It’s GentooPlayer.
I see. Best to reach out to the author to find if they will consider supporting a kernel that supports disabling EEE. Perhaps they have already added a patch to one of the other kernels, for example.
I’ve gotta say this, this thread has been a splendid read - I never imagined my video playing the role it did.
More than anything else, @David_Snyder I love seeing the deep-dive with measurements to back up the subjective claims and then using those results to take the methodology as far as possible.
Since this thread is so focused on pushing the limits, I thought I’d share what I’ve been experimenting with recently (which I’ll hopefully be covering in a video soon):
- Fiber vs Copper: Using SFP between the host and target. I’ve been exploring whether the “specialized” hardware required for fiber yields a justifiable “increase” in subjective performance
- 10Mhz Master Clock: Integrating a 10MHz clock input on the host, target, and DDC/DAC to see if “syncing the chain” changes the Diretta noise floor.
- Protocol Updates: Testing Diretta 148/149 for its support of jumbo frames and DDS
I realize that not all of these are strictly “Diretta” changes, but it’s all in the same spirit of seeing just how much subjective performance can be squeezed out of the streaming system by addressing (in my mind) potential bottlenecks in the chain.
As David has already pointed out, I am hitting a hardware wall with jumbo frames on my Pi 4 Compute Modules, but it’s a minor hurdle in the grand scheme of things.
I am more than happy to share updates here as I make progress in my testing.
Cheers,
Matt - The Everyday Audiophile
4 Likes
Will be interesting to hear about your results. I’ve been using a Cybershaft master clock with my Uptone Audio etherREGEN switch and it’s made a small difference in timing. However, there are numerous other clocks in the system that do not sync up with this master. So, the change may all be in my head, as it were.
1 Like
Hi Matt,
Welcome to the Diretta tinkering thread.
Piero recently added the jumbo frames patch to kernel 6.12.59 (for AudioLinux). Since then, a number of us have been enjoying configurations enabled by jumbo frames (and “baby” jumbo frames) with a range of recent Diretta software versions, including 147_09, 147_13, and 147_14/_19 (the latest, as of this morning, is 14 on Target, 19 on Host).
If either Host or Target is an RPi4, my automation in Appendix 9 sets the MTU (only) on the point-to-point link to 2032 bytes. However, if both are RPi5, we can use the full 9000 byte MTU.
I tried inserting an iFi SilentPower LAN iPurifier Pro between my Host and Target but did not love the sound.
So far, the best I’ve heard for the point-to-point link is a standard Blue Jeans BJC CAT6a (bonded-pair Belden stock) with the shield disconnected on both ends.
I’m interested in your 10MHz OCXO clock experiment results and anything else you care to try. Thanks for joining this thread and being willing to share your experiences here and on YouTube. Cheers!
Welcome to the thread, but the deep-dive measurements do not back up the subjective claims. Indeed, quite the opposite:
2 Likes
Thanks! I think it’s great to see the audio community quantitatively test Diretta’s claims.
While I haven’t run any measurements, I can share some “blind’ish” results from about a dozen members of my local audio group across four different systems. Listeners knew we were comparing Diretta to another source but never knew which was active during the A/B testing.
The setup was very similar to the post you linked, with USB coming out of the host and target to a DAC that could be switched between inputs on the fly.
Even without strict scientific controls (we allowed for group discussion), the consensus was consistent: everyone reported a clear subjective difference across all four systems, preferring Diretta every time.
I fully understand that our testing methods may not have been perfect, but the clear group preferences were enough to validate what I heard on my own.
Anyways, I’ll still update this thread with my personal findings in the coming weeks. Or it’ll be in a video, who knows 
3 Likes
Given that there is no evidence that the Diretta protocol impacted the performance of the connected DAC in the measurements @Marian did, there is only one possible path to establishing that your group subjective experiences were real and not the product of biases/wishful thinking: your DAC has poor resilience in the face of potential noise on its USB input (I keep proposing psychic powers combined with quantum coupling but the group keeps pushing back!)
So when you test again, it would be great if you could replicate the same type of objective measurements with your DAC in the path. If there is an impact of Diretta that, in turn, translates into improved sound quality, it should be obvious in distortions and/or noise emanating from the DAC. If there isn’t, yet you can get a modestly scientific-like listening test outcome, then I get to re-propose my psychic powers idea.
This is very exciting for the psionic/spectral-ability community!
1 Like
So rude! Is this better?
Given that there is no evidence that the Diretta protocol impacted the performance of the connected DAC in the measurements @Marian did, there is only one possible path to establishing that your group subjective experiences were real and not the product of biases/wishful thinking: your DAC has poor resilience in the face of potential noise on its USB input.
So when you test again, it would be great if you could replicate the same type of objective measurements with your DAC in the path. If there is an impact of Diretta that, in turn, translates into improved sound quality, it should be obvious in distortions and/or noise emanating from the DAC. If there isn’t, yet you can get a modestly scientific-like listening test outcome, then there is a remarkable mystery that needs further investigation!
1 Like
While we ponder psychic powers, let me report about my extraordinary experience yesterday after updating both RPi4’s to the latest configurations, including 2032 MTU link, with David’s help:
Thrilling! The most natural clarity and 3D imaging I’ve ever enjoyed at home has made my self-delusion complete.
1 Like
Thanks for the shout-out, though I’m not sure why, and glad you are happy with your system! The incorrigibility of this belief system in the face of evidence is a fascinating micro-laboratory given how technically astute everyone is. I look forward to future developments, especially further tests to find some kind of evidence for why folks are claiming to have these experiences!
Hilarious, but let me assure you and those following along at home that the 3D imaging and clarity is very real.
A friend offered a test he uses. He moves the USB connection to his WiiM Ultra and listens for a few days. Then switches back to Diretta. He told me this is the best way to appreciate the impact a better transport has on how a DAC sounds.
Bits are bits? Sure. But the manner delivery seems to matter for most of us. Thanks, @Brad_Burnside.
4 Likes
Who is “us”? The people for which delivery matters? There’s a lot of “us” who think, based on bias-free testing, that it doesn’t.
1 Like
Maybe read through this thread, there’s plenty of ”us”.
1 Like
Given all this incorrigibility, I ask the following of the (“most of us”) proponents of the proposition that transport processing activities impact sound quality (let’s call it the “Diretta Improvement” (DI)): how do you think it does so and under what circumstances?
Here are two quandaries:
(A) If a DAC is connected via optical does that completely eliminate DI?
(B) Does a well-engineered USB submodule within a DAC eliminate DI? (per one possibility given the necessary limits of the results of @Marian)?
Finally, then, if (A) and (B) are true, is the only value proposition something like: if you have certain older, poorly-measuring DACs you may find some improvement by using Diretta, though you might try optical first before bothering?
1 Like
Without any “older, poorly measuring DACs”, I can’t answer your question. Mine are well-regarded, current models from S.M.S.L. (cheap) and Meridian (expensive) that produce greater detail when subjected to DI. Incorrigible.
1 Like
I’ve read each and every post in this thread. You probably don’t believe in measurements, so let’s try statistics. I looked at the most frequent posters (2 or more posts, as reported by Roon at the bottom of the thread) and tried to classify their position into 3 buckets: favorable, critical and neutral.
Total: 24
Favorable: 10 (42%)
Critical: 8 (33%)
Neutral: 6 (25%)
I tried to be fair and put people into the ‘neutral’ bucket if I wasn’t sure, so it’s not clear cut, but I don’t think the favorable side qualifies as “plenty” by any stretch. I’m frankly a little surprised this hasn’t tilted more towards “favorable”, considering David’s initial efforts to silence some critics, including myself.
2 Likes
Hi Marian. I appreciate the time & effort you’ve invested in evaluating this project. Really impressive. In counting up the contributors to assign buckets in order to determine the percentage of believers, however, you may be overlooking the “why” of a forum reader choosing to post, or not.
My own motivation is pretty obvious: I’m an audio evangelist (with nothing to sell) who is compelled to talk about over-performing, lesser-known bargains in a world of very expensive pinnacle products. I just love finding, and sharing, a great deal on something special.
The humble Raspberry Pi certainly qualifies here, as an army of inventive hobbyists works to make it do amazing things.
Most readers aren’t evangelists. They might try an idea if it’s easy and inexpensive, but may never post about it even if they are thrilled like I am.
Please tell us what your motivation is, and speculate about why others in the critics bucket are likely to post, or not. I’ve heard that people speak out to complain about perceived problems much more often than they take time to confirm their satisfaction…
3 Likes
I won’t speculate about other folks’ motivations at all!
I’m simply interested in the structure of ideas, reasoning, history and philosophy of science, engineering, modern technology, literature…um, well, pretty much everything. In this particular case with arguments about sound quality influence in bit-perfect transports and streamers there is a fundamental disconnect between the goals of engineers in providing perfect fidelity and a subset of the audio enthusiast community who strongly want there to be a secret, discoverable world of tweaks.
The psychology and sociology of these kinds of hopes is something like conspiracy theories. There is a resilience to data and demonstration. In fact, disconfirmation often leads to increased belief! The unique feature of audio-equipment belief structures is that one can always simply claim that the differences can be heard, regardless of any rational support for the claims.
I try a range of approaches to critiquing these ideas and try to always maintain sufficient humility about my own and others’ knowledge limitations. So when I propose quantum coupling to human brains (which is a real, quasi-scientific theory that has some experimental support) it is to show the fundamental explanatory gap facing proponents of something like Diretta.
It’s also just fun! 
One of my goals is to seed basic skepticism about outlandish claims so that when new readers come to these forums they see clear counterarguments that can help them navigate the truth or fiction of the ideas. I also just try to help people in other threads.
Ooooo, this is getting long but I’m just multi-tasking while watching football, which doesn’t take much attention. 
4 Likes
Thank you for your kind words.
I have no intention to question the reasons people post here and I don’t see how that’s relevant. I’m just interested in the merits of the opinions expressed, and unless there is evidence to the contrary, I believe they are genuine.
I don’t know what that means exactly. My intention is to fact-check what is being said here when I think it goes against evidence and/or logic. As I said numerous times before, we are talking about machines, designed and built by engineers based on scientific knowledge. If anyone wants to evangelize - as in, convince others of - something that goes against what is understood about the said machines, they have to offer more than biased sensory experiences. And when no explanation is offered or the explanations are not credible, I’m not going to hold back. There is no harm in that. O the other hand, allowing unchecked opinions to be presented as facts can be harmful.
I am 100% behind that. That doesn’t however detract from what I said. Cheap snake oil is still snake oil.
I couldn’t agree more. I never used any other streamer and I don’t intend to as long as the Pi is supported by whatever digital audio solution I use.
That goes both ways. I guess there are a lot of people who think fiddling with digital transports is a waste of time and never post about it, probably to avoid wasting time. I just wanted to show that the “most of us” who believe in “bits and magic” in this thread is not that overwhelming a majority.
Let me know if that’s not clear by now. As for speculating on other people’s reasons, I will do no such thing.
4 Likes