Plugins and Upsampling. Wrong order

Upsampling apperas before plugins. Should be processed after plugins.
Should be: Dirac-Plugin → Upsamling to 784 → Chord DAC
What is: Upsampling (max 192) → Durac Plugin → Chord Dac

The order of processing is as expected (see screenshot)

I use it this way for all PCM playback:
Playback buffer → HRTF (AU plugin) → r8Brain (modulating all PCM to 5.6MHz DSD128 via DoP1.1) → Output to DAC

:notes: :eye: :headphones: :eye: :notes:

1 Like

Mine is like that picture also…

@Ralf_Napierski you have a picture to show us
or you can see something that we don’t ??
or that graph preferences is just a show?

1 Like

Thank you, but the order on the sreen is the same. The order in audirvana is wrong.
I am using Dirac plugin. It allows max 192 kHz input. If i upsample in audirvana to 768 (for my Chord Qtest) the Dirac plugin does not work. If I upsample to 192, it works. So the chain is upsampling → plugin → output

Maybe when the DAC is set at 768 for upsampling, before playing a song, the plugin will see it also and won’t work??

@Damien or @Antoine will be best to answer this.

The plugin should receive an audio stream prior to upsampling. Subsequent operations should not impact the plugin. I believe Audirvana utilizes the plugin(s) before the upsampling process. The developers would have the definitive information on this.

In fact, concerning the order displayed on the screen, it is a bug.

This is my experience… The signal is buffered, presented to the Plug-in architecture and then sent to the sample-rate converter. for output… I know this because when I feed a PCM 352.8kHz file into the Plug-in architecture using an AU plug-in that is certified to 96kHz PCM… some of the parametric functions of the 96kHz plug-in are not available due to sample-rate variance. After passing through the plug-in, the 352.8kHz PCM file is then modulated to DSD128 (5.6MHz) and delivered to my DAC as DSD128…

Fact: The plug-in architecture is unavailable when playing DSD files…

(Note: my screen-shot above was taken without my DAC connected, and using the macOS default output device(s) so r8Brain was deactivated in this playback scenario)

:notes: :eye: :headphones: :eye: :notes:


Of course you know that the Qutest signal path is:

1FS to 8FS input > 16FS WTA1 filter (49,152 taps) > 256FS WTA 2 filter > 3rd order 2048 FS filter > pulse array noise shaper at 104 MHz > analogue out…

If you send the Quest 768kHz you will bypass that WTA1 filter… Do you really think that r8brain is better than the WTA1? Or do you have something else up your sleeve? :sunglasses:

1 Like

R8brain seems to do quite nicely, see for example its passband or transition graphs vs. an ideal filter here:

Are there similar graphs/measurements involving the internal filtering of the Qutest?

In the context of this all, does it really matter? You won’t be making the ultimate comparative playback assessment… @Ddude003 is posing a juxtaposition that may have not been considered, with the question: “Do you really think that r8brain is better than the WTA1?” … Do you know otherwise about the OP’s grasp of the Chord DAC design?

:notes: :eye: :headphones: :eye: :notes:

1 Like

Do you ever get the feeling that Jud is stalking you? I have had enough of him and ignored and muted him… He will have to do his own due diligence on Chord products…

1 Like

Think you are being a little over sensitive in this case. The @Jud post was merely saying that the R8 brain upsampling algorithm is pretty good. Bypassing the WTA1 and upsampling with R8 brain results to my ears in a very subtle difference in the sound arriving at my Focal Utopia headphones. As an owner of originally a Mojo2 then a Hugo2 and Hugo TT2 I have upsampled with R8 bypassing the 1st WTA stage and also left the upsampling entirely to the Chord DAC, the conclusion was to me personally that although different I have no significant preference, as such (after spending a considerable amount on Mr Watts designs) best to turn off upsampling and listen to the PCM file in the way Mr Watts and the Chord team wanted me to hear it.

Of course that is my perception. There is never a definitive answer in any A-B comparison, charts are meaningless, technology is meaningless, the only thing that matters is what the individual perceives is the best to his ears and brain. If it sounds the best to you, who cares . I assume the OP prefers R8…


I don’t believe this was the rationale for the post… Why did it matter in this context? The OP would make the assessment regarding the contextual value of @Ddude003 's question and insights.

I understand your comment but reading a rationale into peoples posts is prone to misinterpretation. Making things personal in posts tends to detract from the value of the forum. Just my opinion…


Many derailing threads getting boring here now on this site, better not answering back just to make it stop, because it will never stop :roll_eyes:


Agreed, I have seen this too… Feel a need to make the comment once though.

1 Like


1 Like

Well, Yes… It’s a matter of interpretive and connotative semantics… Based on the logic you pose above, we can say that it was presumptive of @Jud to think some additional information was needed to defend the OP’s employment of r8Brain in juxtaposition to @Ddude003 's question and insights in his post. As If the OP is incapable of making an assessment themselves… And it is presumptive of you to make the statement:“I assume the OP prefers R8…”

There was no reason to assume or presume anything until it is revealed as a relevant issue or concern by the OP… If the OP has an opinion, it is will be made clear, but we needn’t make judgements veiled as advice, in my opinion… Your experiential insights are plainly more relevant to the question regarding r8Brain and Watts Transient Alignment (WTA) technology… We know nothing about the OP’s reasoning for employing r8Brain in their playback scenario. And the question posed regarding WTA allows for the OP alone, to respond with a rationalization, based on their personal understanding of why that question was posed.

I totally agree with this perspective and this is why I responded with that same sentiment, to @Jud 's post:

1 Like

Audirvana claims to operate in a specific order, but in reality, the order is different. If I want to test software upsampling and bypassing Chord filters together with Dirac (or any other plugin), I can’t.