Near the end of my Audirvana trial and have only one question. What audible improvement does upsampling bring?
I’m running it on a M1 Mac mini (Sequoia with 16GB of RAM, USB to a Topping D10s) and tried the Power of 2 upsample type with the r8brain and SoX Algorithms (default settings only so far) and am at a loss as to what changes I should be hearing.
Are they subtle and beyond my hearing capacity (as an elderly person with concert related hearing loss) or do I need to make major adjustments to the settings?
First thing that I would do is take a screenshot of your default settings and then dig into it a bit. You can always go back to default. Here’s a recent very interesting topic on just this subject.
Maybe you could share some information on your rig?
Simple apartment appropriate main rig - Substantial Music file collection on 2 & 6 TB hard disks USB to the M1 Mini > USB to Topping D10s > NAD Integrated > ProAc Tablette 8 + Totem KIN 8 Sub. Headphone rig Sennheise HD600 plus a variety of IEMs into an old AudioEngine D1 DAC - Schiit Modi amp.
Also stream Apple Music to the Mini and various HomePods throughout the apt.
The simple answer…
Up-sampling of lower-resolution PCM files, reveals the inherent dynamic range of the encoded signal and presents a more refined signal waveform to the Digital to Analog (D/A) circuitry of the DAC… It moves the Nyquist Frequency (Fs) aliasing of the signal beyond human hearing… No contextual harmonic, dynamic, or spatial information is added or subtracted in the process… the encoded elements of the source remain intact… What is revealed is these contextual elements are now unmasked that were hidden in the dynamic noise-floor of the lower bit-depth file… The frequency and resolution of the source file remains as was codified in the master encoding. What is added to the file to increase the sample-rate resolution (smoothness) is a bunch of digital zero’s (0).
In the case of modulation (up-sampling) PCM to DSD, the signal becomes more refined and is easier to translate into analog voltages as the signal more resembles an analog waveform.
The more you understand the influence of the filter parameters of the up-sampling algorithms the better you will be able to fine-tune the process for your aesthetic.
In addition…. Using the computer to do up-sampling offloads any up-sampling process that the DAC chipset would normally be engaged-in processing the signal for output, reducing operational noise (jitter) that is generated in that processing.
Hi, some of the individual upsampling settings result in subtle improvements. If you don’t hear a difference and enjoy the music Audirvana plays, then I wouldn’t worry too much about it. As I “cleaned up” my system over time (by better power cables and linear power supplies, for example) I started noticing the improvements that upsampling can bring much more clearly.
As you try individual settings, as I’ve done to optimize the sound of my system as documented in the review I wrote, I would not hesitate to ask AI what you should focus on. There are some forum members here who don’t like using AI, but I have found asking questions about what the individual SoX and r8brain settings do very helpful. You can Google them, but I have found that Claude.ai is quite knowledgeable. I recommend trying the “filter phase” setting first, as I have found it quite easy to hear the difference (even when I first optimized my original Audirvana setup years ago).
I personally love what Audirvana brings me, but if you don’t hear a difference with upsampling, you can also try JPlayer if you have an iPad or iPhone. JPlayer doesn’t do upsampling but it sounds very clean like Audirvana does without upsamplling. Once you discover what can be achieved with Audirvana Studio’s more fine-grained control, though, I think you’ll find the slightly higher price of Studio worth it. Good luck and don’t hesitate to ask if you have more questions. It’s all about enjoying the music you love. I have found Audirvana helped me bring my system much closer to how I have experienced instruments and voices “live”.
The Audirvāna Knowledge base link (How upsampling works in Audirvāna?) to the description of the functional parameters of both SoX and r8Brain are clear and definitive, there is additional information in the online User Guide found in the pop-up window of the upsampling module… An excellent starting point for more detailed and focused research if so desired… There are many Audirvāna users on this forum that can contribute very qualified insights into contextual technical functionality of the filter design elements used in the sample-rate and bit-depth conversion algorithms… No reason to engage a machine to get quality input on the subject of sample-rate conversion. … What have you contributed to the understanding of the functional benefits of upsampling beyond your “magical” revelation in finally applying upsampling in your playback strategy?
Experiment and see what you like. From what I can gather you may not have tried DSD128, so see how that sounds to you. (Note: DSD is going to play at a lower volume if you keep everything else constant, and psychologically louder sounds better to us. So you may want to try to equalize volume if you compare DSD to PCM. There’s a free AudioTools app for iPhone with an SPL graph capability that works pretty well for this.)
The typical native DSD mastering level is done at -6dB… However, the modulation of PCM to DSD can add gain due to the increase in dynamic range (hence the ‘reduce gain before up-sampling’ function parameter). I’m not certain about this, but it appears that the algorithms normalize the output of the processing.
I have decided to go ahead with my subscription to Audirvana.
Much of that decision is based on the generosity shown here. Many of you are so much farther down the Audirvana path than I and yet still take the time to offer assistance to a digital audio neophyte. I’ve realized there is much to be done and in retirement, I have that time.
Cheers to you all, and thanks again for all your assistance.
