Francois Stark
Forum Replies Created
-
Hi Chrispy
I have not gone over to tiger yet. I have one installation that runs FCP 5 on os X.3.9 by cheating: I’m booting a G4 1.25DP on it’s own internal os X.3.9 partition, which is still running FCP 4.5. Then I just run FCP 5 from a firewire drive, where it was installed on a tiger partition! It works! I can actually switch between FCP 4.5 and FCP 5 without rebooting, both using the SAN as media storage! This machine uses the Kona LS card, and it can do a multicam cut between 7 layers of DV footage in real time – on a G4 in FCP 5!
It can still see my standard LU mapping which all the other os X.3.8 machines can see. I have 17 LU’s on the ADTX, and can send you a email with a screencapture with the LU mapping when I’m back at work on Monday. I also had to run another trick to get the G4 to see the ADTX box at all on anything over os X.3.4 – email me if you need more info on that.
Back to your question:
I have to say that I think every type of HBA has it’s own set of drivers, and for each of those you might need a different setup of your LU mapping.As I said, I’m on all apple HBA’s, and using those on os X.3 I have found I need to adhere to two rules:
1. LU 0 must be mapped to every port on the ADTX box. If LU 0 is not “visible” on a port, none of the other LU’s on that port will be visible to the clients.
2. Each internal LU in the ADTX box must be mapped to only one port. If you map one LU to two ports, those two ports will show up the same LU twice on the desktop, since the client sees both ports through the fibre switch, and thus sees the same LU twice.Of course, these rules could only apply to the combination of os X.3, Apple HBA’s and an ADTX LH series drive array. Change any one of them and you could need different LU mapping to work.
My preliminary tests with Tiger ont he ADTX box did not work properly, but that was about three months ago. I have since received a new firmware for the ADTX box for use with tiger. Bruce from ADTX support has also told me that I need to map all LU’s to all fibre ports for use with tiger (once again, he also specified that this is what is needed for the apple HBA – other HBA’s need different mapping).
Since that will cause chaos on our current installations by breaking my “rule 2” above, I have no option but to wait until I can ungrade all my machines to tiger at once. That’s going to be difficult, since I have two G4’s and four G5’s in a mixed hardware SAN. I will have to test all these cominations seperately before committing.
So it seems I can not really help you since I’m running different HBA’s.
Regards
Francois -
I’ve been using a DSP1500-AP (PAL version) for about 18 months – no problems.
I can control it via firewire or 9pin. Firewire control is MUCH slower than 9pin. You have to go into the deck’s menu to choose between 9pin and firewire control – it’s displayed on the front panel as “9p” or “ilink”.
Hope it helps.
Regards
Francois -
So: os X.3.9 and FCP 5 is happy with sanMP?
Have you investigated the transition to Tiger with this setup? It seems you have to set up your drive array’s LU mapping differently depending on your fibre channel HBA’s.
Luckily I’m running all apple hba’s. I don’t know how you will handle it if you’re running mixed hba’s, since one hba’s LU mapping will not work for the other type…
regards
Francois -
Yep, the dealer apologised and still took 10 days to replace the card…
So the new card works fine. In a G4 with os X.3.8. Good.
Thanks for listening!
Francois -
There are two ways I get FCP to be where I want it to be:
– use a marker (m key) like Walter suggested
– add a cut while playing – even multiple cuts.The one key I map in my keyboard is the b key – the default is blade mode. I change it to “add cut” which does an immediate cut on all enabled layers – while scrubbing or playing.
Using this, I can do multiple cuts on the beat of the music while playing, and it’s pretty accurate – it only leaves markers while the timeline plays, and as soon as you stop, it cuts enabled tracks up at the precise timepoints. It’s similar to using markers, but for editing it cuts out a few steps – pardon the pun.
Regards
Francois -
Well, I went and bought a new G5 to bolster our offering. Also thought – I need to get rid of the single CPU G4 since that is giving the LU problem. I swopped the fibre card from the G4 to the G5 and the same fault appeared there – in a G5 where I have another fibre card working yesterday!
So I inspected the fibre cards more closely: The faulty card is a LSI7202P with no heatsink, and the working card (I checked two working cards – they are the same) is a LSI 7202XP – with a heat sink on the main IC. Must be a new model. I think my dealer somehow sold me an old version, since the faulty fibre card was bought last, about three months ago, seperate from the other five cards.
I’ll have a serious word with him tomorrow…
Regards
Francois -
The black frame at the end of the insert edit is misleading – the problem is at the start of the insert. The FCP sequence starts playing one frame too early, before the edit deck is in record mode. When it reaches the end, it’s played out what it needed to play and then goes to black. The deck still records that frame. The reason it sometimes starts early could be that you need to tell FCP to start half a frame later.
Try:
Go into your deck dontrol presets and duplicate your main deck control preset ( NTSC or PAL). Try changing your playback offset and test it by recording BITC to tape. There’s a timecode filter you can apply to slug for that.
Regards
Francois -
I have been buying non-apple RAM since the first G4’s came out.
Ram either works or it’s unstable. You need to know your system is stable before the upgrade. Then: upgrade only the RAM. You will know if your system becomes unstable within one day.
I have upgraded 7 systems – 4 G4’s and three G5’s with whatever our most convenient PC supplier has in stock, and have never had any problems directly attributable to RAM. Other guys have had problems, but as I said – if your system is unstable with the new RAM, send it back and get other Ram.
Simple. Only change one thing at a time. That way you simplify the problem so you can trace the fault.
Regards
FrancoisPS = strange – here in South Africa Samsing is the cheapest Ram, cheapest hard disks also…
-
I’ll ignore the nasty comments since I can see you’re fed-up and apologised already…
I have had this problem (1 frame duration) repeatedly when trying to insert edit with either the in or out point on a transition – whether it’s been rendered or not.
Try setting your in and out points on the timeline on clear pieces of footage or slug or a cut without transition. The in and out point must not be on a transition.
Works for me every time.
Regards
Francois -
Somehow I have a feeling all of you have it wrong…
This sounds like a lopsided description of a classic G5 crash. It happens occasionally – just like sh*t.
All that happens is that the G5 has a solid crash, or freeze-up, so the on-board systems controlling the fan speed defaults back to full speed to protect the CPU. So the fans spin up (immediately) AFTER the G5 freezes. Not the other way around. Remember, the fans take about two seconds to spin up to full speed, so it could seem the other way around. Could this be a good description of what happened in this case?
Which means the crash had nothing to do with the room temp. It was just another crash. I must say, it’s been at least two months since I’ve had one of those fan-speedup crashes – on 7 macs (5 G5’s and 4 G4’s).
If it happens regularly you have a faulty motherboard or CPU or Ram – take it back to apple – they will look at the crash logs and decide what caused it. We had one G5 that crashed regularly for about three months while we swopped everything else (all non-apple hardware – pro tools HD etc.); eventually Apple replaced the motherboard and CPU (Single 1.8 G5) and only then did the problems stop.
Regards
Francois