PDA

View Full Version : OMFI MediaFiles



SJ_Lee
04-16-2004, 06:56 AM
This could be in techincal but I think Michael can best answer this question.

1) Let's say I'm recording directly from the DVX to the computer. Is this any way different than recording from tape?

2) Can you explain a the Avid OMFI file structure? For instance, how are the meta deta constructed?

I would like to know what is happening when I import/export files into Avid. I also know that 601 only uses bits 16-240, what happens to those bits when it is exported out to a Quicktime file?

I'm asking because I'd like to know the best way to export files out from Avid to raw data format that will eventually be turned to uncompressed media.

Thanks

MichaelP
04-16-2004, 09:32 AM
OMFI is a wrapper for metadata to an exisiting JFIF media structure. It does not have mich bearing on the actual color space and quality since it only contains what you have set the quality for during cqpture, import, export, etc. If working at DV25, exporting to a TIFF or other graphic file format, the compressed DV25 material gets decompressed (to uncompressed media) then exported.

As far as luminance range, this is a factore of color correction etc. The default mode is for digital telelvision where 16-235 is the standard. If you are exporting out as TIFF for example, you can choose to export as 601 where 16 will remain 16, or as RGB where the 16 will become 0 and 235 will become 255. This is done as an even stretch across the whole range.

When importing, it is up to you to know what is inside the file, 0 or 16. If 16, import as 601 and it will remain 16. If 0, selecting RGB will raise it to 16 and take 255 back down to 235 squeezing the range to standard digital video. The inverse of what was done on the export.

If 601 is selected, then 0 will remnian as 0 and 255 will remain at 255 since we willassume that the file meets 601 standards.

Michael

SJ_Lee
04-16-2004, 11:48 AM
Hrmm,

But that begs the question, whether DV records beyond the 16-235. I'm pretty sure this happens, which means that if you decide to use the full lattitude, you will get the top and bottom end crushed/clipped if you chose not to output them as 601s?

I'm pretty sure that if I output BMPs/Tiffs from Quicktime players, it will not streach that space.

So should outputing it to Raw DV QT (in Avid DV codec), yield the best quality possible for uncompressed media?

MichaelP
04-16-2004, 11:12 PM
The capture process changes nothing to what was actually shot with the camera, it leave it alone. I have found that digitizing the very same media into Xpress Pro and FCP that QUickTime will take what was 16 and make it 0 and what was 235 and make it 255. SO what happens there is that any detail you may have had in those ranges have been clipped during the capture process for any type of file export.

In Xpress Pro, it is perfectly fine to have information in the 0-16 and 235 to 255 range but the decision to keep it or not or how much of it is done in the color correction process. Depending on your final delivery requirement you can correct to 16-235 to meet digitsal TV standards or 0-255 for film out and computer output.

I am still confused by your comment of uncompressed media. When capturing native DV over firewire as DV25, the media is a compressed 5:1 since that is what DV25 is. For all subsequent renders, the files must be decompressed, effects added and recompressed to DV25 to meet that standard.

Working in a full uncompressed format would mean decompressing during the capture stage (via SDI) and working in a 1:1 resolution (that is not DV).

Michael

SJ_Lee
04-16-2004, 11:42 PM
Right, I guess I kind of glossed over what I was thinking.

I figure that the best method of getting DV25 media from the DVX over to digitized computerized form is:

1) Capture using Avid
2) Export to QuickTime DV with Avid Codec
3) Convert exported QT DV to uncompressed media
4) Do as much effects with that media on the computer side

Provided my delivery method is a digitized form (Win AVI, Windows Media, etc) wouldn't doing the above workflow be the best way to go?

By keeping my master source, uncompressed, I can do many things to it without worrying about recompression problems.

I know I'll need to use After Effects in some form or another so I figure that is the way to go. Am I wrong about this?

MichaelP
04-17-2004, 06:40 AM
THw workflow you suggest would work. Once all the effects were done in AE or whatever, it would then have to recompress to DV 25 to come back into the system.

Michael

SJ_Lee
04-17-2004, 02:48 PM
Now, I should really re-ask some of the question properly.

1) What would be the best method of outputing that video out to Windows format? Is it exporting it as QuickTime DV with Avid Codec like I said above and conver that DV into uncompressed via QT, or should I convert it to QT/AVI media striaght from Avid?

2) I'm not too familiar with SD, is this captured via the RCA cable in? Otherwise I don't think the DVX has SD in.

3) You also mentioned that the media I port back out to Avid will need to be recompressed into DV format. Yes, but doesn't this happen anyway in the NLE workflow anyways? When the effects are added to the DV media it should conform the DV media must be comformed to RGB, then recompressed to DV.

Thanks