-
HDV from FC 5.0 to Shake (or AE) and back
hello!
i’m currently editing a music video shot on HDV. i work natively. everything goes fine.
color correction within final cut works very well and i must say that i’m really impressed by the quality i get!to do some masking, retouching i have to leave the “fc-microcosm”. shake will do the job.
– well, no reference clips – they become reencoded anyway and get unusable.
– running through the original source clips and manually finding the corresponging in-points may look like the solution, but why don’t go….…uncompressed from now on!
1. the quality of the exported clips (10 bit uncompressed) looks fine, but they have a slight gamma shift. i recorrect the gamma within shake (1.027 looked near enough). at this time i did not care about the little banding effect this re-conversion began to show…
2. i do my work and render out to uncompressed 10 bit quicktime
3. after importing to final cut and putting in the timeline i apply the color correction (color finesse): very crappy result after rendering to HDV. ok, seem logically – but only to 50%: why is the second generation of HDV so good when i render color corrections on the original clips???
4. sequence settings to 10 bit uncompressed. that should solve the problem!
5. well, it should, but now with the color correction applied the loss of fine shading details become clearly visible! the price for all that re-re-reconversion orgy…. 🙁who knows the ultimate way to get the clips out of final cut without that unnecessary data loss???
thanks a lot!
alex
-
9 Replies
-
I think the ultimate solution for you is to notwork in HDV at all. Take the HDV files and capture them as DVCPRO100 clips and you’ll be miles ahead I’ll wager. If you don’t have the faciltiy to do this, use HDVxDV and get there that way. https://www.dvdxdv.com/products.html
HDV isn’t my favorite format… and your post only solidfies my feelings.
Jerry
Apple Certified Trainer
Author: “Jerry Hofmann on Final Cut Pro 4” Click here
Dual 2 gig G5, AJA Kona SD, AJA Kona 2, Huge Systems Array UL3D
-
Compressing HDV as DVCProHD only makes them worse. DVCproHD is also highly compressed and exhibits many of the same problems as HDV, especially in the shadows, although not to quite the same extent.
Best go to an uncompressed workflow asap. Going to 10bit is probably overkill for an 8bit format, and Shake can work in float anyway, so I’d go out 8bit uncompressed into Shake, and that might avoid the gamma issue.
I’d also do your colour correction in Shake at the same time, rather than use Color Finesse as again you’re avoiding a translation and if you can use Shake, you can do everthing you need then more.
Graeme
– http://www.nattress.com – Film Effects and Standards Conversion for FCP
-
Agree with you here Graeme, but of course this all requires a card, and a lot faster disk drives… Don’t you think that DVCPROHD would still be better than HDV? at least SOME improvement? seems to me that it would, but have to admit I’ve not tried this, just read posts about it.
A whle back I saw a demo of what you are proposing, and indeed it really did a fine job. But the system was comprised of a Kona 2, and an analog converter, and a set of FAST hard drives. SCSI 320 at the time I recall…
Jerry
Apple Certified Trainer
Author: “Jerry Hofmann on Final Cut Pro 4” Click here
Dual 2 gig G5, AJA Kona SD, AJA Kona 2, Huge Systems Array UL3D
-
Going from one compressed format, to another very slightly less compressed format offers no real advantage. The only time I’ve seen it work is when I did a 1080i60 HDV to 720p24 DVCproHD test via my converter, and even then, I’d be getting better results going to PhotoJPEG75% (a very lightly compressed codec) or uncompressed.
You don’t need any kind of fancy HD card to do this, you can just use FCP (not quicktime as it mucks up the 4:2:0 chroma sampling of HDV) to convert HDV to uncompressed, and then use that as your intermediary to go to shake, AE etc.
To view it, you’d need an HD card, but to view HDV you need an HD card, so you need it anyway in my book.
I’m of the view that although DVCproHD is much more edit friendly and less compressed than HDV, it’s still TOO compressed for serious work. I mean, really, it’s more compressed than DV and people don’t consider that serious.
Graeme
– http://www.nattress.com – Film Effects and Standards Conversion for FCP
-
There is some advantage I think: renders will be faster, and especially a down convert to SD MPEG would be a TON faster. Making SD MPEG files from HDV native is a really long render for sure.
Jerry
Apple Certified Trainer
Author: “Jerry Hofmann on Final Cut Pro 4” Click here
Dual 2 gig G5, AJA Kona SD, AJA Kona 2, Huge Systems Array UL3D
-
Absolutely!
But at that point I’d be suggesting PhotoJPEG75% as an intermediary also.
Graeme
– http://www.nattress.com – Film Effects and Standards Conversion for FCP
-
hi,
thanks for your suggestions.
cannot really edit in uncompressed mode because of the missing fast array.
but i would have nothing against setting the timeline to uncompressed at the end of the work and then render all effects again – to avoid recompression in HDV.also i would go in “uncompressed mode” through external applications.
but. if i export from this project in whatever format i’m not able to get the clip back without having some luma/gamma/whatever change.if someone of you exports true uncompressed hd-footage and reimport it – what says your waveform?
greetings
alex
ps: under http://www.schallundrauch.tv/exchange there’s a .psd with exporting examples.
-
Try using uncompressed 8bit – that should avoid luma / gamma changes. You can select a portion of the HDV timeline, and file -> export -> quicktime movie…. and select uncompressed HD from the presets at that point, for generating your intermediaries.
Graeme
– http://www.nattress.com – Film Effects and Standards Conversion for FCP
-
hi
that SHOULD – but it doesn’t….
it even changes the colors – the face gets more magenta
on 10bit there is no color change, but the luminance gets affected.final cut/quicktime does something on conversion that does not happen if the source files were uncompressed (i did a series of tests with uncompressed 10 bit sd footage).
the current project is kind of a research project on how much can you get out of that hdv stuff.
conclusion at this time:
– surprisingly good base for grading
– high resolution
– many artifacts but in practice they are mostly ingnorable – you don’t notice them on playback
– quite noiseless (z1)
– well, forget about keying…what i would do better next time:
– don’t edit natively if there is some (external) vfx to do
– realtime converter from hdv to hd-sdi should resolve the problems mentioned
– or – if possible – capture via component-hd directly on the set (probably the best way at all)
– buy an array 🙂as if i did not know since a long time: uncompressed rulez! 😉
greetings
alex
Reply to this Discussion! Login or Sign Up