[Dave Francombe]

Correct, camera originals on the drive that crashed, and believe me we have the correct tools to ensure this didn’t happen, including a raid drive with redundancy (two drives failed at the same time) and a Quantum LTO4 system to archive everything as a disaster recovery backup. What we didn’t have, was anyone wrangling the media in the system correctly or in a timely fashion. Brutal ! Drives are at Disk Savers hoping to recover the data, but we recovered a lot of media on our own from a variety of places. This media does indeed have TC, so we will try FCPreconnect.

Dave

[Dave Francombe]

Hi Shane, thanks for picking up on the post.

The JVC HD GY 700U records straight to cards, but I believe the files it records are wrapped in QT, so as soon as the files are dropped from the cards to a drive they are ready to edit in FCP immediately – no log and capture needed.

So what we did is exactly that. We shot, dropped camera original material to drives (didn’t bother renaming them) then imported them into FCP project and started editing. So they were called exactly what the camera labeled them during recording, both on the drives and in the time line. Those drives containing the camera original files went down, but we have been able to recover many camera original files using software, however the recovery software happens to rebuild and then automatically rename these camera original files. So now I have files named differently in the timeline, to the recovered files (which are the camera original files but named something new by the software). The new files seem to have the correct time code information, but the names are mismatched.

Dave

[Dave Francombe]

I have a very similar system to Walter, also set up by Bob Zelin, we are about to add another 16tb drive array, but we shoot about 150 hours, and produce 65 shows a year, so scalable storage is a real concern, not just for edit, but archiving camera material. I’d love to hire a consultant, if you can suggest a good one who knows work flow, from the field through delivery, and can update our edit rooms, that would be great. Bob, is the first to admit work flow is not his thing. I just like to try and educate myself a little before I turn everything over to consultant. As you are aware, different folks like different formats, and with so many on the market I’m interested in opinions from a variety of people with real world experience. Thats why I turn here first, consultant next.

Dave

[Dave Francombe]

Jeremy,

Thanks for the post ! Things are becoming a little clearer.

“After your edit, you can dump the ProRes files and you can easily retranscode the files from the AVC-I MXFs. FCP makes that pretty easy actually”

After the transcode, and the package is complete, I could media manage, trash the prores and reconnect back to the MXF files later, so although the transcode has been made it always references back to the original files, even if names etc have been changed along the way ?

From what I can gather from your posts you, use LTO to backup field material. Can you tell me how that fits in with your work flow ? The archive of material is what worries me a little. I want a scalable system, individual TB drives dont feel secure, raids are costly and will fill up our san system after a while. I was looking at the drobo as an LTO alternative. This lets you put 4-8 2tb drives in its enclosure, and gives them raid protection. I can yank them out as a bundle when they are full, drop them on a shelf, and they will instantly have your raid available when you load them back into the machine to gather material. I can hang this on our SAN through the server. I’d just dump material off it onto our raids to edit. Good, bad or terrible idea ?

I have a lot more acquisition work flow questions… sorry and thanks.

Dave

[Dave Francombe]

So AVC-Intra is a great space saver initially, but then you have to transfer to Prores to edit, unless there is a work around, suggested in the MXF4Mac tutorial. Does that then turn a small file size into a pretty big one ? Thats what I don’t understand, I have one file type AVC-Intra, I transfer it to prores to edit, now I have two file types of the same material ? I have the original file, plus the new prores file, have I now doubled the amount of material I have, or is it just wrapping the original file to play as prores ? Just can’t wrap my head around this….sorry !

Dave

[Dave Francombe]

Thanks for the advice. We will dump SD next year and never mix it into the new series (assuming I can get my head around this tapeless deal)

I love the idea of recording the Prores 1080i possibly on the new JVC cause its ready to edit in FCP, but the storage chart indicates you get about 15 hours of storage per 1T. We already own a SAN and 12tb raid 5 to edit in our current SD configuration, which is expandable by adding more raids. We will shoot, and archive about 150 hours per season, and a adding a 16tb raid is about 9k. This means we’d probably get enough archive storage for a season on that drive, but it also means 9k every year for every season, and then we max out the san at some point. You mention you saved xdhd files, any idea how much hours of material you can get on 1tb at 1080i 60fps ? Is there considerable space savings over prores ?

LTO4 sounds nice and very safe but 300gb per cartridges are perhaps a little small in capacity, but love the price per tape after the initial investment, but would love to hear how this format fits into your workflow.

My other thought is using the drobo as an archive since I can get about 3.5 TB of protected storage, from 4x 1 tb drives, pull them out from the drobo enclosure when they are all full, put them on a shelf as a bundle, then put them all back in the drobo when I need something stored on them. Drobo tech support said there is no rebuild time after re-insertion, you still have the raid protection, but access is instant. I’d hang this onto our san via the server computer and FW800 port. Its only an archive system like the LTO, I don’t expect to edit from it.

This seems pretty scalable, since I can just keep buying HD’s, if one or even two fail on the shelf, they can be rebuilt. I was thinking of adding another 16tb, that we would transfer only material we need from the drobo, edit the show packages (we are essentially a news style show), media manage them, then pass those along to our 12tb drive, which is where we string the packages together with anchor lead ins, graphics etc, and finish the show. We keep media managed packages on the 16tb drive, but wipe it clean of the raw material we brought in from the drobo.

On the 12tb we eventually media manage the show, keep it on the 12tb drive for use at anytime, and archive the media managed show back to fresh drives in the drobo, then when they are full, we send those offsite for disaster recovery (can you tell I am worried about losing material!) Go ahead, tell me I am crazy, I’m expecting it, cause there has to be a better way, I’m just not smart enough to know what it is !

Dave

[Dave Francombe]

Thanks Shane, exactly what I needed to know, and the exact work flow I would follow to deliver tape. Being clueless on HD, the 720p 24 gives the film look right, but our show is more “newsy” in genre, and my delivery is 1080i at 59.94, I noticed your delivery was 1080i but at the same frame rate as your original material. Do you think the convert through aja to that frame rate gets rid of that filmic look, would it create the strobing type of effect I have seen when some converts SD 24p to SD 30fps, any thoughts ?

I’m also interested in learning how you are archiving p2 cards and your shows. We do a minimum of 45 x 21 minutes a year on Lifetime, and our archive of raw tape or p2 will be pretty large, thats whats worrying me most about our transition to HD.

Dave

[Dave Francombe]

Your right I am confused, but not about how a good or bad camera or DP can produce god or bad images. I have to deliver 1080i HDCAM 59.94. What I am trying to understand is how DVCprohd 720p looks after the conversion to that format, frame size etc. I like that I can archive more material onto 1TB of drive space in that format than I can using the 1080i DVCprohd, but all things being equal i.e. good dp, same camera, which native format handles the conversion better ? Or am I totally missing something altogether here, which is very likely. SD on digibeta was always a pretty easy concept, HD, not so much ! Your help is appreciated.

Dave

[Dave Francombe]

Thanks again Ben.

I think I am going to try the HD qt from my computer. The monitor is 1080p so I will look through Qt settings to see if I can find the loop functionality. So the key to getting rid of the crappy looking render is finding the closest setting to the logo settings in apple pro rez, and just accept that the render I am seeing looks crappy because the canvas doesn’t view it clearly and the monitor is SD.

[Dave Francombe]

Thanks again for the advice.

This system I believe is fcp 6 running on a dual 2.5 intel based mac with an FW 800 external drive.

Couple questions when you say to render out a pro rez movie. Should I render it out and send a movie to the drives before rendering out the timeline?

Can I composite the 1080 logo with a stretched and blurred Sd background on layer 1.

3. To get best quality, playing out an HD movie on a computer to a monitor is better than from an SD monitor.

Again thanks for your help. We have a lot of SD broadcast delivery experience but zero in HD, and none in creating graphics for a monitor. So its a painful process made worse by the fact I am shooting on Tuesday !