[r010159]

I know I am resurrecting an old thread, but I do not know where else to place this post.  I recommend both Fidelia and Audirvana Plus. I find Audirvana+ to sound a little better than Fidelia. I think it is because of its "integer mode" that bypasses OS X internal drivers.
 
By the way, I found an interesting limitation of Audirvana+. Normally it buffers the entire track before playing it. When I used an Audio Unit plugin, the buffering only just managed to stay ahead of where the player was in the song.  When I used two AU plugins at the same time, there would be frequent dropouts. Next, when I played 24/96K audio files, there was stuttering for every little bit that it played.
 
What this means is if there is any problem with Audirvana playing a track, and you are using plugins, then look closely at them as the source of the problem. I do not have this problem with Fidelia.
 
Bob Graham

 

[edawja]

I use Fidelia Pro, the library system is less than good on the mac, but if you have the remote on an IOS you can at least search the Library by album.
 
For IOS i use AmpliFlac, loading Flac files is not straightforward, but once sorted its a nice player.

 

[brhfl]

  I know I am resurrecting an old thread, but I do not know where else to place this post.  I recommend both Fidelia and Audirvana Plus. I find Audirvana+ to sound a little better than Fidelia. I think it is because of its "integer mode" that bypasses OS X internal drivers.
 
[...]

 
I'm genuinely curious how any bit-perfect playback system can sound any better or worse than any other. By nature, bit-perfect output should guarantee that the exact same data is getting to your DAC.

 

[edawja]

As far as I am aware Fidelia Pro hogbacks the OX internal drivers.
 
I am no technician, but I have just sold my Wadia 830 cd player, which sounded good, because to me Fidelia Pro sounds better.
 
I wish i knew the answer about bit-perfect playback, i'm sure it would save me a lot of cash.
 
The Fidelia plays mainly over my loudspeaker system!!! hope that doesn't get me into trouble!!! But it sounds fabulous on my B&W P7' and  my Grado GR10's.
 
I haven't got the FHX, but would be interested to hear any thoughts.

 

[brhfl]

  As far as I am aware Fidelia Pro hogbacks the OX internal drivers.
 
I am no technician, but I have just sold my Wadia 830 cd player, which sounded good, because to me Fidelia Pro sounds better.
 
I wish i knew the answer about bit-perfect playback, i'm sure it would save me a lot of cash.
 
The Fidelia plays mainly over my loudspeaker system!!! hope that doesn't get me into trouble!!! But it sounds fabulous on my B&W P7' and  my Grado GR10's.
 
I haven't got the FHX, but would be interested to hear any thoughts.

I didn't exactly mean to call you out on that question, it just puzzles me hearing people ascribe sonic differences to software that is feeding the same bits to one's DAC. I can see if there's upsampling or the like going on, and output therefore is not bit-perfect, and whatever filtering is in place in the upsampling process agrees favorably with one's ears. 
 
I'm nearly always listening over (nearfield, but still) speakers as well... no harm, no foul!

 

[r010159]

[question regarding bit-perfect audio processing]
 
Yes, I thought the same thing. I believe this is true for most music players. The algorithms are well-established for the processing of audio signals. For one thing, I think some players perform additional processing on purpose to make the result *sound* better, but not actually be better. For instance, certain high frequencies can be emphasized to give the sound an appearance of being cleaner and more detailed. The bass frequencies can be processed to provide for a "warmer" sound.
 
Also there are well-documented problems with Core Audio and its drivers for Mac OS X. The "integer mode" from what I understand bypasses Core Audio and even the lowest HAL levels of the drivers.  My player directly interfaces with the DAC allowing for more efficient and reliable processing of the sound data. Integers are passed to the DAC instead of floats. From what I understand, this means no distortion due to rounding errors and a better control on the latency (jitter) of the audio. Also anytime the signal is processed, like you previously stated, how the audio is up sampled, dithered, and filtered can influence the reproduction of the audio.
 
I do not know the net effect of these optimizations. But now there are no pops and clicks and the music *sounds* more detailed and pleasant, for lack of better words. And the apparent distortion due to clipping of the audio has definitely improved. So there is some filtering going on. But I have not performed any *blind* A/B testing to absolutely prove this to myself.
 
Bob Graham
 
PS: I do enjoy the Fidelia player, which was my first purchase.

 

[brhfl]

Jitter issues are about the only (potential) thing I've heard regarding this matter, which boils down to a lighter CPU load, which... by all accounts doesn't seem that heavy to begin with. I'm not sure I necessarily want CoreAudio resampling on the fly, so I can see possibly implementing a better driver for that, but otherwise, if you're sending bit-perfect audio through CoreAudio... you're getting bit-perfect audio. So bypassing CA for jitter is the main potential benefit here? Not trying to start a big thing, and I know this starts heading toward the muddy path of audiophile USB cables, it's just very curious to me.

 

[r010159]

Don't forget the additional signal precessing that can be taking place. Sure, most of it is used to make the music *sound* better. But there is definitely filtering going on here. You see, I was not expecting this to happen. My ears are bothered by some high frequencies. This particular passage in the music was sounding uncomfortable too my ears. Upon someone else's suggestion, I pulled the track into an audio editor. There was clipping during this and other similar passages.  Later in time, I just happened to notice the clipped passage sounded much better to me. I did not understand this until I realized the music was being played by the Audirvana music player instead of the previous one that I used. I went back to the original player and found the problem was still there. This is the big reason I switched. I find that clipping is not uncommon with recorded music. It also appealed to me the occasional pop and click disappeared. When I also found out that the player handles the DAC directly, this appealed to the purist in me. So I decided to stay with Audirvana.
 
Bob Graham
 
PS: CPU availability to the driver is not the only thing that can affect the efficient, reliable processing of the audio data. This I discovered when I wrote my own devices driver on a project with Windows. Professionally speaking, direct handling of the hardware is always preferred where efficiency and reliability are concerned.

 

[brhfl]

If there's additional signal processing taking place, then we're not talking about bit-perfect audio, and I fully admit that at that point there are plenty of filters and colorations and whatsits that could easily make music sound better or worse to one's ears. I was more wondering about the bit-perfect PCM audio file -> DAC, pure, unchanged data stream that (many? most?) seem to strive for. 

 

[r010159]

  If there's additional signal processing taking place, then we're not talking about bit-perfect audio, and I fully admit that at that point there are plenty of filters and colorations and whatsits that could easily make music sound better or worse to one's ears. I was more wondering about the bit-perfect PCM audio file -> DAC, pure, unchanged data stream that (many? most?) seem to strive for. 

 
The defects with Apple CoreAudio are well-documented, which do result in pops and clicks among other abnormalities. This is particularly true with Maverick. Passing data to a DAC in its native form (integers) can also be a benefit. But beyond the effects that I feel reasonably certain are there, I do not know what else can be happening to the softwares attempt to preserve bit-perfect audio and its timing. I am thinking for some it may be more a goal than a certainty. FWIW

 

[joseph69]

I use a MacBook Pro with either the Schiit Modi or Meridian203 D/A Converters. I set my MIDI set up to integer mode.
By doing this I believe when I'm using either D/A Converter either bypass the internal clock/converter on the MBP, is this correct???
Thanks.

 

[r010159]

Integer mode was taken away for a couple of versions of OS X, now it is back. All I can say is that I would not be surprised if this feature allows the data to bypass a layer of CoreAudio. Just my opinion, of course.

Bob Graham

 

[joseph69]

Integer mode was taken away for a couple of versions of OS X, now it is back. All I can say is that I would not be surprised if this feature allows the data to bypass a layer of CoreAudio. Just my opinion, of course.

Bob Graham

By "Core Audio" do mean the MBP's converter?
If yes, by "layer" do you mean it doesn't bypass the entire MBP's converter?
Thanks.

 

[r010159]

  By "Core Audio" do mean the MBP's converter?
If yes, by "layer" do you mean it doesn't bypass the entire MBP's converter?
Thanks.

 
I am not familiar too much with the architecture of CoreAudio on the Mac. Most of what I said I am passing on from an engineer who is more familiar with this aspect of OS X. However, I do have a good idea about device drivers and how system services are presented to the software running on the computer, in this case an audio player. So I am making an educated guess here. I would not be surprised that integer mode does bypass some of the processing involved, because this approach should not require additional conversions between different types of number formats before the end result is passed to the DAC. But I doubt a system setting like this one will bypass the entire audio processing stack of the Mac OS. That is why I use the term "layer", meaning only some software of the OS is probably bypassed. And this is why I like what is termed as "direct mode" and "integer mode" on Aurdirvana. Supposedly Audirvana goes directly to the DAC and feeds it numbers in integer format.
 
Would this result in audible differences? According to one EE that I know tells me that the bypassing of CoreAudio completely will have audible benefits. But he has not addressed the benefit of integer mode with me. Perhaps one more familiar with CoreAudio should comment here about the intricacies of this part of Mac OS. Who knows. I may have put the proverbial foot in the proverbial mouth.  
 
Bob Graham

 

[joseph69]

   
I am not familiar too much with the architecture of CoreAudio on the Mac. Most of what I said I am passing on from an engineer who is more familiar with this aspect of OS X. However, I do have a good idea about device drivers and how system services are presented to the software running on the computer, in this case an audio player. So I am making an educated guess here. I would not be surprised that integer mode does bypass some of the processing involved, because this approach should not require additional conversions between different types of number formats before the end result is passed to the DAC. But I doubt a system setting like this one will bypass the entire audio processing stack of the Mac OS. That is why I use the term "layer", meaning only some software of the OS is probably bypassed. And this is why I like what is termed as "direct mode" and "integer mode" on Aurdirvana. Supposedly Audirvana goes directly to the DAC and feeds it numbers in integer format.
 
Would this result in audible differences? According to one EE that I know tells me that the bypassing of CoreAudio completely will have audible benefits. But he has not addressed the benefit of integer mode with me. Perhaps one more familiar with CoreAudio should comment here about the intricacies of this part of Mac OS. Who knows. I may have put the proverbial foot in the proverbial mouth.  
 
Bob Graham

Thank you very much for your input, this is what I had thought you meant, I just wanted to be sure I was understanding you correctly, thanks again.