Star Trek TNG Seasons Remastered on Blu-Ray - Page 5

They almost never use prints as the basis of a video transfer. Interpositives/internegatives or (rarely) the OCN is used.

Prints may be used and projected to give the colorist an approximation of what it's supposed to look like theatrically.

If DS9 gets a proper blu ray release and not simply uprezzed like the tribble ep on the TOS set, I will happily watch each and every moment again. I loved it when I did the whole shebang on DVD a few years back. Was so into it I watched it in 2 different states, Cali and NY via PJ and even laptop with headphones.

with all due respect, i don't know who you are or what you do and I haven't knowingly seen your work, so i won't take your opinions as gospel when I've read interviews and posts from DPs and post-production guys who say there IS a distinct benefit from scanning at 4K even if you're going to downsample it.

In general I have no idea what equipment was used to make a transfer. Criterion provides such information and they have quite a few Spirit 2K transfers made from OCNs that look dated and mushy.

Yes and no.

While it's not identical technology, I can tell you the optical quality, bit depth, noise level and color accuracy of my 90's era drum scanner pales in comparison to my 5 year old model even at the same resolution. The new one was half the cost, too.

I will say, that generally the difference shouldn't affect the final compressed image in a way that is noticeable despite the obvious difference in the masters - with one big caveat: it depends on the material. Some material compresses very poorly if the material you are compressing from starts out at reduced quality.

Further, it doesn't matter what era your scanner is from. If you don't use the correct settings or don't clean the friggin' thing, you aren't going to get a good scan at any resolution. Every scanner used to digitize a negative is capable of adjusting the image in software and those settings can wreak havoc on your final scan, resulting in all kinds of suck.

Unfortunately, you can't regulate stupidity or incompetence.

Again, the source image is at play here.

I'd wager a good vintage 65mm negative would show definite improvement at 4K after compression verses 2K. 35mm, not so much, and a poorly archived negative of any size would likely not benefit at all.

However, if there is restoration work to be done, you'll get a better result at 4K, which will hide itself better when compressed.

------------

That being said, the scan is likely to be the least of the worries here. Despite improvements in technology and optical clarity, scanners even from 20 years ago do a great job of reproducing the material, particularly since it's still meant to be seen on a TV, not a giant screen (or occasionally on a billboard in my case).

The issue that I think a lot of people are confusing with "the scan" is "the digital master", which is what the product is compressed and created from. If that digital master was created to exist as a 1080i DVC tape, anything created from that will be affected by it - especially since some of those early master tapes were often 1440x1080 or worse (and interlaced).

Of course, that was when drives were expensive compared to now. Now days a high rez master can be kept "cheaply" as a digital file allowing for a higher quality master to compress the final product from. There's little compelling reason to dump off the master to a format that may diminish quality. We actually pull the drive(s) we use for a client, and lock them in a vault. We never use them for anything else.

The client pays handsomely for all aspects of the work we provide. The drives are cheap compared to that and it allows us to regularly swap out old drives for new before they get into the failure zone. We take data loss very seriously.

http://www.efilm.com/publish/2008/05/19/4K%20plus.pdf
Pages 21-25 are the greatest interest.

http://www.efilm.com/publish/2004/09...n+for+film.pdf
Page 3.
"Resolution – The first generation of scanners generally scanned a 35mm frame at 2K (2048x1566 pixels) or 4K (4096x3112) (Figure 2). A common misconception is that a high-quality 2K resolution image is scanned at 2K. According to the Nyquist Frequency Rule, the theoretical maximum resolution a digital capture devise can resolve is only half of its sampling rate. Therefore, to achieve a true high-quality 2K can, the film image must actually be “double over-sampled” at 4K, and then mathematically “down-sampled” to 2K."

http://www.cintel.co.uk/bmt_mymedia/...solution_1.pdf
Older paper but it covers again why 4k will provide clearer 2k releases.

http://reduser.net/forum/showthread....l=1#post646904
This summarizes the key points from this pdf** from ARRI.

**http://www.arri.de/fileadmin/media/a...gyBrochure.pdf


Thanks for the clarification PeterTHX, my use of 'print' lends more to interpretation of say a release print rather than an IP/IN, which is more to what I was referring to. And quite a few of those can be in rather poor shape.

Best Regards
KvE

A bit from Sony on 2k/4k projection along with an anecdotal experiment.
http://pro.sony.com/bbsccms/static/f...K_WP_Final.pdf
http://pro.sony.com/bbsccms/assets/f...rch_Sheet8.pdf

PS This maybe blasphemous for some but I actually would prefer DS9 redone in HD first. With that said I love TNG especially season 2-3 onward yet I find the story/character arcs more compelling with DS9.

Again you can crib all the stuff off the net you like. People who actually work with this material are not quite so easily swayed by the marketing hyperbole over years of experience.

RED is not a film camera and is totally irrelevat to this discussion

I'll wager nothing that differentiates a good transfer from a bad transfer that you see on BD is primarily the result of a 4k scan in the mix. A new scan over a junk telecine from before the years of modern film scanning ie about mid 90s... yes , 2k vs 4k given good quality scanning...no.

http://www.avsforum.com/avs-vb/showt...6#post10619206

2k is a "half res" 35mm scan

"half res" 65mm is about 4k in terms of pixel count. I can't remember the exact number off hand but I have posted it on here before and yes I've dealth with them before too.

35mm Motion picture film scanners started out as 4k native , 4k scanning itself is not new and exotic . It hasn't just arrived as an innovation. The scanners responsible for 2k films scans right from the advent of the industry were on the whole 4k scanners.

Fairly early on people realised that for motion picture work you could get away with 2k (half res 4k) . 2k was even originally referred to as "half res" with 4k being "base"

65mm scanned at "half res" ie the equivalent sampling rate to a 2k 35mm scan is actually around the 4k mark anyway. A "base res" 65mm film scan was always about 8k from when it became possible to actually scan 65m film to the sort of quality you get on a modern film scanner. (again mid 90s)

And compression and downsample are seperate stages , downsample happens first from the 4k/2k DI then compression happens to the 1080p video master generated from that.

I am a firm believer from first hand experience working with scanning and resizing digital photographs that scanning at a higher resolution and downsampling leads to better results in the final digital picture. The main reason is there is a lot more detail that is captured during the scan that can be included in the final downsized pixel.

For example, imagine a single tiny white pixel surrounded by red pixels in the higher resolution scan. In the lower resolution scan, you may end up with either a white pixel or a red pixel. Having the higher resolution scan allows you to know that that pixel should be almost completely red, but slightly pink. If it was solid white, it would be just plain wrong.

Another example - aliasing. If you scan an image at lower resolution and you have a diagonal edge like the corner of a black table next to a white floor in the image, the edge might look blocky/rough with the high contrasting pixels next to each other. If you scan at a higher resolution, you can average out the black and white pixels along the edge and make it appear smooth (anti-aliasing). Even 3d video cards have or had used this method to fix the appearance of edges. I think it is called supersampling anti-aliasing.

So, I say, the higher resolution the scan, absolutely the better result that is attainable in the final blu-ray image.

Mike

2k film scanning is a downsample from a 4k scanner.
I've converted 2k and 4k 10bit log fim scans to 1080p video and the actual difference is something that is measurable in terms of generating slightly different pixels , however the actual visual difference is just not visible in my experience. Thats before you even factor in chroma subsampling and compression.

Additionally most 1080p film masters tend not to be resized from 2k to 1080p but are cropped 1:1 pixel from the 2k scan after a video colorspace conversion. Again i know this by comparing the actual 2k scans of various feature films I've worked on against the version that ended up on BD.

Again...actual experience with 2k and 4k film scans not just repeating some stuff I cribbed off the internet.

4k is a useful marketing term in terms of 1080p and BD generation and nothing more in my opinion.

I've even posted an example of this process in the post link above , and that image (which is extremely well known in post production and film circles) is a scan that dates from about 1993 , as scanned on one of the first film scanners that set the standard for modern film scanning and is still used for test purposes today ( so much for old 2k scans)

Well in your link, the 2K image is downsampled from a 4K scan. I don't think anyone is saying there is some remarkable difference between a 1080p master derived from an actual 4K image and a 2K DI made from a 4K scan or internally oversampled 2K (the Arriscan and Northlight both oversample, I think?). That doesn't make much sense. From what I know, the initial analog to digital conversion is where it's most advantageous to oversample.

Isn't this thread supposed to be about Star Trek TNG?

The 2k is a seperate scan , the scanner scans at 4k and then resamples on the fly to 2k , its not the 4k scan rendered then resampled to 2k. Its a "super2k" scan (fancy name).

This is indicative of how 2k scanning is carried out these days. Thats my point its already downsampled from 4k (some people unfortunately add some nasty sharpening at the same time).

I do actually say this in the post.

I fail to see how scanning motion picture film for 1080p mastering for BD is not relevant to the thread topic

... not anymore...

Anyone with a desktop scanner or a high quality DSLR camera can prove unequivocally, in the comfort of his own home, that 4K is always superior to 2K - before and after compression. It's not bleeping rocket science. More pixels = better.

TNG, DS9, Voyager, and all the ST films 1-10 should be redone. I'll pay top dollar. That is all.

hehe

If you don't have a degree in rocket-science, you cannot participate.

If this is actually the case, then enough said although there is still room for argument about the algorithm used for the resampling. However, typical scanners (not necessarily talking about the ones you are referring to) don't resample on the fly. They scan much faster when scanning at lower resolutions.

Not to open another can of worms, but actually this is not unequivocal especially when it comes to digital cameras. Packing more and more megapixels into the imaging sensor often has a negative impact on image quality. Assuming the pixels are captured perfectly, then your statement would be true.

http://www.cinesite.com/services/scanning-and-recording

It's relevant for discussion related to the film, not to veer off into the technology solely.

Well I'm quite happy to delete my posts if you feel they are irrelevant , no skin off my nose. Pearls before swine and all that.

Peter,

Are you referring to the Laemmle Royal Theater screenings that were held in the summer of last year? I was there for II, III, IV & VI and TWOK was a 1982 print without the "II" on the title card and it had significant blue layer fading on all the reels (resulting in a magenta hue to the image). So much so that they had to attempt to correct it by using a blue filter in the projector. But it wasn't entirely successful. The projectionist said that it was due to shoddy Metrocolor stock used at the time.

If this is indeed the screening you're referring to, I'm wondering how you can say that the Blu-ray matched the colors of the 70mm presentation? I would think that you would need so see an unfaded print to make that evaluation.

Did you attend a different screening than I did? It sounds as if you're referring to a screening later in the year with a better print than I saw that night.

HDTV | Blu-ray



The Enterprise is supposed to be white, not aqua

Aren't the exterior shots if the Enterprise leaving spacedock in Wrath the same shots from TMP, only edited to a few seconds instead of the 53 minutes they used in TMP?

One on the left looks magenta to me . One on the right looks as if they took the decision to swing the shadows colder ( notice the area of the spotlight cuts through the shadows and the blue) , Probably to give a more interesting color composition relative to the red supports.

I kinda like the one on the right. It also has a lot more color information and subtlety than the one on the left which looks desaturated in comparison.

This is often true of consumer-grade digital cameras, which is why I referenced specifically DSLR cameras. Shoddy lens quality combined with tiny sensors is never good. Perhaps I should have specifically stated "full-frame DSLR cameras" instead to be more clear. I'm not a professional photographer, but it is an intense hobby of mine. I have yet to see any of Canon's full-frame DSLRs look worse with more MPs.

Isn't this always the hope and goal of image capture?


Indeed. The Blu-ray shot makes the Enterprise look like a submarine. The warmer color in the HDTV shot would also be consistent with the type of lighting used (hint: it wasn't LED or CFL) as well as reflected light from the red superstructure of spacedock.

It kinda was always meant to be a submarine

This is irrelevant. The shooting lights would be balanced appropriately for the filmstock used , therefor the end result wouldn't notionally have a cast one way or the other. Additionaly this is a plate thats been through an optically printed compositing process , and then the final shot has been scanned , color corrected and converted to video colorspace including a rebalance from D.55 to D65.

You are unable to differentiate creative intent from shooting conditions in terms of colour on any of the material you would watch on BD let alone a optical composite. There is really nothing to infer about notional "correctness" at the end consumer delivery beyond the creative intent of the video grade. (or the grade in video).

Cool, well I hope you enjoy your ST film Blu-rays, I'll be waiting (perhaps a long time) for proper ones.

So anyway... I just scoured Google and couldn't find any more details regarding the TNG Blu-ray rumors. One good point brought up by a random ST geek on another forum was that at the time when TOS was rumored to be going HD, CBS had already made significant progress. So it is possible that CBS is already quite far along with TNG. I'd still love to see an official announcement, but perhaps we'll see one after all the ST series hit Netflix? I sure hope we don't have to wait for ST12...

I don't readily have an opinion on the transfers one way or the other based on those stills . I'm clarifying observations that would apply to any video mastering of something originated on film.