You are missing the point I’m afraid. The problem is not that systems are not capable of playing high frequency content like 60kHz.
The problem is, that that content generates additional and avoidable intermodulation distortion in real-life devices, in the sub-20kHz spectrum. Do you understand this sentence, or not ?
(Normally no filtering on the output is applied by the way, but let’s call this not too relevant. It does make me wonder about the supposed benefit of high sampling rates, though…why encode stuff, which you’ll be filtering out eventually?)
That is strange. The very article that is linked, shows a very clear measurement of this. I’ll link you directly to it here : https://people.xiph.org/~xiphmont/demo/neil-young.html#toc_1ch . You can clearly see sound energy below 20kHz, on occasion at -70dB. Which should not be there at all, given the signal is only at 30 an 33kHZ.
(There are many more sources like it, and if you appreciate, I can reproduce such a measurement for you).
I would be interested to read something that contradicts point 3. If you have something to share, please do.
Just to be sure : don’t feel like you’re being attacked. I’m really not doing that.
I understand what you are saying, but as I have never seen anything to support your claim, I totally disagree with it. And the article you refer to does not have anything to back it’s claims.
It appears that I’m doing it again, engaging in discussions that I much rather avoid . But let’s just try and see where we can get.
Not ever having seen supporting evidence, does not mean that it doesn’t exist of course. What would you need from me to be convinced ? Additional links with similar measuments ? Should I recreate such evidence ? Something else ? You can even do it by yourself if you wish so?
For any serious audio engineer, this is simply a known and very logical thing to happen. What would the general public need, as acceptable evidence ?
Of course, you are entitled to your own opinion. And I respect that. But please let and opinion be an opinion, until most reasonable doubt can be removed.
Regarding the lack of ‘sources that supportthe claim’ : the article has linked 24 sources. Even without checking the integrity of the sources, it is not fair to make that statement. Even a few sound clips are shared, which can help you determine whether your system generates said nonlinearities not not. Have you tried yet ?
It’s now the 2nd time that you are using ‘I don’t know, and thus I disgree’ as an argument.
Be honest with yourself. Is that really an argument that follows valid reasoning and logic ?
I cannot continue a discussion that fails to be logical. Those will never have useful output, and only cause frustration on either side of the discussion.
If you have never seen an example of such device : all I need from you to convince me, is to show one chain of amp+speaker, with zero nonlinear distortions.
I don’t care about the linear ones or other defects; they can be as screwed up as you like :). For bonus points, describe me the way to achieve that. I will make us both filthy rich.
Even if slightly non-zero (like the linked example @0.09%), you with get the described effect. This is not a matter of opinion. This is a matter of pure mathematical definition (and confirmed in real-life as well).
I have to say : what @Joel wrote in the link above, is excellent. I fully agree with everything there.
Yes, it is insanely hard to test for the audibility of hi-res differences. And yes, current testing methodology is not perfectly adequate for it.
Even in case you can prove a perceived difference, it is hard to prove what exactly caused that perception difference. Was it the intended improvement, or the introduced defect that was heard?
What bothers me, is the logical fallacy in the conclusion.
If you are unable to determine if your solution actually solves a problem. But you are sure that it introduces avoidable defects.
The only logical conclusion can be ‘well, just don’t apply the solution then!’. Well, there is another one : ‘fix the defects’. But we’re not there yet; not in the foreseeable future.
The funny thing is, that this conclusion is not the situation that we’re currently in.
There is no reason for such high sampling except digital processing e.g. for mixing, DSP and so on. Otherwise, the only part that will make some audible difference is the 24 bit length, that will reduce the noise floor a little. Which would only be audible at 0 db on my amp, where the noise of the amp may be higher. But including any ultrasound content makes absolutely no sense, and it should be filtered anyway, to prevent intermodulation etc. It is true that filtering at 96khz is better, but not audibly so. Today’s filters on most DACs do a great job at any rate.
A more practical reason would be “why not”. There’s a bit more to this than… because we can. Some people mistrust mastering, that includes downconversion to e.g. 16/44.1 It should not be the case, but it is reasurring to get the original master, without extra steps. Why go through extra mastering when you are NOT listening to a CD? Space is not an issue anymore, neither is streaming. For the same reason, hi-rez material should be cheaper or cost at least the same.
Finally, technology has a habit of raising the bar beyond perception, and on the flipside it’s useless to keep things at the level of human perception just because that’s where we are. That would also be silly. But it is even more so to claim superhuman perception to justify hi-rez. We can embrace it while enjoying peak audio without claiming too much about it.
I agree with this. I have a whole range of downloads and rips from my own cd’s. In my experience a 24bit 192khz music file that has been badly recorded or (re)mastered is in no way any way better than a well recorded and mastered 16 bit 44.1khz track. But I can hear the difference between them if both are equally well recorded. Mastering plays a huge part too. I took the trouble to buy the more expensive Sony Japan versions of Miles Davis’s mid 1970’s live recordings of Agharta, Pangea and Dark Magus. Wow! What a difference these make over the Sony USA mastered versions.
Yes, mastering beats resolution, imho. Which is why I care enough to get the masters I want; and why I can never trust a streaming service to maintain access to the mastering I enjoy.
So he’s a programmer. So what. and I have always know about Xiph since it’s founding. but that in no way shape or form gives even the tiniest bit of credence that the article isn’t total bullshit.
For those who didn’t know, “the guy” talking is Monty Montgomery. Here’s his wikipedia bio, and here’s the list of stuff the non-profit he founded maintains. For those who can’t be bothered, his work includes Opus, the codec WhatsApp uses, and the foundation is the maintainer for FLAC.
I’ll leave how likely it’d be that someone with that kind of background would not have access to a “high end, high resolution system”, or would not know how to properly conduct a critical listening test, to people’s common sense. My feeling would be “not very likely at all”, and that, not that it makes him right in this particular case, “that guy” is generally more competent at assessing audio quality than pretty much anyone here, myself included, but believe what, and who, you will.
It shouldn’t even matter who “the guy” was. The entire article is substantiated argumentation, in itself.
If you want to object against that, that’s fine. I’m sure it is even welcomed. You can do so, with arguments. If done well, we can all learn from it.
You can not object, by merely questioning anyone’s authority. No matter how competent. "An uneducated man once explained to me that the earth is spherical. Therefore, the earth must be flat.’
I find that the higher resolution recordings are almost always worth it. For instance, the Beethoven 1963 Von Karajan Symphonies in 24 bit are better, to my ears, than any other available version. It is not about frequency extremes, but much more about an overall ‘musicality’: to my (going on) 67 year old years.
Many 50s-60s classical and jazz recordings were not well remastered for their original CD re-releases. Some of them were remastered later with greater care at higher digital resolution for SACD or other purposes. It’s just very convenient for labels to charge more for those remasters because they can use numbers (like 192/24) to “demonstrate” the superiority of the re-issue, but the reality is that the improvements are in the remaster, which would be audible even at 44.1/16. BTW, https://www.highdeftapetransfers.com/ has some outstanding tape-to-hi-res transfers outside the main label re-issue channel.
Almost rhetorical question to no one specific: How do you know the “betterness” you are hearing can be attributed to the higher resolution and not a different master? You must be comparing with something, but are you sure what it is you are comparing and not changing more than one factor when doing this?
Well, let’s just say it’d appear that more people have statistically improbably superhuman hearing than have access to a null tester (and yeah, what’s said about wires there applies to masters, and resolutions, here).
For those who don’t want to be bothered with a soldering iron, but who’d nevertheless like to determine how much of their DNA comes from Chiroptera, it might be possible to apply this to HiRes files in the digital realm.
And before someone brings this up, I’ve also seen PSAudio’s response to Winer. I’d give it credence proportional to their engineering abilities, but here again, suit yourself.
Someone I know told me he heard more details from the DSD cut off 150kHz filter against the 50kHz filter of his ADI-2 Pro FS playing the same DSD256 music from his HQPlayer PC. He uses Benchmark AHB2 amp.
. But let’s just try and see where we can get.