Option to enable RAM memory playback [to avoid caching problems]

Does roon preload tracks into ram before playing? If not please add this feature. Audirvana and Vox have it.

Could you explain the benefit?

Audirvana’s feature list says “File loading, processing, converting to DAC native format are all done before playback to minimize any possible interference”

Why would there be any interference, sounds like just faster loading and processing? Would it sound better?

Hi Bill,

I’m the same person as Zeny over at the Mcintosh forums. Good to see you here.

The positive feedback review of audirvana states:

It also does memory play where the files are loaded, expanded if necessary, and processed all in RAM before going into, in my case, a Sonicweld Diverter HR USB-to-S/PDIF converter.

I don’t have any technical knowledge or background but I suppose there is a sonic benefit for adding this feature? Otherwise Audirvana or Vox would not bother.

Guys, be careful of falling into the ‘me too’ trap. The competition amongst software players and servers is rife and audio forums are filled with discussion re the perceived benefits of all kinds of OS and hardware tweaks. Roon has had the benefit of working intimately with Meridian over many years and I think we can safely assume that the Roon team intimately understand what it takes to optimise sound quality.


Fact of the matter is that in SQ terms alone, A+2 sounds better than Roon, be it through memory play via SSD or RAM, but more likely because of the optimisation and enhancements offered through the settings. I bought a license for A+2 a little while (this year) before Roon was released, but I stick with Roon now because I really like the whole package.

I have been listening to both to compare SQ, and currently on my system, A+2 is better, no question.

But I would love to see ‘player’ enhancements within Roon. Hopefully the talks and testing of HQ Player code brings something to the Roon table.

@Andy , HI! :smiley:

Hi Allen,This is not a proven fact, nor am I able to refute this declaration with evidence. @audiomuze nailed it.

Why would we want HQ Players Engine (I know about it and it’s converting features)? Most music is recorded on Tape or digitally. Look at both of those mediums and their limitations. 12bits (of signal to noise(SNR)) for Tape and the 24bits PCM for digital, 70ishdb and 144dB SNR, respectively.

I am all for a better sounding product but can you convince us that we truly need it. I am all ears :smiley: and want the same endstate that you do.


1 Like

Better to my ears and fact for me. It’s all about opinions, I am not pushing this on anyone, but the beauty of any of these enhancements is you don’t have to engage them. That’s what settings are for.

I agree with you on the SQ with A+2 and I think I’ve put it down to A+2 using iZotope instead of Apple CoreAudio and also being able to use the power 2x, 4x upsampling and 16bit to 24bit.

I could be wrong but doesn’t HQPlayer also use iZotope over Apple CoreAudio too? I tried using HQPlayer but found the interface overly complex to use for me, but I’ve heard very good things about it’s SQ and very much looking forward to it’s implementation into Roon (if it does happen).

Sure, that’s a fact :smirk:

Don’t be afraid to push it on someone because if you know something that is better, we’ll never know. I want to know why HQ Player’s engine is so sought after- and if you look at the forum here It may come to fruition. Then you can say you were right all along :smiley:

Sure, you can turn the features on and off - that’s one good thing for sure!

I have Audirvana too, as it sounds great, but what are you comparing it too? Roon or something else?

I haven’t done a direct a/b comparison with Roon yet. But I have over the last couple years, compared JRiver MC20 with A+ and Audirvana wins hands down every time (especially with DSD files).

I’m quite happy with the sound of Roon, but I think having a DSP that will convert 16bit to 24bit and provide upsampling may just take it too another level along with memory playback.

I’ve also been thinking of replacing the hard drive in my MacMini to an SSD, I’ve heard that there may be some audio improvements in doing so, although I have no idea yet but for $100 or so, I think it’s worth it to try. It may also speed up the loading time of Roon which is roughly 4-6 minutes to completely open, not sure about that?

Which tracks/albums? Maybe I have them and we can compare notes?

I do not follow here, here I will convert 16 bit to 24bit for you right now.

16 bit 00000111 11011111
24 bit 00000000 00000111 11011111

In short, increasing bit depth just pads the 16bit signal with zeros, no?

There are no tools out there to test this concepts validity. This is pure subjective opinion.

That’s your subjective opinion Bill. You seem to have some axe to grind. No?

What are you referring to?

It’s been awhile but I think it was all DSD albums, playing them in A+ sounded more natural especially in the mid range and bass response… keep in mind my DAC is an Oppo 105 so it’s doing DSD to PCM conversion and that could likely be where the difference lies.

As for the 16/24 question. IDK, all I know is I have the 24/44.1 album of Vista Chino - Peace (from HDtracks) and I had a CD rip of it and the 24/44.1 had a better sound than the 16/44.1 (we’re not talking huge leaps and bounds better, but the bass sounds a little tighter and the overall soundstage was just a little fuller than the 16bit version)

Nothing wrong with subjective opinion at all. Again, I haven’t done this yet… just thinking about doing it and there is nothing wrong with that.

A spinning magnet will create a Magnetic Field [EMI]

A large magnet(s) [disk platters], spinning at 7,200rpm will create an even larger magnetic field…all within the PC it’s spinning in

And Time Clock’s, DAC’s etc., don’t play well with high levels of RFI / EMI

So there is certainly something to be be gained from using SSD in a computer directly connected to your DAC or Endpoint

Having said all of that, using a Network Audio Endpoint obfuscates these issues from the Endpoint…eliminating any Jitter coming from SPDIF and Async USB connections

So while spending on SSD can have merits…looking to “decouple” the Audio Endpoint from the “PC Source” by using Ethernet is something that is much more likely to result in an optimum result IMHO

Basically, almost every sentence you write ends with a ? Perhaps you are just the inquisitive type?

In addition, Upsampling to 2x Rates [e.g. 44k to 88k for CD ripped files] is known and measured to reduce the Jitter on SPDIF & AES / EBU connections by half [50%]

So Upsampling on its own is known to reduce the deleterious effects of Jitter on a SPDIF stream

Resolution Enhancement from 16bit to 24bit also allows more “headroom” for the Processing and that is what RE is seen in so many high end units as well as Upsampling

Adding ‘Apodizing’ or Minimum Phase filters to eliminate the effects of “pre-Ringing” caused by the brick wall filters used to produce CD 16/44 files is also Audible and desirable to many to have heard it

If the 24bit / 44kHz version was a digital download (say from HDtracks), there is a good chance that you have a better mastered version of the original master file. 24bit is better than 16bit. BUT It has to start at 24 bit, not the other way around (upconvert to 24bit) to potentially have a different sound/version. Look at what most music engineers employ in the music creation process. 24bit PCM.

I ask questions every day, much of what has been discussed is popular “norms” in the hobby of the audiophile however I have never really found evidence to be true.

Especially Memory playback. Hardware today is robust enough to retrieve the local audio files and play music without buffering or drop outs. This “me too” feature, in my opinion distracts, the Audiophile/Music Lover/Listener/Anyone by pointing to an issue that doesn’t have a meaningful impact. If it were a fact that not having this feature ruins Roon then I am all for it. I just do not see evidence to move forward. and I will certainly eat crow if the developers include it just to make the community happy-- that happened with DSD.

Myself, I like the Minimum Phase filter and use it in my Audiolab M-DAC / Squeezebox Touch used as a headphone amp with my Sennheiser HD650’s.

Right- Roon does this specifically with Volume normalization. However it doesn’t improve the recording, just matches the volume from track to track.