When/Why did you become an OT purist? — Page 7

danny_boy said:

The irony is that the duplication process used in the fotochemical workflow  tended to "even out" the discrepancy between the parts of the o-neg that were 1st generation and the other parts that were optical duplicates.

So by the time you got to that 4th generation release print the film looked relatively seemless.

But if you strike a positive digital "print" from a 4K scan the difference in granularity between all the constituent parts that make up Star wars will be jarring.

I know very little about all this, but in a way I've come to the same conclusion, which I think is sort of the same point you're making here:  filmmakers account for generational loss, and it is in fact part of the intended look of the theatrical print.  (And therefore, wouldn't a scan of a theatrical print be the definitive way to watch SW, or really, any movie?)

Not sure if you saw it, but have you seen the thread where I asked about this:

http://originaltrilogy.com/forum/topic.cfm/Trying-to-understand-film-preservation-perhaps-a-stupid-question-but-shouldnt-digital-masters-be-struck-from-theatrical-prints/topic/14944/

I'd be interested to hear your thoughts.

danny_boy said:

Peer reviewed International studies conducted in theaters across the globe concluded that the average release print has roughly 500-800 lines per picture height.

I understand the difference between negatives and release prints, but 1) you're still only talking resolution (a single aspect of "quality"), and 2) averages hide the fact that there's some variation between release prints, and even within a single print.  And what about colorspace?  What about audio?

I certainly wouldn't say Star Wars theatrically was 4K equivalent, nor did I intend to imply that.  Certainly some of the scenes with multiple optical effects are sub-DVD quality in the resolution department, but luckily that does not apply to the entire film, though it does bring down the average, perhaps even down to the levels you quote.  Audio is still Blu-ray quality (lossless, let's ignore the channel variations) throughout, and colorspace is better.  I'd say resolution exceeded 2k at the peaks, but neither of us has any data to back that up or refute it, so feel free to consider that claim dropped.  FWIW, my >2k claim is simply from the observation that when the local theatres switched from film to digital 2k, everything looked worse resolution-wise, and it didn't start looking good again to me until they upgraded again to 4k.  Certainly this was with a non-random sampling of films, possibly not average ones like yours, and like yours they were also not Star Wars, so it doesn't really apply.

danny_boy said:

Peer reviewed International studies conducted in theaters across the globe concluded that the average release print has roughly 500-800 lines per picture height.

Citation needed!

zombie84 said:

Yeah, I don't buy that study either. I've done extensive lab-quality tests of my own using Panaflex cameras, resolution charts and 35mm film, and I wasn't getting those results. 800 is a believable number, but 500 in the average? Sorry, there's a reason why we don't screen really good VHS tapes at theaters. There is great generational loss, but 500 discernable lines is pretty crappy, I find that hard to be typical, plus there are things other than resolution, which was the problem with early HD.

That's the problem with "average".  Assuming a credible study, they were averaging in hundreds of crappy 16mm slasher flicks shot in low light.  That's what could have been at the average theatre at the time, right?  And "globally" average too?--ugh, so Bollywood comprises most of the sample, great.  See the problem?  Averages are meaningless when trying to make a statement about something specific.  Ask any woman who wants to be a firefighter, and is told the "average woman" isn't strong enough, but nobody bothers to check if she's strong enough...

It's amazing to me that people actually believe moviegoers used to watch films at just-better-than-Laserdisc-quality.

Are these the same people claiming that GOUT is the best possible preservation of the OOT due to "70's film limitations"?

I agree that 500 lines of resolution could not POSSIBLY duplicate a release print, even if the prints were very hasty and wound up coming out blurry. 

Obviously the optical copying process gives the 35mm a kind of "fuzzy" look, but I've got to think there is way, way more resolution and picture information there, even though it is not going to be razor sharp like the negative.

From what you read, too, a Star Wars print in 70mm is an amazing experience. 

In any case, the quality and dynamic range of the audio would have been awesome and also the dynamic range of the colors would have been far greater than any commercial DVD.

SpilkaBilka said:

danny_boy said:

The irony is that the duplication process used in the fotochemical workflow  tended to "even out" the discrepancy between the parts of the o-neg that were 1st generation and the other parts that were optical duplicates.

So by the time you got to that 4th generation release print the film looked relatively seemless.

But if you strike a positive digital "print" from a 4K scan the difference in granularity between all the constituent parts that make up Star wars will be jarring.

I know very little about all this, but in a way I've come to the same conclusion, which I think is sort of the same point you're making here:  filmmakers account for generational loss, and it is in fact part of the intended look of the theatrical print.  (And therefore, wouldn't a scan of a theatrical print be the definitive way to watch SW, or really, any movie?)

Not sure if you saw it, but have you seen the thread where I asked about this:

http://originaltrilogy.com/forum/topic.cfm/Trying-to-understand-film-preservation-perhaps-a-stupid-question-but-shouldnt-digital-masters-be-struck-from-theatrical-prints/topic/14944/

I'd be interested to hear your thoughts.

Thanks for pointing me to that thread---I had not seen it before.

back in 1977 Lucas compensated for the quality differential between dupes and 1st generation by actually  using 2nd generation inserts to mask these same differences!

Other pieces weren't even original negative, but intentionally degraded duplicates that Mr. Lucas had stuck in to avoid emphasizing the quality of adjacent optical effects, some of which were so crude as to be almost unacceptable.

http://online.wsj.com/article/SB854660380658056000.html?mod=googlewsj

RickMcallum said that the CGI inserts used in 1997 had to be degraded to match the lower quality of the original negative.

All the new and enhanced stuff, which cost about $15 million, then had to be "degraded" to match the original images, McCallum added.

JOURNEY TO THE CENTER OF A FILM
`STAR WARS SPECIAL EDITION' PRODUCER TRACES THE RESTORATION AND RE-ENVISIONING OF THE ORIGINAL

SOURCE:    By Deborah Peterson

Of the Post-Dispatch Staff PUBLICATION: St. Louis Post-Dispatch
SECTION: GET OUT

DATE: January 30, 1997
EDITION: FIVE STAR LIFT
PAGE: 31

It is these compromises that pushed Lucas to go digital.

danny_boy: Where did you find the following?

Peer reviewed International studies conducted in theaters across the globe concluded that the average release print has roughly 500-800 lines per picture height.

This seems very wrong to me, so please provide the necessary citation(s) if you're claiming that it's correct.

CatBus said:

danny_boy said:

Peer reviewed International studies conducted in theaters across the globe concluded that the average release print has roughly 500-800 lines per picture height.

I understand the difference between negatives and release prints, but 1) you're still only talking resolution (a single aspect of "quality"), and 2) averages hide the fact that there's some variation between release prints, and even within a single print.  And what about colorspace?  What about audio?

I certainly wouldn't say Star Wars theatrically was 4K equivalent, nor did I intend to imply that.  Certainly some of the scenes with multiple optical effects are sub-DVD quality in the resolution department, but luckily that does not apply to the entire film, though it does bring down the average, perhaps even down to the levels you quote.  Audio is still Blu-ray quality (lossless, let's ignore the channel variations) throughout, and colorspace is better.  I'd say resolution exceeded 2k at the peaks, but neither of us has any data to back that up or refute it, so feel free to consider that claim dropped.  FWIW, my >2k claim is simply from the observation that when the local theatres switched from film to digital 2k, everything looked worse resolution-wise, and it didn't start looking good again to me until they upgraded again to 4k.  Certainly this was with a non-random sampling of films, possibly not average ones like yours, and like yours they were also not Star Wars, so it doesn't really apply.

Thanks for the clarification and I agree with your points.

Although we have to remember film has other problems too---the accumulation of dirt,tears,dye blobs and the sometimes unstable/jitter/jumps as it passes through the projector---all of this affects the perception of sharpness as experienced by the viewer.

All of the above would have been prevalent to audiences who hit the theaters to watch Star Wars in 1977-1991.

There is also a problem with tonal reproduction as you go through the generations.

Film did have a slight advantage in dynamic range over the first generation of digital cameras/projectors but even this has been matched or exceeded by the  latest iteration of digital cameras provided by the likes of Arri,Sony and RED(be they 2k or 4k).

Understanding this photochemical degradation is at the heart of why Lucas has done what he did with regards to the digital alterations.

zombie84 said:

Yeah, I don't buy that study either. I've done extensive lab-quality tests of my own using Panaflex cameras, resolution charts and 35mm film, and I wasn't getting those results. 800 is a believable number, but 500 in the average? Sorry, there's a reason why we don't screen really good VHS tapes at theaters. There is great generational loss, but 500 discernable lines is pretty crappy, I find that hard to be typical, plus there are things other than resolution, which was the problem with early HD.

AntcuFaalb said:

danny_boy: Where did you find the following?

Peer reviewed International studies conducted in theaters across the globe concluded that the average release print has roughly 500-800 lines per picture height.

This seems very wrong to me, so please provide the necessary citation(s) if you're claiming that it's correct.

Film theoretically has very good resolution capabilities. What is delivered to the theatre is another story. If we believe the ITU tests, then images captured at almost 2400 lines per picture height on the camera negative deliver significantly degraded on screen resolution through the projection system – in the range of 500 – 800 lines per picture height. 500 lines corresponds to about 9 line pairs per degree from 2 screen heights.

 http://www.etconsult.com/papers/Technical%20Issues%20in%20Cinema%20Resolution.pdf

I see you forgot to include the following:

It should be noted that this data is widely discredited by the film community – the numbers that resulted from looking at test patterns did not measure up to the resolution that is widely believed to exist on film. On the other hand, none of the discrediting parties has published any test results that show better performance. Politics cloud the science.

I guess I just don't understand how/why you think that one published report is the same as:

Peer reviewed International studies conducted in theaters across the globe concluded that the average release print has roughly 500-800 lines per picture height.

Across the globe? WTF? The table you provided clearly shows data from four cities, all on the same continent.

Please explain to me how the source you provided is:

1. Peer reviewed

2. International (conducted in theaters across the globe!)

3. Conclusive, at all

Bingowings said:

I see tables, but I refuse to believe mere facts without bar graphs, pie charts and Venn diagrams.

I need some numerical statistics, box plots, a frequency distribution, a line plot, and a scatter plot with regression data as well as a hypothesis test and a 98% confidence interval, and a written comparison of all of the above.

AntcuFaalb said:

I see you forgot to include the following:

It should be noted that this data is widely discredited by the film community – the numbers that resulted from looking at test patterns did not measure up to the resolution that is widely believed to exist on film. On the other hand, none of the discrediting parties has published any test results that show better performance. Politics cloud the science.

Ummmm----this proves my point!

On the other hand, none of the discrediting parties has published any test results that show better performance. Politics cloud the science.

AntcuFaalb said:

I guess I just don't understand how/why you think that one published report is the same as:

Peer reviewed International studies conducted in theaters across the globe concluded that the average release print has roughly 500-800 lines per picture height.

Across the globe? WTF? The table you provided clearly shows data from four cities, all on the same continent.

Please explain to me how the source you provided is:

1. Peer reviewed

2. International (conducted in theaters across the globe!)

3. Conclusive, at all

Sorry--I confused that study with this one:

http://www.cst.fr/IMG/pdf/35mm_resolution_english.pdf

Which was conducted in a couple of US and Canadian cities---aswell as Milan and Paris

And they reached the same conclusion:

The highest resolution that the assessors could still discern in the sharpest part of the screen(not neccassirily it's center) in the most performing movie theater was about 875 Lines/PH

danny_boy said:

AntcuFaalb said:

I see you forgot to include the following:

It should be noted that this data is widely discredited by the film community – the numbers that resulted from looking at test patterns did not measure up to the resolution that is widely believed to exist on film. On the other hand, none of the discrediting parties has published any test results that show better performance. Politics cloud the science.

 

Ummmm----this proves my point!

On the other hand, none of the discrediting parties has published any test results that show better performance. Politics cloud the science.

So a lack of further research "proves" your point? I must have learned the scientific method incorrectly, then.

Tyrphanax said:

Bingowings said:

I see tables, but I refuse to believe mere facts without bar graphs, pie charts and Venn diagrams.

I need some numerical statistics, box plots, a frequency distribution, a line plot, and a scatter plot with regression data as well as a hypothesis test and a 98% confidence interval, and a written comparison of all of the above.

Well how about you publish the above stats in favour of film!

In the meantime take it from Robert Harris(restorer of Lawrence Of Arabia and who offered to work on Star wars for free)-----regarding the Jaws Blu Ray

Bear in mind that Harris owned a 35mm release print of Jaws:

"The new Blu-ray is more highly resolved than would have been seen theatrically, with more accurate color and densities."

http://www.hometheaterforum.com/t/322696/a-few-words-about-jaws-in-blu-ray/60

danny_boy said:

AntcuFaalb said:

I guess I just don't understand how/why you think that one published report is the same as:

Peer reviewed International studies conducted in theaters across the globe concluded that the average release print has roughly 500-800 lines per picture height.

Across the globe? WTF? The table you provided clearly shows data from four cities, all on the same continent.

Please explain to me how the source you provided is:

1. Peer reviewed

2. International (conducted in theaters across the globe!)

3. Conclusive, at all

 

Sorry--I confused that study with this one:

http://www.cst.fr/IMG/pdf/35mm_resolution_english.pdf

Which was conducted in a couple of US and Canadian cities---aswell as Milan and Paris

And they reached the same conclusion:

The highest resolution that the assessors could still discern in the sharpest part of the screen(not neccassirily it's center) in the most performing movie theater was about 875 Lines/PH

 

This is a much better source. I will read this closely.

AntcuFaalb said:

danny_boy said:

AntcuFaalb said:

I see you forgot to include the following:

It should be noted that this data is widely discredited by the film community – the numbers that resulted from looking at test patterns did not measure up to the resolution that is widely believed to exist on film. On the other hand, none of the discrediting parties has published any test results that show better performance. Politics cloud the science.

 

Ummmm----this proves my point!

On the other hand, none of the discrediting parties has published any test results that show better performance. Politics cloud the science.

So a lack of further research "proves" your point? I must have learned the scientific method incorrectly, then.

On the contrary the research has concluded the following:

Lucasfilm technical director Mike Blanchard says, "Almost all of the resolution that’s lost is through the printing process. It’s really funny about technology and the film business right now. People get caught up in these numbers games that are flat-out ridiculous. They say, ‘Film is 4k,’ but it’s not 4k. It’s 4k on the camera negative, but no one has ever seen a camera negative projected. Countless studies have shown that what is shown in U.S. theaters [via the interpositive/internegative photochemical printing process] is between 700 and 800 lines of resolution when you get to the release print. We get that easily(using 1080*1920 digital capture and projection).

http://mixonline.com/sound4picture/f...ars_episode_2/

You show me official data that indicates that 4th generation 35 mm release prints exceed 1K-2K and I will offer a retraction.

Good luck---"You are gonna need it!"