Jump to content
IGNORED

Audirvana 1.5 Beta


Recommended Posts

Having a powerful open format for audio filters is a great addition to A+. There are a number of possible AU plugins not yet mentioned here:

 

- Digital x-over

- HRTF

- Multi-channel down-sampled to stereo

- Simulated multi-channel for stereo or headphones

- Ambisonics

- Etc.

 

For all of these uses, it is desirable to have bit perfect data fed to the plugin. Garbage in, garbage out. Right?

 

The point of bit perfect, is not to simply have unmodified data hit the DAC chip. That is a valid approach, but it is certainly not the only one or even the most popular. After all, as Jud pointed out, most DACs up-sample the data.

 

In my opinion, the point of bit perfect is to not loose or change the data unnecessarily or without benefit. iTunes can change your data with no benefit to your listening. For example, it will down-sample if you have incorrect Audio Midi settings. This is the sort of data change that bit perfect is intended to address.

 

Bit perfect is not a life philosophy. It is a data processing methodology.

Link to comment

 

So, several of us A+ users would be grateful beyond words if A+ had a software means to force 24 bit depth signals in integer mode.

 

 

 

I've just released 1.4.9.7 that brings:

  • Audio Units presets load/save
  • Option to limit bitdepth to 24bit for bridges the offer both 24 and 32bit bitdepth!!
  • Fix issue leading to crash in Direct Mode on some Bluetooth devices
  • and other minor fixes

 

Remains the per sample rate upsampling settings to have the feature set for 1.5

 

It is available at the same place: Download Audirvana Plus BETA | Audirvana

 

Damien

 

 

 

 

enhanced-buzz-6522-1330456635-9.jpg

Link to comment
Problems found:

 

2) The sorting mechnism does not work - klick # or Artist and nothing happens.

 

I often have this problem with the A+ playlist. Sometime when I drag an album from iTunes it isn't ordered sequentially, and sorting via clicking the GUI does not work in those cases.

 

P.S.: SQ is much better than 1.4.6. Thank you.

 

P.S.S.: Also, I like that you added the selectable 24 bit limit. That may be helpful for people using digital volume controls be preserving the word length headroom for attenuation.

A Digital Audio Converter connected to my Home Computer taking me into the Future

Link to comment

1.4.9.x betas are 15x slower than 1.4.6 to load buffer

 

Perhaps this problem is unique to my DAC's Firewire driver:

 

In 1.4.9.6 and .7 (I did not test prior betas), the buffer loads so slowly, per the thermometer display in Audirvana, that sometimes playback overtakes the buffer and crashes the system! This happened with iZotope up sampling from 44 to 176, using the maximum filter length of 2 million samples. With all settings identical, 1.4.6 loaded the track into memory 15 times faster (32 seconds to load an 8 minute track at the end of the playlist).

 

Also, 1.4.6 fully utilized both cores of my dual core CPU, whereas 1.4.9.7 uses one core (per Activity Monitor).

 

My Mac Mini has 8 GB RAM. It made no difference whether I set the maximum buffer size to 3 or 5 GB in Audirvana. Activity Monitor showed at least 2 GB free memory, so that wasn't the problem.

HQPlayer (on 3.8 GHz 8-core i7 iMac 2020) > NAA (on 2012 Mac Mini i7) > RME ADI-2 v2 > Benchmark AHB-2 > Thiel 3.7

Link to comment
1.4.9.x betas are 15x slower than 1.4.6 to load buffer

 

Perhaps this problem is unique to my DAC's Firewire driver:

 

In 1.4.9.6 and .7 (I did not test prior betas), the buffer loads so slowly, per the thermometer display in Audirvana, that sometimes playback overtakes the buffer and crashes the system! This happened with iZotope up sampling from 44 to 176, using the maximum filter length of 2 million samples. With all settings identical, 1.4.6 loaded the track into memory 15 times faster (32 seconds to load an 8 minute track at the end of the playlist).

 

Also, 1.4.6 fully utilized both cores of my dual core CPU, whereas 1.4.9.7 uses one core (per Activity Monitor).

 

My Mac Mini has 8 GB RAM. It made no difference whether I set the maximum buffer size to 3 or 5 GB in Audirvana. Activity Monitor showed at least 2 GB free memory, so that wasn't the problem.

 

Playback has slowed down for me, too. Between selecting an album in iTunes mode and getting a song to play, like 20 seconds more for stuff to happen! PM did that, but now it's wait...wait...wait...wait...in A+ latest beta?

 

Sounds great...real slow with Mytek.

Tone with Soul

Link to comment
1.4.9.x betas are 15x slower than 1.4.6 to load buffer

 

Perhaps this problem is unique to my DAC's Firewire driver:

 

In 1.4.9.6 and .7 (I did not test prior betas), the buffer loads so slowly, per the thermometer display in Audirvana, that sometimes playback overtakes the buffer and crashes the system! This happened with iZotope up sampling from 44 to 176, using the maximum filter length of 2 million samples. With all settings identical, 1.4.6 loaded the track into memory 15 times faster (32 seconds to load an 8 minute track at the end of the playlist).

 

Also, 1.4.6 fully utilized both cores of my dual core CPU, whereas 1.4.9.7 uses one core (per Activity Monitor).

 

My Mac Mini has 8 GB RAM. It made no difference whether I set the maximum buffer size to 3 or 5 GB in Audirvana. Activity Monitor showed at least 2 GB free memory, so that wasn't the problem.

 

The difference is that the multichannel capability of your DAC is now used. This means much more work for the SRC to do upsampling, and this is what you have observed.

Now, if you use your DAC to feed only a stereo amp, the option "Use only as stereo" (Audio System page of A+ prefs) will revert the behavior as before: working only on two channels. And bring back the speed you were used to. I may look into making this automated for stereo tracks in the future though.

 

SRC, and most effect filters need to have their filters initialized to process signal. This means the previous input signal is used to produce the current output. And thus, the processing needs to be done in pure sequential manner.

This means for ensuring gapless playback, when SRC and/or Audio Units are in use, loading is made sequentially, as opposed to when none of these are in use where two tracks can be loaded in parallel.

 

Damien

MBP 15"/Mac Mini, Audirvana Plus, Audioquest Diamond USB, AMR DP-777, exD DSD DAC (for DSD), Pioneer N-70AE, Audioquest Niagara balanced/Viard Audio Design Silver HD, Accuphase E-560, Cabasse Sumatra MT420

Link to comment
The difference is that the multichannel capability of your DAC is now used. This means much more work for the SRC to do upsampling, and this is what you have observed.

Now, if you use your DAC to feed only a stereo amp, the option "Use only as stereo" (Audio System page of A+ prefs) will revert the behavior as before: working only on two channels. And bring back the speed you were used to. I may look into making this automated for stereo tracks in the future though.

 

SRC, and most effect filters need to have their filters initialized to process signal. This means the previous input signal is used to produce the current output. And thus, the processing needs to be done in pure sequential manner.

This means for ensuring gapless playback, when SRC and/or Audio Units are in use, loading is made sequentially, as opposed to when none of these are in use where two tracks can be loaded in parallel.

 

Damien

 

Thanks very much Damien. Will adjust prefs.

 

Hope you are enjoying success with this great product!

Tone with Soul

Link to comment

P.S.S.: Also, I like that you added the selectable 24 bit limit. That may be helpful for people using digital volume controls be preserving the word length headroom for attenuation.

 

Josh would the benefit only be realized on materiel that has higher bit depth or across the board? what's the trade-off?

Link to comment
Josh would the benefit only be realized on materiel that has higher bit depth or across the board? what's the trade-off?

 

Try up sampling something or playing a DSD file with this feature selected on your McIntosh D100 and see if it plays back at 24/176. When I played DSD with my D100 a few months ago it wanted to playback at 32/176.

A Digital Audio Converter connected to my Home Computer taking me into the Future

Link to comment
I have a 24/96 file up sampled to max it's playing at 32/192 and limited to 24 it's at 24/192 so it working right?

 

Nice. Nice. Yeah looks like it.

 

You should A/B compare 24 bit content with it on and off. I'd think the digital volume control would be much better with it limited to 24bit. That way you should have 6 bits (32-24) of headroom. Assuming each bit is = 6dB than that is 36 dB of attenuation before resolution gets impacted. On your MC452 amp that is about ~0.1W. Definitely let us know if you hear a difference with and without forced 24 bit at low listening levels.

 

P.S.: Make sure you upsample 2x or powers of 2x only. Not max.

A Digital Audio Converter connected to my Home Computer taking me into the Future

Link to comment

Won't be able to do any a/b listening as kids are invading my space. From what I'm understanding, I have to upsample most of my music to take advantage of this feature. The few times I tried upsampeling I didn't like it. Maybe I can set it up this way if I'm listening at low volumes.

Link to comment
Won't be able to do any a/b listening as kids are invading my space. From what I'm understanding, I have to upsample most of my music to take advantage of this feature. The few times I tried upsampeling I didn't like it. Maybe I can set it up this way if I'm listening at low volumes.
If your D100 is indicating 16 or 24 bit then no need. However when you play DSD files or use 2x up sampling it should be very helpful when using the digital volume control.

A Digital Audio Converter connected to my Home Computer taking me into the Future

Link to comment
The difference is that the multichannel capability of your DAC is now used. ...

Now, if you use your DAC to feed only a stereo amp, the option "Use only as stereo" (Audio System page of A+ prefs) will revert the behavior as before: working only on two channels. And bring back the speed you were used to. I may look into making this automated for stereo tracks in the future though.

 

That fixed it! Thanks!

HQPlayer (on 3.8 GHz 8-core i7 iMac 2020) > NAA (on 2012 Mac Mini i7) > RME ADI-2 v2 > Benchmark AHB-2 > Thiel 3.7

Link to comment
Not to quibble, brlawyer, but this lawyer thinks that your position is almost wholly undermined by the statement you excerpted from Damien's white paper. In sum, Damien appears to argue that bit perfect is the beginning but not the end of digital audio reproduction. Based on what you excerpted above, I'd suggest that offering plug-ins, designed to combat that which is "detrimental [to] the sound quality," is entirely consistent w/ his stated goal.

 

It's funny that you said that, because I thought about exactly the same point before posting - and I am positive that the logic contained in my message still stands.

 

Damien's white paper is clear in defending a very simple principle: that of non-adulteration of the source/music. The quote above further confirms this fact: A+ is supposed to make sure that whatever you play is subject to the least possible amount of interference, be it on a "bit-perfect" level or a "source" level.

 

Unless he is able to explain further, the use of AUnits DOES defeat that purpose. Nothing that causes me a headache of course, since I won't be using them anyway - I am sure many people appreciate the option; I just prefer to stick to the premise that moved me away from ordinary players in the first place.

Link to comment
It's funny that you said that, because I thought about exactly the same point before posting - and I am positive that the logic contained in my message still stands.

 

Damien's white paper is clear in defending a very simple principle: that of non-adulteration of the source/music. The quote above further confirms this fact: A+ is supposed to make sure that whatever you play is subject to the least possible amount of interference, be it on a "bit-perfect" level or a "source" level.

 

Unless he is able to explain further, the use of AUnits DOES defeat that purpose. Nothing that causes me a headache of course, since I won't be using them anyway - I am sure many people appreciate the option; I just prefer to stick to the premise that moved me away from ordinary players in the first place.

 

As already noted in my previous message, bit perfect, or even bit perfect with integer mode, is not the "least adulterated" way the signal might be sent through the conversion process. For the vast majority of DACs in use today, the least adulterated path makes use of the iZotope oversampling bundled in Audirvana Plus. This is paradoxical - an added feature creating a simpler path - but true. So your "logic" is faulty because you continue to insist on a demonstrably incorrect premise.

 

I said in my previous message that there were potential ways for A+'s new plugin capability to enable the same paradoxical trick. Consider, for example, a system with a sub, and/or active crossovers. (This is a slightly more generalized version of what Superdad talked about upthread.) The active electronics can be taken out of the circuit by the use of plugins, simplifying the electronic path. Meanwhile, the output from the plugin is dealt with by Audirvana Plus in the same way oversampled output is - converted to native integer mode for the DAC and loaded into memory. Thus the signal path from Audirvana Plus through the DAC remains as simple as possible.

One never knows, do one? - Fats Waller

The fairest thing we can experience is the mysterious. It is the fundamental emotion which stands at the cradle of true art and true science. - Einstein

Computer, Audirvana -> optical Ethernet to Fitlet3 -> Fibbr Alpha Optical USB -> iFi NEO iDSD DAC -> Apollon Audio 1ET400A Mini (Purifi based) -> Vandersteen 3A Signature.

Link to comment

I've just released the 1.5 beta 1.4.9.8 that brings:

  • Option to set upsampling per track sample rate
  • Sorting of the playlist by clicking on the header columns (standalone mode)
  • Option to prevent screen from going to sleep
  • Audio Units presets load/save also during realtime configuration
  • Fix mono files playback
  • and some bug fixes

 

It is available at the usual place ( Download Audirvana Plus BETA | Audirvana )

 

Damien

MBP 15"/Mac Mini, Audirvana Plus, Audioquest Diamond USB, AMR DP-777, exD DSD DAC (for DSD), Pioneer N-70AE, Audioquest Niagara balanced/Viard Audio Design Silver HD, Accuphase E-560, Cabasse Sumatra MT420

Link to comment
Option to set upsampling per track sample rate

 

:-) :-) :-)

Qnap HS-264 NAS (powered by an HD-Plex 100w LPS) > Cirrus7 Nimbini v2.5 Media Edition i7-8559U/32/512 running Roon ROCK (powered by a Keces P8 LPS) > Lumin U2  > Metrum Acoustics Adagio NOS digital preamplifier > First Watt SIT 3  power amplifier (or Don Garber Fi "Y" 6922 tube preamplifier + Don Garber Fi "X" 2A3 SET power amplifier, both powered from an Alpha-Core BP-30 Isolated Symmetrical Power Transformer) > Klipsch Cornwall III

 

headphones system:

Cirrus 7 > Lumin U2 > Metrum Acoustics Adagio > Pathos Aurium amplifier (powered by an UpTone Audio JS-2 LPS) > Focal Clear headphones

Link to comment

  • Option to set upsampling per track sample rate
  • Sorting of the playlist by clicking on the header columns (standalone mode)

 

Thank you good sir. You really have set the bar incredibly high with this build. The SQ is just superb.

Link to comment

Bug Report: The sorting by track # does work with the GUI list, but unfortunately playback still follows the non-sorted order. Forward (>|) & Previous track (|<) also follow the non-sorted order queue.

A Digital Audio Converter connected to my Home Computer taking me into the Future

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...