Your response is so Orwellian I felt compelled to respond, if only to alleviate the false concerns you raise.
I think you should leave Orwell out of this because he didn't write what you think he wrote.
The situation is in fact the opposite of what you state. It was due to the perceived risk of Windows in much of the general community that I investigated it, provided objective evidence to its soundness and provided a simple cost free solution for avoiding it's possible artifacts. I also hope that I eased minds about the resampler, which "common knowledge" stated to stay away from based on 5+ year old tests .
You still seem confused about existence for a problem, potential harm, risk and quantification of risk and side-effects of solution.
Here is how science works:
A virus capable of harming a human being can exist. Its existence can be shown with a test/experimentation.
The potential harm it could cause can again be shown in a lab/experimentation regardless of the risk posed by it.
Its risk to humans depends on its mode of transmission, ability to survive enough to harm, natural resistance of humans, the impact of being infected, availability of vaccines, etc., etc.
Quantification of risk is the probability of a person to get infected by such a virus which is known to exist. It could be very low based on the factors above or very high.
A vaccine that can fight the virus can have side-effects.
Not taking anything that has side-effects is the default state for same human beings unless the risk posed by the virus can be quantified (say as a probability) and the cost of infection can be estimated (from being ill to death) and there is a trade-off between the side-effects and the risk.
Saying some people would not have side-effects is not an effective argument for a vaccine to be taken pre-emptively without judging whether one is likely to get it or not and what the effects of it are.
I leave it as an exercise for you to do the appropriate analogies to the above and which of those you have actually done and see why I don't agree with your suggestion for the prophylactic 4dB attenuation.
If there is one area where the risk is truly not knowable , its "Hidden APOs". I view this as a bug on my machine. I presented all the evidence and a clear assessment of the risk at my disposal: "My music system and work PCs didn't show this issue…." Indeed my computer may have been 1 in a hundred, or 1 in 10, who knows? However I provided a solution to those that may be concerned about Windows and so are unnecessarily avoiding Windows audio for their HiFi rig.
You are confusing two different things here -
1. The effect of processing in the chain whether by the Windows Engine or any APOs inserted to alter the content-bits. No one is denying this. The minimal -0.2dB which we already know will fix a guaranteed and known problem exists in Windows is justified for this reason and sufficient. There is no disagreement here.
2. The problem of inter-sample peaks. While a theoretical possibility exists of this problem, there is no evidence that any APO introduces this, or if there is such APOs are found in any general Windows system. So this is just a "could-happen" conjecture. Some content in some badly-mastered content can have inter-sample peaks. However, there is very little evidence that it causes audible problems necessarily. There is no reason to suspect that this is some widespread problem that it needs to be fixed. Your prophylactic solution here is one of suggesting a vaccine without clearly describing the risk of not taking it and saying that the vaccine side-effects don't occur to some people or can be managed. It is a good thing you are not in charge of public health policy (hopefully).
Your response is so Orwellian because you are in fact raising perceptions of risk with no supporting evidence, unnecessarily sewing concern over a mere 4 dB loss in a 24 bit system, ignoring the evidence to its lack of harm, and providing no evidence of your own all the while having the shame to hide behind the cloak of "science".
I am not sure what you mean by a 24-bit system. Taking 16-bit content (most CDs) with inter-sample peaks in them and adding zeros to them so you can attenuate them by 4dB isn't a solution. If you are saying use only real 24-bit mastered content, then you are prescribing a solution that is worse than any potential problem. There is very little music content easily available in 24-bits. Downstream DACs already do floating point calculations to worry about whether they are 16-bit or 24-bit content coming in and that does not fix the inter-sample peak problem by just doing that. So, don't know where you are going with this other than obfuscating the issue.
My goal in responding isn't to argue with you, as you seem to have some axe to grind and have stopped being rational on this. It's to present a balanced view of the risk to others so that you're unsubstantiated fear mongering doesn't drive people away from trying Windows and saving a few bucks.
By Orwellian, did you mean to say accusing the other of what one is doing themselves? If so...
I am done with this thread.
