arrestthisman@comcast.net
Carbonite Member
Hello all. Long time reader, first time poster.... I think....
Anyway, I'm preparing for a feature length film I'll most likely be doing this Spring, and the HVX is an obvious choice for me since I'm an Avid DVX user, and am about to finish a documentary using a DVX.
I've currently been weighing 1080pa versus 720PN Pros and Cons, and due to storage and the vast amount of effects we intend to use for the film, I've decided on 720PN...mostly.
HOWEVER, there will be a lot of tracking shots and even with a mere 10-20% increase in definition with 1080pa it may make a good deal of difference.
I went through all of the acedemic motions to try and come up with a simple solution for the film; but real world usage got the best of me. And, I think the eye ball does a better job telling you what to do than a text book/document.
I knew all along I wanted a set timeline of 720PN or 1080p, because the idea of mixing timelines just isn't appealing for final out with my considerations. if 1/3 of the material is 1080, and the master out will be 1080, that is 2/3 720 that has to be upprezzed, and I want to uprez in shake, it simply looks worlds better than any NLE export filter. Lastly, the idea of going through hundreds of cuts individually for effects, and uprezzing before final out...
I learned through the doc I just did, that the more simple the workflow, the better... for reasons I'm sure I don't need to remind the veterans here.
Long story short, I conducted a series of tests using two TIFFS of the same image, one SHOT at 1080PA and one 720PN. The 720 lacked a good bit of detail, about 15% or so by eyeballing it. I was 50/50 whether the shot would be worth using 1080. The hassle, the space, etc. I got the idea that perhaps a 1080p shot would still retain superior detail if resized down to 720p, than if SHOT in 720PN in the first place.
The result was quite amazing. (And this applies to resizing in Shake only, I cannot vouch for other processes.) in shake, if one takes a 1080P shot, resizes it to 720p, and then resizes it back to 1080p, the resulting image is virtually indestingushable from the original shot. (A side note: we noticed that the original 1080p shot had a bit of aliasing, and that the aliasing was noticably improved in the resulting shot after processing, however the crucial points of detail were 99% preserved. My graphics guy and myself were shocked and amazed at this outcome.)
This means (in my amateur conclusion at least) that it is feasible and advantageous to shoot 720 for the majority of scenes, and 1080 for landscapes, wide shots, or any shot which needs a bump on detail, AND retain a homogenous timeline that preserves the merits and detail of each originally formatted shot.
This is usually not much of a concern for most people, but for those who will be heavily intermixing the material, and utilizing a lot of effects, a homogenous timeline is VERY helpful. I hope this helps any one out there, or provides food for thought, at the very least.
I'm more than happy to answer any questions on methodology, or if you'd like me to post images. (although I recommend doing your own tests, since I was not convinced until I did it myself, with my OWN equipment.)
Take care,
- Alex
Anyway, I'm preparing for a feature length film I'll most likely be doing this Spring, and the HVX is an obvious choice for me since I'm an Avid DVX user, and am about to finish a documentary using a DVX.
I've currently been weighing 1080pa versus 720PN Pros and Cons, and due to storage and the vast amount of effects we intend to use for the film, I've decided on 720PN...mostly.
HOWEVER, there will be a lot of tracking shots and even with a mere 10-20% increase in definition with 1080pa it may make a good deal of difference.
I went through all of the acedemic motions to try and come up with a simple solution for the film; but real world usage got the best of me. And, I think the eye ball does a better job telling you what to do than a text book/document.
I knew all along I wanted a set timeline of 720PN or 1080p, because the idea of mixing timelines just isn't appealing for final out with my considerations. if 1/3 of the material is 1080, and the master out will be 1080, that is 2/3 720 that has to be upprezzed, and I want to uprez in shake, it simply looks worlds better than any NLE export filter. Lastly, the idea of going through hundreds of cuts individually for effects, and uprezzing before final out...
I learned through the doc I just did, that the more simple the workflow, the better... for reasons I'm sure I don't need to remind the veterans here.
Long story short, I conducted a series of tests using two TIFFS of the same image, one SHOT at 1080PA and one 720PN. The 720 lacked a good bit of detail, about 15% or so by eyeballing it. I was 50/50 whether the shot would be worth using 1080. The hassle, the space, etc. I got the idea that perhaps a 1080p shot would still retain superior detail if resized down to 720p, than if SHOT in 720PN in the first place.
The result was quite amazing. (And this applies to resizing in Shake only, I cannot vouch for other processes.) in shake, if one takes a 1080P shot, resizes it to 720p, and then resizes it back to 1080p, the resulting image is virtually indestingushable from the original shot. (A side note: we noticed that the original 1080p shot had a bit of aliasing, and that the aliasing was noticably improved in the resulting shot after processing, however the crucial points of detail were 99% preserved. My graphics guy and myself were shocked and amazed at this outcome.)
This means (in my amateur conclusion at least) that it is feasible and advantageous to shoot 720 for the majority of scenes, and 1080 for landscapes, wide shots, or any shot which needs a bump on detail, AND retain a homogenous timeline that preserves the merits and detail of each originally formatted shot.
This is usually not much of a concern for most people, but for those who will be heavily intermixing the material, and utilizing a lot of effects, a homogenous timeline is VERY helpful. I hope this helps any one out there, or provides food for thought, at the very least.
I'm more than happy to answer any questions on methodology, or if you'd like me to post images. (although I recommend doing your own tests, since I was not convinced until I did it myself, with my OWN equipment.)
Take care,
- Alex
