Welcome to our community

Be a part of something great, join today!

  • 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 :)

Feature Film storage for RC42 RC80 and RC100

Ok, so aside from archive... I think I got it... the 4TB Raid was with 1GB drives! So we get two more to begin with so 6TB RAID where you have two full copies and the benefits of RAID0. I have to confirm that with Peter regarding whether that's the type of RAID he knows would handle 2k.

So the real size might be about 5.5TB hopefully a bit more. But still that means a little over 40hs at 2.25GB/m, so that puts it back into action... 40hs starts sounding better... but that is again the estimate for RC42 so no S35 or FF35 lens action.

Assuming that the numbers were to have a correlation, then these 5.5TB actual capacity RAID would hold half the footage of 5k RC80 coming in at 20hs per Enclosure.

5.5TB
RC42 ~ 40hs
RC80 ~ 20hs

Well if the actual numbers turn out any better than these speculations, I will be more than happy!

This sounds too easy... where's the catch? No catch?
 
Interesting idea... but couldn't a HDD be fine so long as you read each file once every so often? I really hope that we can keep a long term solution conversation as a secondary subject though. But I guess you bring up a very important point in that, I would have to consider that too before I decided if a second RAID enclosure would be wise.

A HDD could be perfectly fine for many years, but then again it might not be. With something like LTO-4, your data will be safely archived for 15-30 years. It's a proper archival medium, and that's what your footage should be archived on.
 
A HDD could be perfectly fine for many years, but then again it might not be. With something like LTO-4, your data will be safely archived for 15-30 years. It's a proper archival medium, and that's what your footage should be archived on.

I have had several hard drives go out over the years from all different manufacturers. (Raid is a good thing!) Many hard drives fail when trying to revive them after sitting for years. Tape in all it's forms is a proven media over time. I have restored audio and video on tape from 20 years ago. I wholeheartedly agree: for safety's sake, archive out to tape. (And hopefully in a format that is common so that you can read it back down the road.)

paul
 
Yea, but that was an underfunded project we did with $5k total. Give me a break please, I kept it rolling the whole time.

And it was DVCproHD 1080p, so it didn't matter much.

I was kind of joking. I done worse than 18:1 on a shoot. :)

The other thing to remember guys is the difference between the stated space on the drives and the formatting. Your 4Tb raid may be 3.2Tb once formatted. In addition, 1080 proreshq roughly equals the data rate of 4k RC36, meaning a 4 minute shot will be a 8Gb R3D file and a 8Gb quicktime 1080 proreshq file.

Yeah I actually forgot about that factor.
 
That's why I was pointing out actual capacity when I took it into account, but the 4 x 1.5GB setup should give us about 6TB with a likely 5.5TB actual... it shouldn't be much less than that if you do get a lower number.

Cedric, you made me feel like an arse, thank you, I should never do that again. :)

So I guess from the lack of clarification that no numbers were even mentioned from the Red team as to actual file size for RC42/80/100, right? All I heard was a rumor I think, can't remember where I read this, but that possibly the size of the files would be a bit smaller than RC36 in relation. I guess they could have come up with an other little way to pack the data... of course if something like this were to happen we shouldn't expect much at all, but any improvement is welcome if they are at all possible.

And secondly, I'd like to learn a bit more about the workflow... I am going to try to keep things simple as possible for budget reasons and cause I will do most of the work without a lot of help. I am aiming myself squarely at a gorgeous 1080p, no need to redefine what Red cameras can do, but I do want a good example of Red's capabilities... you know, no need for a masterpiece, but good enough for a B movie you would enjoy watching. So I was thinking that 2k would be perfect cause that would give me a tinie bit of wiggle room for stabilization and re-framing... not a lot but it helps. I was thinking about trying to get a FCP working that could let me cut right out of the originals with the 1080p proxie... then if I needed to stabilize something I'd grab the 2k and crop in to a 1920 x 1080... well I will actually end up shooting for 1920 x 800 or somewhere around there, so plenty of vertical wiggle room, an other thing I like about 1920 x 800. HAHAHAAH... I haven't even stated my question yet! haha... Ok, here it comes: I haven't used R3D files yet, and I understand that there is a first light step... I understand you do minor color adjustments to the RAW to approximate it to your intended final look, so I suppose there is a possible step where I would have to render out the files to like ProRes 4:4:4? Not sure what but something that still retains the full color and luma info so we can finish the grading and CC... could you guys tell me a little more about the different output possibilities, and explain the need for this step? Or if I could just finish the movie with the 1080p proxie in Color?

I imagine that Color's non support of 12 bit is the reason why you might want to do a first light then hopefully only minor adjustments in Color... anyway, I need to learn more about this part of the cake.
 
Cedric, you made me feel like an arse, thank you, I should never do that again.

Sorry about that. That wasn't my intenssion. But as for your question I am still getting used to working with the r3d file from redrelay. That is the route that I would try first. When I first tried working with the r3d files I found that my computer wasn't ready. That reminds me I need to look back at some of the workflow sections that Peter had done on Scarletuser. I guess I could look at the workflow section on here as well.
 
I haven't installed FCS yet but I checked out Redcine with some of the footage posted and it seemed straightforward. I guess I would render a 1080 4:2:2 to cut in my hackintosh with a 3.2GHz Quad and 4GB of fast RAM which would cut that easily, but I would sure give 4:4:4 a try hoping I might be able to edit that big, but I wouldn't bet on it... that would be all I need though if I can do 444 1080 I am goooood... Blu-Ray... especially if my color grading can be applied to the RAW stuff without losing all of the work, in the event it were needed... how does that work?

Which threads should I look at? I suppose I should read Peter's threads... I have been procrastinating. I think he has it all in there, hope it focuses on FC.

EDIT: Well Just checked them and they looked abandoned, probably from the lack of info. I'll keep checking those too hoping for an update. Can't wait what Red has in store for post... they promised surprises. :watsup:
 
Shooting ratios above 10:1 are not a good sign.

It is said that good, professional directors manage to stay around a 4:1 shooting ratio.

My own ratios are somewhere around 8:1 or 10:1... Working` on it, working on it!... :)

Of course this is all very relative.

My personal scenario for shoots is this:

1. ~10:1 ratios -> 900 min. [for a feature]. Nothing is thrown away, not a single clip.

2. Three on-set backups, using external HDDs, or [even better] - using 2x external RAID-1 enclosures [which practically gives you 4 backups of the footage].

3. [Immediately] Dump one of the RAIDs to LTO tapes, keep the other one. I don't intend to purchase a LTO drive, as it's very expensive. Just buy the tapes and transfer them in a company that offers such transfers.

4. Transfer the footage on yet another work RAID[5], which is faster and intended for the post work.

5. Work. :biggrin:

6. Archive the final rendered master too, on [another] three backups.

7. Be happy. :)
 
Shooting ratios above 10:1 are not a good sign.

It is said that good, professional directors manage to stay around a 4:1 shooting ratio.

Tell that to Stanley Kubrick, David Fincher or Terrence Malick. :)

Not that there isn't some value in keeping shooting ratios low in most cases, but 4:1 seems like a pretty ridiculously low shooting ratio. That's two setups for each scene, and two takes each setup. 10:1 seems much closer to what a typical production does.

Regardless, though, you can (and in most cases should) get your shooting ratio down in most cases by meticulously planning everything out beforehand; the last thing you want to be doing is scrambling around on the day of the shoot trying to figure out what you want. Not saying you shouldn't be flexible in reacting to conditions on set, but you know what they say: fail to plan, and you plan to fail.

That said, don't be so concerned with your shooting ratio that you don't actually get the footage you need by doing a few extra takes here and there. In the long run, that does nobody favors.
 
I believe shooting ratio is per scene, not per setup, which is where some of the confusion is coming in. Shooting ratio is a method to figure out if the director is budgeting the project's feet of film properly, cause you can only expose film once.

In the digital world we have a delete key and the ability to make copies (like to tape...) which frees us up in terms of shooting ratio. You can pay a PA to down-rez your footage for editing and archive every shot to tape at night, but then your conforming process will involve wading through a lot of tape, and may add weeks of post production time...

My tact is a bit different: Under the covers Apple is Unix, and as such it has NFS built in. I'm going to take advantage of this by building a Big Linux file server using a 3ware RAID card, RAID6 arrays, and ~16 1.5TB drives. I did the math and because I can do this work myself this is a lot cheaper than most of the Apple plug and play solutions. Plus is gives me a place to mount the internal LTO-4 drive (which is a lot cheaper than the networked version). I'll have 2 arrays at 7.8TB each that can both loose up to 2 drives and still be available, and I'll be able to buy a third array and drop it in very quickly if I need it.

This looks like a lot of space as it would support a shooting ratio of 75:1, but this will be a central repository for all of the work, including holding the 4K tiff files for compositing (and possibly color correction, which would be the whole 100 minute film dumped out to tiff).

Workflow is this: Shoot to 16GB cards--> ingest to on site Mac Pro raid array --> Copy to external disk --> ext disk to server (then off-site ext disk) --> server to LTO4 tape (off-site tape). Of course I've budgeted for a technical PA to do this work. This is VERY redundant as the footage will always be on 2 or more medias after leaving the card and will have a target destination of three. If I'm going to take investor's money to make a film I'm not going to allow a single media failure to kill the project.

Bob
 
Well the shooting ratio is the total amount of coverage in minutes divided by the final output in minutes. That ending number is the actual shooting ratio. If you have 10 hours, 600 minutes, of total footage and your final output movie is 2 hours, 120 minutes, then your shooting ratio is 5:1.

I just wish I could get my shooting ratios to this level.
 
2. Three on-set backups, using external HDDs, or [even better] - using 2x external RAID-1 enclosures [which practically gives you 4 backups of the footage].

Be careful of this thought process - because even though its physically on 4 drives (2x raid 1), its not on 4 partitions. If you loose the partition info for a since raid set (by accidentally deleting the raid set, for example), you have now lost "two" of your copies with one operation. Instead, it is always best to think about the number of different partitions you have the data stored on as the number of copies you have.
 
I like to try to keep everything under 6:1, sometimes I go a bit lower but never less than 3:1, and that's if I am sure I saw it on the monitor or with my head close to the lens, framing in mind. However with DVCproHD I am rarely afraid to shoot more so long as the talent has the energy, and there is a potential benefit. It all depends on how you are covering it though, depends on the scene.
 
So I am assuming Raid 1 is the best way to go. Right? This is something I know nothing about. I have built my own computer but I have never done a raid. I have been looking at the Wikpedia definitions of each Raid description. Raid 0 and raid 5 are great for speed and raid 1 is better for security.

I did not consider DVD backups. That should be a pretty good backup system if you keep your disks in secure places. I can't think of any instance where a DVD, free of scratches, that fail to read in a peoperly operating drive.
 
only if it's double layer blu-ray, i can't imagine sitting around and burning 300 or more 8 gb dvds. If you figure in the cost of blu-ray discs and the time to burn them and then look at an LTO drive ( which I just checked is only $3k ) and another couple of hundred for the tapes, the LTO option isn't so bad after all...
 
I hope RedRay might be the answer for us non-rich customers... Jim, Jared, red guys hope you are getting this one... I hope it is an archival solution and that it turns up at a price low enough for everyone that buys a Red to be able to consider it. That way you sell more of them too... I am sure this is under heavy consideration and I hope you can find the way to make it possible.

What was the last capacity spec we heard?
 
Back
Top