Shooting for HDR deliverable (3 replies and 1 comment)
I am in pre-production for a feature film and we will have SDR and HDR deliverables. Since this is my first time shooting for an HDR deliverable, I was wondering if anybody has some tips and tricks, eg. anything I should pay particular attention to? I will have an HDR as well as SDR monitor on set. Any thoughts and advice are welcome. Thank you.
I'm not an expert on HDR, and of course Mr. Deakins and Mr. Mullen (and also other great DP's on this platform) can be more specific than me.
I think that the HDR delivery is related only to the "post-processing signal". A RAW file of a high-end camera (as Alexa and Red) but also Prosumer (as B-Raw and ProRes Raw) is already an HDR as contain something from 12 to 14/16 STOP of Dynamic Range.
So you have to light your scene for the mood you want, if you think about lighting your scene in HDR lighting you could lose the mood you want for the scene and the story.
The HDR then is a re-mapping process of all the tonal values contained into the RAW file for an HDR monitor (that should have the pure white around 1000 nits but I could be wrong), so your "near clip and clip" highlights moves from the nits of an SDR monitor to the nits of an HDR monitor.
I think that the double monitor on-set can help you to see the double version of a re-mapping of the tonal values, but I think that you have to light your scene in the same way of any SDR or film stock delivery.
This is only my thought but I will be very happy to learn from masters that are here.
I wish you a nice day!
Hi Max, just figured that I should have posted this as a comment to your reply, rather than as a new reply. So let me refer to the next reply in this thread.
thank you very much for your reply. My understanding is that, indeed, HDR relates to the post-processing signal, and that means that the color grade will have to be adjusted to the HDR version of the film.
My main concern is that, after reading up on HDR a bit, with the added sharpness that comes along with an HDR master, things like strong backlights might look harsher than on SDR. I am wondering whether DPs out there try to take that into consideration when lighting a scene, or whether that's something that is mainly dealt with in color grading. I am concerned exactly about what you mentioned in your reply, that I might lose the mood I want due to thinking about the HDR "look" on set.
Again, thank you very much for your kind reply, your thoughts are very much appreciated.
Have a great weekend.
Going to talk from the point of view of a tech (as a DIT).
I recently got off a shoot where our final deliverable was HDR Dolby Vision.
I'm not too familiar with the ins and outs of the post workflow. However, on set we had a Flanders DM240 SDR Monitor a CM251 OLED and an XM551.
Yes theoretically you can 'tone-map'/'re-map'/transform your SDR signal to take advantage of the full range of a HDR signal but most DP's I've worked with never truly want to utilise full range of HDR viewing but rather let bright highlights go a little brighter than standard.
One would also note, XOCN, ARRIRAW is typically captured as a 12-bit RAW file. Now, yes they advertise the ability to reconstruct a full 16-bit linear signal from there container (in a visually lossless manner) and while this is somewhat true when dealing with a differentiation of linear values higher than 512 it would exceed the highest capacitance that (for example) the ARRI Log-C container can hold in terms of per-stop values. Yes... one can easily interpolate values between the 512 values assigned to each stop above a linear value of 1024 but at that stage you are interpolating data... I guess. You would never be able to tell, but, in terms of taking advantage of your full signal that's the point in which you are, around the 1500 nit/value mark.
The exception is if you want to capture true 16-bit linear or REDCODE due to prior patents have a superior compressed RAW codec allowing a more accurate reconstruction of a 16-bit linear signal... apparently...
In terms of visual characteristics, as you're not compressing the top end of your signal as much differentiations in highlights are a lot more pronounced. High-light roll-off to clipping can be a lot harsher, differentiations in highlights like glares, flares, halation so on and so forth can appear more pronounced. As again, you're no longer compressing the high-end of the signal noise can also be more pronounced (as of course with more data values and compression, your noise ratio drops as you compress signal).
However, this entirely depends on your DI. You can of course just select or window certain highlight features you'd like to go above standard. I know with us we did have to over-all expand and utilise a certain range of highlight signal for a HDR master but that was in a more sausage factory environment.