December 19, 2011

How Do the Canon C300 and Sony F3 Compare in Low Light?

Cameras have been getting more and more sensitive, so it follows that low light tests are a nice showcase of the latest in camera technology. Here, Sebastian Wiegärtner pits the Canon C300 against the Sony F3 (with S-Log) and Canon 7D:

Thoughts? I think one of the reasons people were initially disappointed with the C300's announcement wasn't just the RED announcement that followed -- it was the fact that the Sony F3 has S-log (for a price) and the C300 doesn't have a similar upgrade option. Will a C-log picture profile be enough? Well, yes. It's a hell of a camera. Without handling the C300, I still suspect I would prefer the F3 with S-log, but... it's not like anyone's going to say no to either of these cameras if they're available for a project. They're exactly the kind of post-HDSLR camera we were hoping for from Sony and Canon -- though they might not be at the FS100/AF100 price point people were hoping for.

As you read this I'm actually out with Raafi Rivero and Timur Civan shooting a test short on my SCARLET-X, and in the weeks to come I might conduct some with other cameras as well (F3, ALEXA, C300 if I can get my hands on one). Regardless... all of these cameras are appropriate for shooting low budget features, and I would expect a high percentage of festival submissions to be shot on the C300/F3/SCARLET over the next few years. Don't get me wrong -- DSLRs are great -- but XLR audio inputs, timecode, and better monitoring options (to name a few factors) all make a huge difference on set.

Link: CANON C300 C-LOG AND SONY F3 S-LOG COMPARISON - Sebastian Wiegärtner

Your Comment

41 Comments

Thanx Koo for posting this video. Thanx also to Sebastian for the time he took to do the test. I think the F3 and C300 are both good cameras. I particularly like the looks of the F3 s-Log for the possibility It will give in post Prod.

December 19, 2011

0
Reply
Panasonic AF300...

"Don’t get me wrong — DSLRs are great — but XLR audio inputs, timecode, and better monitoring options (to name a few factors) all make a huge difference on set."

You know, these things are nice, but I think I'm against the 'have to have' nature that usually follows this comment (you stop short at "huge difference"). I'll take a DSLR, great prime or two, zoom h4n, mid-grade audio gear, and rent some lights, over the price of what it takes to buy either of those cams. All without xlr (use a mini adapter and a rode mic for your cam audio, clap your sync to the zoom off-cam), timecode (I edit just fine without it), and better monitoring (though I can go hdmi-out just fine too).

December 19, 2011

0
Reply
Rev. Benjamin

There is little I hate more than being tethered to audio on set. For purely cinema and hobby shooting, I actually prefer no audio. Audio and time code come more handy in doc, eng, live events, etc.

December 19, 2011

0
Reply
James0b57

not everybody needs professional gear, to each his own

December 19, 2011

1
Reply
Ryan

I'd quip Ryan (one replying on my post, not Koo), that a dslr is every bit as professional as one of these cameras, save xlrs (no big deal) and timecode (granted, though again I don't necessarily work with it). You can get a similar or better image, just happens to cost thousands less.

You know, I think there is one edge these cams have that dslrs will not have I suppose - the appearance of being a bigger professional camera to the untrained eye (eg, a client). Show up on set with a c300 vs a gh2 and ask the client which one they're more comfortable working with, I would wager they'd say the c300 bc it looks bigger, period.

December 20, 2011

-1
Reply
Rev. Benjamin

I can't really respond to that, so much misinformations that I'm sure if I refute with facts, they will not be received well

December 21, 2011

0
Reply
Ryan

It was painfully obvious that the two "pro" cameras have considerably better image quality than the 7D. Why not just rent the better cameras, when you need them, and keep the 7D for your personal, need to shoot right now moments?

December 21, 2011

1
Reply
Jonathan

Now see there's a quandary for you, Ryan McSnarky, because your attitude isn't being received well either.

GH2 shoots 200mbs 1080p24, unbelievable ISO range. Period. You can use whatever fantastical lens your heart desires. That's not possible for these other cameras. I'm saying let audio guys take care of the on-set audio, with their fancy mixers, and just get something you can sync to. And in the process - save yourself thousands. Or - pad b&h's pockets a little more, whatever.

Also, "image quality" is in the eye of the beholder and isn't measurable.

December 21, 2011

0
Reply
Rev. Benjamin

And here's a fun source for you over with our pal Philip Bloom. Can YOU guess which cam is which? Hint: the yucky looking one (pre CC but still) costs much, much more.

http://philipbloom.net/2011/12/15/christmas-shootout/

December 21, 2011

2
Reply
Rev. Benjamin

Which one is the yucky one?

December 21, 2011

0
Reply
James0b57

The GH2 has been blowing me away recently.

December 21, 2011

0
Reply
James0b57

Do we keep forgetting the RED Scarlet does not have XLR inputs?

December 20, 2011

-1
Reply

The Scarlet has balanced audio via two 3.5mm jacks...when converted to XLR, it should provide the same audio (including phantom power) as other video cameras...though I'm not sure it's activated in the firmware yet. This was designed to convert the 3.5mm jacks on the Epic and Scarlet to XLR: http://woodencamera.com/a-box-epic-scarlet

December 20, 2011

0
Reply
Gabe

Phantom power is not active in firmware yet, but yes the 1/8" inputs on the SCARLET/EPIC are balanced. Which is a blessing and a curse -- your standard 1/8" plugs will NOT work. More soon...

December 21, 2011

-1
Reply
avatar
Ryan Koo
Founder
Writer/Director

Hi Gabe,

Audio is indeed enabled on the EPIC since quite some time now. When was the last time you used one?

The Minijack audio inputs of the EPIC/Scarlet can not and will never be able to provide Phantom power, unless you use an accessory that converts them to XLR and provides 48V phantom.
That Wooden camera accessory is good and well-made but does not provide Phantom power. There are some XLR/DSLR adapters from Beachtek that do provide Phantom power.

December 22, 2011

0
Reply
William

Oops, my bad: Phantom power should indeed be enabled in futur builds.

December 22, 2011

0
Reply
William

The F3 looked awesome in this test. A clear winner IMO. The C300 looked like a slightly clearn 7D to me. The F3 was like...damn

December 19, 2011

0
Reply
Nagato

I have to say I'm surprised with how substantially similar all three cameras seem, at least with "normal" ISO. It's hard to believe that the only light on her face is an iPhone screen.

Obviously the 7D is worse than the other two, but aside from the nose at high ISO and some of the weird artifacts (like stripes), it's quite reasonable. It really looks like you could inter-cut the 7D footage with footage from either of the other two without it being distracting. There are so many of these tests where the 7D comes out looking devastatingly awful so I'm pleasantly surprised to see it perform reasonably well.

December 19, 2011

-1
Reply
cows

I've been trying to get this question answered a few times. I've gotten some answers (form factor) but nothing compelling. Why would someone buy a c300 over an f3? Thanks

December 19, 2011

0
Reply
Ryan

Jeez crucify the guy just because he wants to shoot one step above a DSLR. If DSLRs were so damn good why doesn't Hollywood just drop Panavision 35mm, RED and Alexa and just shoot on DSLR's. I think DSLR's are great but like always fanboys are not. Most of the features I have seen shot on DSLR's look like crap when put up against a Hollywood film. Like Crazy, I Melt With You, and Tiny Furniture were all shot on DSLRs and all cannot not even compare to 35mm film or some of the newer high end digital cameras . Do you really think Apocalypse Now or Days Of Heaven could be shot on a DSLR? If you're just going to limit yourself to one camera for the rest of your life, you're probably not going to shoot on anything else.

December 19, 2011

1
Reply
Rob

The mistake people make is thinking that it's just the camera that accounts for the look of a Hollywood blockbuster. It's also the huge budgets and crew dedicated to lighting, sets/props, and rigs.

That said, you don't need all of that to shoot a good movie. They have bigger budgets to work with, and have a need (back then as now) to justify people paying for the price of a movie ticket versus staying home.

December 19, 2011

0
Reply
moebius22

Just have a look at Shae Hurlbut work with DSLR´s and them tell me it´s not possible to shoot a Hollywood blockbuster with it...

December 27, 2011

-1
Reply
Augusto Alves d...

Did anyone catch Sebastian's comment in Vimeo. "Please, don't judge the picture quality, it's absolutely compressed (H264). C300 and Sony F3 has a nice grain in the blacks."

:)

December 19, 2011

0
Reply

good, I was actually going to say something about how crappy the signal looked in the shadows...

December 19, 2011

0
Reply
sean

Looking forward to details on Scarlet shoot. There's very little test footage online from early users...

December 20, 2011

1
Reply
Jonathan

Seems like ever since Koo ordered a RED each post is becoming more and more bias, and he continues to downplay DSLR's/users when thats what brought most of us to this site. Please keep that in mind.

And lets face it, the best film in the world can be shot on the C300 but Koo would never see it because we're too stuck on comparing it to the Scarlet.

December 20, 2011

-1
Reply
*sigh*

* relax and be patient, there's no conspiracy
* to each his own
* right tool for the job

December 20, 2011

0
Reply

there is no new about dslr, and the small insignificant stuff like discount pricing on dslr, koo has an article on it. what are you talking about? Plus every article on nofilmschool is for dslr shooters, but perhaps just the ones who are trying to do bigger and better things.

December 20, 2011

-2
Reply
Ryan

Ahh, so someone else too noticed that ; )

December 20, 2011

0
Reply
Bob

Seems to me like someone is trying to justify a pricey purchase with every consecutive post. Should have rented an Epic and Scarlet.

December 20, 2011

1
Reply
moebius22

- Where is the big DSLR news that I've missed?
- I've posted as much about the C300 as I have about the SCARLET.
- I still have my 5D.

Saying something makes a "huge difference on set" is not turning my back on DSLR shooters. A proper 1080p monitor output makes a huge difference versus a scaled 480p output. I've been covering the GH2 quite a bit lately because I love the idea of an inexpensive camera that can be hacked (like, oh, I don't know, the 5D with Magic Lantern years ago). If there was a brand new DSLR that was making headlines you'd see more coverage about that, but at this particular time there's not. As Samuel said... relax, there's no conspiracy.

December 20, 2011

0
Reply
avatar
Ryan Koo
Founder
Writer/Director

I love how this blog has grown big enough that conspiracy theories are starting to develop... : D. Yeah you have done a great job covering gh2 lately btw, thank you.

December 20, 2011

0
Reply
Rev. Benjamin

I'm just saying man, that you sound like you are trying to convince yourself you made the right purchase. Maybe I'm wrong, but that's how your C300 posts are starting to sound.

Thanks for that free book link.

December 20, 2011

0
Reply
moebius22

Dude, really? I don't want to come across combative at all - believe me, I'm just curious - but what about the C300 posts has seemed biased? This one, for instance, check out the quotes: "It’s a hell of a camera", "no-one's going to say no to either of these cameras if they're available for a project", "they're exactly the kind of post-DSLR we were hoping for". Pretty positive. I have to say, I'd spring for the C300 if I had a spare $20k lying around, largely from the info made available through this site. Aside from being mindful of the C300's admitted planned obsolescence - and that's a pretty big deal when you're in the market to buy - I'd say Koo's been very positive about the cam, and certainly fair.

But I'm a NFS fanboy, so take it with a grain of salt. :)

December 21, 2011

-1
Reply

Thanks Lachian.

Moebius, I'm not trying to convince myself of anything... We actually just shot a micro-short on my SCARLET the other day and it looks great (coming soon). I think the RED is designed for a very specific set of needs and those just happen to be my needs. The C300 is a different camera with a lot of positives, so when I praise it I'm really not worrying about justifying my purchase.

December 21, 2011

0
Reply
avatar
Ryan Koo
Founder
Writer/Director

Hi,

"the Sony F3 has S-log (for a price) and the C300 doesn’t have a similar upgrade option. Will a C-log picture profile be enough?"

The C-Log is not a normal "picture profile" but a real log curve similar to the S-Log. There is even a "cinema" mode when you use the camera that locks all the settings but the more basic so you are sure to shoot in the good color/gamma setting that will get the more info out of the camera without messing things up (more like a film camera). The difference is that it is included in the C300 while you have to pay extra to get it from Sony.

December 21, 2011

1
Reply
William

My thanks to Sebastian - this was a very interesting and valuable test that I've been wanting to see. I've been a very happy user of the Sony F3 with and without S-log. I've also been an admirer of the Canon C300 - especially after seeing all of the Canon sponsored short films projected on a large screen by a Christie 2K projector. I was very impressed with all of the films - especially the green screen footage. The C300 may only be 8 bits color, but the results that I saw projected were quite impressive. Even light blonde hair keyed very cleanly. As for this test, it's obvious that both cameras perform extremely well in low light. I noticed on Sebastian's blog that many people thought that the Sony F3 was more light sensitive because it produced brighter images than the Canon C300 at similar db levels. I would like to point out something for viewers to consider when comparing. The Sony S-log raises the effective ISO of the F3 from a native 800 ISO to 1600 ISO (1 stop). So, if you want to make an apples to apples comparison between the Canon and the Sony, one should look at the C300 at double the db level of the Sony. So, for instance, the C300 at + 6db is ISO 1600 because it is a native ISO 800 camera even with C-log. This should compare to the Sony F3 with S-log (ISO 1600) at 0db. When you compare like this, the brightness differences are very close between cameras. However, there is a different look overall. I believe that the Canon C300 begins to outshine the Sony F3 when the ISO gets above 3200. The noise in the Canon C300 looks more like film grain which is more pleasing than the blocky noise exhibited by the Sony F3. But this is only apparent at very high ISO. IMO, for a 1080P deliverable film, one can't go wrong with either camera.

December 22, 2011

0
Reply

Randolph,

"The Sony S-log raises the effective ISO of the F3 from a native 800 ISO to 1600 ISO (1 stop)."
This in incorrect, the F3's ISO in Video gamma is 400, and it is 800 in S-Log mode.

The F-3's (and any camera's) "native" ISO depends on the gamma curve used, (gamma curve being the equivalent of the characteristic curve of film stock). It is a rating, which means that the people who built the camera believe that the camera will give the best results at that ISO. In 8-bit compressed Rec-709 style recording, you do not have much latitude so you will always rate the camera at the ISO provided by the manufacturer, because the amplification of the signal occurs in-camera and is hard to change in post. But in S-Log, you decide what ISO you use IN POST. Yes, just like in raw. This is for this reason that Sony introduced the S-Log EI mode, which are in fact a set of LUT's that match different ISO ratings. All this means that in fact, you are free to rate the camera at the ISO you want, the camera is still recording the same "simil-raw" image.
This is also true with film stock is not rated in ISO but in EI (Exposure Index), which you can push/pull in post.

Now about dBs, 0db is the nominal level of amplification of the signal coming from the sensor according to the engineers. That's all what it means. When comparing two cameras, one should start comparing them at 0dB to have an idea. Then one should compare using the ISO! (both the F3 and the C300 can display sensitivity in ISO or dB, which makes things easy)
Comparing the dBs like in this video doesn't mean much. Then of course they compare the images that have a similar luminosity but that's not precise.
Also the video does not explan how they graded the S-Log footage.

December 22, 2011

1
Reply
William

William, I agree that a camera's native ISO depends on the gamma curve used. I will also concede that perhaps 400 ISO may technically provide the cleanest signal for the Sony F3. However, I think that you will find that most DP's are rating the Sony F3 at 800 ISO in the Rec 709 gamma. At 400 ISO, there is just so little highlight dynamic range. Although still not equally balanced between shadow and highlight detail, 800 ISO is much more workable because the camera is so noise free. I based my own ISO evaluation by placing the 18% gray on a chip chart at 50 IRE on a waveform monitor and measuring the corresponding stop and ASA rating with both an incident and a spot meter. Abel Cinetech came up with the same conclusion using even more accurate measuring tools: http://blog.abelcine.com/2011/03/11/abelcine-tests-the-sony-f3-iso-rating/. If you look on Sony's website there is a report on the F3 written by Jon Fauer: http://pro.sony.com/bbsccms/assets/files/mkt/cinema/articles/36FDT-SonyF.... On page 8, he quotes Sam Nicholson, ASC that the F3 had an EI of 800 and that Sony conservatively estimates the S-log at an EI of 1600. There are many user forums which report the same numbers. Even Sony's field reps are stating that the F3 is an 800 ISO camera. Yes it does depend on the gamma curve used. The Cine Gamma 1 curve is much more effective at holding more highlight detail than Rec 709 and the middle gray will fall much lower as a result. For purposes of comparison, I would rate that gamma setting at 400 ISO. But these numbers are all somewhat subjective because it all depends on how one wants to expose an image and what part of the dynamic range scale one wants to emphasize or protect. The same was true of film stock EI ratings. Different DP's used different EI's for the same film stock depending on their exposure and printing preferences.

December 23, 2011

0
Reply

Wow! I learn so much from you guys - nothing beats experience. Thanks.

In choosing between these cameras, I guess the bottom line is to pick a camera that delivers the beauty of the color, the details delivered in the highlights and shadows, least amount of noise for night shots (do a lot of that), the clean crystal clear image shown on the theater screen after post, color corrections, SFX etc. the burn to blu-ray, the copies, and which one will leave the audience with a sense of wonder and awe. I know it is no 4K unit, but then again, neither was the $200k Sony F35 either and the DP were drooling over that unit. And what about upgrades, firmware releases, third party software guys like magic lantern. Yes, the C300 is sure small, like about the size of a 5D, where the Sony is the size of Mack Truck. But who cares, the final result is what the audience sees and how easy is it to get those great moments. As it sits now, with Sony's 3 lenses and S-log, that is a hellova deal, seeing that Canon wants $6,200 a pop (ouch).

I heard somewhere that these Sony's lenses were made by Cooke - nahhh - can't see that. Another question, are the 3 lenses of 35, 50 and 85, a 35mm equivalent? Will they work on a 5D?

Tough calls, eh? But sure is fun watching it all unfold.

December 29, 2011

0
Reply

Check out this ultimate low light field test with the Canon C300 using an iPad for key lighting and an iPhone for fill lighting.

http://eduardoangel.com/2012/01/28/lighting-with-an-ipad/

February 4, 2012

0
Reply
jason krasnov