juanjomora
New member
DSLR Cameras and the DI Process
DSLR Cameras and the DI Process
Hi.
I've been reading this thread with a lot of interest.
Full disclosure: I work at a major DI facility in Mexico City and have some experience (~7 years) in the so-called 'high end postproduction' market.
From my personal experience you need to have a lot of hands-on practice and a good knowledge of the film production and post workflows to get good results from the DSLRs.
I'm speaking about doing a feature film or documentary for theatrical release in 35mm prints or DCPs.
Most DSLRs record to some flavor of the H.264 codec, nice for viewing or delivery. Not so nice for acquisition and post. It's very processor-intensive. You need to transcode it to a codec more suitable for editing and processing (color grade, compositing, etc.) Apple ProRes 422 is the more popular choice. You need to take that into account.
Personally I don't like the quality of the image: if you load a clip from a DSLR into a good color grading system (i.e. Resolve, SpeedGrade, Lustre, Scratch), visualize it in a proper monitor with a porper LUT and isolate one of the color channels you can see big ugly blocks (from both color sampling and compression schemes). If the material is not exposed carefully there's not much you can do to rescue blown highlights or sub-exposed areas.
For broadcast or electronic delivery one has to be very careful: the risk for concatenation artifacts is high. Yes, I know about the episode from "Dr. House" but if you read carefully you'll find they spent a lot of money in post just to make it look decent...
In the other hand we have finished 5 documentaries that were recorded in a DVX-100B and they looked great in the final 35mm positive print; of course the DoPs in charge of these projects knew the camera and its limitations very well, and the editors were very careful to maintain frame integrity through the whole process.
Of course DV video with its limited color sampling and small frame size also has severe limitations; but the workflow is pretty well understood and (IMHO) very robust.
Also, when working with DSLRs you need to be very disciplined and have experience in what is called separate audio system production: if you are not careful you'll end with takes with no sound or with very serious lip-sync issues.
From the production standpoint (I used to be a camera operator man years ago) I believe the video camera package is more compact and robust: if you don't have a good crew (necessary for separate audio work) you'll encounter a lot of trouble.
In the end, it all boils down to the kind of project you want to make and the resources you have access-to.
Sorry for the long post. English is not my native language so please excuse my poor grammar and orthography.
Regards,
Jj
DSLR Cameras and the DI Process
Hi.
I've been reading this thread with a lot of interest.
Full disclosure: I work at a major DI facility in Mexico City and have some experience (~7 years) in the so-called 'high end postproduction' market.
From my personal experience you need to have a lot of hands-on practice and a good knowledge of the film production and post workflows to get good results from the DSLRs.
I'm speaking about doing a feature film or documentary for theatrical release in 35mm prints or DCPs.
Most DSLRs record to some flavor of the H.264 codec, nice for viewing or delivery. Not so nice for acquisition and post. It's very processor-intensive. You need to transcode it to a codec more suitable for editing and processing (color grade, compositing, etc.) Apple ProRes 422 is the more popular choice. You need to take that into account.
Personally I don't like the quality of the image: if you load a clip from a DSLR into a good color grading system (i.e. Resolve, SpeedGrade, Lustre, Scratch), visualize it in a proper monitor with a porper LUT and isolate one of the color channels you can see big ugly blocks (from both color sampling and compression schemes). If the material is not exposed carefully there's not much you can do to rescue blown highlights or sub-exposed areas.
For broadcast or electronic delivery one has to be very careful: the risk for concatenation artifacts is high. Yes, I know about the episode from "Dr. House" but if you read carefully you'll find they spent a lot of money in post just to make it look decent...
In the other hand we have finished 5 documentaries that were recorded in a DVX-100B and they looked great in the final 35mm positive print; of course the DoPs in charge of these projects knew the camera and its limitations very well, and the editors were very careful to maintain frame integrity through the whole process.
Of course DV video with its limited color sampling and small frame size also has severe limitations; but the workflow is pretty well understood and (IMHO) very robust.
Also, when working with DSLRs you need to be very disciplined and have experience in what is called separate audio system production: if you are not careful you'll end with takes with no sound or with very serious lip-sync issues.
From the production standpoint (I used to be a camera operator man years ago) I believe the video camera package is more compact and robust: if you don't have a good crew (necessary for separate audio work) you'll encounter a lot of trouble.
In the end, it all boils down to the kind of project you want to make and the resources you have access-to.
Sorry for the long post. English is not my native language so please excuse my poor grammar and orthography.
Regards,
Jj