[Krutsch]

FYI ... the Roon remote app for iPad was released yesterday (last night?). The app is hard to find on the store (searching Roon doesn't yield the app), so use the download links from the Roon download page for iOS and Android.
 
The iOS version requires newer iPads, as does the Android version (which seems very restrictive; neither my Note 4 nor my older Nexus 7 are eligible).
 
I am running RoonServer an an old Mac Mini (2009) and it's blazing fast. My iPad Mini 2 beautifully replicates the desktop experience. You can tell this is a was design goal from the very beginning.
 
For me, this ends my search for an end-game music library, playback engine and browsing interface.

 

[Lohb]

  I compared the SQ of Vox to Audirvana Plus yesterday over the course of three hours and prefer the latter by a fairly large margin. It offered greater resolution, more natural timbre, greater rhythm and pace, and simply a more involving listening experience without any increase in listener fatigue. The former is free and good for what it is. Audirvana Plus is not free, but worth it. I do have a caveat; I did not get the chance to use it with a large library and have read that it is very slow loading one. 

Yep. 2.2 is superb, even if it was a paid upgrade for older owners of 1.5.x.
It all starts at the mastered file's quality and the player's decoding/handling quality of that before it gets pass up the line and for whatever reason, all these players have a quite distinct signature.

 

[Zoom25]

I decided to give Roon a try and have been comparing it to Audirvana Plus. I am very impressed by how clean Roon looks and navigates. I've yet to give their iOS app a try. I'm using Direct Mode, exclusive, integer on both software. Everything is optimized to the max on both software.
 
Soon looks like it's been designed by professionals, whereas Audirvana looks slightly amateurish. Although I've gotten used to Audirvana Plus and it works fine. I haven't tested Tidal integration either.
 
Sound impressions:
 
- Only based off HD 800 atm. Will involve more headphones and monitors later.
- Both are very close sounding. I'll easily take either over Amarra. 
- Three differences noted: bass, mids, soundstage depth
- Bass: Audirvana has slightly bit more heft in the lower octaves. It's honestly negligible. Both are very good at handling bass.
- Soundstage: There is slightly more depth on Audirvana Plus. The music feels more immersive on Audirvana. It draws me in more.
- Mids: This is where I am absolutely sure there was a REAL difference between them. Audirvana simply adds magic to vocals. Audirvana handles vocals with control that has been unbeatable so far on OS X for me. Roon is a bit dry and flat in comparison, however, with excellent resolution. It's just that Audirvana gives you the last 5%, the magical and involving 5%. The higher dynamic range and subtleness in vocals is where Audirvana Plus is golden. It transforms from a recording to hearing a real person without any digital artifacts.
 
These differences are very subtle. Either player would be fine sound wise and could live with. If I stopped using Audirvana Plus and continued with Roon, it would easily become my reference player. It's only in direct comparison when the differences can be truly observed.
 
I will most likely end up not pursuing Roon even though it's better in every regard minus SQ. It's just that at the end of the day I care about the sound more. Audirvana Plus is the only player so far that has made me feel actively emerged in the music. With all other players, I'm simply hearing it passively.
 
EDIT: The difference is VERY REAL. Going from Audirvana to Roon felt slightly weird but got used to it. Instead of going back and forth, I stuck with Roon for an hour listening getting used to the overall sound. Next, I wanted to try the same tracks (files) from various artists and albums on Audirvana Plus. This time around, the difference seemed much more pronounced. It was like several layers of veil was removed.

 

[joeexp]

+1 Similar experience I have got. Since Version 2.2, Audirvana is just in a class of its own. It even sounds better then Amarra 3.0 YMMV.
The iPhone app is fab for remote controlling! Although as you said the UI could do with some improvement. SQ is king.

 

[Zoom25]

I've been doing another experiment. It's one I've done a couple of times over the years (at least since 2006) on both Mac and Windows with different setups. It's comparing WAV vs FLAC (compressed). I have never found a difference between them in real time playback. They are both lossless, however, FLAC is compressed. Although there is FLAC uncompressed, which I look forward to add to the equation. The output should be the same as both are bit perfect copies of each other. However, the audio streams of various players (Amarra, Roon, and Audirvana Plus) are also bit perfect, as measured by developers themselves, yet many here can hear very clear differences. The developer of Roon actually tries to tackle this on Computer Audiophile.
 
For fun, I decided to take a few of my albums from my catalogue (which is 95% FLAC) and convert them to WAV. I used XLD for this. I batch converted 30 albums in less than a minute. It's ridiculously fast! I put all the files in the same directory on my Macbook Pro from which Audirvana reads.
 
(In the past I've talked to a couple of guys from Pioneer representatives at event talking about why they don't support FLAC on their CDJ's. This was around 2011. They also hinted at real time performance issues and that it's noticeable when played on massive top end club systems like Funktion One, which I can vouch for its sound. Nonetheless, I thought they were being somewhat snobbish. I understood their argument about real time performance and conversion, however, most computer guys would argue that decoding and processing of FLAC is negligible at this point with computers advancing. So in that respect, my pimped out SSD Macbook should make any negativity of FLAC decoding process even less than before, i.e. FLAC and WAV playback should be essentially the same. On the other side of things, having a very quiet and optimized computer might actually let you hear differences more easily. 
 
All the original FLAC tracks had various bit rates. However, their converted WAVs all had 1411 kbps, by default.
 
SysOptimizer on extereme, with direct mode, exclusive, and integer mode engaged. NO EQ, DSP, or audio units engaged. All OS X volume levels at 100%.
 
I have been testing to see if there any differences by two following methods. Letting an entire song play in FLAC and then in WAV, or vice versa and seeing if there was an overarching difference in sound. Another was to compare certain complex sections of songs, usually 5-10 seconds interval at max. Just go back and forth a good 50 times to see if I can pinpoint what I've been hearing. I've especially been focusing on decays, reverbs, and sustains on trails.
 
The results so far have been confusing and beautiful. I would've laughed reading this post a few years back. More (scientific) experimentation remains to be done. 

 

[Krutsch]

  ...most computer guys would argue that decoding and processing of FLAC is negligible at this point with computers advancing. So in that respect, my pimped out SSD Macbook should make any negativity of FLAC decoding process even less than before, i.e. FLAC and WAV playback should be essentially the same. On the other side of things, having a very quiet and optimized computer might actually let you hear differences more easily. 
 
...

 
Intersting post; thanks for taking the time to write it.
 
I performed similar experiments with my Bryston BDP-1, which is basically a low-powered AMD CPU -equipped, purpose built Linux box.
 
When I measured CPU utilization, it was eye-opening to see how close FLAC was to WAV and AIFF (which is basically WAV with a byte-swap required during decoding). FLAC was so close to AIFF that I decided to just stick with that format. Apple Lossless, however, required an order of magnitude more CPU to play, on average. MP3 and AAC also required significantly more CPU than FLAC.
 
If you believe that electrical noise caused by a higher activity CPU can affect audio playback (and I personally do), than FLAC is a great choice.

 

[Zoom25]

I will also add in AIFF and FLAC uncompressed to the comparison to WAV. I'm done comparing WAV to compressed FLAC for now. Results were fairly consistent over different tracks. I don't think this experiment would've been as useful if I didn't utilize Audirvana Plus. It removes veil upon veil.
 
EDIT: End of Day 1 Results
 
Format Preference based on SQ:
1) WAV
2) AIFF
3) FLAC (compressed)
 
Have yet to try FLAC uncompressed. Should throw in ALAC as well for good wholesome comparison. WAV is a nuisance in terms of storage and metadata! 

 

[Zoom25]

Figured out how to get metadata and tagging with WAV through Audirvana Plus when converting with XLD. It's a very simple step.
 
XLD -> Preferences -> General Tab -> Output Format: WAV -> Click on the "Option Icon" and check off the "write tags" box, and for Format specify "Both," and Encoding should be UTF8.
 
In Audirvana Plus, now both WAV and FLAC albums appear with same tags, i.e. the Audirvana Plus experience remains the same as before with FLAC album directories. For this particular computer setup, this negates the need for AIFF.
____________________________________
 
Regarding FLAC (compressed) vs WAV (of the same file), I noticed that when I click on the FLAC file there is a brief (1/100s) buffer delay which shows visually on Audirvana Plus with the line loading from left to right on the track bar. With ALAC, this process is even slower. WAV on the other hand loads instantaneously. You don't see a left to right movement where the line darkens, instead it appears as if the line simply turns on upon clicking. It's impossible for me to capture the motion. I should add the respective AIFF file to the mix, but I have currently deleted all of my AIFF set from previous day's set. Maybe later.
 
Just a random observation.
 
Day 2. End of transmission.

 

[Zoom25]

   
Intersting post; thanks for taking the time to write it.
 
I performed similar experiments with my Bryston BDP-1, which is basically a low-powered AMD CPU -equipped, purpose built Linux box.
 
When I measured CPU utilization, it was eye-opening to see how close FLAC was to WAV and AIFF (which is basically WAV with a byte-swap required during decoding). FLAC was so close to AIFF that I decided to just stick with that format. Apple Lossless, however, required an order of magnitude more CPU to play, on average. MP3 and AAC also required significantly more CPU than FLAC.
 
If you believe that electrical noise caused by a higher activity CPU can affect audio playback (and I personally do), than FLAC is a great choice.

 
Nice. How did it go with BDP-1? In the past I also tried a BDP-2, BDA-2, BHA-1 stack with LCD-3 non fazors. Everything connected to PS Audio power supply. Unfortunately, I wasn't able to get much out of the combination at that time. I think the BHA-1 and/or the LCD-3 held me back.

 

[Krutsch]

   
Nice. How did it go with BDP-1? In the past I also tried a BDP-2, BDA-2, BHA-1 stack with LCD-3 non fazors. Everything connected to PS Audio power supply. Unfortunately, I wasn't able to get much out of the combination at that time. I think the BHA-1 and/or the LCD-3 held me back.

 
I've personally never heard anything like it and I'm done with using a Mac as a source. It was game changing for me. YMMV.

 

[bixby]

   
I've personally never heard anything like it and I'm done with using a Mac as a source. It was game changing for me. YMMV.

WOW, just curious, did you do all the mega tweaking to the mac mini?  
 
You are not the first person who says they like the BDP-1.  I wish I could hear it.

 

[BirdManOfCT]

  [...]
 
The results so far have been confusing and beautiful. I would've laughed reading this post a few years back. More (scientific) experimentation remains to be done. 

Looking forward to hearing your results!

 

[Krutsch]

  WOW, just curious, did you do all the mega tweaking to the mac mini?  
 
You are not the first person who says they like the BDP-1.  I wish I could hear it.

 
I did... as much as I could (i.e. disabling Spotlight, Time Machine, extra processes, blah, blah, blah).
 
Don't get me wrong; I still use an upgraded 2009 Mac Mini on my 2-channel system on the living room, and it sounds good. I use it to run RoonServer and control playback with an iPad Mini 2.
 
I experimented a bit with different DACs, USB isolators and connections, finally settling on the optical out into a Wyred4Sound Remedy re-clocker, then into an NAD MDC DAC. Music is on a Firewire 800 G-drive, bus-powered.
 
TOSLINK seems to isolate the playback chain from electrical noise and the Remedy takes care of any jitter issues. Honestly, the Remedy is amazing - the imaging with speakers surpasses anything I was able to hear with a USB connection, even using Schiit Wyred or an iFi iUSBPower. Granted with a higher-end DAC, this may be less of an issue.
 
But, for my desk/Head-Fi system, it's all via the BDP-1 or the SACD player (I still like to spin optical plastic). Just no comparison, IMO.

 

[Zoom25]

Day 3 Results:
 
- Comparing FLAC (compressed) vs. AIFF vs. WAV
- All listening via Audirvana Plus with all system optimizer settings on max. No EQ, DSP, or Audio Units engaged - just like in the past
- I ditched the HD 800 and my DAC/headphone amp (Dangerous Source) and played music straight out of my built in Macbook Pro speakers and with KRK KNS 8400 speakers plugged out of the headphone jack of my Macbook Pro Retina.
- Have yet to try my Mackie monitors via Emotiva DC-1
- My preferences still held up in favour of WAV > AIFF > FLAC (compressed)
- In fact, the differences were easier to pick up on both the Mac speakers and the KRK KNS 8400 over HD 800 via Dangerous Source.
 
- Another interesting fact I should mention is that the sound benefits and changes in sound as a result of Amarra SQ (without EQ) is much easier to perceive without my DAC. When outputting to DAC straight from the System Preferences audio selection vs. having Amarra SQ send it to DAC, the difference in sound is much closer.
- However, when my headphones are plugged into the headphone jack of Macbook Pro, the effects of Amarra SQ vs. without Amarra SQ processing becomes much more easier to perceive. It's a day and night difference. With the DAC in place, the difference drops.
- I noticed something similar with Audirvana Plus as well with respect to difference in WAV vs FLAC vs. AIFF. Not sure why this is the case. I'll let others take a crack. Perhaps the DAC in my Dangerous Source is good at compensating, whereas the built in DAC on my Macbook Pro doesn't care much.
 
Sound impressions
 
FLAC - When I compared FLAC to AIFF and WAV, it sounded like I went from using Audirvana Plus to Roon or using Foobar2000 on my old Windows machine. The sound seemed further apart and less involving. It's a more passive experience for me. It's dynamically flatter and the vocals are somewhat strident. Vocals sound inherently choppy, as if there is a grain. Imagine a pure river being contaminated with very fine grain of sand that is not visible to the eyes. So when you drink it, even though you cannot see the grains, but you can still taste it. You know it's there, but can't prove it from simply looking (It's the best I could think of on the spot). That's how I feel about FLAC. It's still lossless with all the details, but just not as pure and SMOOTH.
 
AIFF - Vocals are smoother and vocals are practically clean, however, the sound is still dynamically flat. The soundstage is more involving than FLAC. The vocals are cleaner and you can hear everything cleanly, but they still sound flat. It's still not involving for me. Better than FLAC for sure, but IMO not worth it considering the gain in SQ isn't that drastic and WAV is right there at the same size with tagging available.
 
WAV - Fiji crack glacier water. This is where music sounds like proper music, rather than just hearing all the intricate details. This is where I get musically involved each time. Dynamics, micro detail, lack of grain, clean vocals, proper sustains and reverbs - it's all there. Soundstage is also more diffused and places you in the room where you can be active. WAV breathes life into the track and vocals. The resolution in vocals is enhanced to the point where you can see what shape the vocalist's mouth and lips are making when singing. This is especially true (and kinda hilarious) when listening to hip hop and rap because I can now guess the artists' faces in the studio while singing the song, by comparing it to how they sound in music videos, interviews, live performances etc. Really! It's finally gone to that point for me on a computer setup.
 
Summary:
- WAV is king in every sound aspect. Every track is more enjoyable and lifelike. It's impossible to not physically react. I have already started converting everything to WAV. I'm getting all the tagging and cover art like before with FLAC. Only drawback is hard drive space, which is not an issue in 2015. Mad props to Damien for taking Audirvana Plus this far.
- After becoming more familiar with all the formats and knowing what to look for, it's very easy to hear the differences between all the formats. The differences are very real for me! Also, the change in sound and presentation is very consistent in every track. This has become a regular pattern and impossible to ignore. Apparently, many others on Computer Audiophile have experienced the same with WAV vs. FLAC or AIFF over the past 5-6 years.
- Why WAV? - DYNAMICS. RESOLUTION. RHYTHM.
 
My next experiment should be to see if I can replicate the same results through a player other than Audirvana Plus.

 

[Zoom25]

   
I've personally never heard anything like it and I'm done with using a Mac as a source. It was game changing for me. YMMV.

Nice. My next purchase will also be a dedicated server/player/streamer. I have a few in mind ranging from the new Cambridge lineup (CXN, 851N) to Auralic Aries. In the past I had considered the BDP-2 and NAD M50. I still might consider BDP-2. I really want to audition a bunch of these streamers with my setup.