Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Why 4K workflow?

C

caseyhayward

Active member
Joined
Sep 9, 2007
Messages
42
Reaction score
0
Points
0
I have been following the recent threads on native support for r3d in Premiere and all the complaints about proxies but I am left scratching my head.

Let me first say that I understand completely why someone would shoot 4k.

It seems like most productions (aside from film-outs and projection on 4K projectors) would not benefit from being 4k. So for most work, why is native 4k important? It seems like it would slow down machines that already have trouble with 1080 and 2k.

What am I missing? Why at this point in time do we need 4k finishing? Certainly not for broadcast, Bluray or web.
 
well, if you have access to the SONY's 4K projectors then it would be interesting.... otherwise no business at all..

But also, if you want to print out to film for 4k distribution, someone might think it is a good idea....

But all these, can be done on mastering/online and not editing... Editing in 4k seems to be no point... except for VFX work... like AFX, shake, nuke, flame etc.... but editing...?

And just for the fun sake; I discovered that a company in my city has two, yes you are reading right, two SONY 4k projectors stacked in a lab...

So now I just have to get some time of, do do some real testing...

But what about playback, they have a big PC, with a dual output DVI card and a splitter box, for taking all the different signals to the 4 DVI inputs on the projector.... so the next few weeks will be quite interesting...

cheers. Helge
 
Casey,

Imagine working with a 4k file that will debayer and play and color correct in real time right on the timeline of your NLE. If you want to make an Indy Film, imagine a 4k edited master that can sent for distibution/film out. Image only needing 1080 so you open a 1080 timeline and import your 4k file, it scales to 1080 in real time, plays and color corrects with 3d LUTS all in real time. And if you want to reposition or zoom in, that can be done in real time too because your original file is 4k, and thereby plenty large enough for repositioning and pushing in. Imagine all of this done just with the CPU in real time non destructively, never altering the original 4k file. Always ready for real time changes in all the above. That is the power and flexibility of a 4k workflow. Now stop imagining and go to cineform.com. Scratch $70k plus or Scratch light $30k plus offers real time playback and color correction to 1080 and I think 2k, but your color correction is baked in and you have none of the before mentioned flexibility.
 
I'm going to give you a film analogy: in commercials even though the delivery format up until recently and surprisingly often still today the delivery format is/was puny sd. Yet, when finishing the commercial, after editing, fx, graphics, etc., one would always go back and work off the original film to do the final color, not a tape to tape. Why? Because in color correction you decide what information to keep that will make it through the limitations of the delivery format and what information to throw to the wayside. Oversampling is the cornerstone of a quality image.
 
Still, the answer to why would your workflow have to be 4k is not clear. Even if you produce for HD you dont need a 4k workflow. Besides, where would you monitor/grade such produce.
 
Why 4K workflow?

caseyhayward said:
What am I missing? Why at this point in time do we need 4k finishing? Certainly not for broadcast, Bluray or web.
Click to expand...

There are several reasons for making a 4K DI rather than 2K DI, particularly when the DI is 48bpp (16bit RGB TIFF) uncompressed all the way through up to mastering for output:

1) When you send a 4K output uncompressed to the film recorder you get 4 to 16 pixels from the TIFF file being seen as 1 pixel on the projected film, this results is color depth greater than what the D2A converter in the film recorder can produce reducing tone banding and improving the color and tone quality with less pixel to pixel noise since the pixels are mixed, this results in something closer to film-to-film optical printing.

2) When you resize images from 4K down to broadcast, HD, SD, and DVD if you do it properly you also average several pixels to get smoother lower noise results. By coming from a 4K compressed image down to the lower resolution compressed formats the double encoding artifacts will be less since the 4K encoding artifacts from the RED CODE (tm) are smaller to start with.

3) When you have a 4K DI you have something that will look "current" in the near or far future when all production switches to 4K+ projection, disk, and broadcast. If your DI is only 2K or HD your product will look out of date sooner.

4) When you do CC on uncompressed 4K files you save tone values in groups of pixels, this can help give smoother tones and fill in gaps in the histogram caused by additional CC and gamma adjustments done later with LUT in the film recorder or other distribution since the pixels get averaged together so the tone resolution in your eye is greater than the bits value because you cannot see single pixels at 4K at normal viewing distances.

5) Alaising artifacts can be smaller. The way the images are resized is important since poor methods can make alaising worse. It is better to downres as the last step after the 4K uncompressed images have been CC, edited, and had titles added, this hides some artifacts since the artifacts are smaller in the 4K frames, and get blurred in the down res process, a reduction print looks better than a contact print.

That is why I am working on my "freeish" 4K uncompressed DI system to mix digital camera and film camera footage and do the DI uncompressed the whole way through, that way you can archive the best quality for future use without having to have the high cost roadblock of redoing post. Better to just do it right while you have the time, things that look good may have better legs.

If compressed files are used for the 4K DI then the individual pixels do not have their own true value, so the gain in accurate tone rendering could be less than going all the way uncompressed.
 
If you have the means to deal with a 4k DI, go for it. Not everyone has the resources to do so. That's why a majority of even studio DIs are 2k.
 
I think the best answer is in the panavision series of videos on digital cinema.

They talk about MTF, and it really makes sense to take that principle all the way through. Its a bit complex to answer here, but go and watch those video's, because they wil enlighten you on many things not just MTF.

So 4k DI is really important if you can do it.
 
One way to think about it:

When you (re)sample something, you will always have one or two of these artifacts:
Aliasing (seen as moire or spurious bands of detail that aren't in the scene)
Loss of frequency response (seen as blurriness)
Ringing (halos)

Check out the pictures and article here:
http://www.cambridgeincolour.com/tutorials/image-resize-for-web.htm

You can think of the world as being very high resolution. When you sample that with a camera, you'll have to tradeoff between resolution and aliasing. You can't have ringing artifacts because it's optically impossible. Now to make a good HD image, it would look best if it had a certain level of ringing since we have to pick our poison and ringing looks the least bad when downsampling (usually*; but the human visual system does weird crap).

The only way to get a good HD image with ringing is to sample at higher than HD resolution and then digitally downsample to HD.

2- Hmm I don't think I answered the OP's question.

Why at this point in time do we need 4k finishing?
Click to expand...
4k digital cinema, large public signage, and RedRay might be possible distribution means.

Obviously if you're delivering on Bluray it won't do 4K. As to whether it's worth going to the expense of doing 4K or if current 4K projection actually looks better than 2K projection (4K has inferior blacks) is up for you to decide.
 
If the performance is there I certainly have no problem being 'forced' to edit in 4K. The lovely thing is that once I'm done with my 4K edit I can derive easily a 2K version, HD 1080p and 720p and of course SD. It seems like the data rate of Redcode ought to put it into the category of usable performance on a fast machine the same way native DVCPROHD, HDV and XDCAM EX editing is done, not too mention ProRes HQ.

Of course you don't need to see 1:1 pixels 4K on your NLE just as I suspect most folks are not actually viewing their canvas in FCP at 1920x1080 for HD work, more like 50% or even less. Should be fine- again as long as playback and effects handling is robust.

Noah
 
Just to add to my post with some real world evidence.

Watch a SD DVD shot with 35mm film. Then watch a SD DVD the was shot on DV. Notice the Huge difference. Now if, when you down sampled to a lower codec or whatever you lost all that info, the 35mm would look the same as the DV DVD. But it doesnt. The higher you acquire and Finish at the better the delivery format. as i said for more in depth look at the panavision Vids that have to do with MTF.

Its amazing. Really informative.
 
4K over 2K...

4K over 2K...

chocblu said:
Notice the Huge difference. ...MTF.
Click to expand...

With my system it does not matter if you are working at 6K, 4K, 3K, 2K, HD, SD the time to process the files is just the difference, the editing is all done with low res proxy frames of the same res anyway. The CC works with the full res key frames, but displays at 1K since that is close to what you see on a film print in a movie house (the projector is almost never in focus anyway). The sharp/soft filter has a 1:1 display for a part of the frame so you can see what the result is, you cannot judge filters with a reduced resolution display, but you should make lower res copies of the filtered frames to check to see how the filtering results look. It is best not to sharpen at 4K since later display can have built in sharpen on the down res and the two combined can make the result look like bad video. Since the monitor is limited to about 2048x1536 when you look at keyframes in any graphics program you see 1/4 the 4K frame, but that should be enough to judge the noise and artifacts level.

There does not seem to be any "brick wall" between doing a 2K DI or 4K DI it just takes four times as many Harddrives to store the 4K frames, and it takes about 5 times as long to process the data. If one has worked in film then you would not have a problem with processing times of several months, people from a real-time video background might need to relax a bit. If you cut a negative on a cement splicer it takes a month or more (negative not workprint) but doing it with a DI you do not get fingerprints dust and scratches.

For myself, I would like the DI to be as close to good Film-to-Film optical printing as possable, and going uncompressed 4K DI helps in that, without many of the drawbacks film has, even if it takes 6 months+ extra or so.

Note, Some tasks or programs used require Windows ME (tm) or Windows XP Home SP3 (tm) with the FreeDOS FAT32 (tm) disk mounted as a slave drive.
 
WHAT "uncompressed" and "freeish" workflow?

WHAT "uncompressed" and "freeish" workflow?

"That is why I am working on my "freeish" 4K uncompressed DI system to mix digital camera and film camera footage and do the DI uncompressed the whole way through, that way you can archive the best quality for future use without having to have the high cost roadblock of redoing post."

As far as I know, all Red One footage is compressed in a wavelet algorithm. Where is the uncompressed coming from and where is the freeish coming from?

Cineform is the only 4k online editing I have heard about. And that is definitely compressed.

So, what is this new workflow?
 
Check out dancad3d.com for information on the software Dan is working on. I believe that is what he's referring to.
 
This theme has been beaten to death on these forums before.
The people who actually make movies still use film, 2K, etc.
The amateurs who think buying a camera body will make them Steven Spielberg, keep talking about 4K, etc.

The bottom line: Go buy a Sony handycam and try to make something, anything, instead of endlessly debating 4K V 2K.
 
Uncompressed 0K-6K+ "freeish" workflow

Uncompressed 0K-6K+ "freeish" workflow

mail4joeg said:
"That is why I am working on my "freeish" 4K uncompressed DI system to mix digital camera and film camera footage and do the DI uncompressed the whole way through, that way you can archive the best quality for future use without having to have the high cost roadblock of redoing post."

As far as I know, all Red One footage is compressed in a wavelet algorithm. Where is the uncompressed coming from and where is the freeish coming from?

Cineform is the only 4k online editing I have heard about. And that is definitely compressed.

So, what is this new workflow?
Click to expand...

Note, Some tasks or programs used require Windows ME (tm) or Windows XP Home SP3 (tm) with the FreeDOS FAT32 (tm) disk mounted as a slave drive.

I will answer this here rather than just by a PM since it may be of interest to several people who want to make feature films.

I have been working on a uncompressed DI system for several years, and am adding some extra features so that it can be used with the cameras from RED ONE (tm)+ or other Digital Cinema cameras that shoot uncompressed to get high quality results.

Now I am adding TIF import/export and have import working with footage shot with RED ONE (tm) using these workflows (up to output frames, the steps past that have been done with film scans or are being investigated).,


1) RED ONE > R3D > REDCINE > TIF > DANCAD87.EXE > output frames+WAV mix > (do what you want with the edited CC frames)

1b) RED ONE > R3D > REDCINE > TIF > DANCAD87.EXE > output frames+WAV mix > VirtualDUB > uncompressed AVI > TMPGEnc > DVD

1c) RED ONE > R3D > REDCINE > TIF > DANCAD87.EXE > output frames+WAV mix > VirtualDUB > uncompressed AVI > blue-ray encoder > Blu-Ray disk.


2) RED ONE > R3D > REDCINE > TIF > DANCAD87.EXE > output frames (WAV mix goes to sound flow) > DANCINEL > 35mm FILM print (sound from optical negative) or IMAX 70mm print (interlock sound).


3) Any FILM SCAN > TIF > DANCAD87.EXE > output frames (WAV mix goes to sound flow) > DANCINEL > 35mm FILM print

3b) Any FILM SCAN > TIF > DANCAD87.EXE > output frames (WAV mix goes to sound flow) > Any Film Recorder > 35mm FILM print


4) 8/16/35/70mm film > DSLR or USB/FireWire camera downloader + DANCINES > RAW BAYER > DCRAW > TIF > DANCAD87.EXE > output frames (WAV mix goes to sound flow) > DANCINEL > 35mm FILM print (print SMPTE time code or optical sound on print)


5) DSLR > RAW BAYER > DCRAW > TIF > DANCAD87.EXE > output frames (WAV mix goes to sound flow) > DANCINEL > 35mm FILM print

5a) DSLR > RAW BAYER > DCRAW > TIF > DANCAD87.EXE > output frames+WAV mix > VirtualDub > uncompressed AVI > TMPGEnc > DVD


6) 8/16/35/70 film > RED ONE > REDCINE > TIF > DANCAD87.EXE > output frames > DANCINEL > 35mm Film Print

6a) 8/16/35/70 film > RED ONE > REDCINE > TIF > DANCAD87.EXE > output frames+Wav mix > VirtualDub > uncompressed AVI > [something?] > 4K projector


Or many variations on that workflow. Uncompressed avoids double compression lossed added to the RED CODE (tm) losses when the RED ONE (tm) is used as the source camera. Right now to test the TIF would need to be converted to BMP with something like XnView (tm), but that step should gone soon as I have things going on unreleased versions...

Yes RED ONE is compressed, but working uncompressed from that point on is better since the pixel values between 12 bits form during CC and working uncompressed avoids double compression losses. Uncompressed TIF files can hold more of the tone values generated than some other file formats. For archive of video gamma 2.2 graded shots I have a smaller uncompressed format that saves 11 bits RED, 11 Bits Green, and 10 bits blue. For working I have my own 48bit file format that can a little smaller than TIF and is uncompressed and lossless and may load/save faster on a PC.

My system is not just a random access editor, it does complex feature film sound mixing, you get 10000 shots, 512 tracks per shot, plus 512 tracks for the whole length of the film project, there are built in noise gates, and compressors, and it can output to up to 16 channels in one pass automatic sound mixing. The system was originaly written for mixing feature film sound tracks automaticly so one person could make a quality sound mix without limits, TB of sound frames can be used, and special code reduces the disk space requires without compression of the sound information.

The DI image part was added to support the sound mixing features, and has film like color correction that works with any source, color negative or color positive, video frame files, or Digital Cinema frame files, DSLR frames, scanner images, or paint program images, etc. It automaticly resizes images, so it is better to resize everything up to 4K rather than down to 2K to hold the best quality for all sources before the CC is finished.

I hope to have the TIF version out soon, and am working on some XP issues now.

You can set up a "render farm" using FreeDOS FAT32 for low cost 4K frame processing and audio frame storage on cheep PCs without fancy hardware.

If you have any questions feel free to ask me what I am doing, this is our "in house" DI system that works with uncompressed 35mm film scans made on a modified Oxberry optical printer using my DIY DANCINES.EXE (tm) program, and the filmout will be made using my DIY DANCINEL.EXE (tm) program.

For people who want to use a non-DIY film recorder TIF output will let them keep good quality, as was mentioned above a 4K uncompressed filmout where the film recorder reproduces the pixels in the TIF file 1:1 will give more tone values than the same output at 2K because the D2A converter in the film recorder limits the color tone depth, four pixel groups give you two more efective bits over what you get from the D2A converter at 2K, for results more like film-to-film optical printing.

The results I am getting from the RED ONE (tm) footage look quite good when you start with the 4K TIF (REDLog) files, much of the blown out footage shown here seems to be due to the way the grading has been done, and is not the fault of the RED ONE (tm) camera when exposed properly, you cannot judge the quality of the RED ONE (tm) from the messed up JPG files, especially since when they are uploaded here they seem to get re-encoded. In my opinion the RED (tm) people are doing a disservice to their camera by not allowing posting of 1K, 2K and 4K BMP frames rather than JPG since the uncompressed BMP would show fewer artifacts, artifacts that are not present in the RED ONE (tm) footage when processed well.

The "freeish" part comes from the fact that right now I am working on my programs and am not charging for there use in making feature films, you can install them on any reasonable number of computers you own yourself for use in parallel computing of the many shots or reels. I do ask that you tell me if you have any problems of any kind since I goal is to have zero bugs, and to have software that runs for months on end without a single error or fault. You would be a "Beta Tester" and would have to live with lost data if that happens, so making backups you be a good idea.

If you want to try my system it is open, the frames are all out on the harddrives to be worked with by other programs if need be, you can make a workprint on video tape with a SVGA to NTSC/PAL converter box and load that into any other editor with the SMPTE frames burned in to keep track of the changes.

Each shot goes in its own folder, frames are numbered with frame zero as the slate close frame, audio frames are clap or beep on frame zero as well, that lets you use both double and single system sound from any source. I have a way to make sub-titles and rolling titles, as well as macro commands that can be used for simple image composites (these are 24bpp now but there may be 48bpp or such commands later.) This is a digital film editing system, not a video editing type system, it has low res proxy 24, 25, 30 fps payback for workprint editing, that quality does not effect the end result, it is just for checking sound sync, like a movie viewer for editing the frames and picking the "trim" points for each shot. Color CC Grading is done on a keyframe from each shot. This is a filmmakers tool, not a video real time editing system, just metioning that to avoid confusion. I am interested in Digital Cinema cameras for input since they are getting better than film as far as sharpness, no dust to clean off (if the sensor is clean) with solvents, and no high film or processing cost for the negative, the time required to go a good DI is the about same for film scans or Digital Cinema camera, so that is not much of an issue with our "in house" projects. I just needed to be able to use the same workflow for any source, and to reduce the labor over Film-to-Film printing, and in some ways get better results since the digital CC gives control that Film-to-Film printing cannot come close to for on location footage shot with bad light.

Feel free to email me at the email on the top of each page at my web site if you have any questions... Things are being worked on to get the quality to be as good as needed for filmout work with optical track, so if you are interested, it is in your interest for me to know what you want to do.

If you are going to spend 5 years of your life making a film I would rather shoot with a RED ONE (tm) than a handycam since the cost of the camera body is only a small part of the many costs involved, you still need lights, actors, locations, sets, better to have a good DI than a fuzzy POS for all that very long and hard work, if for no other reason than to be able to look at your answer print and enjoy what you have done knowing that you did the best you could.

If you cannot make up you mind 2K or 4K you can work at 3K ;)
 
Doing its thing...

Doing its thing...

mail4joeg said:
So, you're basically creating your own Premiere?

That's pretty ambitious stuff. What system requirements are there, and how fast does it do its thing?
Click to expand...

Note, Some tasks or programs used require Windows ME (tm) or Windows XP Home SP3 (tm) with the FreeDOS FAT32 (tm) disk mounted as a slave drive.

If you are going to do a uncompressed 4K DI you would want several cheep PCs to do the frame processing, i.e. CC, resize, letterbox, etc., and to filter with my program or other programs.

The process computers would have removable harddisks or at lest some kind of network to move files around for backup and to organize the workflow.

You can have DANCAD87.EXE (tm) working on many computers at once, each one doing a sub-range of the total number of shots, like if your feature has 750 shots, and you have 10 computers for processing each one would do 75 shots it would go 10 times faster.

The process computers do not need a monitor, mouse, or keyboard, you can use a switch box to look at what is going on from time to time. Any cheep video board like SIS can be used for the process computers, no sound is required for them, just CPU and TB harddrives...

For the edit computer you should have two computers, one or both can be WIN88SE, the fastest one runs DANCAD87.EXE (tm) to play the frames, it might be good to have a solid state HD of 32GB or more for the proxy frames to play out of on that computer to get fewer dropped frames in the proxy playback, burps from the HD do not affect the 4K frames, but the playback is better on a realy fast computer so you can use higher resolution proxy frames, the range of resolution goes from EGA 320x200xM16 to 1920x1440x32 (it goes to 2048x1536x32 but I do not have a video board that does banked memory mode yet, anyway I do not think there is a computer fast enough to do better than 1280x1024 if that uncompressed, but the software supports proxy frames at many resolutions and mono or color to find one your computer can play, i.e. from 133MHz '486 and up). The second computer runs Magix DeLuxe Music Studio 7 (tm) (costs about $10 on ebay) as a WAV editor to prep the WAV files for each shot (slate clap at start) and as a stereo playback device for SMPTE LTC coming from the computer that shows the proxy frames.

The two edit computers are hooked together by putting a MIDI cable on the soundcard's game port of the Magix (tm) computer and plugging that into some used MIDI MTC to/from SMPTE LTC boxes (about $20-$50 on ebay) the SMPTE LTC goes to a Telcom T102 SMPTE time base (about $50-$100 on ebay), that allows the sound computer to master or slave the proxy frame computer. The external Telcom T102 time base has a video sync input, but I have not tried that for 23.976+fps since I just work with 24fps now, but it might be possable to lock the sync to a video source.

You do not need two computers for editing, there are other options, it is just better.

Using the external time base overcomes some issues with the crystal in the PC, and lets you get pure 24fps film like exiting playback with uncompressed proxy frames of "workprint" quality. With a Y cable and a SVGA to NTSC/PAL converter box you can watch the proxy playback on a TV or output it to a VCR with SMPTE or other frame burn in. You can add a SMPTE burn in box to the NTSC/PAL with a Y cable on the SMPTE LTC to get a burn in of both the frame count on the proxy play back computer and the SMPTE LTC code itself as double check on the sync, i.e. you can look at the video tape on single frame and the SMPTE time code should match.

You can slave more than one proxy frame play back computer to look at multi camera shooting with the playback in sync (although not of broadcast quality, just proxy workprint frames resolution).

Likewise you can slave more than one sound computer using Y cables on the SMPTE LTC to get up to 16 channel sound playback in sync with any number of Proxy frame images, one computer acts as master and the others all as slave.

If you shoot three camera, you can have three computers showing the three sync takes, and you can have three sound computers playing three stereo tracks for 6 channel sound. The proxy image quality would be workprint, but my programs and some "junk" computers would let you get the job done.

===

System requirements for DANCAD87.EXE are about, (edit/mix/cc program)

OS=FreeDOS FAT32, WIN-ME, XP Home SP3 (XP can has some issues on some computers, like with graphics modes, looking into that, try WIN-ME first),

Memory=256MB+.

HD=FreeDOS has a FAT32 mode that may work up to 2TB per HD. All disks should be EIDE and FAT32 to swap between OS, use the freeware program FAT32FORMAT.EXE (tm) with XP to format 500GB+ disks for ME and FreeDOS use.

CPU=2GHz+ for proxy playback, 700MHz or better for frame processing. For DANCAD87.EXE (tm) you need a CPU with math co-processor built in under DOS, if not you use DANCAD3D.EXE, almost the same program but works without math co-processor.

You may need to upgrade the mother boards to work with 500GB+ disks (if you get Panic fnode error, etc.)

video=VBE1.2,VBE2.0,VBE3.0 VESA SVGA like Geforce 6200 or SIS315.


System requirements for DANCINES.EXE, (film/printer shoot program)

OS=FreeDOS (?), Win-ME (?).

Memory=256MB?

HD=none required can be run from 1.44MB floppy disk under DOS.

CPU=PII or better?

video=text only required.


System requirements for DANCINEL.EXE, (DIY filmrecorder)

OS=Win-ME (tm), maybe XP Home SP3 if you can get graphics to work on your computer.

Memory=256MB?

HD=Being able to store TB for the film recorder you need enough disks to hold 2000' feet of frames to do a reel at one time, if you are going to make splices every 400' then you can use a smaller TB limit and just store 400' worth of frames at one time, i.e. swap the data disks for the next mag of 35mm film to shoot.

CPU=2GHz or better to get the work through the film recorder ASAP?

video=needs to support high resolution VBE2.0 or VBE3.0 linear modes like 2048x1536x32bit, or 1600x1200x24bit. Geforce 6200 or better maybe, you need to see what modes work on your monitor, under XP you may have some problems that do not show up under FreeDOS etc.

Since video monitors working with standard video boards are limited to 2048x1536 if you want more than that, for 70mm output, you may be able to tile more than one monitor and use croped frames in more than one computer, I will be looking into tile printing later. You can get better results maybe to make three color seperation negatives on pan stock, then print on a contact printer from the R,G,B rolls.

All specs are subject to change without notice, the programs are in "Beta Test" phase and being worked on from time to time as things come together.

===

Depending on the computer used the 4K frame processing is maybe about 30 seconds to one minute, most of that time is in the DISK I/O so having a RAID 0 array with 8 solid state disks might make things go faster on a 4GHz CPU. XP seems faster on some Disk I/O.

It is good to have several computers so that you can be trimming and filtering WAV data, while you are doing the keyframe CC on another computer rather than wait for things to cook on a single coputer, if you are making a feature yourself, every dead minute is wasted time. My CC stuff can run itself from a range selected in the edit list, say you want to CC shots 200 to 250, first you do the keyframe CC then transfer the KCC files to the computer you are doing the batch processing on, and let it rip, come back in a few days and TB of nice CC frames should be there ready to feed into the film recorder...

No matter what software you are using you still need to sync the sound for all the tracks, adjust the levels and filter the sound, do the keyframe CC values, and such, and that would take months for a 1200 shot feature film no mater what program you are using, that is why it is better to have things going on more than one box, since a multi core processor will still bog down if you have the HD going 95% on something else.

Sorry for the long reply, but some people here would like to make a feature, so this way of working with RED ONE (tm) footage may have general interest.

Things are in place, and for my own interest should progress in time, if there are any bugs I would like to know about them, I tried putting some footage from REDRelay through REDCINE and into my system and things seemed to work for the most part, I am making some revisions today and working on the TIF I/O, so we will see where this goes... The RED ONE (tm) footage looks very good to me coming from the REDLog TIF files, compared to the RAW 35mm film scans I am also working with for development, you can mix and match if you like.
 
4k post and filmout are out of reach for most independent filmmakers. They are just too expensive at this point.

The most visible advantage of shooting on RED in 4k vs. 2k is reduced depth of field. An average viewer will hardly notice aliasing improvements, lower noise, etc. but they will notice depth of field that looks like the material was shot on 35mm.
 
You must log in or register to reply here.
Back
Top