Bitzie USB-DAC Preview

Haden2866 wrote: I think I tend to fall into the trap of associating higher sample frequencies with necessarily 'smoother' or less saw-like digital rendering. It sounds like what you are saying is that most of those bits will be inaudible on playback, but the extra sampling frequency over and above the 44.1 kHz, specified by Redbook to satisfy Nyquist, provide 'headroom' for any aliasing to take place well outside the audible spectrum. Hmm, not sure if I have fully grasped the relationship between frequency of sound and sampling frequency. I will keep at ,

I principle, the higher the sampling frequency, the higher the frequency of the sound. A 96 khz music file has an frequency of about twice of what a human can hear. Useless for improving sound in other words. This idea of smoother or less saw-like digital rendering, well, forget about it. Digital sampling don't work that way. There are texts on the internet explaining the benifits of so called hi- res files in that way. It is false. Period. 24 bits are very nice when editing music, but it is VERY questionable if it has any benifits for the end product, ie the music we listen to as music consumers.
Most DAC.s upsample the input signal, but this is because the designer want to put the filter well above the hearing threshold.

GoSUV wrote: The original specification for Red Book CD format is 44100 Hz for the sampling rate, which came from the Nyquist frequency theory that to reproduce 20000 Hz of upper frequency limit, one has to sample approximately twice. Since human hearing has a limit of what frequency we can hear, depending on age, 20kHz is a safe cut off point. So Red Book CD chose 44100 Hz sampling rate, slightly more than the Nyquist frequency limit of twice the given frequency. Of course it is just a theory. If you have a 1Hz sine waveform, by sampling twice, the result is a sawtooth waveform at best after digitization. This is a poor representation of the original. Red Book CD makes up for this by allowing slightly more room, but still to reproduce a smooth sine wave of the original, one has to sample many more times than 2 to get a fairly accurate reproduction. The second factor is the bit depth. Red Book CD is 16-bits, or 65,536 levels. More bit depth means more granularity to store the resulting information. 24-bits gives you 16,777,216 levels. Now whether your end equipment and everything in the chain will fully utilize all these is another matter. This is just a theory. Of course., one cannot forget the source. If the so called "hi-res" 96/24 source you bought was simply up-sampled stuff from 44/16, it will completely defeat the purpose of "hi-res" in the first place.

This makes me feel like a kind of warrior trying to bring light into medieval darkness. This is so full of missunderstandings. What you say here is a contradiction.. You are in principle first giving a correct picture of the Nyquist theorem, but then you end up in a very strange conculsion based upon the idea that Nyquist are wrong.

You have to understand that we are talking about a theorem here, wich can and have been proved mathematically. This is not just theory as good as any theory. This theorem and the way digital sampling works is well known and etablished facts. It is used in a lot of other applications, not just for music. Medecine is one of the fields where digital sampling is used, and sometimes in life depending processes.

The problem when comparing digital and analog probably has very much to do with the filtering. There is no perfect filter.

Edited by Humboldt - 11 May 2013 at 11:02am

mitch65 wrote: The problem when comparing digital and analog probably has very much to do with the filtering. There is no perfect filter.

I think this illustrates the problem with digital music, you can easily get caught up in the theory (which is way over my head) and miss the point....listening to music. I am of the opinion that, for most people, 16/44.1 will be more than enough. The better the equipment I play 16/44.1 on, the more information I can hear. If you have crap equipment it does not matter how high the resolution is because you won't hear the, so called, benefits. Just my view you understand.
[/QUOTE]

I ageee, but I also think the sound quality is very much up to the recording engineer and the record companies recording philosopies..
I would not care about the thecnical details and the missinformation beeing spread, but it is a bit sad, because basically it says that the Bitzie is not good enough because it will not play anything above 48 khz. Well it plays it, but it downsample, as I understand it.