How many bits, how fast, just how much resolution is enough?

[sdolezalek]

I'm assuming for this post that 16/44 captures all of the frequency range and transient information we need (you may disagree, but make that assumption for a moment). Then the real reason for hi-res (or upsampling) is all about the digital-to-analog conversion process and the benefits we get by moving filters and other conversion artifacts as far above 20kHz as we can.

 

That results in

a) our wanting to feed our DAC hi-res (call it DSD128 or higher or DXD) information (irregardless of whether it started as 16/44 that we upsampled or an actual DSD128 or DSD512 or DXD file);

b) then applying less aggressive filters (so as to avoid ringing and other artifacts in the 20Hz-20kHz range only); and

c) feeding all equipment downstream of the DAC considerable content above 20kHz (unless the DAC itself introduces a 20kHz cliff filter).

 

If we are doing a,b and c, then how important is it for us to know how each piece of downstream equipment will interact with those higher than 20kHz frequencies (i.e. pre-amp, amp and tweeters)?

 

Some have made the case for having pre-amps, amps and speakers that are linear out to 40kHz or more such that feeding them information not brickwalled at 20kHz doesn't introduce new problems that we were trying to solve with the hi-res and gentle filters solution at the DAC. Is it therefore possible that the reason some of us do/don't prefer (or hear) the benefits of hi-res is the limitations of the downstream equipment?

[sdolezalek]

You're missing one piece, which is that moving the sample rate higher permits exclusion of near-20KHz ultrasonics without a "brickwall" filter. For DSD, ultrasonic noise is moved further from the audible range.

 

Jud: Thanks for adding that. Could you explain a bit more? What causes the "exclusion of the near 20kHz ultrasonics?" I assume that is the same difference that tailspn is referring to in terms of less >20kHz content in DSD256 than DSD64?

 

(Sorry, posted too quickly; the graphs you cross reference in your post above, do explain it. Thanks)

[sdolezalek]

I guess the question comes down to: assuming A) recording at DSD256, and then B) state of the art conversion to Redbook, what percentage of people can tell the difference when upsampled to DSD512 and fed to the DAC? Assuming optimal playback equipment/conditions.

 

Isn't that the wrong question? If your DAC is doing internal upsampling anyway; whether to DSD512 or to something else, isn't the real question: "Can I tell a difference between where the upsampling occurs, how it is done and what filters are applied in that process?"

[sdolezalek]

I think a big part of what makes this discussion difficult is that we are now into such a minute level of differences that many systems don't reliably pick them up and depending on what equipment we have in the chain may determine how much of the difference we actually hear.

 

For example, I can fairly reliably tell a difference between a Tidal 16/44 file and a digital 16/44 copy of the same file on my NAS. I can also tell the difference between what I listen to at 16/44 and what I uprez to DSD128 in HQPlayer. I also believe I hear differences between well recorded DSD256 files and a 16/44 version of the same file.

 

BUT, I only hear those differences on my main system, my others are simply not resolving enough to matter, AND, better recordings always trump lesser ones irregardless of resolution. That being said, adding all of the incremental improvements together means what I'm listening to today is a hurge improvement over what I heard from the best CD playback in the same system just a few years ago. So I applaud the ongoing work of Jussi and others to continue to make tiny adjustments that may be insignificant on their own, but that ultimately add up to greater listening enjoyment.

[sdolezalek]

Yuri and Jussi: Based upon the last two posts, and without asking either of you to disclose any trade secrets, is it fair to assume that how you treat these ultrasonics is a difference between using HQPlayer and Audiventory for upsampling files to DXD or DSD256+ resolutions?

[sdolezalek]

2. Optimised

0…[20 kHz + transient band]. Rest band is filtered.

I recommend it as «DAC helper».

Optimised mode have linear and minimal phase filters.

 

I suppose, optimised mode is single case, where I recommend perform hearing test before spectrum analisys :-)

 

Yuri: Yes, thank you, that is what I remembered, that in Optimized Mode, in addition to upsampling and using "better" filters you also applied a second filter to remove noise and other other information beyond the 20kHz limit.

 

I believe that when I use HQPlayer to do upsampling "live" Jussi allows the user to choose filter types and characteristics (including linear and minimal phase filters) but is not applying a second filter to cut off information above 20kHz. That would make the results of doing an upsampling from 16/44 to DSD256 different between the two programs.

 

That was what I was trying to clarify.