Media 100HD – 10bit Uncompressed to MPEG-2 HD File?

[Kieran Matthew]

If you could do the compression on the machine that has the media100 HD system on it, then you could do a ref file export to the compression software.

Bearing in mind the issues you’ve raised over file size and cost of a physical drive to put it on, it would probably be cheaper & easier to buy whatever software the client was intending to use and do the compression yourself.

Is that possible?

[Michael J c]

Well, yeah… and that’s what I’m considering. They aren’t sure what software they’re using yet. They’re trying to do it for as cheap as possible.

I’m not sure I want to get into the encoding business (just because I don’t know a lot of the technical details). I also don’t know what a fair rate for software encoding is and worry about time consuming troubleshooting and re-encoding.

But the software I linked to above is what I’m considering getting. I’d actually prefer a hardware encoder for HD mpeg encoding, but again… not sure if I want to get into that.

But I am considering trying to figure this one out because of the fact that I can use a reference file.

[Stephen Bull]

Hi Michael,

1.5TB seems a lot for only 16 mins of uncompressed HD.

We have been capturing uncompressed 1080i into Media 100 HD at approx 500GB per hour. Bearing this in mind i would estimate you would be looking at approx 125-150GB for your 16 minute show, which should fit onto a portable hard drive for delivery to your client.

Hope this helps

Steve

[Michael J c]

That sounds better, but how are you doing that math?

The way I got to the 1.5tb number was that I exported a self-contained M100HD file from our 1080i timeline. It was exactly 10 seconds in length and size ended up being 1.58 GB.

Is my math wrong? Have you exported a long HD file as a self-contained movie, or are you talking about the size when media 100 captures something in real-time?

[Michael J c]

WHOA!!! Now that’s what I get for doing my math at 3:45am. 🙁

OK… I now am expecting a 150gb file… not 1.5 tb… Much, much better and more manageable.

oops.

[Michael Slowe]

I exported 29 minutes of 8 bit uncompressed HD to a LaCie firewire drive and it took up 167 of the available 250 Gigs. Took two hours to do it!

[Michael J c]

Thanks for the real world info Michael…

On that note, does anyone know if 10bit would make a difference considering it will end up an mpeg 2 hd file?

also, do you imagine 10bit being A LOT more time on that export, or just a fraction more time?

[Michael Slowe]

Michael C, this might be the blind leading the blind here but I do remember seeing a chart somewhere and there was a big difference in drive requirements between 8 and 10 bit uncompressed. I read in the Media 100 HD manual (page 2-15) that a 10 bit image carries 4 times more data per channel than an 8 bit image so it all depends on whether your system can handle it.

[Floh Peters]

[Michael Slowe] “I read in the Media 100 HD manual (page 2-15) that a 10 bit image carries 4 times more data per channel than an 8 bit image so it all depends on whether your system can handle it.”

While a 10 bit image has 4 times the resolution of an 8 bit image (1024 values for each color channel/pixel instead of 256) it only needs 25% more drive space and performance. Usually it should not make a big difference for going to MPEG, but it can if you e.g. want to colorcorrect the footage.

[Michael J c]

I think I’ll be using the M100 to do any minor color correctiong (if necessary) since the sources have all already been color corrected (they’re a myriad of feature film).

I don’t think drive performance on the client’s end will be an issue since they’re just doing a software encode of the file… They won’t have to play it back realtime before it has been made into an mpeg-2 hd file.

We just finished another project in 10bit M100HD but went to tape (and then a color correction session outside) and our drives have handled it with no problems. But wow… uncompressed hd sure takes up a lot of space… especially if you’ve got a few effects to render.