One obstacle...
Converters. While you may be able to sample convert up to that resolution, you certainly would be hard pressed to find any DAC capable of resolving that combination.

Also, what would be the advantage of that over DSD or DSDx2?

Cucco

Hi cucco,

I hope to record at this frequency and will not be upsampling the data.

I appreciate your question about what the advantages could be over DSD or DSDx2 but hope it's ok that I give a short answer - although there's also a lengthy one...

The short answer is resolution and the ability to chose the noise shaping differently from what Sony/Philips recommend. Also, ENOB of DSD is quite low and although I believe in the simplicity of one-bit conversion as a sampling process, I believe higher sample-rates would be beneficial.

I will not be using a DAC but plan to use simple analog filtering. This is where I believe the one-bit process has its advantages.

Do you know about Blu-ray for audio?

Regards,

Jesper

How would you record at this sample rate? I don't know of any native converter or any ADC chip (unless recently released) that will do this? Will you be designing your own chip?

The resolution will not increase with a SR increase, only the available bandwidth (and storage required). Additionally, the noise figures on 5.8Mhz DSD is sufficiently high that a 96kHz LPF filters the vast majority of it out.

Also, it's not possible to take a single bit stream straight into an analog filter and make it work. You must still use a DAC - again, I'm not aware of any that exist that are capable of this frequency.

I am familiar with the capabilities of the BluRay media, but it's still just media. If the technologies aren't available to make it playable, you can put anything you want on there, but if no one can play it what benefits do you gain?

Also, it's not possible to take a single bit stream straight into an analog filter and make it work. You must still use a DAC - again, I'm not aware of any that exist that are capable of this frequency.

... You seem to know something that I may not know... Can I ask you why it would not be possible to use an LPF filter only for conversion? Please be specific - I have discussed this with lecturers at the Danish Technical University (specializing in digital and analog technology), and they did not have comments to this approach.

And, yes, I (or hopefully as part of a collaboration) aim to design custom circuitry.

Best regards,

Jesper

(copied and pasted by mod)

Regarding ENOB I understand that:

ENOB = (SNR – 1.76 dB) / 6.02. As the SNR for an identical noise shaping order increases with higher sampling rates, the ENOB also increases.

Again, I'm open to feedback ...

Best regards,

Jesper

gentlevoice wrote: ... You seem to know something that I may not know... Can I ask you why it would not be possible to use an LPF filter only for conversion? Please be specific - I have discussed this with lecturers at the Danish Technical University (specializing in digital and analog technology), and they did not have comments to this approach.

And, yes, I (or hopefully as part of a collaboration) aim to design custom circuitry.

Best regards,

Jesper

What you are suggesting (taking the bitstream and filtering it) is a home-made DAC, but a poor one. There would be no amplitude reference, no rejection of power supply noise and no clock regeneration to minimize the effects of jitter.

Also, why the fixed 11.289MHz bit rate? This is a 44.1KHz-only equivalent word rate. I would have thought that you would want at least the choice of using 12.288MHz bitrate to allow 48KHz and 96KHz equivalent word rate replay.

If you want to construct something yourself, you could try using one of the DACs from the Burr-Brown (TI) DSDxxxx range. The DSD1792A is pretty versatile, but you would need to be able to communicate with the registers in the device though the 4-wire serial control port.