Podcast HTF Podcast Episode 5: Visiting Criterion Collection (1 Viewer)

[Josh Steinberg]

I have yet to observe anything problematic in those discs.

There are so many places in the chain where artifacts can be introduced to the display that aren't present on the disc.

For instance, if you're viewing these on a 4K/UHD monitor, something has to upscale that 1080p signal to 4K. Depending on what equipment you have and how its set up, that conversion could be done by your disc player, by your receiver, by your television or by some unholy combination of those entities. And it's certainly within the realm of possibility that any one of those upscaling devices could be introducing artifacts as it attempts to output a lower resolution signal onto a higher resolution display.

 

[JohnRice]

I have yet to observe anything problematic in those discs.

There are so many places in the chain where artifacts can be introduced to the display that aren't present on the disc.

For instance, if you're viewing these on a 4K/UHD monitor, something has to upscale that 1080p signal to 4K. Depending on what equipment you have and how its set up, that conversion could be done by your disc player, by your receiver, by your television or by some unholy combination of those entities. And it's certainly within the realm of possibility that any one of those upscaling devices could be introducing artifacts as it attempts to output a lower resolution signal onto a higher resolution display.

That’s exactly why I specified the details of my playback. Not even 3:2 pull down.

 

[Josh Steinberg]

That’s exactly why I specified the details of my playback. Not even 3:2 pull down.

Exactly - I don't think by and large that it's a global problem with the discs. I think it's a local problem with how playback equipment is interacting.

 

[PMF]

[...]I see nothing at all wrong with Pan's Labyrinth.

Which one is better? The Labyrinth that Pan's or the Labyrinth that Scans?

 

[bugsy-pal]

In defense of Michael Hoffman, I agree that Criterion's compression has been problematic in some instances. There have been instances where fine compression artefacts are evident on Criterion discs yet absent on corresponding UK versions of the same transfer, eg. Time Bandits. I'm sure that David Mackenzie would back this up - he is one of the best in the business at the intricacies of encoding.

This is not to say that these encoding problems are always detrimental to the viewing experience. I would say that the superior source material and all of the other work that Criterion puts into their releases makes them the preferred version to watch in most cases.

 

[Michael Hofmann]

(...) the long-term commitment and services from such a group as Criterion requires an overall round of applause, appreciation and gratitude - without qualification.

Sure they do. But they also deserve constructive criticism where it is due, in the hope that their encoding process will improve.

There are so many places in the chain where artifacts can be introduced to the display that aren't present on the disc.

I agree, but my equipment is set up to minimize this kind of processing, and, since I know a thing about image processing and video compression or two, I'm very confident that these kinds of artifacts are not introduced by anything in the chain. Output device is a calibrated LG OLED, where the banding/macroblocking artifacts on some of the above mentioned Criterion discs are revealed more so than on other equipment (e.g. they're far less visible on any output device that's backlit, or has a higher black level lower bound).

But that doesn't change the fact that they're there. They're the artifact of a compression algorithm (or setting) that assumed this loss of detail would never be visible to a viewer. Other people see them, too: https://www.blu-ray.com/movies/Pans-Labyrinth-Blu-ray/156565/#UserReviews.

There have been instances where fine compression artefacts are evident on Criterion discs yet absent on corresponding UK versions of the same transfer (...)

Thanks for backing me up -- indeed that is the case. Actually, it is the case on many, many Criterion discs. On some of those, the problems only become apparent when pixel peeping, so the overall viewing experience is not compromised. But on many others, these artifacts are clearly visible in motion, and at regular viewing distances.

Another example: Don't Look Now. People have praised the grain structure, and yes, it's much better than the previously DNR'd-to-hell UK disc. But what people think is grain actually contains myriads of tiny compression artifacts on the Criterion disc; see e.g. https://www.blu-ray.com/movies/screenshot.php?movieid=42646&position=12. All these small square blocks with chroma noise are actually visible in motion! A proper grain structure would look much better.
And if someone thinks this (https://www.blu-ray.com/movies/screenshot.php?movieid=132737&position=9) is a beautiful sky, then I cannot help them.

That said, these kinds of compression artifacts are a somewhat different problem than the banding/macroblocking in dark scenes (Pan's Labyrinth, Blue Velvet, ...), which only shows intermittently on some discs, and only on equipment that brings these flaws to light.

I'm sure that David Mackenzie would back this up - he is one of the best in the business at the intricacies of encoding.

Yes. I don't know David personally, but I'm 100% convinced he would back me up, including my opinion on the Pan's Labyrinth disc. (I'm just not sure if he'd want to, since that would mean publicly criticizing the competition.) Is he on this forum?

Cheers,
Michael

 

[Robert Crawford]

If you're seeing something on your display then it's there from your perspective. If people are not seeing the same thing on their displays then the same is true for them too. It sounds to me like you guys should agree to disagree because you're not going to change each other's mind. Opinions vary and people can discuss the differences in opinion, but most of the time in these kind of disc discussions, opinions don't usually change from one perspective to the contrary one.

 

[Sam Posten]

Yup, let's move on.

 

[Detour (1945)]

Did Criterion happen to comment in the podcast about their decision-making process regarding the inclusion of commentaries on their releases?

 

[Michael Hofmann]

It sounds to me like you guys should agree to disagree because you're not going to change each other's mind

Unfortunately it does not work like that when it comes to provable technical shortcomings; it's not an opinion, or, say, a discussion about religion. And I have tried to give ample evidence that it's not just my own opinion.

I do understand that some people may not see these artifacts as clearly. This can be due to their equipment tending to hide these technical flaws (yes, I'm serious about this), or it can be due to them not really knowing what to look for.
Example for the latter: A few years ago, lots of reviews raved about the "intact grain structure" of certain Italian film masters from LVR, to my strong dismay. Today, after several releases from fresh scans saw the light, it's a widely recognized fact it was all just ugly scanner noise.
There might still be a lot of confusion today about what a well-mastered film should look like.

I truly wish a community of what I thought were technically minded home theater enthusiasts would take these kinds of quality issues more seriously. I also remain disappointed with Criterion for not taking their compression issues more seriously. And since I don't want to debate against further semi-hostile comments, this shall be my last comment on this matter.

EDIT: The following comment was edited after I had posted message #73. What a welcoming community!

 

[PMF]

My gut is telling me that it's not.

 

[Robert Crawford]

Unfortunately it does not work like that when it comes to provable technical shortcomings; it's not an opinion, or, say, a discussion about religion. And I have tried to give ample evidence that it's not just my own opinion.

I do understand that some people may not see these artifacts as clearly. This can be due to their equipment tending to hide these technical flaws (yes, I'm serious about this), or it can be due to them not really knowing what to look for.
Example for the latter: A few years ago, lots of reviews raved about the "intact grain structure" of certain Italian film masters from LVR, to my strong dismay. Today, after several releases from fresh scans saw the light, it's a widely recognized fact it was all just ugly scanner noise.
There might still be a lot of confusion today about what a well-mastered film should look like.

I truly wish a community of what I thought were technically minded home theater enthusiasts would take these kinds of quality issues more seriously. I also remain disappointed with Criterion for not taking their compression issues more seriously. And since I don't want to debate against further semi-hostile comments, this shall be my last comment on this matter.

Well, have it your way then and I'm sorry we didn't measure up to your "home theater enthusiast" expectations.

 

[Michael Hofmann]

Well, have it your way then and I'm sorry we didn't measure up to your "home theater enthusiast" expectations.

Thanks! Allow me a very last comment.

Pan's Labyrinth, Blu-ray edition from the Trilogia de Guillermo Del Toro. Cat. no. CC2686BD-3. Photo of my LG E6 OLED TV. Pretty much the first scene at 02:24. I didn't have to look far.

If this is not really bad macroblocking, then I don't know what is. Just one example of many, but a particularly egregious one. Effect is only slightly exaggerated due to choice of camera exposure. Black levels of this disc are milky, but this is not really what I'm complaining about.
And before people tell me that there's something wrong with my equipment: 1) Look at the disc more closely. The artifacts are there. 2) The UK 2016 release (compressed by David Mackenzie, from what I've heard) has no such issues. 3) All properly mastered Blu-rays have no such issues.

Thanks for reading. More comments only when explicitly invited.

 

[Sam Posten]

Allow me a very last comment.

My gut is telling me that it's not.

WINNAH.

 

[Sam Posten]

Spine 1000 is a monstaaaaah!
https://www.hometheaterforum.com/co...-godzilla-the-showa-era-films-blu-ray.362987/

 

[JohnRice]

OK, I took an even more in-depth look at Pan's Labyrinth. I played the exact same sequence at the beginning of the film as the recent post. I played it at regular speed. I played it frame-by-frame, repeatedly. I used calibrated brightness. I turned up the brightness. I tried both copies I have, because I had bought the single release before I got the three movie set, and I still have the single release.

Then I tried all of this... again... on my cheap living room system that has an old, low end Panasonic BR player and a new $200ish TCL 43" 5 Series UHD. I wanted to see how its low-end upscaling would do. No upscaling with the player, and scaling is deactivated in the Marantz SR7012 receiver. (The video in that system is cheap, the audio is not.) I went through the entire process again. FWIW, I can definitely see the full shadow detail on both TVs.

I looked and looked, wanting to find something wrong. There was ZERO blocking. None. Anywhere. Not even a hint.

Clearly there are people who are experiencing blocking, but I am positively confident it is NOT on the disc. My only guess is... maybe it's sharpening. Maybe. I despise any amount of artificial sharpening, so I try to find the "null" setting on my TVs. That's the only guess I can make, but clearly something is happening with some people's systems. It is NOT on the disc.

I am truly putting this out of my mind so I can just enjoy some movies.'

One other thought, both systems are wired entirely with new, certified high-speed Monoprice hdmi cables.

 

[Worth]

Maybe it has something to do with the way OLEDs process the image.

 

[Sam Posten]

Or maybe.....

 

[Robert Crawford]

Or maybe.....

That's not fair.

 

[JohnRice]

If anyone is curious, this should be the answer.

I missed the detail that Michael is in Europe, which uses PAL. North America uses NTSC. What's important with that is black levels. NTSC has a "blacker than black" region, which PAL doesn't. In other words, with 8 bit digital video, which has levels from 0-255, NTSC black is at 16. Everything below 16, logically, is also black. In PAL, black is at 0. After taking a direct, digital capture from Pan's Labyrinth and manipulating it in Photoshop, I found there was blocking, but it's entirely in the 0-15 "blacker than black" range. So, it's invisible on a properly calibrated for NTSC monitor.

EDIT: I think I got some details wrong here, but the theory stands all the same.