Moniitor/TV/Projector gamma issues are a mess, I have been trying to work these out with the system I am working on, for now I export files at gamma 2.4 which is the "natural" gamma of many CRT TV sets and un-corrected monitors. You read the monitor gamma wrong if the black and white lines or dots on the gray background are vertical since the monitor peaking or cable bandwith make the small high frequency patches too light or two dark, I use horazontal lines in my grading gamma reference, they are larger than the uneven part of the monitor bandwidth. Apple seems to pick gamma settings that are not on the natural CRT/TV curve, so you need to shiift the tones on each import/export. If you are using 8 bits for each color, 24bpp, changing the gamma on each import/export may produce gaps in the tonal range giving the image a slight posterazation. You may want to pick a reference image such as the LAD, and load that before you grade you footage to be sure that the monitor and import/export gamma settings are giving you the same look. Also I would use a CRT monitor rather than a LCD for grading since if you move your head around the LCD can change the look. Also the windows should be blacked out, and the room light constant since the color and brightness of the light in the room can change the way the monitor image looks, black level and gamma change a little with lights on or off. If you save an load from uncompressed 48bpp file types the constant gamma changes going from PC to MAC programs is less of a issue, but the files are twice as big...Originally Posted by Akube
To work on your project on a PC you may be able to down size the frames to 320x240 resolution. The color correction and editing will be the same, and you will be able to store a feature in less than 100GB. You can then later edit the full 2K, or 4K footage a few minutes at a time by matching the "proxy" frames. Doing color correction on uncompressed 4k 48bit frames will take quite a while even with fast software, with 126720 frames in an 88 minute feature, and each uncompressed 4096x2048x6 frame being 51MB, you get about 6.5TB of files, to process that at one minute per frame would take about 88 days on a single PC.
To edit on film, conform the negative, cut the mag film, and such yourself would probably take longer than a year, so if you have several computers going 24/7 using a uncompressed Digital Intermediate will give you a much better result with probably less work and cost. From my experience the last few years working with film scanning, if you can shoot digital you are far ahead for someone who would have to do the bulk of the negative cutting both in cost and quality.
People who have sponsors have a different view of things since they can use the sponsors money to have work done in labs and such. If you are going to do the work at home, and want a quality result you will have to put some time into getting the results you want.
While there is a strong temptation to compress the image frames on import and export to save disk space, use fewer backup tapes, and speed loading and saving of the frames, I would hope that everyone understands that if you uncompress JPG2000 type images with their artifacts then recompress them after changes the artifacts develop more artifacts.
If the RED ONE (tm) footage is compressed in the camera at 4k then downsized to 2k for export as 48bpp uncompressed then the losses are not to large if you use enough fill light on the subject and crush the dark areas later. But if you do not keep the frames uncompressed through all further processing the compression artifacts will accumulate, and making gamma changes or sharpening will make the artifacts more noticable, as will resizing the images.
With TV going digital in compressed formats, and distribution to the TV stations and DLP cinema being compressed, the end image gets compressed two or three times at a minimum. If the image is reduced in size each time it gets compressed the loss may not look a bad, but with HD TVs being 1920x1080 the reduction is not as large as when going to a 720x480 DVD, so having less compression used in the image generations during editing can only help.
As for people selling their RED ONE (tm) because they cannot tolerate the workflow, that seems silly, if you want a quality result the RED ONE (tm) camera is a good if not the best choice now.
There should be a easy way, or maybe the RED (tm) team should work on an easy way, to edit a DV proxy on any editing system and match the editing later with the high resolution stream. You could use OCR to read the SMPTE time code off the edited frames, or something, to do the match editing. (with my system I just suggest that people inch through and type the start and end SMPTE into the edit list manually to conform to an edit done on some other system) Could someone speak more about the current situation with regard to the use of "proxy" frames at lower resolution for editing and grading in the RED (tm) workflow to save disk space?
In my system I am just doing the grading with single key frames from each shot, and low rez proxy frames for the editing of the reels...
Bottom line, if you like DV for its ease of use, maybe you can shoot DV side by side to the red, and do your edit on DV, then get a sponsor to have someone else wrestle with the 4k, but if I had a RED ONE (tm) I would not sell it on ebay if I ever thought I was going to try to do a quality project just because of the workflow...
Anyone know how long it takes to go from footage ingest start to film out (if you had a celco) using the RED ONE (tm) for a single user on a single PC to make an 88 minute feature at 4k? i.e. 90days...2years...
