June 6, 2011

A Refreshingly Practical Shootout: Six Digital/Film Motion Picture Cameras Compared in Available Light

Just because the latest single-sensor digital camcorders allow you to shoot with available light does not mean you should shoot with available light. However, the importance of being able to should not be overstated. To that end, John Brawley put together a great practical test utilizing six different cameras in the same setting, in order to evaluate which format would work best for a forthcoming feature to be shot mostly on Parisian streets at night. The cameras/formats were Aaton 35mm, Aaton Super 16mm, ARRI ALEXA, RED ONE MX, Canon 1D Mark IV DSLR, and the Sony F3. Here's the video of the test, which is refreshingly devoid of test charts and instead focuses on the devices as storytelling instruments:

For more details on what you're seeing, here's John on the test shoot:

We made a definite choice to NOT edit them exactly the same way, but to favour each edit to be the best it could be with the available material. This extended to the grade. Instead of trying to match everything to one camera, we simply tried to make each camera look the best it could... Now, as an extra challenge, I’ve only identified the cameras by a letter code. So you’ll watch all 6 sequences and then your challenge is to try to pick which camera is which. After the 6 sequences, I’ve put selected shots from each camera back to back so you can more readily make a direct comparisons. I’ve also then done a 200% blow up on a selected shot, and the theory with this was to try to show how the resolution (or lack of resolution) stands up in a cinema environment.

More details on John's blog below, where he makes the all-important distinction of this test being for emotional as opposed to scientific reasons. To me, this the best test I've seen in recent memory thanks to its focus on storytelling -- it's about resonance, not resolution. Which camera's images did you respond to most?

Link: John Brawley - Available light tests are finally online

[via ProVideo Coalition]

Your Comment

12 Comments

the RED and the ALEXA look great, worlds better than anything else (I must be getting used to the "digital look" but I actually liked them better than 35mm film)

BUT: it has to be wrong: look at the "selected shot comparisons": are you telling me that the F3 creates those huge aliasing artifacts in the bridge, whereas the 1D is nearly clean? no way!!

the aliasing is a dead giveaway, but I also think I see issues with skin tones and highlight detail in that "supposedly F3" footage

they answered to my comment on vimeo, saying it's not flipped, but still, I refuse to believe it: unless that F3 had no anti-aliasing filter on its sensor at all, the footage is flipped, there's no other explanation

June 6, 2011

0
Reply

I would tend to agree with what you're saying based on looking at it, but the 1D footage clearly has more rolling shutter and is much, much softer on the 200% blowup.

So I don't believe they've mixed anything up, I really have to believe that the aliasing on the F3 has to do with the pre-production model they had, even though they said it may not be fixed in the production model.

They also missed focus on the F3 footage near the bridge - that could be why we don't see any aliasing on the 1D, they aren't focused on the bridge in the 1D footage.

June 6, 2011

0
Reply
Joe

well, if it's not flipped, then I have to disregard the F3 results: from other tests I've seen, I just can't believe it has so much aliasing, and so any further problems I find with it may also have been corrected in the final version

taking that out, my conclusions will be:
* the RED and the Alexa look absolutely amazing
* I don't like film grain anymore; maybe it's just me
* F3 and 1D have much more codec artifacts than I would like, deliver uglier skin tones, and can't cope as well with highlights

June 7, 2011

0
Reply

Yeah, I don't think the F3 fares as well in this test as it could -- the highlights are blown, it doesn't look like it was using a cinegamma, and it was a preproduction unit. Not to mention the lack of S-LOG availability at the time of the test.

You know you've been watching too many camera tests when you can tell without the key exactly which cameras are which...

June 6, 2011

0
Reply
avatar
Ryan Koo
Founder
Writer/Director

The F3 was indeed shot with cine gamma. S log still isn't available.

June 6, 2011

-1
Reply

"Practical" shootout, very impractical presentation... i'd rather not have to sit through the entire 18 minutes in order to compare them, especially when they don't even tell you which camera is which.

June 9, 2011

0
Reply
Phil

Yes, they do tell you right in the video -- and if you're serious about using any of these cameras, 18 minutes is nothing in the scheme of things.

June 10, 2011

-1
Reply
avatar
Ryan Koo
Founder
Writer/Director

ALEXA look amazing, then the red and then 35. this is one of the best tests i've seen

June 10, 2011

-1
Reply
patriot

Great test for the most part ! For my taste, the best over all ? Come on, it's the Aaron 35mm, hands down. But I must say, the Sony F3 has the most realistic daytime look (park scene.) Alexa would be the best overall digital, next to the Red. But if you look closely the 16mm was the only one the held the detail and color in the small sign that was panned down on before he starts running in the wide shot ! And quite frankly the Canon D1 - sucked, but if that's all you can afford . . . it will do in a pinch. Remember these are just tools the real test is in the person behind the camera !! :-)

June 12, 2011

0
Reply
A Madsen

............. also 35mm (in any situation) still gives the feeling you are engaged in the story, it has texture, unlike digital where it leaves one cold and unattached. Digital makes everyone look the same and shows you a perfection that is untouchable, unattainable by the viewer. I'll always prefer that beautiful element called grain - it's there for a reason.

June 12, 2011

0
Reply
A Madsen

great test, I knew both 35 and 16mm off the bat, guessed A was Arri but guessed the DSLR was the Sony, due to the bad aliasing.

I am very surprised at the Sony's aliasing and lack of proper skin tone reproduction.

I also believe that using a 5dmkII with technicolor's new profile would of resulted in much better DR and less noise. The 1dmkIV does not hold up to a 5dmkII for color and skin tone reproduction.

I cannot believe what people are saying on the Vimeo page regarding the Red as a true winner. Even the Alexia was more natural and less 'harsh and plastic" than the Red.

I believe people need to watch it again and look at what really matters: forget noise and grain. Look and skin tone, color and dynamic range. The film blew all the camera's out of the water by showing clouds and beautiful reflections in the water, the skin tone was so lively and orange it looked the most beautiful.

35 mm is the clear winner hands down. Let's see, would I rather have bluish grey faces with light spots and partial blown out flesh with completely blown out skies or bright, lively skin and detail in the sky?

I'll take that over noise and tact sharpness ANY DAY!

I am curious to see how the 5dmkII with technicolor would of held the sky detail.

1d mkiv was better than the Sony but had the most codec compression obviously. I'll take 16 mm grain over blocky codec compression, thank you very much.

June 13, 2011

1
Reply

This was a very good test!! Bravo!! I rated the camera's in 3 categories as I watched...yes only 3 because I did not want to be so distracted. 1-Color 2-Low Light 3- Rolling Shutter+ (the + being: Bokeh, Sharpness)
As I watched the first video I thought to myself..."Amazing"...ok be fair...give it what you think!"
I scored CamA a perfect 30. CamB 16. CamC 25. CamD 21. CamE 19.5. CamF 17.
Now I am not at this level of affordability when it comes to Camera's. We shoot on 3 JVC GY-HD110's which we love, outside of still having to deal with tapes. But in no way does any of the other Camera's perform (at least in this test) from top to bottom as well as the Arri. Look at the "Sparkler" segment, and the beautiful Bokeh behind the girl as she waits for her man. And the way the camera handles going from one scene of bright to dark full of colors...the Camera never seems to be fooled and simply captures what is in front of it.
The bottom line...I wish I had one!
Again great test!!!

July 19, 2012

1
Reply
Craig Shamwell