Best Output Option For Shake Into Fcp

[user_deleted]

If I start a comp in Shake, what is the best render out option for retaining DV quality. Also, if going back and forth between FCP 5 and Shake 4. Not sure if I export from FCP, does it automatically choses the best/cleanest format or not. What is the optimal choice...DV, Apple Animation, Uncompressed 10-bit? Can I/should I have any transitions in FCP or will force another compression onto the footage....First post for me thanks for the help...

[user_deleted]

QUOTE(vashifilms @ 12/08/05, 08:47 AM)
If I start a comp in Shake, what is the best render out option for retaining DV quality.  Also, if going back and forth between FCP 5 and Shake 4.  Not sure if I export from FCP, does it automatically choses the best/cleanest format or not.  What is the optimal choice...DV, Apple Animation, Uncompressed 10-bit?  Can I/should I have any transitions in FCP or will force another compression onto the footage....First post for me thanks for the help...
[snapback]221919[/snapback]

Since I work everyday with 10 bit uncompressed I have a habit of converting everything and outputing everything to 10-bit uncompressed. I've see this kind of workflow at feature film editing and many different places. you can export with ntsc codec, but if you want to be safe, export uncompressed 10 or 8 bit. Perhaps someone with more insight on file compressions can justify my reasoning.

[user_deleted]

QUOTE(vashifilms @ 12/08/05, 08:47 AM)
If I start a comp in Shake, what is the best render out option for retaining DV quality.  Also, if going back and forth between FCP 5 and Shake 4.  Not sure if I export from FCP, does it automatically choses the best/cleanest format or not.  What is the optimal choice...DV, Apple Animation, Uncompressed 10-bit?  Can I/should I have any transitions in FCP or will force another compression onto the footage....First post for me thanks for the help...
[snapback]221919[/snapback]

The main problem is changing constantly between RGB and YUV color spaces. When you capture video into FCP, it brings it in as YUV unless you use a codec that is RGB such as from KONA2 card where you can specific capturing the footage as such. If you do native DV its YUV space.
On the other hand, all compositing programs work in RGB space, so once you are in Shake, you are working in RGB space. The good thing about the new path from FCP to Shake is that you don't need to render out of FCP, so you don't only save on storage space, but also on rendering to another codec. Work with the footage in Shake and then render it back to FCP as YUV (DV, 8bit uncompressed, 10bit uncompressed - depending on your needs, quality and storage space). You want to do this back and forth as little as possible, once only would be best.
To get the highest quality possible, you should capture the footage as 10bit and go from there. That would give you the biggest headroom. Consider that you can do an offline using dv and then re-capture only the footage being used in the seq again with 10 bit and that way you save a lot of storage space and can do your offline edit faster.
Hope this helps...

[user_deleted]

QUOTE(algit @ 12/09/05, 02:57 PM)
To get the highest quality possible, you should capture the footage as 10bit and go from there. That would give you the biggest headroom. Consider that you can do an offline using dv and then re-capture only the footage being used in the seq again with 10 bit and that way you save a lot of storage space and can do your offline edit faster.
Hope this helps...
[snapback]222020[/snapback]

Question: what do you have your 'Globals' bit-depth set to? 8-bit, 16-bit, or float?

thanks,
david

[user_deleted]

QUOTE(boscablanca @ 12/15/05, 11:05 AM)
Question: what do you have your 'Globals' bit-depth set to? 8-bit, 16-bit, or float?

thanks,
david
[snapback]222497[/snapback]

As the Manual puts it:

"Sets the default bit rate for Shake-generated images. The defaultBytes parameter has no effect on images that are read in via FileIn nodes, nor does it affect the rendered output from your script. "

So this means that if you bring your own footage, it will hold its bit depth. 10 bit images are treated as 16bit inside shake. It would depend on what you want to do and what your final delivery is. Going with greater bit depth is great, but you pay for it in time. So... be sure to know why you are doing it and don't just have it at float because you can.
A.

[user_deleted]

QUOTE(algit @ 12/15/05, 06:35 PM)
So this means that if you bring your own footage, it will hold its bit depth. 10 bit images are treated as 16bit inside shake. It would depend on what you want to do and what your final delivery is. Going with greater bit depth is great, but you pay for it in time. So... be sure to know why you are doing it and don't just have it at float because you can.
A.
[snapback]222525[/snapback]

Good point. My goal is get back to 10-bit 4:2:2 YUV (for laying back to D5 HD) without going down to 8-bit RGB. After Effects lets you set your working mode to 16-bit, but sometimes the image goes haywire (artifacts.) I don't know of a similar switch for Shake, other than for Shake-created graphics, as you quoted; is there a way to color correct in float other than going through a 'float' node first?

thanks,
david

[user_deleted]

QUOTE(boscablanca @ 12/16/05, 07:17 AM)
Good point. My goal is get back to 10-bit 4:2:2 YUV (for laying back to D5 HD) without going down to 8-bit RGB. After Effects lets you set your working mode to 16-bit, but sometimes the image goes haywire (artifacts.) I don't know of a similar switch for Shake, other than for Shake-created graphics, as you quoted; is there a way to color correct in float other than going through a 'float' node first?

thanks,
david
[snapback]222577[/snapback]

If you need to work in Float, you have to put a Bytes node. You can have multiple sections of your tree in different depths... so you should take the section you want into Float, do what you need, and then take it back to 16 when you combine it with the rest of your stuff. This way you only work in Float in the area you really need. What is your source footage?