Dino Sanacory
Forum Replies Created
-
Actually the new Avid local storage is sourced from Promise. The picture on the website does not look at all like what they had out at the recent New York New Thinking event. It seems they have just completely tossed the entirety of the old Medea line and handed the job off to Promise. Sort of like what one of their competitors recently did…
Looks like a good box though. Interconnect through SAS. All RAID and stripping management and intelligence is inside the box so it can present to the computer as a single volume with no additional management burden placed on the CPU.
Management interface available via browser over ethernet. No more bouncing serial cables through multiple controllers. Claim the single 16 bay outperforms the previous double 10 bay setup. Don’t know what sort of RAID configurations apply in that comparison. As it’s super new couldn’t get too many questions answered.
-
If you shot 720P varicam for 24P, your fine, digitize overfirewire and the Avid reads the flags, bringing in only the relevant frames out of the 60 it records. If you shot 1080 as 24P plus pulldown then there is no easy way back to 24P. As the format makes no nod at 1080 as to the way an image was acquired, there is no information letting the Avid know which frames are keepers.
If the intention is to compare this to what Final Cut does, it does this. After capturing the interlaced media, you then go through a conversion process where every file is processed by removong the pulldown and generating a whole new file. A time and drive consuming process.
the best way to deal with this is to shot on a format that gives you what you actually need for you intended goal.
-
MediaComposer 2.5 and above will work with media of the following types: DNxHD, DVCProHD, HDV, XDCam HD, uncompressed MXF.
DVCProHD, HDV, XDCam HD is ingested as native files from the original acqusition media. The Media Composer does not create new media of these types.
With the DNxCell hardware, HD can be digitized as DNxHD. the hardware also outputs the above as video.
DNxHD and Uncompressed MXF media can be created via import, render or mixdown. Any DNxHD or uncompressed media generated in say, a Symphony, can be used in a Media Composer. the limit is the hardware. uncompressed media can neither be created nor output through the Adrenaline.
My experience with HDV on the Media Composer was quite positive. there were certainly some limitations, and I wouldn’t try to finish back to HDV, but it ran and a lot of nice work got done by a non-technical editor.
-
Even with 1/2 inch chips, recording at 50Mbit, 4:2:2 is a much better use of that camera. A 2/3 inch imager would give the price a nice bump as well. One of the great advantages of XDCam is it’s cost. Taking the price point between HDV and HDCam, the 1/2 inch chips fit.
Dino Sanacory
Edgeworx, NYC -
DNxHD, especially, at 220 should yeild noticably better results than DVCProHD. Two main factors drive this:
1. Data rate. DVCProHD at either 720/59.94P or 1080/59.94i runs at 100MBps. DNxHD is at 220 MBps.
2. Sampling structure. DVCProHD sub-samples the 1280 of 720 to 960 and the 1920 of 1080 to 1280. DNxHD maintains the full raster of the relevant format.I would consider the DNxHD 145 format to be for offline work.
-
Edgeworx, 212-334-3334
I have G4s and G5s with either AJA Io or Decklink and digibetas.
Dino Sanacory
Edgeworx, NYC -
As I don’t yet have FCP 5, I can’t comment on setting the audio bit depth. However, Digibeta is 20bit. An odd choice on Sony’s part but all the decks have been 20bit going back to D1, introduced in 1987. Possibly, further back on the BVH-2600? If you are sourcing from digibeta, you might as well work at 24bit.
-
If you have a deck that plays the tape, and it has an SDI out, that should be compatible with regular SD for capture by the Io into whatever codec your want your sequence to be.
-
You can use the composite output of the AJA Io to feed the REF in on the Beta deck ONLY for output to tape. You cannot use this configuration to digitize from the tape. With no actual BB source, the Io locks to the output of the beta for capture. The composite output of the Io is locked to the input. You are in effect referencing the Beta to itself. This does not work.
-
If your talking about real uncompressed HD, you’re gonna have to render. There are a few potential real time effects depending on which i/o card and computer you use but I wouldn’t trust them for final output. To get decent real time in HD you have to step up to something expensive like the DS from Avid, a bizarre but somewhat impressive system. And I still render for final output on that.
Something else to think about regarding what is real time and also overall render time is the growing range of settings and options in the systems (not just Final Cut) that are available today. Easiest is the choice of media at 8 or 10 bit. 8 bit is faster, 10 bit is better. Then there is the choice of render precision: 8 bit, 10/16 bit, and 32 bit float, as you may have already guessed, the bigger the number the more accurate but slower the render. I usually keep rendering set at 8 bit while I’m working then shift to a higher setting for final render.
Keep in mind (and your clints mind as well), HD is a LOT of data. It just isn’t going to be all real time and anything that is (other than a linear suite) is cutting corners. Let them be aware of the render time up front. Even the top Discreet (sorry, Autodesk) product need to render.
Also, Media 100 does have an HD product now. Not trying to sway you away from FCP, for the money, it’s damn impressive, just not as good as the hype.
Good Luck