|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
151.197.245.69
Hey guys,
I posted this at audiogon and havn't seen much interest, so sorry if you've read this already.Can any benefit be realized by changing the word length of a 16/48 recording to 24/48? The conversion was done via a Tascam DV-RA1000 if you're curious. I would like to know any negative or positive effects of this process and, theoretically, what is going on during playback that can account for the perceived change in sound.
Follow Ups:
Yes , there are benefits especially if any other processing is done on the signal. As an example , if you had the no 23 and divided it by 2 , the result is 11.5 , if you only had 2 spaces to store that result you could either store 11 or 12 , both of which are incorrect. Should you have an extra space , you can store 11.5. any subsequent operations would use the correct result.
You would still have to downconvert at some stage back to 2 spaces but the figure you get will not be as innacurate due to multiple rounding up/off due to the lack of "space" - this is simplistic but it does illustrate the point.
No operation on a digital signal is trivial , there are definite sonic changes in changing word length , if you have a unit like a Meridian 518 , you can change all sorts of things like wordlength , dither alorithyms etc and store these and directly compare. Differences are obvious in this case.
Rodney Gold
Does anyone know how to deciper each word. It is 16 bits and there has to be some code to it.
Thanks,
Depends on where you intercept it. Look up I2S or S/PDIF. Tons of references.
"Can any benefit be realized by changing the word length of a 16/48 recording to 24/48?"None whatsoever. If the word is padded with zeros, there should be no effect sonically. If the word is somehow "augmented" with random values, it would change the sound, but never improve it. For the "resolution" being added has zero correlation with the resolution lost during the encoding process.
in addition to what KR said about dithering word length some devices will also perform noise shaping so there could be an increase in percieved SNRatio.
"in addition to what KR said about dithering word length some devices will also perform noise shaping so there could be an increase in percieved SNRatio."While what you stated is true, the dither/noise shaping is done in the recording (A/D) process, and is contained in the data encoded to the media. It does not apply to the playback (D/A) process, aside from the digital filter itself using the dither-enhanced data from the media to reconstruct the low-level signal information below the media's LSB. Providing an output signal truer to the pre-digitized signal's low-level information compared to playing back media of the same signal encoded without dither.
Well, you can change word length by padding with zeros or you can up/over-sample to spread the same data over more bits. In the former, there should be no change in sound. In the latter, any change will be due to the sampling algorithm (or digital filter).
This post is made possible by the generous support of people like you and our sponsors: