pro-bassoonist wrote:Moshrom, I don't form an opinion by looking at screencaptures. I like to test the discs and see exactly what types of changes have been made.
You imply that screen captures fail to faithfully represent video. Would you prefer that I create a video instead?
I see no reason why a screenshot cannot be used to examine discrepancies in contrast and luminance. The AE image may be unstable, but these two properties remain constant in motion.
pro-bassoonist wrote:Now, the gamma balance is off on the AE release -- there is artificial elevation that flattens the image across the board.
You continually praise the virtues of the Criterion transfer by denigrating the Artificial Eye disc. Please understand me: I am
not saying the luminance and gamma on the AE faithfully represent what this film should look like. I’m arguing that while Shochiku’s restoration of the video offers a substantial improvement overall, it is flawed.
pro-bassoonist wrote:What you see on the Criterion release isn't toned down highlights, but a correct balance.
There’s the word! (I was totally waiting for it.) We appear to be at a standstill. I say A, you say B, and no real way to objectify matters. But I could prove to you that the Artificial Eye image (let’s pick the last screenshot) can be digitally manipulated to look—objectively—more like the Criterion. And I’ve actually just done this by
toning down the highlights. (I can share the result if you’d like, but I suspect you would reject it on the grounds of it being an inadequate representation of video, or some such nonsense.) Anyway, the purpose of this test would be to prove that de-emphasising clipped highlights (i.e. blown-out whites) to a significant degree loses visual information.
Again, I didn’t say the AE is superior. Pick any matching frame from both discs and chances are I’ll
agree with you that the new restoration is better. I already asserted as much. My problem is with just a few shots from the new restoration.
But why are you arguing with me over the
video's presentation? I’ve already said that I’m quite satisfied with it. The dreadful audio is a much bigger deal here.
pro-bassoonist wrote:The audio is so badly damaged that on some of the films in the Mizoguchi box set the distortions/pops/hiss can be seriously distracting. It appears that you think that the audio has been so badly compressed on the new restoration that more has been lost than gained.
One would hope that a professional reviewer of digital media would understand how compression works, or what it
is for starters: it’s ~never a good thing, and your phrasing of this last sentence above implies that you believe it can be. Or perhaps you simply misread what I wrote? I haven’t mentioned compression anywhere.
pro-bassoonist wrote:Even on the AE disc it is pretty clear that the audio is already too thin and overwhelmed by various cracks, so the attenuation/rebalancing work was almost certainly a series of compromises.
The aim of any film restoration is to restore a film to its original theatrical presentation, or to approximate this state as closely as modern technology will permit. Rebalancing an audio track in the way you describe is revisionism. It is scientifically impossible for an optical soundtrack to
gain high-frequency detail that isn’t distortion, and any such detail (which exists in spades on the Artificial Eye) was most certainly present during the film's first theatrical run. It should not be
rebalanced to conform to modern sensibilities and it should
never simply be removed altogether. Optical sound from the ‘30s was perfectly capable of producing a frequency spectrum extending upwards of 10 kHz. The Shochiku restoration has an almost hard cut-off at 4.5 kHz. How is this a compromise, exactly?
Sonic noise reduction is a mastering practice that was mostly phased out from album reissues in the late ‘80s and early ‘90s. That it’s still used to restore films (and so aggressively, too) is utterly ridiculous. Granted, these albums were mostly recorded on magnetic tape, which naturally has a lower noise floor than optical sound—true—but what about 78s? An apt comparison – 78 RPM shellac records are in many cases the best surviving analogue sources for any music pre-WWII, and the most well-regarded mastering engineers are now wise to the destructive qualities of heavy-handed noise reduction as a means to restore these materials.
pro-bassoonist wrote:Depth is very limited, but balance is decent.
Could you define what you believe these two words mean in the context of digital media? All this would be much easier if I had some kind of legend to decrypt your reviews!
pro-bassoonist wrote:I am perfectly fine with the new track.
Disregarding everything else, this is the thing to be disturbed most by – that your opinion of the audio track runs contrary to every other reaction (I count 8... maybe a conservative 7?) expressed in the last two pages of this thread.
And so we can step away expecting more restorations that sound this way – the consequences of having a reviewer for blu-ray.com who praises such rubbish.