Jump to content

How a 2k or 4k works?


Dimitrios Koukas

Recommended Posts

  • Premium Member

Hello,

I need some explanations about 2k or 4k, systems. The terminology 4k is been used to clarify the fact that an image 4k is used (1024kx4=4096), but what this means for the result?

How this is achieved? Frame by frame or real time?

Dimitrios Koukas

Link to comment
Share on other sites

  • Premium Member
Hello,

I need some explanations about 2k or 4k, systems. The terminology 4k is been used to clarify the fact that an image 4k is used (1024kx4=4096), but what this means for the result?

How this is achieved? Frame by frame or real time?

Dimitrios Koukas

 

 

Hi,

 

Negative can be scanned at 1K, 2K, 4K or bigger. It's not a real time process.

 

A S35 full frame 2k is 2048x1556 and 4K is 4096x3112

 

Normal 35 1:1.85 2K is 1828x988 and 4K is 3656x1976

 

These are big files, At 8 bit a 4k S35 file is about 36mb and a 2k file about 9 mb.

 

The greater the resolution and higher bit depth will yield the best results. However it's usually a compromise as any rendering starts taking a long time!

 

So to how it works, shoot as normal (S16,35,S35), scan into computer, carry out post production and output to film!

 

Stephen

Link to comment
Share on other sites

Hi Stephen,

 

Thanks for your explanation.

 

Would you compare the costs involved in scanning at 2K, 4K and 24p HD ? I know 4K is more expensive but ... how more expensive ?

 

And ... do people scan 24p HD for DI ? (I believe for a 35mm negative that's not enough but how about S16 ?)

 

Sorry if this question has been already posted. I did a quick search but didn't find anything.

 

Thanks,

Link to comment
Share on other sites

  • Premium Member

If you shot in 24P HD, there would be no scanning necessary. If you mean, can you telecine film to 24P HD tape, then yes. With HD-D5 or HDCAM-SR tape, it would be 1920 pixels across, similar enough to a 2K scan. With HDCAM-SR, you can also get 4:4:4 color, not subsambled color, and the compression is very low. I would say that an all-2K post and an HDCAM-SR post would produce similar (just slightly lower) quality but with a few different tool sets possibly (plus an HD telecine transfer would not be pin-registered like a 2K scan.) Since a 16mm frame is half as wide as a 35mm frame, then a 2K scan would be about the same resolution as a 4K scan of a 35mm frame. So I think an HDCAM-SR D.I. for Super-16 would probably look fine.

 

Well, 4K is four-times the data of 2K -- would that mean four-times the costs overall?

 

For D.I. work, many of the bigger films are scanning at 4K and then downrezzing to 2K after that for everything else. This preserves a little more fine detail than an all-2K approach.

Link to comment
Share on other sites

If you shot in 24P HD, there would be no scanning necessary. If you mean, can you telecine film to 24P HD tape, then yes. With HD-D5 or HDCAM-SR tape, it would be 1920 pixels across, similar enough to a 2K scan. With HDCAM-SR, you can also get 4:4:4 color, not subsambled color, and the compression is very low. I would say that an all-2K post and an HDCAM-SR post would produce similar (just slightly lower) quality but with a few different tool sets possibly (plus an HD telecine transfer would not be pin-registered like a 2K scan.) Since a 16mm frame is half as wide as a 35mm frame, then a 2K scan would be about the same resolution as a 4K scan of a 35mm frame. So I think an HDCAM-SR D.I. for Super-16 would probably look fine.

 

Well, 4K is four-times the data of 2K -- would that mean four-times the costs overall?

 

For D.I. work, many of the bigger films are scanning at 4K and then downrezzing to 2K after that for everything else. This preserves a little more fine detail than an all-2K approach.

 

 

What I meant was telecine film to 24 HD.

 

Yeah, I read on your book that "City of God" used an HD intermediate (HDCAM if I'm not mistaken), so that's why I asked the question. I wanted to know if that's become a common practice or it was just an isolated case.

 

 

Thanks,

Edited by ropbo
Link to comment
Share on other sites

  • Premium Member

For lower-budget people, HD in various flavors has become one method of doing a D.I. for 16mm/Super-16, even some 35mm (I had to use HDCAM-SR for a 35mm feature due to budget issues; the director & editor had cut the film in such as way as to make a D.I. necessary even though it was not part of the budget, nor had I planned on one. HD was the only affordable option for the post.)

 

Yes, "City of God" used HDCAM, which is not a great format for D.I. work -- it's only 1440 x 1080 pixels, highly compressed, and 3:1:1 color, giving the D.I. more of a video-ish look.

 

In theaters, I've seen HD used for D.I.'s more for documentaries that mix formats (like "Fahrenheit 9-11") than for narrative fiction work so far.

Link to comment
Share on other sites

  • Premium Member
So to how it works, shoot as normal (S16,35,S35), scan into computer, carry out post production and output to film!

 

Stephen

 

Stephen thank you very much,

I was just wondering ''how it works'' meanning the printing method.

Is it a video monitor that is scanned? or a projected image on lcd or DLP , does the procedure comes in a continuous light or by flashes on the negative?

Dimitrios Koukas

Link to comment
Share on other sites

  • Premium Member

A "film recorder" is the device used to transfer digital files to film; the most common types are LASER RECORDERS (like the Arrilaser) and CRT RECORDERS (like the Celco). There was also something called an EBR (ELECTRON BEAM RECORDER) but few were made and are hardly used now.

 

Laser recorders usually record to intermediate stock, producing either an IN or IP. CRT recorders use camera negative stock.

 

http://en.wikipedia.org/wiki/Film_recorder

 

For converting film images into digital ones, a "film scanner" or "telecine" is used. Scanners tend to be slower, pin-registered, and do one frame at a time. Telecines tend to be faster and scan & transfer continuously, and often record the image in a video format of some sort, not as RGB data. You can think of a telecine as a form of scanner, especially a 'datacine' like the Spirit, although usually there is a distinction made at post houses between a data scan and something transferred on a telecine.

 

Mike Most, Dominic Case, Phil Rhodes, etc. can explain this more accurately than I can...

Link to comment
Share on other sites

  • Premium Member
A "film recorder" is the device used to transfer digital files to film; the most common types are LASER RECORDERS (like the Arrilaser) and CRT RECORDERS (like the Celco). There was also something called an EBR (ELECTRON BEAM RECORDER) but few were made and are hardly used now.

 

Laser recorders usually record to intermediate stock, producing either an IN or IP. CRT recorders use camera negative stock.

 

http://en.wikipedia.org/wiki/Film_recorder

 

For converting film images into digital ones, a "film scanner" or "telecine" is used. Scanners tend to be slower, pin-registered, and do one frame at a time. Telecines tend to be faster and scan & transfer continuously, and often record the image in a video format of some sort, not as RGB data. You can think of a telecine as a form of scanner, especially a 'datacine' like the Spirit, although usually there is a distinction made at post houses between a data scan and something transferred on a telecine.

 

Mike Most, Dominic Case, Phil Rhodes, etc. can explain this more accurately than I can...

 

:)

David...

Thanks for the telecine explanation, I believe I ve used one first time back in 1992.

I have made transfers in my life, film to_digital_or_whatever_format way, it's just I haven't ever did the opposite,(video to film) only experimenting thru a crt monitor.

I guess my question was very general, so I am sorry for this.

Thank you , and what about the sound? This comes after I guess?

Dimitrios Koukas

Link to comment
Share on other sites

  • Premium Member

Just like with a photochemical post, you'd take your digital master from the sound mix and get it transferred to an optical negative for printing the soundtrack onto the print.

 

You'd probably want to take a silent check print made off of the digital negative of the recorded image into a mix stage and run it with the digital sound master just to double-check the overall sync if you had done your final mix to a video edit of the movie.

Link to comment
Share on other sites

  • Premium Member
Hi Stephen,

 

Thanks for your explanation.

 

Would you compare the costs involved in scanning at 2K, 4K and 24p HD ? I know 4K is more expensive but ... how more expensive ?

 

And ... do people scan 24p HD for DI ? (I believe for a 35mm negative that's not enough but how about S16 ?)

 

Sorry if this question has been already posted. I did a quick search but didn't find anything.

 

Thanks,

 

Hi,

 

For costs I have looked at www.arri.de

 

They charge a set up charge of 150 euro per roll, up to 1000'

 

2K scan 1.25 euro per frame scanned to log. Log to Lin conversion (TIFF) .20 per frame if you wanted that.

4K scan 1.90 euro per frame

 

A Spirit telecine onto HDCAM SR will be 1000+ euro per hour.

 

A Spirit telecine can also produce DVX files if it has the Datacine option. About 6 frames a second uprezzed to 2K

 

A HD finish is only a small premium to DigiBeta, is real time and is easily understood by people in the video business.

 

2K and 4K is much slower as very few computers can manipulate the huge quantity of data in real time. 4K sounds like a great idea but in reality you need a very big budget to stay 4K all the way.

 

2K is about the practical limit for super 16, however if shooting super 16 to save money then a HD finish is quite likely.

 

Stephen

Link to comment
Share on other sites

  • 6 months later...
For lower-budget people, HD in various flavors has become one method of doing a D.I. for 16mm/Super-16, even some 35mm (I had to use HDCAM-SR for a 35mm feature due to budget issues...

 

What was the aspect ratio of the film? If you do a HDCAM-SR DI for a Super35 2.35 film, does the 2.35 image get optically squeezed to use the whole 16x9 HD frame, or does a fair amount of the vertical resolution of the tape get lost? What about Super 35 with a 2k DI, where the native aspect ratio is 4x3? Again, would the 2.35 image be squeezed to fit into the 4x3 frame for later film-out to anamorphic print, or does some of the vertical resolution of the scan go unused?

 

I am trying to figure out the cheapest way to do a DI for a Super35 2.35 film. I did an HDCAM-SR DI for a Super16 film, but I think it would be silly to shoot on 35 and then end up using less digital resolution than with Super16 just because we want to shoot in the widescreen aspect ratio.

-Josh Silfen

Link to comment
Share on other sites

  • Premium Member

Well, I don't think there is a major difference in resolution -- Super-35 scanned at 2K is 2048 pixels across, in HD, 1920 pixels across. If you had shot in 35mm anamorphic, a "2K" scan would only be 1828 pixels across anyway.

 

In terms of vertical resolution, with HD you are limited to 1080 pixels. So if you had shot in 35mm anamorphic, there would be some advantage in using all 1080 pixels for the scope image since it already has a 2X squeeze and just compress it horizontally to fit within 1920 pixels.

 

But if you had shot in Super-35, then your 2.35 image inside Full Aperture doesn't take up the whole negative anyway, so you might as well just transfer a 1.78 area of the Super-35 neg to fill 16x9 HD, and then crop & stretch for the final output to scope. A 2K D.I. does the same thing -- even if they can scan the whole height of the 4-perf Full Aperture frame, they crop the scan to 2.35 before recording it out to scope. The only advantage you gain from scanning the whole 1.33 Super-35 frame is the ability to reframe vertically or to create the 4x3 TV version with less panning & scanning.

 

I'm not saying that there is no loss doing an HDCAM-SR D.I. instead of a 2K one, just that they are closer in quality than you'd think, especially if you had shot in Super-35 instead of anamorphic.

Link to comment
Share on other sites

  • Premium Member

Re: squeezing a 2.35 image on transfer to fill an HD frame

I've run that one by a number of post houses, and although they've all conceded the possibility of that approach, I haven't found anyone as of yet that has tried it. If you shoot flat, you're probably best to transfer flat, as well, saving an anamorphic squeeze for the film out. (You'll want your HD master to be flat, anyway.) In dealing with this abstraction, the squeeze being done on transfer to HD would be an atypical one, since the usual 2:1 anamorphic squeeze places your image within a 4X3 frame, which would take up even less HD real estate horizontally.

 

As David has stated, the quality of 2K vs. HDCAM SR is very close. Some transfer engineers at major D.I. facilities quietly concede that it's even more than close. Also, the Spirit Datacine is starting to be used for high-end productions (Walk the Line, Lucky Number Slevin, etc.) that chose this method of transfer over traditional film scanning. That, and a growing consensus that log files don't necessarily lead to an easily perceived qualitative difference over linear transfer, is why HDCAM SR is becoming a viable intermediate format.

Link to comment
Share on other sites

As David has stated, the quality of 2K vs. HDCAM SR is very close. Some transfer engineers at major D.I. facilities quietly concede that it's even more than close. Also, the Spirit Datacine is starting to be used for high-end productions (Walk the Line, Lucky Number Slevin, etc.) that chose this method of transfer over traditional film scanning. That, and a growing consensus that log files don't necessarily lead to an easily perceived qualitative difference over linear transfer, is why HDCAM SR is becoming a viable intermediate format.

 

Well, I know that many 2.35 movies have done a 2k DI, and I know there is little perceivable difference between 2k and HDCAM SR, so I guess a Super35 2.35 movie could have an HDCAM SR DI and still look decent, but it still seems very scary to me to actually be using less of the HDCAM SR resolution for a 35mm film than for a 1.85 or 1.78 Super16 film.

-Josh Silfen

Link to comment
Share on other sites

  • Premium Member
Well, I know that many 2.35 movies have done a 2k DI, and I know there is little perceivable difference between 2k and HDCAM SR, so I guess a Super35 2.35 movie could have an HDCAM SR DI and still look decent, but it still seems very scary to me to actually be using less of the HDCAM SR resolution for a 35mm film than for a 1.85 or 1.78 Super16 film.

-Josh Silfen

 

If you scan Super-16 or Super-35 at 2K in the same aspect ratio (let's say, 1.78), you have the same pixel resolution regardless of the film's resolution. And even in a traditional 2K scan of Super-35, you crop the scan vertically to create 2.35. So you're using about 2048 x 871 pixels for 2.35, versus about 1920 x 800 if you use HDCAM-SR.

Link to comment
Share on other sites

Taking in all the issues discussed in this thread, I had a question. If someone was producing a low-budget, independent narrative feature with an eye towards a (hopeful) theatrical distribution, and originating on 35mm would just be too expensive, would your recommend a) originate on Super 16, D.I. in HD and print to 35mm or B) simply originate on HD, keeping everything digital until your final out to 35? I realize there are numerous variables that would come into the question, but just generally speaking and assuming the cost would be about the same to go either way, is one way likely to give you a better overall look and feel than the other.

Link to comment
Share on other sites

assuming the cost would be about the same to go either way, is one way likely to give you a better overall look and feel than the other.
Yes there are indeed many variables that could swing your decision one way or the other.

 

Super 16 film will capture a wider brightness range than any digital format - though top of the range cameras like the Genesis will come close (but then costs come back into the equation). With the wider brightness range you are better equipped to shoot certain types of scene, and you also have more latitude to grade the image in the DI post production stages.

 

If there is a lot of low light, night, or available light work,then you will have the choice of film grain or electronic noise. Each has a look that some people prefer. Even in well-lit 35mm work it is possible to see the difference in the structure of the image - I guess it's a choice of one look over the other.

 

On the other hand, if you are looking to have a very high shooting ratio (documentary style, animals, children, etc, then you can be more relaxed with HD. Even more heavily directed & acted scenes can benefit, as you just leave the camera rolling for another take, which keeps make-up and hairstylists away between takes!

 

And then there are things like overcranking. Undercranking is becoming easier on a number of digital cameras, but most of them are limited in their higher speeds.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...