| View previous topic :: View next topic |
| Author |
Message |
messenger82 Xsan Master

Joined: 16 Jan 2007 Posts: 113
|
Posted: Wed Mar 17, 2010 2:04 pm Post subject: CompressHD and Final Cut Server |
|
|
| We are attempting to get our hands around the processing problem h.264 has presented us. Has anyone attempted to install a CompressHD card into a Final Cut Server system? This xsanity announcement seems to hint that it does (http://www.xsanity.com/article.php?story=20090513115003531&query=compresshd). We purchased a CompressHD card and have had no luck with Final Cut Server integration. If someone has had any luck making this work I would love to hear about it. The Matrox rep I spoke with never even heard of Final Cut Server. |
|
| Back to top |
|
 |
messenger82 Xsan Master

Joined: 16 Jan 2007 Posts: 113
|
Posted: Wed Mar 17, 2010 3:54 pm Post subject: |
|
|
I'll answer my own post.
I called into Matrox support. The CompressHD card will not work with Final Cut Server 1.5. It will work with v1.0 / 1.1.
They have a ticket filed with Apple regarding the issue and would like to support it at some point in the future. They seemed to hint that Final Cut Server is at fault so we may be waiting for an upgrade to FCS as the fix. |
|
| Back to top |
|
 |
messenger82 Xsan Master

Joined: 16 Jan 2007 Posts: 113
|
Posted: Fri Apr 30, 2010 7:22 pm Post subject: |
|
|
I'll reply to my own post one last time.
The recent update to Final Cut Sever 1.5.2 does fix the issue with the CompressHD card. The settings appear to now work properly. |
|
| Back to top |
|
 |
ogminlo Xsan Master

Joined: 29 May 2008 Posts: 149
|
Posted: Sat May 01, 2010 2:54 pm Post subject: |
|
|
Better still, Matrox showed v2.0 drivers at NAB 2010 with which the card can finally produce low-bitrate encodings that actually look good instead of horrible.
They have not been posted yet, but are supposed to drop any day now (at least if the guy at the Matrox booth is to be believed). The big deal is the new drivers will finally allow for VBR encoding as well as a very trick scene detection option that places I-frames where they will be most efficient given the content. |
|
| Back to top |
|
 |
Toby@ITV partially protected

Joined: 02 Dec 2008 Posts: 9
|
Posted: Thu Jul 15, 2010 1:23 pm Post subject: |
|
|
Hi there-
Looks like your problem was solved, but we've been using CompressHD succesfully with FCS since V.1. We did not have to wait for FCS 1.5
With that said, the questions and responses on the Matrox forums in regards to using their products range from mildly misinformed to hilariously false.
Unfortunately, there is such a small user base of people using FCS successfully, there aren't enough people out their to point out this misinfo.
Compress HD is an absolute MUST for us and FCS. I can't fathom how you couldn't use it if you have a moderately busy facility and are generating H/264 proxies. The Xserve is simply getting taxed too hard if it's ingesting, serving FCS to multiple users AND encoding proxies. |
|
| Back to top |
|
 |
ogminlo Xsan Master

Joined: 29 May 2008 Posts: 149
|
Posted: Fri Jul 16, 2010 11:02 pm Post subject: |
|
|
| Toby@ITV wrote: | | Compress HD is an absolute MUST for us and FCS. I can't fathom how you couldn't use it if you have a moderately busy facility and are generating H/264 proxies. The Xserve is simply getting taxed too hard if it's ingesting, serving FCS to multiple users AND encoding proxies. |
Qmaster. Since v1.0, FCSvr has been able to offload the compression duties to a Qmaster cluster. A much more expensive option than CompressHD to be sure, but worth noting.
Also, the CompressHD 2.0 drivers finally dropped this week and they deliver. Now we just need to get Matrox to support more wrappers than .mov and .264... |
|
| Back to top |
|
 |
Toby@ITV partially protected

Joined: 02 Dec 2008 Posts: 9
|
Posted: Mon Jul 19, 2010 10:21 am Post subject: |
|
|
| ogminlo wrote: | | Toby@ITV wrote: | | Compress HD is an absolute MUST for us and FCS. I can't fathom how you couldn't use it if you have a moderately busy facility and are generating H/264 proxies. The Xserve is simply getting taxed too hard if it's ingesting, serving FCS to multiple users AND encoding proxies. |
Qmaster. Since v1.0, FCSvr has been able to offload the compression duties to a Qmaster cluster. A much more expensive option than CompressHD to be sure, but worth noting.
Also, the CompressHD 2.0 drivers finally dropped this week and they deliver. Now we just need to get Matrox to support more wrappers than .mov and .264... |
We have found Q Master to be too glitchy and temperamental for anything but a standalone quickcluster on an 8 core. That's why we rely on Compress HD. We tried Q master, but we couldn't afford to restart the xserve on daily during one of Qmaster's inevitable freakouts when 300 HD clips come down the pike. |
|
| Back to top |
|
 |
ogminlo Xsan Master

Joined: 29 May 2008 Posts: 149
|
Posted: Mon Jul 19, 2010 11:28 am Post subject: |
|
|
| Toby@ITV wrote: | | We have found Q Master to be too glitchy and temperamental for anything but a standalone quickcluster on an 8 core. That's why we rely on Compress HD. We tried Q master, but we couldn't afford to restart the xserve on daily during one of Qmaster's inevitable freakouts when 300 HD clips come down the pike. |
I do not push that kind of volume through our cluster, but once I treated my Qmaster participants like Xsan MDCs in terms of DNS (static address and reverse lookup), our managed cluster (dedicated boss and nodes) runs happily for weeks and months on end. Now and then I need to reset services, but not very often. But like I said, our throughput is not nearly as large as your 300+ HD clips each day. You are an order of magnitude beyond what we do. That is not to say that our cluster couldn't handle volume like that, I just can't claim to have tried. |
|
| Back to top |
|
 |
|