tenia wrote:The question I'm always wondering when stumbling on cases like this is : does actually anybody knows how it should look, rather than how these looks are possibly all wrong in various aspects and levels ?
And I'm saying that as a very general and open point, aiming both sides of the debate (restoration teams and forumers). I mean, it's not as if the restoration teams are aiming to be wrong on the purpose so they can be ranted upon, but it's also not as if we knew exactly what the original photo was.
That is absolutely a valid question to raise, and my sense is that there aren't always clear answers. (I was privy to some of the questions faced on another restoration project.) From what I understand, one of the challenges arises when you start with a scan of the camera negative; it's great for capturing fine detail, but it means that you have to recreate the color timing from scratch. Not only that, but you are using different tools from what were used originally - computer programs rather than an optical printer, filters, etc. That creates an opportunity for inadvertent errors-you might forget to properly filter a day-for-night shot, for instance. Or you could end up introducing bigger changes to the film.
If you have a vintage release print that is not too faded you can use that as a reference, but even that can raise its own set of questions. Maybe the initial release prints of a film were not ideal quality for various reasons.
There is still always some question of taste and creative judgement that ends up creeping into the process, I think. In the case of Conversation Piece, the fact that the screen captures of the Eureka edition are so dominated by grays and lack highlights & shadows is a tip-off that something might be wrong. Svet Atanasov comments in his review that the new transfer lacks vibrancy compared to old transfer, and he is right. The film was shot on studio sets--i.e., careful control over lighting, especially on actors' faces--but the new transfer used in the Eureka edition doesn't seem to reflect that adequately. Many of the images look flat and lack depth; the compositions are not shaped by the lighting in the way that you would expect for a studio film. Especially a studio film directed by Visconti.
I do agree with Svet that the color balance looks better than in the Gaumont Blu-ray, but that becomes a moot point ultimately.
UPDATE: I would like to hear more input from people who have seen both the Gaumont and MoC Blu-rays, when they get the chance to make a comparison. For me, looking at the Gaumont Blu-ray again confirms my initial impression that it does tend towards the greenish and brownish side in some scenes, and the yellowish in others. I think that the perceived color bias is shaped in part by the heavy interiors, and may also be partly a byproduct of the original cinematography. But on the whole the disc is fine and even looks very nice in many shots.
If you look at the whites on the Gaumont Blu-ray, there is in fact a broad range of whites and off-whites in the production design. They are affected in part by the overall lighting scheme and the reflected light of the costumes and settings. To give you an example of the interesting range of whites, look at the white apartment interior that corresponds to the screen capture above. The Gaumont Blu-ray does tend toward the yellowish side in that shot, but it's not extreme and contrast looks fine. The basic design concept of the scene - the subtle play of different whites - is still preserved. So while Gaumont Blu-ray may not have flawless color timing, it may not unduly distort the intended look of the film, either.
I'm curious to hear how the MoC disc plays in motion.
(Raro 2012, Gaumont 2013 and now MoC 2016). I'm extremely surprised.
