|'); } // End -->|
I'm thinking of purchasing the musical fidelity a3.2 cd player. It has a 24bit 96khz dac. My question is, is there really a sonic difference between 96khz and 192khz ?
If it's a CD player, I'd avoid anything that converts the signal to 96 or 192.... I think the "older" 8x oversampling and non-OS are both better options.
Monarchy Audio M24. Uses BB PCM63 true 20bit DAC and DF1704 8x oversampling digifilter. Also, passive IV conversion and no opamps. Output through zero feedback SRPP tube stage. No transistors in the analog section! Wonderful sound.
"Or is that just your experience across the board with players?"
Exactly. Whenever I listen to converters that use asynchronous sample-rate conversion on a high-resolution home audio system, there is a band of noise of constant spectrum cast in the high-frequency components of the converted audio signal. At first, it is perceived as detail, but after a while, I notice a constant signature to the HF, and I ultimately get fixated on this artifact. It's a symptom that I've noticed with ASRC, and it becomes more-apparent when I later listen to a non-ASRC DAC, where this artifact is noticeably absent.
I later deduced technically that ASRC transforms jitter at the input to noise at the output (Item 3 in the link), and while I have not proven this to be what I actually heard, the fact that jitter has a spectrum would cause the transformed noise to have one as well. Which is consistent with what I heard.
You will have to see for yourself by listening.
For every person who says they can "hear it" you have another who says it is "impossible to hear" and "no scientific proof" of anyone being able to hear it.
This has been a huge argument between many people for many years.
LISTEN for yourself and if you think you need it, get it.
If you can't hear it, your choice nevertheless.
I have 192K because I have products offered in that format that I want to use. If you have 192K capability..chances are you will also have 96K
Let your hearing be your guide..not someone elses.
Not sure if I understand what all this means. Are players designed to convert the 44.1 kHz sample rate to something higher? What would the advantage of this be? I am under the impression that there is no more info to be extracted from a CD than 44.1k samples per second, so what does upconverting do?
The only thing I could imagine is that some of these players sound different/better not because of upsampling, but because of the specific dac, the capacitors, power supply, etc.
No. it's not the factors that you are thinking of. I have a DAC that can upsample to 88.2, 96, 176.4 and 192 KS/s as well as to DSD. Obviously the DAC and other internal components remain the same but each of the upsample rates chosen sounds different and it may be argued that the higher rates appear to offer more in terms of detail. timbral accuracy and ease of listening even if there is no additional data over the standard sampling rate. There are,of course, dissenters from this view but that's audio for you. The reason, if the premise is accepted, is partly or wholly to do with the frequency at which the digital filter comes into effect. See the link if you want detail.
I was under the impression that this topic was on CDs, not DSD media. I still am not completely clear (I am new to this!) why a CD with a sample rate of 44.1kS/s recorded onto it would sound any different if played back after being upsampled to 192kS/s or DSD, just seems like dividing 1010 into 11001100 or 1111000011110000 and playing it twice/4X as fast and etc. But I am really not up to speed on this topic (my circuit design course did not cover digital electronics).
Thanks for any help!
Well, what I gave in my response - both in my answer and in the link that I posted - are not exclusively concerned with DSD although it is also mentioned as it is, if you like, another way of achieving a high sample rate. The basic points apply to PCM as well as DSD. Here is another link for you. I hope that this helps. As other posters have pointed out, this is probably really about the effect of digital filters, their intrinsic nature (type) and the frequency from where they are operational. High sample rates allow for them to be placed well above the 44.1 KS/s Nyquist point of 22.05KS/s with advantages IMHO. It is not just raw numbers which would, as you say, show no improvement. However, whether or not the effect is accepted as benefical will, as always, depend both upon its implementation (there are some awful sounding upsampling players) and personal judgement (I think plain vanilla 44.1 can sound great).
Actually I wouldn't buy a player/DAC where the upsampling could not be switched out as my experience has shown some repertoire where upsampling reveals problems that standard sample rate PCM hides e.g. where a lossy codec has been used (stand up DAB radio) or for some CDs where the music is from early multitrack recordings.
"As other posters have pointed out, this is probably really about the effect of digital filters, their intrinsic nature (type) and the frequency from where they are operational. High sample rates allow for them to be placed well above the 44.1 KS/s Nyquist point of 22.05KS/s with advantages IMHO."
Whatever the final sample rate of the DAC, with a CD input each and every 'standard' digital reconstruction filter has its transition band at 22.05kHz. The only relaxation is on the subsequent analogue low-pass filter which indeed can be at a much higher frequency as the preceding digital filter has already attenuated everything between 20kHz and the oversampled fs/2.
('standard' meaning filters obeying to the Nyquist/Shannon definition of reconstruction, as opposed to filters that allow some imaging in order to - somewhat misguidedly - show a faster time response, e.g. Wadia, Legato, ...)
bring bac k dynamic range
...instructive as always. Incidentally, having a number of alternative filters to play with in my CDP my misguided ears always tell me to use filter 4 when using PCM which nevertheless allows for a certain level of Nyquist imaging. I see that John Atkinson found that was his preference too in his review of the dCS Verona. Curious, eh?
Not at all curious when one accepts that the notion of stereo-based audio is (approached) an unbroken chain that brings a musical performance alive in a living room (or transports the listener to the original event, depending on your pov) is wrong.
And it is wrong because stereo-in-a-living-room is fundamentally not capable of doing so for the majority of original acoustic events and the majority of living rooms. (No need for the high-end fundamentalists to jump on my back here.)
So when we accept that the chain is fundamentally flawed, it is no major leap to accept that doing something in the chain that is contrary to theoretical correctness can actually sound better.
Two wrongs making a right, but at a higher, almost metaphysical level.
Hence also tubes, vinyl, SE triodes, horns, LSD, ... we are all hunting for a truth that is not attainable, not even approachable. But that's no reason to make the hunt less than fun.
"Hence also tubes, vinyl, SE triodes, horns, LSD, ... we are all hunting for a truth that is not attainable, not even approachable. But that's no reason to make the hunt less than fun."
I don't want the truth Werner for partly for the reasons you give, but some system's are able to consistently give a very convincing lie and in my experience these systems rely on doing the basics right rather than relying on technological 'advances' to hide cost cutting measures.
For me, the primary requirement of an audio system is that it should connect me to the music and affect me emotionally just as the writer and performer of the song intended, even when I'm listening to an old Aretha Franklin recording with obvious analogue hiss which doesn't allow the illusion of the recording being anything other than a recording.
If, for instance, a singer emphasises certain words and structures phrases to give more meaning and emotion to a song, there's more going on than simple changes in volume or timing but this is what most systems miss in my experience and all that is picked up is the sound - not the emotion behind the notes being sung or played.
Of course it's probably impossible to quantify what is going on but once it is heard there's no going back, and those who appreciate the difference are accused of being seduced by euphony/colouration...........or LSD. :0)
Raising the sample rate allows more freedom in implementing the reconstruction filter. Without raising the sample rate a reconstruction filter that can go from full on to full off between, say 20kHz and 24.1kHz and still not mess with phase too much is well neigh impossible. (i.e. 96 dB in a fraction of an octave.)
This post is made possible by the generous support of people like you and our sponsors: