[Glen Hurd]

I made my wife get out of bed just so I could read this to her LOL. Well done . . .

[Glen Hurd]

Great points, Gary. Faster drives mean better overall performance. What I am dealing with is a “before and after” scenario, where memory, drive speed (different enclosures but same speed), and cpu are constants. What has changed dramatically is the Kona performance.
My scratch disk is one of the attached Satas.
I just heard back from Kona (I mean AJA :P) and they’re saying I may have corrupted a bunch of files outside of the normal Kona set (which I have uninstalled and installed 4 or 5 times to date). So more and more it’s looking like wise-ass . . . I mean wise-ol’ Bob’s advice was still the best. 😉

I’ll let you know . . . and thanks to everyone for helping me out.

[Glen Hurd]

Apparently I suck at communication.
The 5-Bay enclosure was not set up for speed. Just convenient storage of terrabytes of SD footage for an ongoing reality show. Losing the bay has no effect on speed since I can hook up the same unstriped drives within the Mac Pro and connect to the other drives with the PCIe Sata card. My point was that the system itself is not being taxed at all. Drive speeds don’t need to go over 10MB/sec on some sequences. Yet the Kona struggles even there.
There was definitely damage to the enclosure, what I’m suspecting is that there was “software damage” (if there is such a thing) to the Mac Pro.
So my next step is to do as Bob suggested and set up a second system disk with fresh install and see how the card behaves then. Just can’t find the crowbar at the moment . . . might be lying over there next to the port multiplier . . .
And thanks for the tip on firmware and version number of driver. I’ve tried all the latest drivers – 6.03, 6.5, 7.0. (I was just going back to the one that I’d known had worked the longest.)

[Glen Hurd]

1. Yes
2. Yes
3. I wasn’t running striped – no need – currently on an SD project with lots of storage requirements. My point was that the drives aren’t my bottleneck. The Kona freezes even at read speeds of 10MB/sec. The drives are capable of more. (Yes, the Sata port multiplier on the 5-bay is fried.)
4. I’m not arrogant, and will happily buy another drive and do a clean install to see if that fixes anything. Not sure where I came off looking like that, but I do appreciate your input.

‘Bye for now . . .

[Glen Hurd]

Tried different slots, different speeds (4 and 8). No difference.

CPU usage at around 200 % when Kona card locks up on SD material.
CPU usage at around 300 % when Kona card locks up on 720 60p material.
So doesn’t seem tied to computer performance at all.

I wish there was a table somewhere where I could look up Kona performance for a particular broadcast parameter.
If anyone wishes to compare their Kona system to mine, here’s what I’m able to do.

RT Extreme – Unlimited.
Playback Quality and Framerate to High and Full
All clips playing off same drive, which never exceeds 50MB/sec during test.

I add 1 filter to each clip. Apple’s Broadcast safe filter (which I don’t use, but I’m using as a common RT filter to stress the system). It’s in default mode.

For my SD setup (DV50), Kona 3 can display 5 clips with the BS filter, set to 25% size (spread throughout the monitor) before it locks up.
For my HD 720p, Kona 3 can display 1 clip at full size with the BS filter, or 2 clips, if only the top clip is cropped. Without any filters it’ll play 3 clips, all sized to 50%. But on 4 it freezes.
I’ll try talking to Kona direct.
Thanks for looking.

[Glen Hurd]

I tried trashing prefs first. Hardrive I’m working off of for my tests is a SATA that I have internally installed as an extra – I’m not working off of system HD.
When I trashed prefs and removed the card, the FCP was still very sluggish – through several cycles of trashing and rebuilding prefs.
Once I copied prefs from another machine, it was back to normal (without Kona card). I celebrated 🙂
Once I install the card, the Kona was very sluggish, and I was back to head scratching.
Now, when I turn the Kona off (viewing), the computer performs well again.
Perhaps there’s a secret setting in the control panel that I’ve forgotten about? I’ve racked my brain over this. Drives are running at 90 MB/sec and I’m only using SD footage as my test bed.
Since RT Extreme is a big part of working in FCP I am surprised there’s no way to hack it or catalog it or compare it to other machines. I’d love to be able to look at its specs and say, “Hmm, looks like there’s a bottle neck in my memory or cpu or whatever that wasn’t there 6 months ago. Better take it in.”
Without Kona, I’m playing a base layer with 7 color effects,including 3 3-way color filters controlling saturation on 3 different hues. Above that base layer I have 6 alpha-layered motion-graphics dissolving in and out – together, along with a static graphic layer. It plows through these without much of a stutter.
With the Kona engaged, I can only do the base layer. Otherwise the video output just freezes.

As it stands now, I have no idea where the problem is or even why. Other than speed, everything seems to be working normally -except the fried 5-bay Sonnet Tech tower.

I’m going to try putting the Kona in other slots (it’s in slot 3) and see if that makes a difference. Or maybe try it at an 8x speed and see what happens. It does look like a Kona card problem, though. Which helps me at least be in the right forum, LOL.