Coincidentally, those (maximum DAC input rate, believing I hear a tangible difference) are two of the reasons I do inline upsampling. And unless I miss my guess, Reynaldo has similar reasons for doing the inline upsampling he does, and for requesting additional Audirvana capabilities in that regard.
Now considering that we are all doing similar things for reasons that track along similar lines, and also that you havenāt had the opportunity to listen to DSD256 or DSD512 in your system, perhaps youād be a good fellow and stop this constant drumbeat of telling us we shouldnāt be able to hear what we hear?
The premise for all of this was in the context of modulation of DSD64. DSD128 to DSD256, until you insipidly reframed the contextā¦
I have no problem with what folks feel like doing with PCMxxx to DSDxxx, if folks believe there is perceptibly tangible sound-quality difference between a 16/44.1kHz PCM file up-sampled (modulated) to DSD256 or DSD512, versus that same file modulated to DSD128ā¦ I question this perceptibility, and will only argue the cognitive bias that may be associated with the interpretationā¦
In the context of adding a superfluous function to Audirvana (like DSD64 modulation to DSD256), I will always voice my opinionā¦ folks may not like what I have presented to support this position and I do my best to hold up as much fact to support my position, rather than subjective and hyperbolic opinion about audibility of ultra-sonic noise artifact, and reframing of context to get my point across.
Do you realize what you were posting with these graphs?
It is plainly describing a signal level of -144db at 27-30kHz ! ā¦ Tell me you, or anybody can detect the audibility of this -144db signal at those frequenciesā¦ and at 50kHz the signal is still inaudibleā¦ Folks like me are always going to question this hubristic contortion of information to fit a narrative in the way that you do.
Tell me you donāt understand the technical reason why the SACD specification requires a 50kHz filter without telling me you donāt understand the technical reason why the SACD specification requires a 50kHz filter.
It is there to to filter ultra-high frequency noise of a 2.8MHz signal without FIR filtering, from damaging the component output circuitryā¦
The SACD format is capable of delivering a dynamic range of 120 dB from 20 Hz to 20 kHz and an extended frequency response up to 100 kHz, although most available players list an upper limit of 70ā90 kHz,[17] and practical limits reduce this to 50 kHz. Super Audio CD - Wikipedia
Anyhow - since this will otherwise continue indefinitely, back on my ignore list you go. If anyone else wants to know how to put a user on your ignore list so you wonāt see their posts, let me know.
To make everyone happier, I know some super-rich people who bought all the equipment at any price, like $100K DAC and $250K or even %500K speakers. The good news is once you go above $15K for a DAC, the performance becomes lower, especially for speakers. Maximum 20 is the sweet spot. Most products with crazy prices are scams, especially cables. Again, performance between 20K and 2K equipment is just marginal if you know what you are buying.
I highly respect German brands and the opposite of Swiss brands in audio. Some Japanese are fine. The US, the UK, and some parts of the EU are also good.
When it comes to noise, the source is important, but the issue is the other parts of the system, especially the DAC, preamp, and amp. Noise and distortion are there, and it is difficult to identify the sources of distortion and colorization. Electricity and ground loops between equipment are serious factors here, in my experience.
My amp is a modern design that is extremely low noise (-130dB or greater S/N). The preamp is built into the DAC, and it is also very low noise (about -115dB if I remember correctly).
The reason for taking such care with the source is twofold.
First, Iām using a DAC that will take a DSD256 or DSD512 signal from AudirvÄna and pass it through unmolested for conversion to analog. This provides the lowest-noise measured performance of which the DAC is capable.
Second, using optical fiber for Ethernet input to streamer and USB input to DAC eliminates ground loops and electrical noise to the extent itās possible for me to do so in my system.
I agree. Ground loops can come through the power provided through the network hub. Optical is a solution or a network hub powered by a battery or an isolated power supply. I am simply using WIFI.
DAC+preamp is a good choice as the less equipment you have, the fewer sources of issues.
I have upsampled a lot to DSD in the past, and I always enjoyed the DSD sound signature. But since around a year ago, I have not upsampled using any software. I have Chord Mscaler, which provides gentle upsampling.
My experience is extreme; music is more complex than visual. We know that in the age of AI, video upsampling is a very hot area. Music has harmonics (musical instruments). When upsampling, we damage those harmonics, and that impacts the spatial characteristics of the music. For ultimate instrument separation, we need the harmonics untouched. This is how our brain perceives depth in sound.
Thanks for your thoughtful comment. I have speakers that have time aligned drivers and linear phase crossovers for best imaging. I find as long as I use linear phase filters for upsampling, imaging is preserved.
Actually, it is the synergy of contextual Left channel + Right channel, harmonic, dynamic and spatial (timing/phase) relationships that facilitate the perceived soundstageā¦ When these contextual elements are delivered with integrity, focus/detail and spatiality is increased and/or revealed.
My feedback is based on listening and some good background in signal processing and upsampling filters. It would be great if you have some time to explore how harmonics are captured and how they might likely be impacted during the upsampling. Adding bits does not mean better
If you are talking about digitization ADC filters at a low sample-rate, as compared to the original performance spectral and dynamic energy as these correlate to spatiality/soundstage, you are correctā¦ A 16 bit/44.1kHz ADC will have reduced spectral and dynamic energy (harmonics and dynamics as related to spatiality/soundstage) in comparison to the same performance recorded at 24 bit/88.2kHzā¦ The theoretical dynamic range of the 16 bit recording being approximately 96 dB versus 144 dB for the 24 bit recordingā¦ The Nyquist frequency of the 44.1 kHz signal being approximately 22 kHz in juxtaposition to the 88.2kHz Nyquist frequency at approximately 44 kHzā¦ The resolution differences are obvious in the comparative audition.
The process of upsampling does not add or subtract harmonic information, only the sample-rate is changedā¦ The harmonic and dynamic information is codified in the original digitization process sample-rate and bit-depth. An upsampling algorithm works on the signal it is presented, within the interpolation process.
It boils down to what is valuable to the individualā¦ If we stick to output sound-quality, we see here in the Community forum, many former Roon users, are ascribing the greatest value to sound-quality of Audirvana as produced across a wide spectrum of playback systemsā¦
To me music is probably more important than the pure SQ debate so maybe i am not part of the ultra hi fi debate
For a variety of reasons i spend some of my listening time on BT headphones. I use Sony Wh 1000 MX4 and more recently the MX4 IEM. The at home kit is all headphones HD800 Focal Clear Mg and the Aryas i mentioned so hi fi is important to me.
That is another grouse with AS . The remote is just that it cannot play on the hand held device and hence BT in my case. Both Roon and JRiver allow for that
I am currently sitting in hotel with a cold beer watching the later afternoon sun with ear buds . Each to his own
Yes, thatās absolutely a major part of the appeal for people who prefer Roon. I personally donāt mentally organize my music in the way Roon does, and thus prefer being able to use and edit metadata in the way Audirvana allows.
