A few words about...™ Howards End -- in Blu-ray - Page 4

Sean, I'm also in post, as an assistant editor though, so I understand where you're coming from at being dissatisfied with a mediocre/terrible digital image that others think looks terrific. 

I was at the academy last summer for their 'film formats through the ages' night which showed a ton of clips from a variety of films, mostly in large format.  Some were digitally restored presentations of 70mm films, most were 70mm film.  One in particular looked horrid blown up on that wonderful Samuel Goldwyn screen (and I've seen it in 70mm and know what its grain looks like, it wasn't grainy it was noisy and digital).  There were a handful of DPs as part of a panel afterwards, and one made an aside comment that one clip looked particularly atrocious tonight, and shouldn't because the people who were so proud of their image hadn't properly harvested it to begin with, they had got an image that they thought represented film grain but was actually like a pulsing parameceum of digital noise.  He went on that so many people today don't know what film grain looks like (I think this was John Bailey, but I'm not certain it was a year ago!)  His first job had been printing and running dailies before sending the dailies back to the set and he watched film dailies 12 hours a day for a year and or two, he knows what film grain looks like.  And he then said there were certain image harvest machines he simply wouldn't use because they'd produce an image that looked to most people like a beautiful 2k harvest, but when he looked at it all he saw was a "digital floor,"  not film grain at all, a digital floor.

I think that's either his own term, or the in house term he's familiar with.  I've never run across it before or since.  But I'm going to hypothesize that a digital floor would describe what you're seeing.  This was the result of an inadequate image harvest.  The image was pulled, they looked at it, thought they were seeing film grain and didn't look closer, they didn't get that though and since their harvest was calibrated to the digital noise rather than to the grain of the film, everything after that would be off in small but to some people significant ways.  So since film grain is already hard to encode think how much harder it is to encode unfocused film grain that looks like a watery layer of digital noise interfering with the grain.  The codec has issues with such a complex random structure, and compensates by introducing additional artifacting.  I'd hypothesize again that the playback stage introduces another layer of potential problems perhaps some players hardware simply smooth complex transfers out before sending the video output signal to the monitor. And other hardware might more faithfully represent the image encoded or the tape master the encode was produced from.  It's not hard for me to imagine that this could be the source of what you and others are seeing.  Particularly telling--to me--is that it becomes more clear when you blow up the image. 

This isn't an issue of monitor sensitivity, though it partially has an element here, it's more an issue of eye-sensitivity/knowledge.  And sometimes even experts will not notice things they're trained to look for (I watched down a show today five times during various outputs and it wasn't until the last output that we noticed a cameraman on the edge of picture.  Then we had the joy of insert editing on the other masters.) 

I also think this could explain why the DVD looks acceptable, because of the different way it is encoded, much lower bitrate, and the way it resolves detail differently the DVD simply doesn't have the resolution to be bothered by an image harvest whose flaws are only exposed in high definition.  A 2k harvest that looks great on DVD could very well be exposed to be as bad as you describe on DVD particularly if people mistook digital noise in the harvest for film grain and passed the harvest through QC. 

I cannot be anything but bemused by anyone having troubles with the Blu-ray of "Howard's End."  It is one of the most extraordinary Blu-rays in my collection.  The image is brilliant, the sound beyond reproach, and it plays like a dream.

I have a Sony Blu-ray (1350) and a Panasonic Viera plasma screen. 

It's right up there with the Fox "The Robe" and Warner "How the West Was Won" in how Blu-ray releases SHOULD be done!

Sean, I'm also in post, as an assistant editor though, so I understand where you're coming from at being dissatisfied with a mediocre/terrible digital image that others think looks terrific. 

I was at the academy last summer for their 'film formats through the ages' night which showed a ton of clips from a variety of films, mostly in large format.  Some were digitally restored presentations of 70mm films, most were 70mm film.  One in particular looked horrid blown up on that wonderful Samuel Goldwyn screen (and I've seen it in 70mm and know what its grain looks like, it wasn't grainy it was noisy and digital).  There were a handful of DPs as part of a panel afterwards, and one made an aside comment that one clip looked particularly atrocious tonight, and shouldn't because the people who were so proud of their image hadn't properly harvested it to begin with, they had got an image that they thought represented film grain but was actually like a pulsing parameceum of digital noise.  He went on that so many people today don't know what film grain looks like (I think this was John Bailey, but I'm not certain it was a year ago!)  His first job had been printing and running dailies before sending the dailies back to the set and he watched film dailies 12 hours a day for a year and or two, he knows what film grain looks like.  And he then said there were certain image harvest machines he simply wouldn't use because they'd produce an image that looked to most people like a beautiful 2k harvest, but when he looked at it all he saw was a "digital floor,"  not film grain at all, a digital floor.

I think that's either his own term, or the in house term he's familiar with.  I've never run across it before or since.  But I'm going to hypothesize that a digital floor would describe what you're seeing.  This was the result of an inadequate image harvest.  The image was pulled, they looked at it, thought they were seeing film grain and didn't look closer, they didn't get that though and since their harvest was calibrated to the digital noise rather than to the grain of the film, everything after that would be off in small but to some people significant ways.  So since film grain is already hard to encode think how much harder it is to encode unfocused film grain that looks like a watery layer of digital noise interfering with the grain.  The codec has issues with such a complex random structure, and compensates by introducing additional artifacting.  I'd hypothesize again that the playback stage introduces another layer of potential problems perhaps some players hardware simply smooth complex transfers out before sending the video output signal to the monitor. And other hardware might more faithfully represent the image encoded or the tape master the encode was produced from.  It's not hard for me to imagine that this could be the source of what you and others are seeing.  Particularly telling--to me--is that it becomes more clear when you blow up the image. 

This isn't an issue of monitor sensitivity, though it partially has an element here, it's more an issue of eye-sensitivity/knowledge.  And sometimes even experts will not notice things they're trained to look for (I watched down a show today five times during various outputs and it wasn't until the last output that we noticed a cameraman on the edge of picture.  Then we had the joy of insert editing on the other masters.) 

I also think this could explain why the DVD looks acceptable, because of the different way it is encoded, much lower bitrate, and the way it resolves detail differently the DVD simply doesn't have the resolution to be bothered by an image harvest whose flaws are only exposed in high definition.  A 2k harvest that looks great on DVD could very well be exposed to be as bad as you describe on DVD particularly if people mistook digital noise in the harvest for film grain and passed the harvest through QC. 

Thanks for the excellent, informative post, Adam.

I finally got around to giving this disc a brief look -- still haven't been able to give it a good sitdown yet -- but that description of "a pulsing parameceum of digital noise" comes very close to what I saw in a few scenes, especially in red areas.

I didn't see quite all the same issues that Sean and a few others saw on my Samsung LED DLP w/ Panny BD60 for playback, but did see what looks like "a pulsing parameceum of digital noise" that occasionally started to vaguely resemble the tetris-like grid that Sean mentioned w/out being quite that pronounced.  I should note though that how bad it looked depended a fair bit on how close I got to the screen.  From my normal viewing distance (of ~9ft from the 61" display), the problem was not as pronounced, but still quite noticeable.  At maybe 6-7ft, it definitely look very bad.  Also, I should note that I do not have the latest firmware update for my BD60 though it's probably not *that* old since the player itself is only ~6 months old.

Since I haven't actually tried to watch the film as I normally would (rather than look for issues), I'm not sure how much it'll actually bother me and whether it will end up being unwatchable to me.  I get the feeling I'll notice the flickering/swarming characteristic of the noise/artifacts from time to time, particularly in reds (as I often do w/ OTA HD broadcasts, especially during the earlier days of OTA HD, though those tended to be much worse IIRC).  Also, I think I've seen something like this before probably in some other Criterion BD, but can't remember for sure offhand.  Think I'll check my other Criterion BDs to see.

_Man_

Man, I'm about ten feet from my 50" and it looks bad, especially in dark scenes. The grid (not harsh or hard but pale like frequence interference) appears mostly in the dark scene of Jacky Bast on the bed with the train passing the window behind her, and at the end of the film during the long slo mo shot of Charles entering the train carriage. It stretches out on both sides of the 2.35 image in horizonal lines. In any case the shadows are way too light on this blu ray; the dvd shadows are darker and hold up far better.
 

Sean, I'm also in post, as an assistant editor though, so I understand where you're coming from at being dissatisfied with a mediocre/terrible digital image that others think looks terrific. 

I was at the academy last summer for their 'film formats through the ages' night which showed a ton of clips from a variety of films, mostly in large format.  Some were digitally restored presentations of 70mm films, most were 70mm film.  One in particular looked horrid blown up on that wonderful Samuel Goldwyn screen (and I've seen it in 70mm and know what its grain looks like, it wasn't grainy it was noisy and digital).  There were a handful of DPs as part of a panel afterwards, and one made an aside comment that one clip looked particularly atrocious tonight, and shouldn't because the people who were so proud of their image hadn't properly harvested it to begin with, they had got an image that they thought represented film grain but was actually like a pulsing parameceum of digital noise.  He went on that so many people today don't know what film grain looks like (I think this was John Bailey, but I'm not certain it was a year ago!)  His first job had been printing and running dailies before sending the dailies back to the set and he watched film dailies 12 hours a day for a year and or two, he knows what film grain looks like.  And he then said there were certain image harvest machines he simply wouldn't use because they'd produce an image that looked to most people like a beautiful 2k harvest, but when he looked at it all he saw was a "digital floor,"  not film grain at all, a digital floor.

I think that's either his own term, or the in house term he's familiar with.  I've never run across it before or since.  But I'm going to hypothesize that a digital floor would describe what you're seeing.  This was the result of an inadequate image harvest.  The image was pulled, they looked at it, thought they were seeing film grain and didn't look closer, they didn't get that though and since their harvest was calibrated to the digital noise rather than to the grain of the film, everything after that would be off in small but to some people significant ways.  So since film grain is already hard to encode think how much harder it is to encode unfocused film grain that looks like a watery layer of digital noise interfering with the grain.  The codec has issues with such a complex random structure, and compensates by introducing additional artifacting.  I'd hypothesize again that the playback stage introduces another layer of potential problems perhaps some players hardware simply smooth complex transfers out before sending the video output signal to the monitor. And other hardware might more faithfully represent the image encoded or the tape master the encode was produced from.  It's not hard for me to imagine that this could be the source of what you and others are seeing.  Particularly telling--to me--is that it becomes more clear when you blow up the image. 

This isn't an issue of monitor sensitivity, though it partially has an element here, it's more an issue of eye-sensitivity/knowledge.  And sometimes even experts will not notice things they're trained to look for (I watched down a show today five times during various outputs and it wasn't until the last output that we noticed a cameraman on the edge of picture.  Then we had the joy of insert editing on the other masters.) 

I also think this could explain why the DVD looks acceptable, because of the different way it is encoded, much lower bitrate, and the way it resolves detail differently the DVD simply doesn't have the resolution to be bothered by an image harvest whose flaws are only exposed in high definition.  A 2k harvest that looks great on DVD could very well be exposed to be as bad as you describe on DVD particularly if people mistook digital noise in the harvest for film grain and passed the harvest through QC. 

I just remembered a hardware-related issue that happened to me at the time that the remake of The Thomas Crown Affair made it to DVD. I experienced what many others did, the unstable image of the sea when the characters wind their way up the mountain to Crown's bungalow on a Carribbbean island.

Shortly after that my DVD player took a dive. I replaced it with Toshiba's highest-end player for a close-out SALE (!!!) price of $700 (down from $1,200 - and we complain about BD player prices!!!) (it was a beautiful monster, built on a platform of solid copper!).

That 'monster' resolved the problem, making the image solid (my vocabulary of technical terminology is nil, so bear with me).

I was wondering, then, if the reasons some of us are getting good results from our discs is because our players are actually overcoming deficiencies in the transfer, like the Toshiba did with 'Crown', rather than some anomalous 'handshake' issue with certain players. I realize of course that one shouldn't be obliged to have any particular player in order to have a good result.

Curious to know if the grid, screen door and white blocks artifacts you're experiencing are 'architecturally' the same each time you watch the BD (painful as it may be to do that) and if the same artifacts appear in exactly the same places on each viewing. 

Thanks for the excellent, informative post, Adam.

I finally got around to giving this disc a brief look -- still haven't been able to give it a good sitdown yet -- but that description of "a pulsing parameceum of digital noise" comes very close to what I saw in a few scenes, especially in red areas.

I didn't see quite all the same issues that Sean and a few others saw on my Samsung LED DLP w/ Panny BD60 for playback, but did see what looks like "a pulsing parameceum of digital noise" that occasionally started to vaguely resemble the tetris-like grid that Sean mentioned w/out being quite that pronounced.  I should note though that how bad it looked depended a fair bit on how close I got to the screen.  From my normal viewing distance (of ~9ft from the 61" display), the problem was not as pronounced, but still quite noticeable.  At maybe 6-7ft, it definitely look very bad.  Also, I should note that I do not have the latest firmware update for my BD60 though it's probably not *that* old since the player itself is only ~6 months old.

Since I haven't actually tried to watch the film as I normally would (rather than look for issues), I'm not sure how much it'll actually bother me and whether it will end up being unwatchable to me.  I get the feeling I'll notice the flickering/swarming characteristic of the noise/artifacts from time to time, particularly in reds (as I often do w/ OTA HD broadcasts, especially during the earlier days of OTA HD, though those tended to be much worse IIRC).  Also, I think I've seen something like this before probably in some other Criterion BD, but can't remember for sure offhand.  Think I'll check my other Criterion BDs to see.

_Man_

Adam,

Great post!

Which sounds wholly accurate to occasional problems which plague the industry today.  The information that I received re: HE was that the harvest was performed on the equipment (a C-Reality) available at the location in the UK.  And yes, something is there other than the film grain in certain scenes, which appears to be digital noise.

This appears to be a similar situation that occurred with Spartacus in which the entire image harvest was invested with digital noise.

A top facility generally has no problem with a harvest, which needs to be viewed on a very large screen as a test before the entire production goes through the system.  Performed properly, by people who know what the final result should look like, is no longer a smoke and mirror situation, and is now routine -- as long as both the hardware, software and human interface all are of sufficient quality.

RAH

Forgot to say, and the reason why I quote Man's closer viewing of HE as exacerbating the problems he sees: When I viewed HE  through my second player I paused at the places Sean described as having the most egregious problems and stood inches away from the screen and still could see no artifacts (or any that, again, my untrained eye could perceive).

 I wanted to pull quotes from some of the recent posts without actually reproducing the entire posts but just can't figure out how it's done (I've tried the multiple quote/quote facility at the bottom, the quotation marks above right and even copying and pasting, all to no avail.

Andrew:

Just click the "two quotation mark" button for each quote you wanted to pull.  When you've got all that you want hit the single quote button and all the quotes will appear in the Reply box and you can edit them down in any fashion you'd like. 

All right! Everybody's invited to my house to see what I'm seeing!!!