Excellent! All this work should have been unnecessary but it contributes to the community. You can console yourself that at least you learned something about everything from new protocols to the fine details of testing methodologies. Your playbook will help others investigate when confronting novel theories and devices, and ultimately it contributes to building a resource base for countering unwarranted and untested claims.
As a parent of a Kid A, as well, while I am disappointed that this does not support my quantum-consciousness-Diretta-paranormality theory, at least you got some time with them (and pyrokinesis is best achieved with a long fireplace lighter).
Now you just need to run a conclusive test to establish exactly which shade of red produces the most pleasing sound. And would painting cables lift even more veils?
As a matter of fact, I did buy two long orange CAT6 cables, since the one I used in the first two rounds between the boxes wasnât reaching the switch, and I wanted to keep the same looks as much as possible, and I only had black long cables. They didnât check to see where the cables ran, and they also didnât seem to notice the Ethernet dongle changing positions and then going away. It canât be said I didnât try my best.
A big thanks for deducting a thorough test! I am a bit disappointed that the Kids didnât find an obvious preference, but that likely doesnât come as a surpriseâŚ
I tried finding the answer to my question here, but didnât succeed:
When measuring i suppose you addressed each renderer in is singular form, using Roon?
And when the Kids listened, was the RAAT and Diretta route grouped or was the Zone changed in Roon as well as the USB cable connection?
During both measuring and listening, the output device in Roon was switched whenever the USB connection was switched. There was no grouping involved. I made sure the names of the devices remained consistent with the colors of the boxes in all rounds.
I donât see mentioned what Host setting you used or if you configured the Host and Target to proper IPv6, and set CPU and IRQ isolation. A base level configuration of RPi5 will not provide that much. If your serious about testing take some time to optimize your system. It will provide significant improvements over RAAT. Diretta assumes that:
⢠CPU topology matters
⢠IRQ placement matters
⢠Memory locality matters
Rather than hiding these realities, Diretta exposes tuning hooks so advanced users can align:
⢠Network interrupts
⢠Worker threads
⢠Feedback/control loops
When properly configured, Diretta tends to produce:
Stronger image stability
More intelligible lowâlevel detail
More natural decay and ambience retrieval
Reduced listening fatigue
Importantly, these traits do not come from coloration or DSP. They emerge from reduced time-domain error and lower control noise.
I faithfully followed the instructions @David_Snyder provided at the time. I believe the points you mentioned were all addressed there. You can find them in the thread linked to in the opening post.
Thatâs not what my experiment found.
It doesnât seem to me Diretta itself exposes those hooks; rather, the OS does. Itâs possible Diretta is sensitive to them as you mentioned, but, as @Suedkiez points out, any sensitivity to the streamerâs hardware and software or to factors unrelated to the audio payload is a liability, not an advantage. You might as well say that Diretta is sensitive to streaming, which is one thing you can never do away with.
I want to take this opportunity to make it clear that Iâm not going to repeat my experiments every time Diretta or anyone else comes up with new tweaks or recommendations, be it in the protocol itself, the OS, or the hardware. The setup I used was already supposed to âprovide significant improvements over RAATâ. Since I found zero evidence of that, I have no reason to believe there are any veils to remove on the digital side. This tweaking will never end, and will probably be used, ad infinitum, to âproveâ that whoever doesnât hear any differences must have missed some obscure but crucial tuning.
But you have taken the position that youâve busted the myth without an understanding what Diretta is and isnât, and broadcasted that out. Iâm not saying you should re-do your tests, just that your methodology is flawed.
mjw
(See you in the major leagues, Jack.)
Split this topic
80
First, Iâm pretty sure I know what Diretta is and does. Iâm not sure which part of my experiment suggests otherwise. Then, what kind of understanding do you think one requires to perform listening tests? And finally, itâs not only the listening part that busts the myth, itâs also the measurements, both on the power rails and, most importantly, on the DAC output.
2 Likes
mjw
(See you in the major leagues, Jack.)
Split this topic
93
Thanks so much for your efforts Marian! This is the first thread of this sort that I have ever encountered when one of the participants went as far as you to test the hypothesis being discussed. There has been a lot of opining in this discussion but your fact finding effort is monumental in comparison. Thanks also to David, who I believe started the thread in a sincere effort to share something he thought he had discovered. I see he is going forward continuing to work on the diretta concept. I hope he succeeds. Maybe there is something there but the lesson I take from the evidence presented is that any improvements so far are likely to be within the self-deception event horizon and therefore not worth the trouble for most of us to pursue. Good on all the other participants who almost never descended into the ad hominem attacks so common in public discourse these days.
There is a phrase from entrepreneurship, âdue diligence,â that embodies the process of evaluating whether value claims are true or not. With sites like ASR there has been increasing interest on the part of audio vendors to make certain that their products have exceptional value and that their marketing is truthful so the end listener gets what they think they are getting.
In the case of Diretta, I recommend the advocates learn from the exceptional efforts of @Marian and invest in the modest tools needed to perform sufficient due diligence of their own claims (no AP or Klippel needed) so they can be confident and crisp that they are adding value that goes well-beyond the âself-deception event horizon.â
Okay, Iâm shutting this down, @Karl_Karl. I gave you a gentle nudge recently, and you have continued to post provocative messages with no clear intent. This is trolling.
