Clean power for your.... SSD/HDD...?

And I could claim that what you are providing is not logically sound and therefore misinformation (to paraphrase: that a claim without technical explanation/support is automatically confirmation bias.) It could well be that my observations were the product of confirmation bias, however that is not determined by an absence of presence of technical explanation or evidence. That only determines the degree to which the observation is supportable from technical argument or usable data. The extent of scientific knowledge does not place limits on the universe - only limits on what we can explain, measure and reliably work with.

So yes there are strong limits to the usefulness of claims without usable data or technical argument however anyone with half a brain should be able to determine this upfront and take this into consideration. The true misinformation would be to provide a half baked technical explanation, which is what I believe you are asking for.

 

Half baked explanations? There's nothing half baked, it's clearly understood of digital data transfer works. You not understanding does not make it magically possible for bits to not be bits, for error correction to not work, and for noise to be magically created where there is none possible.

 

Or the actual explanation that it's bias.

 

Can't you just put two identical music files on the SSD and HDD respectively then run those using foobar's ABX test, run it a few times and show how there's definitely a difference there? Even if subtle, if it's there and you can identify it then this shouldn't be a problem.

 

Actually that is exactly a half baked explanation - your argument is, if I understand correctly, that audible differences between storage devices are not possible because bits are bits, because there is ECC, and because differences in noise are not possible, and that all these claims are automatically valid and require no in depth explanation because "it's clearly understood".

My argument is that I should not attempt to provide a technical explanation, because chances are it will be as deficient as yours.

 

Yes, please do this Drez. Where do you live, I'm sure we can get someone to administer this test to you.

 

Yes, I think it's possible. And it's subtle.

 

To be honest, I have little confidence I could tell between drives in a blind test. What differences I believe I heard are very small (eg. between SSD and HDD and RAM drive) and in fact I generally don't even bother to use a RAM drive because the effort involved.

In my experience, the kind of small, subtle differences such as I believe I have observed with storage devices are small enough to escape detection with most ABX testing - it is hard enough to blind test DS vs multibit DAC's.

Honestly my advice would be don't bother with linear power to your SSD or HDD. Powering the whole system with a full linear ATX might be worth it with a dedicated music computer if you are crazy enough.

 

If you're really lucky, Tak, the older you get, the less certain you'll be about these kinds of opinions. If you're not, you'll become Analmort.

 

To clarify my original post - when I write "beyond subtle" I meant "more subtle than subtle" rather than "not subtle"

 

Lol right you are...

EDIT: " offer rebuttal to your post" fixed

 

Just a thought, but software like Jriver has the option to play files from memory (presumably the RAM?) instead of disk. This seems like the best way to remove our HDD from the equation. But I've never done an experimentation here personally.

It does seem to me that we want the point where we start transmitting digital audio real time and the point where it ends up (inevitably the DAC chip) to be optimized for performance, or else jitter and other timing issues will flow through the system.

 

Well I think it would be prudent to leave no stone unturned, if we want to embrace the whole open mind thing.

I think it's a good idea to compare different types of flash storage - TLC vs MLC vs 3D TLC. Different programming times of the NAND ought to change the invisible information that hides in the bits that aren't.

ECC vs non-ECC memory is a no-brainer, as well as RAID 1; too many bits getting flipped due to read errors could be a problem. Data can only last for so long on any storage medium before it is irrevocably corrupted.

Presumably RAM, but who knows these days. In any case, the argument seems to be that the badness can seep in even then; once the nasty, spinning HDD, switched-PSU corruption makes it into the data there's no getting rid of it. Scary, when you think about it.

Sure, jitter and noise shouldn't matter when the bits are still at the stage where there is no time domain, but that is just objectivist drivel; clearly doesn't line up with real experiences of real people.

 

Yep I understand that. By all accounts this falls under the "it should not make a difference but for some reason it appears to" category. I would like to think that if my ears disagreed with what some n00b on audiophile kool aid posted I would work it out. I have done so with many useless tweaks over the years.

 

Yes, that's what I suspected. Presumably if it's a real-time process any jitter induced by the HDD mechanism will flow through the system. Just like it would with a CD transport mechanism, no?