Alexa / RED / Super 16 comparison footage

John Brawley said:
Thanks.

Yes it was a hacked GH1. I used for a few key scenes in the series. We had a scene that largely took place in a night club toilet....with three cast....in the cubicle ! So it was perfect for those kinds of one of shots. One of our characters also rides a bicycle a lot and it was a great camera for rigging to his bike. This allowed the cast member to actually safely ride his bike in the show without being on a trailer etc. And you get much better travelling shots that way, cause he can turn through 360 degrees and it looks great on camera.

http://flic.kr/p/9Nz6se

We also used the GH1 for the occasional car rig shot but the AF100's tended to be the preferred car rig cameras. We have very strict road rules that don't even allow for cars to be driven with Hostess trays, or hood mounts. We have to shoot everything from within the car or get a low loader.

The GH1 did struggle a little in the cars, mainly because of it's more limited dynamic range in difficult lighting conditions, but just about any camera has a hard time with these kinds of

I've found with the cast that if I tell them what I'm doing and why I'm doing it then they get right on board. They equate it with more intimate scenes too. I've often used the GH1 / AF100 for closed set type scenarios where it's just me and the cast in the room. It's a lot less intimidating and more discrete for this kind of work. Sometimes I'd retire the A camera operator and the director would operate a second camera. She was the series setup director and is ex-camera department herself. Again, this makes it much more intimate with the cast. Just two camera operators, a boom swinger and the cast in the room. And the director operating with me on the B camera.

I've also given the cast AF100's or GH1's for the occasional POV shot. Our main lead LOVES to do her own POV shots when she can. She's actually become quite good as an operator and asks to do them !

Her character has a lot of fantasies or segues into alternate possibilities so we're always doing speed ramps or rigging cameras in crazy ways. We use lot's of different cameras so the GH1 is just another camera that gets used for certain kinds of shots or setups. The main disappointment with it is that it doesn't have any way of monitoring it's output when it's rolling.

jb
Click to expand...

I just woke up. It's sunday, and this was a joy to read after a camera fight. John, that sounds like the experience we're all after man. Good job and cheers :)
 
Ok, waking up. s16 makes me the happiest, alexa is the first cam i've seen footage off that threatens that happiness...

... but the red? what happened to that window? surely exposing for the window can result in not too noisy shadows? did u not expose for the window? I mean I'm not into red, but still I'm sure it can't pale that bad. It looks like an hd-fake-shoulder next to those two...

reading over to see if i missed any exposure details...
 
I think with HDR-X on the MX will pull ahead clearly and decisively. It's supposed to have about 18 stops when HDR-X is on...from what I've heard. This test is clearly MX without HDR-X, so it's not the best it's capable of and I think a bit more detail could have been pulled out on that window.
 
It should be able to pull more detail without hdrx though, i think the test was to see how far it'll reach into the highlights when exposed for the 'average'. I guess. I may sound like a moron.

But definitely could've pulled more out of the windows. I think at least.

T
 
Please keep in mind that even Jim Jannard himself said HDRx will not work in all situations and should be kept for "needs must" situations.

Basically, HDRx should not be used in any DR tests to help determine real world shoots because certain shots might not work out with HDRx
 
Last edited:
greymog said:
It should be able to pull more detail without hdrx though, i think the test was to see how far it'll reach into the highlights when exposed for the 'average'. I guess. I may sound like a moron.

But definitely could've pulled more out of the windows. I think at least.

T
Click to expand...

I agree. I think they could have pulled out more detail from the highlights as well.
 
Two issues with the Red footage. What settings were used to grade it? Are the windows actually blown out, i.e clipped, in the raw file or is the data there? Because if the data is there in the raw exposure, then the way the scene is rendered for value in the example is a grading choice, whether deliberate or not. A reasonable comparison with Arri log-C footage would start with Redcolor2 color space and Redlogfilm gamma as grading points and refining from there. The major difference between Arri and Red 1MX on a truly comparable exposure and grade would be just a little more noise in the deep shadows with the Red, not blown highlights.
Many of these comparisons look to me like the R1 footage was graded with REC709 gamma and color space as a starting point, effectively throwing away 3 stops of available DR in the Raw file.

Here is a true example of how an R1 Mx can handle extreme dynamic range when skillfully graded.

http://www.zoomfilmtv.com.au/ftp/Dynamic_range/RED_hi_contrast_test2.mov
 
Last edited:
Razz16mm said:
Two issues with the Red footage. What settings were used to grade it? Are the windows actually blown out, i.e clipped, in the raw file or is the data there? Because if the data is there in the raw exposure, then the way the scene is rendered for value in the example is a grading choice, whether deliberate or not. A reasonable comparison with Arri log-C footage would start with Redcolor2 color space and Redlogfilm gamma as grading points and refining from there. The major difference between Arri and Red 1MX on a truly comparable exposure and grade would be just a little more noise in the deep shadows with the Red, not blown highlights.
Many of these comparisons look to me like the R1 footage was graded with REC709 gamma and color space as a starting point, effectively throwing away 3 stops of available DR in the Raw file.

Here is a true example of how an R1 Mx can handle extreme dynamic range when skillfully graded.

http://www.zoomfilmtv.com.au/ftp/Dynamic_range/RED_hi_contrast_test2.mov
Click to expand...

WOW! That is amazing.
 
greymog said:
It should be able to pull more detail without hdrx though, i think the test was to see how far it'll reach into the highlights when exposed for the 'average'. I guess. I may sound like a moron.

But definitely could've pulled more out of the windows. I think at least.

T
Click to expand...

You're correct. I didn't expose for the highlights. That was my choice to do so, to expose them all at the same level. I could have exposed the RED for the windows and traded some shadows for highlights, but that wouldn't be a fair point of comparison would it.



jb
 
Razz16mm said:
Two issues with the Red footage. What settings were used to grade it? Are the windows actually blown out, i.e clipped, in the raw file or is the data there? Because if the data is there in the raw exposure, then the way the scene is rendered for value in the example is a grading choice, whether deliberate or not. A reasonable comparison with Arri log-C footage would start with Redcolor2 color space and Redlogfilm gamma as grading points and refining from there. The major difference between Arri and Red 1MX on a truly comparable exposure and grade would be just a little more noise in the deep shadows with the Red, not blown highlights.
Many of these comparisons look to me like the R1 footage was graded with REC709 gamma and color space as a starting point, effectively throwing away 3 stops of available DR in the Raw file.
Click to expand...

It was graded in Lustre natively from the R3d's. The film was 2K DPX and the Alexa was Pro Res LOG. The end result was for 709 (and this is a test for a TV series so that's the final delivery) but there wasn't any more information in those highlights on the r3ds. I didn't expose for the highlights. I exposed to match what I exposed the film at. No grading was done aside from a little balancing (off a correctly exposed grey card.)

Razz16mm said:
Here is a true example of how an R1 Mx can handle extreme dynamic range when skillfully graded.

http://www.zoomfilmtv.com.au/ftp/Dynamic_range/RED_hi_contrast_test2.mov
Click to expand...

That just looks like an overworked HDR *look* grade to me. It doesn't seem very natural. Compare the top of the first lamp post in the proxy to the graded shot. It's darker than all the others in the bg. And the deep BG buildings hit by the sun are still clipped. The text says sun was metered at F32. But which part of the scene was F32 ? The clouds ? The buildings ? There is also no skin tone in the sun (aside from filtered dapple) to compare how this grade works when applied to skin tones. It's a single shot, not a shot that pans and changes, where heavily rotot'd windows and tricks are much harder to apply.

My shot's aren't graded (like your example). They are only processed and balanced at a common exposure. I know I can expose for highlights and trade around for other DR. That's not what I was wanting to look at.

jb
 
Last edited:
John Brawley said:
It was graded in Lustre natively from the R3d's. The film was 2K DPX and the Alexa was Pro Res LOG. The end result was for 709 (and this is a test for a TV series so that's the final delivery) but there wasn't any more information in those highlights on the r3ds. I didn't expose for the highlights. I exposed to match what I exposed the film at. No grading was done aside from a little balancing (off a correctly exposed grey card.)
Click to expand...

It is acknowledged that Alexa shows about one stop more DR than a Red 1 MX, so I am not suggesting that the test was a misrepresentation. But the choice of ISO for reference exposure and gamma space for grading have a major impact on the information that can be extracted from the Raw file. What were those settings for the Red in this test? Comparing Alexa to Red in a video gamma vs a log gamma for instance.
 
John Brawley said:
You're correct. I didn't expose for the highlights. That was my choice to do so, to expose them all at the same level. I could have exposed the RED for the windows and traded some shadows for highlights, but that wouldn't be a fair point of comparison would it.



jb
Click to expand...

I've had this discussion with other people, and here is what I have noticed about a lot of different cameras. If you expose for say the skin of a person in each clip to be the same, the issue is, that location within the curves being used internally in the camera might not be the same. When I was experimenting with the AF100 as compared to the 7D, this was the first time I realized this. Once I started exposing for highlights and shadows and let the middle be where it was in ONLY the situation of comparing two cameras, I was amazed at just how far off they were. On the AF100, I constantly underexpose the skin more than I would have on a 7D and the final result is something where I push up the skintones, and then get a superior image to my 7D.

Now, granted, I personally think that the Alexa did a FANTASTIC job on its own, hands down, and I think that is a valid point to make as well. But, I think you can also say that what ultimately matters is the whole picture. Cost is one issue, second is, you know that the Red isn't going to have the same lattitude, but if you work more into that lattitude, and expose for HL and shadows, do you get a better picture that is much more usable? Or, as you've pointed out, what happens in the shadows? Or, do you go in and try to tweak the settings more? So many variables!!! :)
 
Barry_Green said:
According to the SCCE, it's more like two and a half stops more for the Alexa. Red One MX clocked in at 11.9, Alexa was 14.3.
Click to expand...

Barry, how do you account for the discrepancies between this camera test and other camera tests performed by Red and Red One MX owners who find that the MX sensor had around 13.5 stops of dynamic range? We are having a pretty lengthy discussion about it on RedUser.

If you look closely at the Dynamic Range tests, especially the highlights test, the scenes are inconsistently lit from shot to shot.

Lot of love for Steve and the ASC members who spent a lot of time on this test, but there are some major mistakes in this test, and I am surprised we could find these kinds of mistakes from people with their resumes.
 
You must log in or register to reply here.
Back
Top