Upsampling the way to go? ?

Not to prolong the point, Tim, but why would an interpolation process be needed to generate an extra 8 bits of *random* dither? As I understand it, this dither acts as sort of a "digital bias", and is not related to the actual signal. I think the increase in word length simply helps the DAC remain linear deeper into the noise floor. It does seem to me, however, that interpolation would be appropriate to aid in generating the "extra" samples needed to raise the sampling rate from 44k to 96k, or whatever higher rate is chosen. With only a finite frequency of voltage samples available off the disc, some sort of algorithim that interpolates new sample voltages in between the recorded ones would have to be implemented to raise that frequency, no? So I agree that upsampling is separate from increasing word length, but wonder if you might be confusing which of the two processes incorporates interpolation algorithims. (I already know that *I'm* confused!) And since I'm certainly no digital engineer either, I hope someone who is can come along in this thread, and soon!

Zaikesman--I certainly don't mind this point being prolonged because I think there are a lot of CD players and DACs out there that claim to upsample to 24/96 or 24/192 and really don't increase word length in the more strict and proper sense(whatever that is--help). I too would appreciate anyone who knows more about the process by which upsampling is done vs. increasing word length, and maybe even give us a way to identify the "pretenders" from those that truly do increase word length along with upsampling--there seems to be a lot of misunderstanding out there regarding this point.

My sense is that upsampling by itself will not make a huge difference and could do harm if not implemented with care due to the added circuitry involved(some of the above comments seem to bear this out to some degree), but that combining it with a longer word length can bring about even more dramatic changes if done properly. Where's an engineer when you need one?

Tim

Tim is correct ... upsampling refers to increasing the sampling frequency from 44 kHz to a higher number, and using the extra samples to interpolate.

The benefit is that the anti-aliasing rolloff filter is much simpler to implement since it can rolloff more gradually (since the sampling frequency is now much much greater than the maximum frequency recommended ... it's all to do with Nyquist's theory of sampled signals).

At the same time one can also increase the word length of each sample by adding dithered noise. Not extra information .. noise. Dithering is a wierd concept where adding some noise increases the perceived clarity. I'm not so familiar with it ... but as an experiment try looking at a pixelated image and then blur your vision and notice how the image becomes much more recognisable.

Neither upsampling nor adding dither and extra bits can add any information that was not originally present in the recording, neither can they increase the theoretical maximum dynamic range ... instead I think they help to mitigate some of the real-world problems associated with extracting the signal (the roll off filter being the biggest problem.)

Of course a well implemented non-upsampler will outperform a bad upsampler ... probably as much to do with the electronic design, analog output stage, noise from digital bleeding into analog .... it's just good design vs bad design. However since there are plenty of people out there who just have to own the latest buzz words upsampling will sell. Whether a good upsampler will outperform a good non-upsampler ... I really don't know ... probably down to individual taste.

This reminds me of the bitstream vs multibit arguments of about 10 years ago. Both can be done well .. both can be done badly.

All this said (and I have neither heard an upsampler nor SACD) I cannot imagine that upsampling can beat SACD since SACD encodes more of the original information and upsampling appears to "fake it". Perhaps SACD real world implementation has not yet quite caught up with the theoretical possibilities ?

I am biased as I own the Delius and Purcell however dCS does have some good information on their web stie.

dCS White Papers

Cheers,
Dan

Sean T., sorry to second-guess, but your first sentence strikes me as being just a *little* bit confused itself! :-) Tim and I don't disagree about what upsampling is (as opposed to increasing word length), we are just, again, a *little* confused about how it happens!

You say the upsampling "...uses the extra samples to interpolate." That seems backwards to me. Wouldn't upsampling have to *use* interpolation to *arrive* at those "extra" samples? Then the DAC would use *all* the samples, including the "added" ones, to reconstruct the wave? Or, yet again, do I "misunderestimate" this process completely? My new MSB Gold Link DAC mentions in the manual that "oscillators" are used to upsample the signal. My dictionary defines "algorithm" as any mechanical or recursive computational procedure. "Recursive" is defined as descriptive of a mathematical expression like a polynomial, each term of which is determined by applying a formula to preceding terms, or of the formula that generates the successive terms of such an expression. The definition of "mechanical" that would seem to apply in the case of an oscillator is "governed by mechanics". Sounds simple, doesn't it?

The Gold Link comes with two oscillators to choose between, and seems to sound better on CD's with the upsampling set to use the one at 132KHz instead of 96k, which is recommended by MSB since 132k is an even multiple of 44k. The math has got to come out better this way in a "mechanical" process, no? Still confused, Z.