Pass the word along!

Regarding what I would maintain to be the important questions I think the answers are predictable. For me film looks a fair bit better, having an aesthetic that no other format can really match. Red is definitely cheaper and more convenient, but not by that much. The camera body might be cheap as chips but if you’re looking to hire a camera package with good lenses (no one’s mentioned lenses) you’ll probably get abetter deal for 35mm. I’m working on a movie right now where the transcoding of Red files for the offline takes almost as long as film processing. Generally the cost of working on Red in post-production, especially at 4k, is very high.

Don’t get me wrong though, Red is an amazing step forward; but it’s not yet knocked 35mm out of the water.

Best,

Guy

1. film has a higher latitude then the red. know your format, protect your highlights and test your pipeline.
2. 4k red means 4096×2304 but resolves to an optical resolution of 3.2k.
3. your data rates off off the chart. Redcode 28 @ 4k 2:1 to 8gig gives you approximately 4 1/2 minutes. less at redcode 36 and less at 16:9.
4. 2k is cropping the sensor and only capturing 1/4 or your sensor real estate similar to 2/3″ or 16mm. you can do the maths for exact figures if you wish. the data rate is like a pipeline so when you limit your capture area by 1/4 then you open up the speed at which the sensor can capture meaning your multiplying the FPS by 4. it all comes down to the datarate.
5. transferring data off Reddrive to FW800 should be around 100gig/hr which is around 1hr of footage.

a 320 gig REDDRIVE is about 3hrs and should take 3 hrs to transfer with out a MD5 checksum. double the time if you are using checksum.

@ FW400 it should be a gig a minute. 100min=100minutes. :(
you are limited by the smallest connection so if your using a macbook pro you only have 1 FW800 slot. you need to get a PcExpress card reader to FW800 (usually dual) for more connections or a Esata PcExpress. If you dont have either. Use a Fw800>Fw400 cable. Plug your REDDRIVE in and read off the FW400 connector then your FW800 Ext HDD on the FW800 slot.

The fact is data transferring takes a long time and it should be done in staggered steps during the shoot. But these times are way too long. Also dont let them shoot more then 10% on a red drive before transferring. 10%=32gig=20mins footage x 2 cameras x2 copies = 1hr of transferring (x 2 for checksums)

6. overcranking means adjusting your shutter to compensate. always check this.
7. AFAIK overheating hasnt been an issue since build 8, setup you camera to turn fans on when not recording. i have had issues in the sun on a 40degC day on long takes over 10 mins and the cam shuts down. :( but thats been 1 day in 2 years.

8. sync sound. agreed. gen lock master TC from lockkit box. get sound onto it early. do not take TC from camera.

9. Hmmm personally dont use log and transfer for FCP. its sucks balls. if you dont have a red rocket card and your generating offlines, go buy one, you can afford a mac you can afford a red rocket, sorry but it will save you hours of your life and you can see your family every now and then. otherwise use red rushes, render go to bed wake up and check your footage.

you can easily load all of the _h.mov proxies into FCP and then relink to newly generated offline renders from red rushes. I use red rocket so i tmp grade everything in rocketcineX then render to ProresLT. I am spoilt with a new 8 core nehalem and FCS3, but im currently getting 1.5xRT to ProresLT (so 1hr footage=1 1/2 hr render)

check render: load all proxies and preresLT into FCP. drop Prores first, yes to auto settings, then drop proxy over the top of the prores clips and you will see any incomplete clips.

10. quality loss in film for each generation print made but also generation loss in digital depending whether there is compression or not. a digital copy is a digital copy. if every byte is the same and this can be verified by a checksum then it is technically identical. Your image loses quality if you render out from R3D to another format. The process for generating assets for DI ie. 10BIT DPX log is an extremely important process. If not setup correctly in the intermediate tool/program ie Red Alert/RedCine then you could quite easily lose colour information. ie highlights and shadows. make sure whoever does this knows what they are doing. ive had producers do this on their macbooks and then wonder in the grade why their footage looks so bad.

Re colour. im a scratch guy so i stay away from colour if i can. but i do recommend that if you do use colour to get your self a wave tangent panel. around $US2500 i think. not sure look it up but grading without a panel is like flying without wings. And if you dont understand the difference between colour space and gamma space then go look it up. the world is a clearer place once you get those two things.

one last point. Your data guy needs to know what hes doing and not be a copy paste / jnr from the office guy. No offence to office jnrs, but they are essentially your Lab – onset. If they know what they are doing they are a valuable resource to both the planning and preparation to the technical outset of your production, safely backups all your data to 2 copies yep 2. even 3, with one remaining onset on a Raid1 set of Drives.

Also archive off to something such as LTO3 or 4

And please provide your crew with enough media to cover your own asses. i would suggest at the minimum of 2 REDDRIVES 320gig and 4 16gig CF cards per camera per day. Give your guys more then the turn around before your formatting your previous days media to shoot your next day. Not saying your Data guys dont do their job im sure they do but you want to give yourself as much time between formatting and potential drive failures.

oh and how cool does scarlet look :)

cheers
Chesh

That’s true if you are making a copy of a digital file.

What about when you shoot digital and you want to manipulate it. This is almost always the case since some color correction, vfx or touch up work usually needs to be done.

Anytime you manipulated pixels and rendered out a new version, it’s always possible to introduce some degradation. It all depends how much manipulation is going on, the workflow, and what digital formats are involved.

Final Cut Pro 7 and COLOR 1.5 now FULLY SUPPORT REDCODE 4K Files.

Details can be found here:
http://images.apple.com/finalcutstudio/color/final-cut-integration.html

Forgive my ignorance, but in the still photography world it seems that the megapixel equivalent to 35mm film is much higher than 2 or 4MP (see for example http://www.clarkvision.com/imagedetail/film.vs.digital.1.html). Is there a reason why megapixel equivalents is different for film? Or is it just that projection output is always less than 2k, so input differences don’t matter?