Imagined you just have half of your RAM free for streaming, then a file (in case it’s fully loaded into RAM for playing) would still have to be nearly as big as 4GB to have any problems. A normal DSD256 or 24/96 file would be around 250~500 MB. So I think there’s enough free space to load and unload sequential playing files into memory. But that’s just theory, if you trust people who heard something, that’s always better to listen to.
Regarding other measures, there’s so much…maybe optimizing the USB connection as a first measure, or better having an own power cirquit for the stereo (that’s a more elaborate tweak) but I leave it to others maybe recommending more.
These are people who don’t understand how RAM is used for computing. Also, different operating systems optimizes RAM and virtual RAM differently. Don’t apply a “healthy” criteria from, say, Windows to OSX.
You want to avoid things like page ins / outs. The “easy” way to see if this is happening is by looking at something OSX calls “memory pressure”.
Does Audirvana still support “memory play” option? It’s been many years since I’ve used it. Loading giant files into memory might be the only reason you’d want a large amount of memory overhead.
*Note, I never recommend less than 16GB when spec’ing a modern PC but that has 0 to do with audio.
100% agree with this entirely, both on the people saying it would require 16Gb not understanding how RAM is utilised and the difference between OS types.
Typically OSX will require around half the RAM that a windows machine will, it’s just a far more efficient OS in that respect. Same with Linux really.
RAM is only used if it’s called by the application in use, it’s not something you just randomly add more of and = better performance, a lot of people don’t understand that.
If there isn’t enough RAM available, then all that will happen is the system will instead write to the OS drive the information it’s unable to store in RAM, so you’ll see a slow down in performance, but these days with SSD’s it’s fairly minor with a use like this in mind.
Well, as to the last 2 replies…although I thought I ordered one with ssd, my invoice is ambiguous, and I generally upsample to 128. We will see what happens. I have 30 days to try it out once I receive it.
And to all, yes your answers are in line with my opinions prior to reading the naysayers.
macOS Catalina 10.15 minimal hardware is Mac mini (Late 2012 or newer)
Latest macOS is 12 with 13 in beta.
You really don’t want to be running an OS this old with no easy way to upgrade.
If you want to run old hardware you need to stick with generic intel / amd and run Linux. Audirvania says minimal OS is 10.13 (high sierra) but buying hardware that could become obsolete any day isn’t saving money.
I wouldn’t recommend going below that. With my NuWaveDSD I was using JRiver to convert PCM to DSD 2x. My bedroom HTPC has 6GB RAM and it also keeps up. It does the same conversion with a NuWaveDSD. It’s an Acer with Intel Core3i.
Most audiophiles uninstall software that wouldn’t be used in an HTPC. I used a Dell Win10 Intel Core5i with 8Gb RAM and it played JRiverMC reliably. Plus I streamed content from youtube, Prime Video, TV episodes from network websites, etc.
I shut down my power regenerator without powering down the HTPC. So that was the end of it after 7 years. I recently replaced with an HP Win11 Intel Core5i with 16GB RAM. Works wonderful as an HTPC.
Good decision going with the 2012 mini. I own a 2012, 2014, and 2020 M1 mini and feel that the 2012 clearly sounds best. I’m running Audirvana 3.5 via USB into a DSJ. My 2012 is running Catalina with a 1TB SSD and 16GB of RAM. If I had to rank all three mini’s by sound quality via USB from best to worst it would be 2012, 2014, and lastly the 2020 M1.
I have an Intel i5 NUC as my only music source, over the past 4 weeks I have upgraded the SSD to a quality 2 TB and upped the RAM to 32GB, any differences?.. not any I can hear. Adding a Matrix Audio X-SPDIF 2 32Bit/768kHz DSD513 USB Interface next week