would > > help to make it clear that SSRC in ZOH and linear) modes sells a lot of > > quality to get good speed. >I can assure you that those two converters are only there for completeness > (so I can show people how bad linear interpolation is for audio) and > for people wanting to mess up there sounds when trying to emulate 1980s > sample playback synthesisers.
Some converters rated on that page are so bad that I'm convinced they used something like ZOH. If you don't believe that people actually do that kind of thing (and try to sell the idea as "high-end"), read this: http://stereophile.com/reference/104law/index1.html Now that method would surely make for a nice graph ;-). Regards, Andor
Erik de Castro Lopo wrote:
> Marc Brooker wrote:
<snip>
> >> A scale or key would have been good, too. > > For the color scale? Its on the right hand side of the page. You may need > to scroll over. > > Erik
It's moments like this when I regret giving up coffee.
Andor wrote:

> Some converters rated on that page are so bad that I'm convinced > they used something like ZOH. If you don't believe that people > actually do that kind of thing (and try to sell the idea as > "high-end"), read this: > > http://stereophile.com/reference/104law/index1.html > > Now that method would surely make for a nice graph ;-).
To be fair, the author does admit it's not good enough for rock'n roll at the end ;) Martin -- Quidquid latine scriptum est, altum videtur.
Martin Eisenberg wrote:
> Andor wrote: > >> Some converters rated on that page are so bad that I'm convinced >> they used something like ZOH. If you don't believe that people >> actually do that kind of thing (and try to sell the idea as >> "high-end"), read this: >> >> http://stereophile.com/reference/104law/index1.html >> >> Now that method would surely make for a nice graph ;-). > > To be fair, the author does admit it's not good enough for rock'n > roll at the end ;)
Which begs the question: what is it doing in a stereophile magazine? Steve
Steve Underwood wrote:
> Martin Eisenberg wrote: >> Andor wrote: >> >>> Some converters rated on that page are so bad that I'm convinced >>> they used something like ZOH. If you don't believe that people >>> actually do that kind of thing (and try to sell the idea as >>> "high-end"), read this: >>> >>> http://stereophile.com/reference/104law/index1.html >>> >>> Now that method would surely make for a nice graph ;-). >> >> To be fair, the author does admit it's not good enough for rock'n roll >> at the end ;) > > Which begs the question: what is it doing in a stereophile magazine?
Begs an answer, rather! :-) http://begthequestion.info/ Sorry: I'm one of those it bugs. http://www.worldwidewords.org/qa/qa-beg1.htm From the tenor of the article, someone wrote in claiming that his new method works just fine, so why the fuss, and the author responded with the equivalent of "Well, maybe ..." as politely as he knew how. What's more, writing a monthly answer column must occasionally hit dry spells, and this was the best he had. Jerry -- Engineering is the art of making what you want from things you can get. &macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;&macr;