Modo stuff:
I put a stop to sunjam's hijacking of this thread to post the exact same stuff he's been proved to be dishonest about in the other thread.
If a square wave is recorded with a proper band limiting, it then would have the ringing looking wiggles, and that's how the output should be if it tries to maintain fidelity for the recorded signal.
The cherry-picked line from Monty is something where I already showed that Monty gives the answer just a few lines below when he stops talking about digital domain and discuss the imperfection of a reconstruction filter. Yet here it is again, quoted out of context on purpose to make a guy look bad.
The 1.000 and 1.001 in the above post is yet another thing that he already tried and was proved wrong about, but here it is again, of course out of context because otherwise he's the one looking like a fool.
It's a level of manipulative dishonesty that I personally had never seen on the forum.
Anyway.
@knownothing2 . Sorry for even giving the guy a chance and turning your thread into a mess. Hopefully, now those who have information can share some.
Castleofargh stuff:
I don't know what Robs machine really does. IMO, upscaling seems weird and not weird at the same time for audio. Scaling is a common term for signal processing, it's applying a multiplier on the time axis. And if you then play that at a sample rate also modified by the multiplier, you get the revolutionary technical wonder we casually all understand as resampling.
What we know is that resampling is not regenerating data lost when sampling. That's not a possibility.
But there are various uses to resampling, some that do lead to improved fidelity, like why most DACs oversample a great deal. I often wonder about the merits of doing it before the DAC itself, and which DACs are really fine with that.
Historically, I only considered doing that to a NOS DAC, which in itself felt strange because if you understand the merits of oversampling, you wouldn't get a NOS DAC in the first place IMO.
Maybe I'm missing something. Anyway, the audibility question can only be answered by an actual listening test. On that, I really don't see the point of asking people what they think after doing it wrong. It has no factual value.
I also often consider an underlying question when audibility is established, and that is to confirm the cause instead of making one up on based on vague correlation. Rob's a sort of model example of that issue. He talks about doing things that even he recognizes shouldn't be humanly noticeable. But when he concludes there is an audible difference anyway, he doesn't look for other causes, like his test being flawed, or other causes that are within the domain of humanly audible changes. No, instead, he concludes that there is more to learn about human hearing. Which is when it becomes clear who Rob is and who he isn't. He is an engineer, he is a very active marketing guy, he isn't much of a researcher. I "watched" a conversation about research in astronomy and the statistical level of confidence they require before bragging about discovering something. It's incredibly difficult for them and they spend much more time checking for errors that could give them the special results, than they do getting the data.
Most businesses don't need anywhere near that level of rigor, and for us in audio, a good example of that would be the metanalysis about hires being audible. With just barely over guessing stats, they conclude in favor of an audible difference. Just about any different choice in the selection and evaluation of the papers used in the analysis would have altered the stats in a way, bigger than the margin for the entire final result being called a positive one.
And that still counts as somewhat serious research in our audio world.
It comes down to what it always comes down to, what do you consider enough to call something a fact? I tend to stick with IDK and I don't care until I get clean, well documented data convincing me to give a damn.
But in this hobby, the consensus is that the amount of data does not affect the number of conclusions being reached/jumped to. You know everything on a question, and you conclude something. You know nothing at all on a subject, can't actually test it in a conclusive falsifiable way, and you conclude something nonetheless.
It doesn't exactly elevate our hobby or make me trust people in it. That's just my general point of view. Again, I don't know that particular magic box, maybe it solves a problem the already expensive DAC had, maybe his method of spending more money and computational power into the things that most brands and chip manufacturers have worked hard on to save money and use simplified stuff that still were fine for human ears, gets the predictable improvement it can get. Is it audible? Again, not something we can actually answer with opinions. And of course most people will anyway because our hobby isn't a very serious one when it comes to facts.
