C300 vs. Scarlet-X lowlight

[Robert Andriessen]

"...the Silicon Imaging SI-2K camera played a dominant role, but other formats included 35mm film (3-perf and 4-perf), S-VHS, a Canon Elura (mini-DV) camcorder and a Canon EOS 1D Mark IV digital SLR. The Canon 1D was used for some HD video, but also for stills photographed in a burst mode. This allows the camera to shoot a series of full-resolution, sequential stills at a frame rate less than 24fps. These clips were assembled by Union VFX in London as motion media files at a variety of frame rates."

http://digitalfilms.wordpress.com/2011/01/01/127-hours/

unfortunately journalism isn't what it used to be. That's a badly written article by a blogger.

original article link:
http://www.definitionmagazine.com/journal/2010/10/15/127-hours-cinema-tech-in-small-places.html



“The other cameras that we shot with were the Canons. We shot with the 5D MkII as well as my favourite camera the 1D MkIV. Anthony was very frustrated with the video mode out of the Canon cameras. There are limitations, it’s a very compressed H.264 file format, it’s not ideal for the type of shooting he’s used to. We looked at it and tried to figure out ways to break the code, get uncompressed RAW out of the camera and there was no way. What he ended up doing was shooting in still mode at 10 frame bursts and in post he would double it to 20 frames. He definitely bends the rules. The 20 frames is an interesting effect and will look really cool.

“The Canon cameras had their own set of challenges as they were really designed for still photography so we had to come up with some rigs to control the focus. We had questions on how you monitor and how you get the Director a video feed, there were a lot of challenges. We used the 1D MkIVs for a lot of high contrast, over exposed sunlight, snow caps, skin tones type shots, this camera could hold those shots in still mode. In the video mode it wouldn’t get anywhere near.

 

[Tim H.]

what happened to this thread? this has nothing to do with the original subject.


oh and btw I don't think there is a single video shot of 5d in 127 hours, most of that movie is sl2k and i'm guessing some go pros. It's definently not one of Danny's best looking films. Darn, can't believe I took the bait

Already answered by others but I'd just add that IMDB is your friend. http://www.imdb.com/title/tt1542344/technical

I agree that this is off topic for this thread. However, what turned out as a casual comparison ignited into a Red vs Canon war as usual. The bottom line is that Canon makes nice freaking sensors. They just need to get the rest of the pipeline to match it. Now you can admit that and still like the Red cameras right? Or is just one or the other?

 

[Stephen Matthews]

I'm fairly sure that does not specify between stills and video mode and there is a huge difference between the two. there has been stills use in many films before the advent of VDSLR's.

 

[Nick Morrison]

"...the Silicon Imaging SI-2K camera played a dominant role, but other formats included 35mm film (3-perf and 4-perf), S-VHS, a Canon Elura (mini-DV) camcorder and a Canon EOS 1D Mark IV digital SLR. The Canon 1D was used for some HD video, but also for stills photographed in a burst mode. This allows the camera to shoot a series of full-resolution, sequential stills at a frame rate less than 24fps. These clips were assembled by Union VFX in London as motion media files at a variety of frame rates.

The Canon DSLR cameras were used in a couple of ways – in a burst mode to gave us a hyper-real, kinetic look that was used in the flashbacks – as well as for traditional time lapse shots."

http://digitalfilms.wordpress.com/2011/01/01/127-hours/

Thanks Antony for the clarity! The more info the better.

 

[Robert Andriessen]

Already answered by others but I'd just add that IMDB is your friend. http://www.imdb.com/title/tt1542344/technical

I agree that this is off topic for this thread. However, what turned out as a casual comparison ignited into a Red vs Canon war as usual. The bottom line is that Canon makes nice freaking sensors. They just need to get the rest of the pipeline to match it. Now you can admit that and still like the Red cameras right? Or is just one or the other?

I like both camera's, different purposes. Have a canon, waiting for the other to be delivered. I think there's very few camera's left that are actually bad in low light. It's not an issue these days. And the Canon c300 seems to do extremely well in this regard. My personal preference is for raw 4k images though.

 

[Gunleik Groven]

I like both camera's, different purposes. Have a canon, waiting for the other to be delivered. I think there's very few camera's left that are actually bad in low light. It's not an issue these days. And the Canon c300 seems to do extremely well in this regard. My personal preference is for raw 4k images though.

For the most situations I work, highlights controll (NDs) are much more of an issue than lowlight performance...

we had two sets of discolored Schneider ND's which we'll get replaced as soon as Schneider can manage.
But this has been a real problem. Having to shoot with ND 1.8+ outdoors, (which would equal ISO 12,5 if it was apliceable...) is the real challenge most of the time for us. NOT shooting whatever is under the car in an unlit nightscene.

If I had a wish for Epic/R1MX, it would be an easy way to add IR calibrated ND's before the matte.

If the DP wants to shoot 2.8-, which seems to be the preference for many scenes here (often 2- with masters...), image degradation through NDs is much more of an issue than anything else...

 

[Daniel Kunz]

@Gunleik, what are you implying with that? I don't really get what your point is on the subject.

 

[Bob Gundu]

@Gunleik, what are you implying with that? I don't really get what your point is on the subject.

I think his point is that everyone seems so concerned with low light capabilities when in fact shooting in bright daylight can be just as problematic. I personally have been spoiled using video cameras with built in ND's. Getting a decent set of NDs is critical using RED cameras.

 

[Timur Civan]

@Gunleik, what are you implying with that? I don't really get what your point is on the subject.

What hes saying is what i've been fighting with the RED/Epic since last year. @800 ISO out doors in the sun you need a T64.0 to get basic exposure on a grey card. So lets say you want to shoot at T4 or a God forbid a T2.8, you need, ND2.1 thats 7 stops of ND. Thats a lot of glass to be filtering through for what should be the clearest sharpest scene in a film, Day exterior. Now i dont know about you, but a lot of movies i shoot take place outdoors, in the day, or at least a Day interior.

I think this race for lowlight is somewhat foolish. I mean my F3 in sLog, its base ISO is 1600ISO. Its INSANE. At night im shooting with ND in the camera to shoot at a T2.8. I mean, base ambient fill is the biggest enemy for night interior, and EXPECIALLY night exterior.

I'd Kill for the RED to have a grainless, high dynamic Range ISO 50 camera for day EXT work on bigger budget productions. Even as a specialty rental item.

 

[Pascal Payant]

Im preparing a web series and i,m still debating if i go with the canon c300 or scarlet. I have the budget so it's not a problem. it needs to be guerrllia style , run with the camera etc. i'm just wondering if the canon is just a better camera for that then the Scarlet. I love the 4k RAW but i,m wondering if it,s too much problem at the end for that type of work. what do you think guys? they are both great cameras I'm sure of that. I'm shooting out in the woods for a couple weeks so

 

[Antony M]

What hes saying is what i've been fighting with the RED/Epic since last year. @800 ISO out doors in the sun you need a T64.0 to get basic exposure on a grey card. So lets say you want to shoot at T4 or a God forbid a T2.8, you need, ND2.1 thats 7 stops of ND. Thats a lot of glass to be filtering through for what should be the clearest sharpest scene in a film, Day exterior. Now i dont know about you, but a lot of movies i shoot take place outdoors, in the day, or at least a Day interior.

I think this race for lowlight is somewhat foolish. I mean my F3 in sLog, its base ISO is 1600ISO. Its INSANE. At night im shooting with ND in the camera to shoot at a T2.8. I mean, base ambient fill is the biggest enemy for night interior, and EXPECIALLY night exterior.

I'd Kill for the RED to have a grainless, high dynamic Range ISO 50 camera for day EXT work on bigger budget productions. Even as a specialty rental item.

+1

There was talk about Red making a user-switchable ND OPLF.

But then things went quiet.
:-(

 

[Sam Taylor]

+1

There was talk about Red making a user-switchable ND OPLF.

But then things went quiet.
:-(

Yes, I would love to see this! Even in addition having switchable OPLFs for whether you want to optimise your shoot for 5K shooting or for 2K shooting would be useful, in addition to the ND question!

 

[Sebastian Wöber]

So many toes stepped by saying we compared two images. Didn't even use the word "test"...
I'll gladly share our findings, but with some of the comments here I'm losing interest.
The intention was certainly not to measure how much difference lies in the light sensitivity of these cameras, but just to get a rough feeling for it.
The request for perfect conditions and measurements to make a predication is absurd to me. As a cameraman I look at the images, then the exposure values and get a feeling for what the camera can and cannot do for me. On a live set I'm not getting around relying on my experience and gut feeling, not everything can be measured and tested.

 

[Sergio Perez]

Sebastian, I'm really looking forward to your drama narrative post you're going to post next. In this thread, you have to understand that people weren't happy with the conclusions because there were significant differences in the weather condition in both shots. I think that many want to see the Scarlet in a real world, low light night narrative situation. THere are other threads with great examples of low light stills and footage- the Di Caprio cigar, for example, or the "fake" Scarlet one. The recent Scarlet MTV of the Fort Knox band also shows great light gradation and available light handling. That's why that comparative shot just looks not representative of the camera low light performance.

 

[Erik Franzén]

So many toes stepped by saying we compared two images. Didn't even use the word "test"...
I'll gladly share our findings, but with some of the comments here I'm losing interest.
The intention was certainly not to measure how much difference lies in the light sensitivity of these cameras, but just to get a rough feeling for it.
The request for perfect conditions and measurements to make a predication is absurd to me. As a cameraman I look at the images, then the exposure values and get a feeling for what the camera can and cannot do for me. On a live set I'm not getting around relying on my experience and gut feeling, not everything can be measured and tested.

It doesn't say "test" but it does say "comparison" and goes on to state that "And the results are probably as expected, but I didn’t really think there would be THAT much difference."

That to me sounds like a test... I do appreciate that you put full disclaimers on each part of the comparison. But the C300 obviously is put at an advantage when the smoke lights up the scene, so not that surprising that people would spring to Scarlet's defense when it makes the camera look bad in an unfair comparison. I assume you though it was valid within the parameters of what you were testing, and I am sure you found out whatever you wanted to find out.

But as a test, or a fair comparison between to cameras lowlight capabilities, it just doesn't ring "fair" at all -- same conditions, same environment and the same lens; otherwise, it just doesn't tell us anything.

 

[Benoît-Joan Clariana-Roig]

Im preparing a web series and i,m still debating if i go with the canon c300 or scarlet. I have the budget so it's not a problem. it needs to be guerrllia style , run with the camera etc. i'm just wondering if the canon is just a better camera for that then the Scarlet. I love the 4k RAW but i,m wondering if it,s too much problem at the end for that type of work. what do you think guys? they are both great cameras I'm sure of that. I'm shooting out in the woods for a couple weeks so

You can't really do a run'n'gun with Scarlet or a C300. You should try to rent a P2 HD cam with 10 bits 4:2:2. The HPX250 from Panasonic is a really nice cam for web series.

 

[Álex Montoya]

What bothers me is that when I load the Scarlet footage in Redcine it does look VERY grainy even at 800 ISO and I don't understand why since I've seen a lot of MX footage at this ISO and it's consistently clean.

 

[Gunleik Groven]

@Gunleik, what are you implying with that? I don't really get what your point is on the subject.

Like the others said:
Controling high intensity light scenes have been much more of an issue than lowlights.
HDRx helps to some extent. Adding 3 stops in the top is cool, but comes at a price (twice the filesize + extra storage and processing needs)

When what I really want is a sensor wt the same (or higher) DR, but with higher tolerance at the top end of the exposure, so I don't have to carry all those NDs on at any time.

As the shutter isn't the best way to controll exposure for moving images, and the aperture changes the DOF, I am left with ND's as the compositional tool for light controll, and they come with the price of softening and sometimes discoloring the image. I think I'd have accepted a 40 asa cam with the same DR (but with 3 stops more highlight tolerance compared to the zeroed/calibrated 320) for 75% of the shooting situations I am in.

Don't take this as Epic/Scarlet is "bad" or that I cannot controll the highlights, but ND stacks are much more of an issue than lowlight performance in everyday dramatical/feature/fiction-shooting than lowlight performance.

 

[David Battistella]

You can't really do a run'n'gun with Scarlet or a C300. You should try to rent a P2 HD cam with 10 bits 4:2:2. The HPX250 from Panasonic is a really nice cam for web series.

I agree. I would not shoot RAW for these kinds of "dosposable" productions.


David

 

[Hrvoje Simic]

So many toes stepped by saying we compared two images. Didn't even use the word "test"...

Toes you stepped on are irrelevant.
With posts like these you spread false information over the internet.

Your post was suggesting same conditions, resulting in drastically different performance.
What you provided as a basis for the claims you make is inaccurate frame of reference and misleading.

The intention was certainly not to measure how much difference lies in the light sensitivity of these cameras, but just to get a rough feeling for it.

Sebastian....really...

The request for perfect conditions and measurements to make a predication is absurd to me.

Request is pretty common in a world where people compare any type of equipment.
"Perfect conditions" for a valid "comparison" are not really that hard to achieve:

Step 1: Point both cameras at the same scene with same light values.
Step 2: Provide detailed info on equipment and settings used.
Step 3: Allow access to raw material from used cameras.
Step 4: Make sure your website is not focusing on promoting DSLR gear made by the manufacturer whose device you claim to be superior.

As a cameraman I look at the images, then the exposure values and get a feeling for what the camera can and cannot do for me.

And as a cameraman one should be aware of the light levels difference between city lights bounce from the cloudy/foggy sky + smoke...and a night with a clear sky.

Especially if posting low-light performance claims and comparisons on public websites.

On a live set I'm not getting around relying on my experience and gut feeling, not everything can be measured and tested.

This has nothing to do with you posting two differently lit scenes shot with different cameras and starting a noise thread on a cine-gear public forum.