|
|
|
|
|
|
by mrob
1476 days ago
|
|
|
>in almost all test cases Common sense suggests that "visually lossless" means no detectable difference by the naked eye ever, not in "almost all test cases". MP3 is a very old codec, and it's possible that there are still some "killer samples" that can be ABXed by skilled listeners with good equipment even when encoded by a modern version of LAME. A better example of something that could reasonably called "audibly lossless" might be something like Opus at 160kbps, for which I've seen no evidence of any successful ABX. But even that is is usually called "transparent", not "audibly lossless", so not only is "visually lossless" a lie, the name itself is propaganda. |
|
|
Common sense suggests no such thing. When you buy a bottle of “water”, it actually has a bunch of stuff in it that isn’t water. How dare they?! When someone says “I’ll be there in 15 minutes”, it is highly unlikely that they will show up in exactly 900 seconds. Such liars! Why are you even meeting them? When people say airplanes are “safe”, you might angrily point out how many people have died, not realizing that “safe” is relative to other things and not an absolute in that context. This is common across basically everything in life. “There are no absolutes.” If you think common sense is to automatically assume every statement that even looks remotely absolute is intended to be taken absolutely… that is not common. Short statements will come off as absolute, when they are just intended to be taken as approximate, but even absolutes are usually meant to be taken as slightly less than absolute.
“Visually lossless” is a description of the by far most common experience with DSC. They’re not describing it as truly lossless, so you know there is some loss occurring. It is natural to assume that in extraordinary circumstances, that loss might be noticeable side by side… but you don’t have a side by side when using a monitor most of the time, so the very lossy human vision system will happily ignore small imperfections.
> so not only is "visually lossless" a lie, the name itself is propaganda.
Your whole comment shows that you don’t understand how communication works. It is “visually lossless” as far as people are concerned. The study shows that! This is not at all what propaganda looks like.
When Apple labeled their iPhone screen a “retina screen” because people would no longer notice the pixels, I suppose you called that a “lie” as well because you could lean in really close or use a microscope? The retina display density achieved its stated goal.
There is literally no point in continuing this discussion when you take such an absolutist position and refuse to consider what alternative communications would look like. How about “99.9% visually lossless”? That would be even more confusing to people.
Communicating complicated concepts succinctly is a lossy process. Language is lossy. As they say, “all models are wrong, but some are useful.”