[Francois Stark]

Hi

I have both.

The Mac Pro is much faster. Renders FCP effects 2-3 times faster than a Dual 2.7 G5, even when reading the media across the LAN.

BUT

It’s not stable yet. We found that it falls over whenever we access it over the LAN, even to get just a small project file – it crashes instantly, or a few minutes later. If you don’t access it, its fine. It can access our G5’s and open complex projects while mounting 12 volumes over tha LAN. We have a Intel Mac Mini which does the same – crashes when you access it over LAN.

We are also running a fibre channel SAN, and the software for it does not run properly on the Intel system yet.

Rgards
Francois

[Francois Stark]

Hi Wayne – thanks for all the answers.

[Wayne Carey] ” It uses and accepts all version of XDCAM SD and HD”
Just for interest sake – when using the PDW-D1 drive in File Access Mode, can it be used to access XDCAM HD footage?

Thanks
Francois

[Francois Stark]

Hi Mark

Thanks for the discussion – few questions:

In the file FCP_XDCAMHD_whitepaper.pdf, I see a FCP plugin that allows you to do your logging from the low-res proxy of each clip, and it also imports to FCP in the background. You can supposedly continue logging or working in the FCP project while it imports in the background.

Why are you not using this? Is this workflow only available for the HD versions? Or is the plugin not stable yet? Why is TOD faster than standard timecode?

We are doing post for TWO seperate daily half-hour soapies now… (one is still on AVID). Sourcing on beta SP currently and editing in DV format. We do a lot of subtitles (about 400 titles per 22 minute program), using inscriber and TGA files… So we mainly do some grading and subtitles – will the native IMX be better, or do we stay with DVCAM? We need a format that is rock solid and without surprises, with good RT performance. Delivery will still be on BetaSP for a while.

Thanks
Francois

[Francois Stark]

Ok, we found that the problem is related to both the FCP version and the PRo Tools version.

We have 2 Pro Tools HD suites – one PT 7.1 CS9 and one PT 7.1 CS10. The CS10 version has a serious problem with OMFI imports, the CS9 version imports the same files with no problem. Funny thing is, it is only FCP projects with audio stored on networked drives, or any off-line audio clips, once exported, that gives these problems. We did not use to have any problems with FCP 4.5, FCP 5.0 and Pro Tools 7.1 CS9.

At least we know how to work around it for now.

Regards
Francois

[Francois Stark]

Hi

I just rashed prefs – no luck. same problem. We also realised that the workaround of consolidating (media manager) to one drive before exporting to OMFI does not help at all.

This is getting serious – we cannot find a workaround that fixes the problem!

Also, we trashed the prefs on a decklink machine and re-tried the export. Afterwards we realised that the video digitiser, and all the decklink capture presets are gone. All the decklink output presets are still there, but the input presets, decklink video digitiser component and the decklink deck control components are gone. So now we have to re-install the decklink software – and find the newest installer is gone. AND to add insult to injury, the decklink website is off. Cannot access it at all.

Getting really frustrated 11am on a sunday morning…

Francois

[Francois Stark]

Nope, it’s not Metasan. I have 4 suites on os X 10.4.7 and have the same problem, using MetaSAN and SanMP. I recon it’s a FCP issue.

I even had it a few times on FCP 5.0, but it seems to have increased since I moved to FCP 5.1. Rendering just stops and it says “file name already exists”.

Regards
Francois

[Francois Stark]

Your in or out point was standing on an transition. FCP needs both to be on any place that is not a transition. Always complains about “1 frame” even if it is clearly not. Been a problem forever.

[Francois Stark]

Hi Ben

I’ve been where you are. I started with a volume locking san (20 months ago), and it works. But we grew, and with four FCP systems and three Pro Tools systems, we are now running 11 volumes on volume locking. It gets really messy. I’m fed-up of closing FCP, umnounting/re-mounting a volume just because one of the other editors captured one clip from DVD for me. Or if I use a faster machine to render while I continue working on a slower system – the FCP shutdown/unmount/remount process almost makes it not worth it.

I am now testing a file locking SAN called metasan. It has two distinct advantages over XSAN: it does not need a dedicated metadata server or metadata volume, and also it does not need a dedicated second ethernet network. I got home about 90 minutes earlier tonight because I rendered on another suite while continiung work. Without closing FCP or unmounting anything.

I’ve been testing metasan for a week now – Bernard will supply you a 30 day eval license for your full san. The only other hardware you will need is a fibre channel switch. Since it seems most of your hardware runs on 2gbit fibre, I would suggest a QLogic 5200 switch, some fibre channel cables and bob’s your uncle.

Once you start realing the efficiency advantage of not having to transfer stuff using firewire drives (yuch!), you really won’t look back.

Regards
Francois

[Francois Stark]

Ok, so this is the first time we are running a file level SAN. Up to now, we stored projects locally, on a second drive in the mac, with the FCP archive set to the media drives. We would mount job directories through the LAN.

How do you store and share FCP project files on a typical file level SAN liek XSan or MetaSAN?

Thanks
Francois

[Francois Stark]

So Bernard from TigerTechnology sent me a demo version of MetaSAN. I installed it today with sanMP and they seem to co-exist quite nicely. I added two Metasan drives to four systems on our seven seat SanMP system, and they seem to work well.

Doing further tests over the next month.

Francois