Sean Oneil
Forum Replies Created
-
The online facility most likely outputted your project at 29.97, since that’s what most tape formats can record at. During this process, 3:2 pulldown was added to convert 23.98 to 29.97. This is “regular” telecine pulldown, which is different from 3:2:2:3 Advanced Pulldown.
In order to remove 3:2 pulldown, you should capture at 29.97 and use Cinema Tools to convert it back to 23.98. When using Cinema Tools, you’ll need to determine the right cadence which will be one of five choices. You can use trial and error to figure it out.
-
I just learned all this myself (the hard way).
Volume Level = Only one client can have write-access at a time.
File Level = All clients can safely have read/write-access at the same time.If multiple computers have write access to the same disk volume, it will cause data corruption without a proper SAN solution. Volume level solutions manage it so that only one machine has write access at a time. Someone has to administer this usually. File level solutions use a metadata controller to allow all machines write-access. XSan and MetaSAN are both file level. Xsan requires a dedicated server for this (plus reccomends a backup). But MetaSAN automatically assigns the metadata duties to any available client machine, eliminating the need for an additional server.
I just started using MetaSAN with an iSCSI solution from Open-E (server) and ATTO (software client). In case you don’t know, iSCSI uses ethernet instead of Fiber Channel. I went with my particular solution because it’s 100% scalable. For example, I can easily upgrade to 10GB ethernet when it’s available and cheap. It hasn’t gone as smoothly as I’d have liked, but as far as MetaSAN is concerned, it’s been great.
Studio Network Solutions offers a complete iSCSI SAN solution that’s supposed to be pretty solid. But it’s pricey and it’s volume-level sharing only.
Sean
-
??? I replied to this a few days ago and it’s gone.
Anyways, I think you’re just misunderstanding me. When the camera is set to 30fps (which Daryl could very well have used), there is no pulldown added. 60i, with every 2 fields making up a frame. So it only requires a simple deinterlace filter to make this kind of footage progressive-scan. At that point, you can scale it and lower the frame-rate using simple tools (FCP) without any side-effects.
But when you have 60i footage shot at 24fps (3:2 pulldown added), then of course a simple deinterlace filter will not work. As you and I both said, you need a Teranex in this case (or a lot of time and patience w/ Cinema Tools).
Both the HDX200 and the Z1U record 1080 60i but can be set to use a framerate of either 30fps or 24fps. So the the same concept applies for either camera. The fact that DVCProHD is far superior to HDV doesn’t have anything to do with this particular situation. In fact, 90% of what we shoot is on 720 Varicam, but a lot of B-roll is shot on 1080 HDV. This is how I figured out this workflow, and it does work.
And yes, I use a HD CRT monitor connected via component to a Blackmagic Multibridge (which is connected via HD-SDI to the Mac).
So again, if Daryl shot it at 30fps, all he has to do is throw a deinterlace filter on the clip, choose the right setting, and then place the clip on a 24fps 720p timeline and render. It works great with FCP 5.1. It does not “fall apart”. If you ‘d like to stop by I’d be happy to prove it to you.
Sean
-
[gary adcock] “While running the “deinterlace” filter may pass in the SD world- in HDTV the converted footage would be very, very noticeable when placed in between other non converted footage, showing as much as 50-60% softer looking on an HD display. Mostly because you do not have the interlaced field order to hide detail in.”
Gary, I’m a bit confused. What do yo mean “you do not have the interlaced field order to hide detail in”. Could you elaborate on that a bit? It didn’t make sense to me.
All I know is that I’ve had 100% success doing this with HDV footage. It only works if the footage was shot at 30fps 60i. The reason is because this type of footage does not contain any pulldown pattern. A simple weave deinterlace will combine the fields back together perfectly and create a flawless 30p video image. No more fields, and no redundant frames. Thus, you are able to scale it, change the framerate – do whatever you want to it using simple tools (such as rendering it in a FCP sequence that has different settings). And the results are pristine.
What I was saying though is that this will not work if the footage was shot at 60i using a 24fps framerate – because a pulldown pattern is added to the video to achieve this. If this is the case, one has to first remove the pulldown using more complex tools (Cinema Tools, Natress, etc.) before scaling it or changing the framerate.
-
SCSI is dead. SAS (Serial Attached SCSI) just came out. SAS is to SCSI what SATA is to ATA.
-
[Chris Poisson] ”
According to section 4 of this article, the render situation is as you describe. Worth the extra time? Pretty subjective call I’d say.”It’s the “better safe than sorry” route. If you want to encode from your “already rendered” media to save time, just export a self-contained quicktime and then run it through Compressor. Exporting a self-contained will not re-render anything if you choose “Current Settings”.
Sean
-
If it’s 1080 60i without any 3:2 pulldown, just add a deinterlace filter (essentiall making it 60p) and then throw it on a 720p 24 timeline and render. That’s it, works great and shouldn’t take very long to render. Export a self contained when you’re done and import it into your project.
But if it does have 3:2 pulldown, it will be a chore to convert it.
Sean
-
[Tom Wolsky] “My point is that there are much more important things to fix and do in the application in my view than MP3 playback. New things are broken in FCP5 like XML export and import, which is a major killer, and long time things have yet to be made to work properly, like merged clip and freeze frame media management, and a true multiformat timeline are just a couple of them.”
I agree. I never said MP3 playback should be the FCP team’s #1 priority. Nor did anyone else to my knowledge. That hardly justifies the notion that people shouldn’t talk about it.
-
BTW,are you sure the actual video footage is 30p? Are you sure it isn’t just 30fps timecode but the video is 23.98? Jog through frame by frame and look for duplicate frames.
I’ve gotten several jobs where they set the Varicams to 23.98, yet the timecode was 30fps SMTPE non-drop. This has caused a tremendous headache.
Sean
-
You can still capture it natively and render the final edit on an Uncompressed timeline. That way the effects don’t get recompressed.
Sean