Audirvana Plus 3.5

[edtsui]

Would like to thank RunHomeSlow again for his advice on the playlist 'Added Recently' and how to edit it and other playlists.  I did not realize that the default 'Added recently' setting is 0 days.  No wander I could not see any albums showing. Once I change 0 to 30 days, all the newly added albums of the last 30 days show up.  Great!  Learning something new everyday!

[bottlerocket]

I successfully updated my license to 3.5, but when I tried to update the software it attempted a music library conversion.  That failed.   It raises the question if my disk is full?  But is not.   I winnowed my drive down to 347 gig used with 652 gig available.  Any ideas?  I did not read the thread so I might have missed the answer.

[Piano Man]

4 hours ago, copy_of_a said:

For those who think they don‘t need dither (like Damien) you could implement a switch to turn on/off dither as the last processing step. Easy.
And above all ... it was available in previous versions of Audirvana (iZotope MBit+). Unfortuately it was abandoned in v3.5...

That may be a question of money. In v3.5, the focus in on SoX. I guess the legacy iZotope SRC is not for free, and including MBit+ would be more expensive, too expensive for something that most of us presumably cannot hear.

[Piano Man]

By the way, looking at Damien’s post #782 and your post #869, there should be the possibility to use an external AU for dithering in real time mode. Didn’t that work? 

[jimdukey]

Does Dither matter when up-sampling?

I also use HQ, which has some dither options, but I never use them, as I upsample everything to DSD 128,

as I do on AV.

[jos]

On 2/12/2020 at 12:18 AM, copy_of_a said:

 

Just one thought... with regard to a common misconception about Dither:

 

Dither does not simply „mask“ distortion (truncation).

It REMOVES distortion!

 

The whole thing about Dither is to PRESERVE the harmonical content of the source when reducing bit depth.

We - always (!) - can only chose between distortion (audible) or noise (inaudible) when reducing bit depth.

Damien decided to go the distortion route 🙄 ... and herewith discards the accurate (and well documented) way digital audio should be processed 😞

 

Now ... I am so absolutely committed to Audirvana and it‘s faboulous sound that I‘ve meanwhile offline upsampled my entire library to 88.2kHz/96kHz and playback without upsampling in Audirvana.

Lots of work!

Sound is great!

 

Still ... what a sloppy/lazy design decision on behalf of Damien! Unbelievable!

Even more unbelievable that nobody cares ... on a forum where you‘ll find for instance hundrets of pages about the „sound“ of USB-cables (that often not even meet the USB 2.0 specs)...

 

I‘m really baffled (and extremely pissed) about this - both the designer and the bunch of unaware users - ... but at least I‘ve found my (personal) workaround.

 

Full stop.

Thanks, very interesting to read! I let my Chord DAC doing the upsampling.

[copy_of_a]

 

14 hours ago, Piano Man said:

By the way, looking at Damien’s post #782 and your post #869, there should be the possibility to use an external AU for dithering in real time mode. Didn’t that work? 

generally it does work. But running plugins in realtime mode also has its drawbacks (also with regard to sound quality).

 

14 hours ago, jimdukey said:

Does Dither matter when up-sampling?

absolutely! Whenever you apply any processing - for instance replay gain or a manual gain change - the source is no longer in tis original bit depth. It has to be dithered back to its original bit depth.

 

Isn't it funny (or maybe not so much) that we talk that much about workarounds and the usefullness of dither ... only because Audirvana does not provide dither ... what every other serious bit perfect player provides.

[Piano Man]

4 hours ago, copy_of_a said:

absolutely! Whenever you apply any processing - for instance replay gain or a manual gain change - the source is no longer in tis original bit depth. It has to be dithered back to its original bit depth.

 

Isn't it funny (or maybe not so much) that we talk that much about workarounds and the usefullness of dither ... only because Audirvana does not provide dither ... what every other serious bit perfect player provides.

@jimdudey The reason is that such processing is done at higher bit depth to minimise rounding errors. That also applies to upsampling or EQ, for exemple. Instead of just rounding or truncating off the bits which are not output, dithering introduces some random variation in the output signal. For example, if a calculated high precision signal slowly moves from 10 to 11 and the bit depth has to be reduced such that the output must be 10 or 11, rounding will switch from 10 to 11 exactly when the high precision signal reaches 10.5. Dithering will provide a varying output including 10s and 11s, more 10s at the beginning and more 11s at the end. After D/A conversion and low pass filtering that sounds better.

@copy of a: May I ask again whether you did a hearing comparison of dithering vs no dithering at 24 Bit output? Would be interesting to know whether according to you we discuss an audible flaw or a theoretical imperfection. Although I know that there is this psycological satisfaction to listen to music processed according to state of the art rules.

[AudioDoctor]

Audirvana 3.5 has recently begun doubling each and every song in my local library...

 

Has this been encountered before and is there a fix?

[RunHomeSlow]

19 minutes ago, AudioDoctor said:

Audirvana 3.5 has recently begun doubling each and every song in my local library...

 

Has this been encountered before and is there a fix?

 

Sometime ago i had that thing happening when looking in the info panel of Audirvana some of my files had the same path, but don’t know why a double was created that will not play… it just had a / before the name…

 

The only way i found a cure, was to delete my database here (you will lose all Playlists):
Go to User… Home Folder… Library… Applications Support… Audirvana Folder…
Trash All the files in that folder (that is your database of music, back it up! or not).

Restart mac, then Audirvana, then let it load your music folder, but i had to reload all my playlists

(that are saved elsewhere on my mac).

 

While writting, i think i have not done that, but just replace my database with an older one that i always keep as backup… that way, i didn’t lose my playlists… just lose the count play of files i did play between the 2 versions…

[RunHomeSlow]

Always backup your .sqlite database... every time you make changes in Audirvana tagging 

[AudioDoctor]

1 hour ago, RunHomeSlow said:

 

Sometime ago i had that thing happening when looking in the info panel of Audirvana some of my files had the same path, but don’t know why a double was created that will not play… it just had a / before the name…

 

The only way i found a cure, was to delete my database here (you will lose all Playlists):
Go to User… Home Folder… Library… Applications Support… Audirvana Folder…
Trash All the files in that folder (that is your database of music, back it up! or not).

Restart mac, then Audirvana, then let it load your music folder, but i had to reload all my playlists

(that are saved elsewhere on my mac).

 

While writting, i think i have not done that, but just replace my database with an older one that i always keep as backup… that way, i didn’t lose my playlists… just lose the count play of files i did play between the 2 versions…

 

Thank you, I did that. I will see if it fixes the problem. SO far it appears so, but it didn't develop immediately last time either.

[AudioDoctor]

This is just a suggestion for Damian about UI continuity. It would be nice if all of the columns were either, left or right justified, not a mix of both. It makes it look chaotic and unfinished.  Sound is still great though.

 

 

[jos]

On 2/7/2020 at 9:17 PM, jos said:

In addition: It's best to watch this video first, I did.....https://www.youtube.com/watch?v=Bh7GJvJmy1w

 

My first measurements and Sonarworks correction afterwards:

 

 

 

My first impression is that it sounds tighter and more balanced. I totally agree with the reviewer that using a stand would better the measurements. Next week I will try a second measurement.

 

 

As promised my second measurements:

 

The blue lines represent the measurements and the purple lines the correction. The measurements where better because the first time I was a bit to far in front of my sweet-spot (a couch against a wall). There are two correction lines now because I activated the listening spot correction. With this function in the off position I would get this:

 

In the on position (clicking 'Adjust To Listening Spot') both values will be 0.0 dB at the listening spot.

Nice to have is that you can adjust all options from within Audirvana! Before I did the measurements, I adjusted my speaker placement a little, but and the bass and tremble settings, which can be done behind my Revel F208 loudspeakers, as you can see below...

One speaker was set to Normal and one to Boundary.

The tweeter level was set to +1 on both sides.

I don't think it makes a lot of difference.

[copy_of_a]

On 2/13/2020 at 6:35 PM, Piano Man said:

@copy of a: May I ask again whether you did a hearing comparison of dithering vs no dithering at 24 Bit output? Would be interesting to know whether according to you we discuss an audible flaw or a theoretical imperfection. Although I know that there is this psycological satisfaction to listen to music processed according to state of the art rules.

I‘ve never compared side by side because I‘ve never listened to undithered audio when bit depth was reduced (to 24 bit or whatever) due to prior audio processing.
I’ve read quite a lot about dither and there are a lot of mixing & mastering engineers and other ‚experts’ (DSP designers and such) that do hear a degradation with undithered 24 bit audio while there are also a lot who do not care.
So for me it would be pretty time consuming to do a comparision as you apparently won‘t hear it at all in certain music but may hear it as a kind of grunchiness (or loss of smoothness and accurate imaging) in another.
For me it‘s really about peace of mind ... set it and forget it and enjoy the music.
This kind of confidence is somehow even part of the enjoyment - free of thinking about the technique and your equipment/workflow ... 

 

I only wonder why anyone would really find reasons to discard dither when it‘s beyond doubt the one and only accurate way that follows the rules (maths) of digital audio and that is always - in any usage case without a single exception - the better way... because it‘s accurate and ALWAYS at a lower level than the truncation distortion of the respective bit depth level.
And to be honest: I hate these kind of „you don‘t need it“ justifications only because a certain product is based on lazy design. I know this very well from other product categories and sooner or later the „this will do“ approach will be busted.
But hey ... since I am the only one complaining apparently all is good ... no need to incorporate dither (which would be peanuts, by the way...).

[buonassi]

On 2/11/2020 at 5:18 PM, copy_of_a said:

Even more unbelievable that nobody cares ... on a forum where you‘ll find for instance hundrets of pages about the „sound“ of USB-cables (that often not even meet the USB 2.0 specs)...

 

 

Yes!  Good point.  I'd of expected more of an "audiophile" outcry here.  Is it possible we're just in the minority of users who digitally attenuate?  I'm convinced my system sounds better using digital attenuation vs accepting the non-linearities of a traditional potentiometer (like that on my headphone amps).  Full tilt just sounds so clean and pristine when the 'pot' is essentially bypassed (resistance removed from the circuit).

 

7 hours ago, copy_of_a said:

But hey ... since I am the only one complaining apparently all is good ... no need to incorporate dither (which would be peanuts, by the way...).

 

I'm with you on this.  It's just generally accepted DSP principle.  Even my $7 USD android player software (Neutron Music Player) does this.  

 

On 2/12/2020 at 1:59 AM, Piano Man said:

- CPU load: I don’t know if dithering is a significant CPU effort as compared to upsampling, but Audirvana is all about decreasing computer activity as much as possible when playing music. Is it possible that dithering to 24 bit creates more noise by CPU activity than it removes?

 

I really doubt it.  the probability equation is an easy calculation for a computer processor.  It's nowhere near the calculation power needed for 4 or 8x upsampling. At least anecdotally this seems to be the case with my battery observations on my DAP (digital audio player).  I'm not 100% certain however.

 

To your point, @Piano Man , how audible it is (especially if you're only attenuating by a few bits) isn't the concern.  Just knowing that that there is empirically more distortion without dither, is.  That doesn't really sit well with an audiophile's mind. 

 

 

[jimdukey]

I look forward to trying YOUR Software Player.

When will it be available?

[mazuly]

To me Audirvana + sounds superb and in my opinion I think we are really nit picking.

 

If you actually can hear the difference, in real blind test, then that is really something we need discuss.  But in real blind testing can u actually hear the difference?

 

Real world is a business of compromise and I am sure there is a reason @Damianhad decided that it’s not relevant.  If it becomes relevant and can be heard in blind testing then I am sure he will incorporate it like all other things he has fixed for Audirvana +.

 

[copy_of_a]

10 hours ago, mazuly said:

To me Audirvana + sounds superb

agreed - as stated above.

otherwise I would use another player (I own quite few).

 

Real world is a business of compromise and I am sure there is a reason @Damianhad decided that it’s not relevant.  If it becomes relevant and can be heard in blind testing then I am sure he will incorporate it like all other things he has fixed for Audirvana +.

Inverted world!

Dither in conjunction with word length reduction does not only follow the theory and the maths of digital audio... it is also de facto standard.

So it would be Damiens job to provide evidence that - without exception for every audio track - dither at 24 bit is superfluous.

 

 

[jimdukey]

Or, it's YOUR job to move on.

If you already think AV sounds Superb, what is the point of your posts?

[copy_of_a]

9 minutes ago, jimdukey said:

Or, it's YOUR job to move on.

If you already think AV sounds Superb, what is the point of your posts?

I‘ve already moved on... months ago.

I have my workaround to deal with missing dither in Audirvana. All good. Peace of mind!

 

On the other hand: I am a long time user of Audirvana and once payed for built in dither.

I don‘t care about the fact that I‘d payed for it... only that it‘s missing today. I’d even pay again for an „upgrade“ with dither built in.

But still: think about that fact for a moment before you ask me to move on.

[jos]

I known, we are critical users, but I want to thank, praise and bless Damien for all to excellent work he has done over the years!! I enjoy 😊 Audirvana almost everyday and where would we be without his involvement supporting our love for the finest music reproduction there is? I’m quite sure version 4 will be even better than expected and I’m certainly willing to pay for further improvements. Hope to hear about his plans for future releases. 🙏

[jos]

On 2/14/2020 at 5:17 PM, jos said:

As promised my second measurements:

 

The blue lines represent the measurements and the purple lines the correction. The measurements where better because the first time I was a bit to far in front of my sweet-spot (a couch against a wall). There are two correction lines now because I activated the listening spot correction. With this function in the off position I would get this:

 

In the on position (clicking 'Adjust To Listening Spot') both values will be 0.0 dB at the listening spot.

Nice to have is that you can adjust all options from within Audirvana! Before I did the measurements, I adjusted my speaker placement a little, but and the bass and tremble settings, which can be done behind my Revel F208 loudspeakers, as you can see below...

One speaker was set to Normal and one to Boundary.

The tweeter level was set to +1 on both sides.

I don't think it makes a lot of difference.

I have to say: after a few days I enjoy my music system much more with the ‘Sonarworks correction settings’ integrated into Audirvana. It is just much more balanced/structured than before. I think it’s an even better investment than the integration of the EtherREGEN in my system for about half the price, although that was a very fine addition too (and a different one of course). I will make one more additional (final) measurement with Sonarworks with my low Frequency Compensation knob on the left side also set to Boundary, and perhaps a change to my Tweeter Level settings too. Conclusion: it’s  a worthwhile investment in most cases, I think! 😊

[thecrow]

a while ago i upgraded to audirvana 3.5

 

(just for the sake of it) i wanted to have a listen to audirvana 3700 or audirvana plus 2 versions that i never deleted from my imac however these would not open tonight

 

are they no longer accessible or become non accessible once the licence is upgraded to 3.5?

 

 

 

[yukio]

I found the pdf manual of previous versions of Audirvana very handy and that I still often use to understand more of this player. Why is it no longer published? Wouldn't you all find it useful?