Part 2 of the fascinating documentary of the SCCE is now up. It’s a compelling watch. Way more entertaining than the SCCE itself and actually really educational, packed with a lot of technical info as to why and how these cameras do the things they do.
Shootouts are funny old things. I think they should be viewed a bit like Top Gear. When you’re racing different cars against each other, of course someone will always say “why didn’t you use that car?” or “why didn’t you use that turbo charged upgrade for that car?” Neither of which were available of the time of the test. It’s a bit of no win situation because there will always be something newer out and someone will always complain about something.
First off…how valuable is something like this? Honestly? It’s of more value to some than others. For the vast majority, it’s rather like camera porn (nudge, nudge, wink, wink!) as most of the cameras we are watching in this documentary are just fantasy anyway and something we just aspire to!! Actually it can be kind of depressing when you see your little baby beaten the crap out of by a camera that costs 70 times as much! What do you expect? Of course there are always surprises. The performance of the F3 in the SCCE blew me away. It did so well considering it’s price, now of course with the S-LOG for the F3, which was not out at the time of shooting of the SCCE, it would have been mind blowing, but we can’t dwell on that too much because that sort of thing will always happen. Much like it will be fascinating to see how the FS100 would perform here. That’s why making a shootout sucks. Especially something as massive as this. That’s why I will just still stick to doing my little unscientific ones!
As I am sure many of you who read my blog will know I am not the most technical person in the world. I learnt so much by going freelance 4 and a half years ago. 17 years of being a staff cameraman left me very uneducated. Becoming freelance forced me to learn much more about how cameras work, and writing this blog even more so. I still have a huge amount to learn and just watching this documentary I still realise I have much to learn…the documentary is very educational in that respect and actually very entertaining. It flew by. Not sure my mum would sit through it though!
What stood out for you? For me, the Alexa looked incredible, especially in the shadow detail. Actually I thought the AF100 looked damn good too. I know this camera has taken a bit of battering recently but I still think that, handled correctly, it’s capable of some cracking images. The F3 looked good. Film sucked in the low light test, but then that is not where film shines. It was in the highlights, shown in part 1, where film showed what it can do. Also as Nino pointed out in the doc, it had zero noise reduction done in the transfer, which it would have had if used on a job. DSLR wise, the D7000 actually looked pretty nice in places. WISH WISH WISH the GH2 was in here. Such a shame it was not included. Really love to have seen how it would have stood up to these tests, as it’s a powerful little beast!
Charts? Well. As expected. No surprises there.
Colour rendition? Film and the F35 maybe…but it’s SO subjective.
Dress detail. Again not that surprising that the DSLRs failed miserably here. All that line skipping, or polka dot skipping! GH2 would have shone here I think! The D7000 is doing weird things there….The Alexa, F3, AF, F35, Red all looked good. In the screen grab below, that noise in the film gets in the way.
Honestly I am just overly picking through things here, and that’s not healthy. I am doing something I am totally against! Pixel peeping! Over analysing shots when that is not what is important!!
The thing is, a shootout like this is a double edged sword. Not the documentary – the SCCE itself. The documentary made by Steve and the Zacuto team is terrific, brilliantly made. Taking the painfully dry SCCE edit and making something very compelling from it. It’s fantastic to see all these cameras compared and it can be very useful in helping you make a purchase for a camera, or in deciding which camera to use for a project…BUT, It can make you think your camera is inferior and not up to the task of making a film. Nonsense. Every single one of the cameras in here can be used to make a film. It all depends on what you can get your hands on. Just because your 5D looks shitty on a chart here (and my god I hate charts as you may already know) doesn’t mean you cannot make your footage look good. There is a quote by someone in the documentary saying the DSLR stuff just looked bad on the big screen. That’s because it is being shown up against way superior cameras. Of course it is going to look bad! It’s not fair! Take it on it’s own. Show the footage without the others, and it will impress. Stick it with everything else, and then it falls apart. That’s the problem with something like this! Oh well!
Still. A great watch. A great production and part 3 in much anticipated.
Special mention must go to the awesome Karen Abad for the amazing edit, Daniel “Skubes” Skubal and Chris Voelz for the graphics and Scott Lynch for the pulling it all together. Immense job by all three…Anyway, those are my rambling thoughts…please share your thoughts below!
LOW LIGHT FRAME GRABS
SHARPNESS TEST
OECF
STILL LIFE BLOW UP
DRESS BLOW UP
WRINGER
54 comments
Thanks to Chris Voelz who did the graphics and a lot of the animations for the tests.
A great thanks to all the team that produced
this wonderful documentary for all of us…
🙂
As with part one I’m a little confused at the 7D apparently outperforming the 5d in the non-chart subjective tests.
While the DSLRs all done particularly badly on the still life shot of the piles of heroin on the plate ~25mins, again surprisingly the 7D seemed to outperform the 1D there despite its smaller sensor sample area.
Great review Phil. I couldn’t agree more. You and I both saw how great your SkyWalker piece looked at the Stag Theater at SkyWalker projected 60 feet shot with your 5D, compared to it’s version transferred to film. Night and day we liked the 5D better, we were astounded and annoyed by the grain.
As Phil says, you are not shooting charts in your film. The key is to use the right camera for the right job. If you have no budget and you own a camera then that’s the right camera for that job. If you have budget, then pick the right camera for the right job. An Alexa or Red is not the right camera for a climb up Everest. The 5D might be a bit easier and allow you to get shots in cracks and corners that you couldn’t do with the Alexa giving you a better film (maybe not better looking) but who cares. Better is Better, it’s all about the story and not the picture quality. This shootout was shot with all different kinds of cameras and nobody seems to care. It’s a documentary and it’s all about message, story and getting the moments.
Personally, I say the same thing on every shootout. My DP Jens (who has been my DP for 25 years) could make anyone of these cameras look great because he would know their weaknesses and strengths and light to compensate for it’s weaknesses and then use its strengths to his advantage.
Remember we stressed these cameras in this shootout. Even Film, RED & Alexa were not looking great in many of the tests in this shootout because we were seeing how they could do in scenes that were 5 and 6 stops over and under exposed. As a DP, don’t over or under expose your scenes that much if you can all help it. Knock those windows down with ND gel which is incredibly cheap if you don’t want to have them blown out.
Keep in my this is just for entertainment and education and take it with a grain of salt. Thanks Phil for taking the time to review this doc.
Steve
Why didn’t the GH2 make it into the mix?
it was decided by the SCCE team that only professional line cameras qualified and the GH2 is a consumer camera.
Then again the GH2 is the most professional SLR regarding Video. In a video test they placed build quality and stills quality above the video specs, go figure. And how is a D7000 considered a Pro tool same as the 7D? Catalogs place them in Enthusiast brackets.
Facts:
GH2 Out resolves Canon and Nikon SLRs by using proper down-sampling.
GH2 has better moire control in charts than even the AF100/101.
GH1 and GH2 have been used in a huge number of professional Productions and its footage shown on HD TV channels. Virtually any lens can be adapted.
Other Thoughts:
Now with the hack and the incredible quality of 1080p 24 at 60+Mbits it would be a total deliberate oversight if they didn’t include it in the tests for Shootout #3 if there is one.
Some consider having the AF 100/101 as a close to GH2 comparison but several owners of both have expressed that the GH2 image is even better with higher dynamic range.
HI james,
I have both AF and GH2 and don’t see the higher dynamic range. I do see a much easier to control image in the AF.
Anyway regarding shootout 3. A hacked camera would NEVER be in it no matter how good it made it. You know it would be a bad idea! You start a precedent. The cameras have to be as the manufacturers sell them otherwise what is stop RED sending a technician to make a super EPIC just for the shootout?
P
Fair points and that’s why we have good knowledgeable people like you that can do unscientific tests like you did with the F3 AF100 and FS100. I found your test a better guide on which camera to buy than the Zacuto tests, a great review.
And maybe show what a hacked GH2 can do in the future.
During film history many camera operators and DPs hacked cameras to support different sized film stocks and lenses to deliver even more that was possible at the time by pure manufacturing. Some like Kubrick custom ordered extremely fast aperture lenses to give his creative vision life.
But I agree completely that a modified camera in such a test would introduce many opportunities to game the system like an EPIC that for small periods of time like that of test scenes, can deliver insane DR and Resolution but not in a production environment. Not that it would need to be overly hacked to win the next test without a sweat as it is!
I am actually planning my next unscientific shootout which will have GH2, 5D, T3i, D7000, AF100, F3, FS100 and RED EPIC in it. I will shoot a series of shots to create a whole piece. What I will not do is tell people what is what and I will see if people can work out which camera is which. I may make a competition out of it!
Phil,
will you start Camera Rentals in UK?
🙂
p.s. There are lots of people buying gold these days…you on the other hand must own at least 5-10 cameras.
Just tested FS100, i really do not like it.Selling MTF to Nikon adaptor if anybody is interested. Well, the best video camera i saw at SMPTE2011 in Sydney was Panasonic 3100!
regards
jiri
what did you not like?
I did not like pictures out of it.Canons have more “organic” look.
I would prefer FS100 to have the same design as af100.VF on fs100 is unstable,tiny buttons,no NDs….and those horrible SONY piggy skintones.
Camera will be “slow” to operate.
I would prefer these days all camera manufacturers designing Large Sensor cameras to have decent 2-3″ EVF.Why so many cables(HDMI) for external recorders and EVF???
Saw EPIC “sitting” next to F3………F3 was almost 1 metre long with all the necessary options, EPIC – tiny,little,nice EPIC on the other hand was just perfect.
I have no idea why people are raving about F3-it is intentionally crippled camera from start-i am sure new F4 will address those shortcomings….
I am hoping this year Canon will release new 5d.
Picture out of what Camera? Organic is bad?
People are raving about the F3 because of it’s price point and the type of image you get out of the camera. I am definitely looking at getting one along with some cp2’s!
Or a fund raiser. Great idea, like that Navy short with the 5D and F35, no one could really tell all the shots apart.
Organic is always good:-)
Canons have nice “organic” look.
“People are raving about the F3 because of it’s price point and the type of image you get out of the camera.” OK
Let me try to explain:
Cost of F3 is about $14,000.
Cost of VF is about $1,000.
S-Log about $3,000
External 4:4:4 recorder is $7-30,000.
Unless You make those spendings – camera is nothing special.
It is much better to invest in EPIC-x or even RED 1.
Once You get all those options, make sure You buy Sachtler 25. Cheers
Regarding CPs..nice lenses but too slow at Wide End.
You will be better off with Duclos 11-16 and RED PRO 17-50mm
Whoaa Nelly! Now if you look at my recent blog post about s-log you will see the camera is exceptional without the 444 recorder. Actually without the s-log saying it’s nothing special is way out…of course personal opinion and all that. I have SLog. Have no plans to get a 444 recorder. I get the gamma curve, the increased sensitivity and still record 422 on my kipromini. Amazing results!
First let me say thanks to Steve Weiss, Robert Primes, Eric Kessler, Philip Bloom, Karen Abad, and all 700+ of you for doing all this work!
This is by far my favorite Shoot out thus far, as much as I love and shoot on my 5DMKII
I have decided it’s time to take the plunge and get the Sony F3 and 3 CP2’s.
The amount of work I am getting with the 5D alone justifies it.
The need to step it up, The Red one was on mind a lot while trying to decide, After this shoot out the F3 seems a perfect git for me price wise until i can get the Epic, Speaking of the Epic, Philip I’m willing to wager $100 USD that I can tell the Epic from all the rest in your unscientific shootout, as long as you let us see them ungraded.
Looking forward to your shoot out Philip, Your Camera lineup is truly perfect for us as most of us can afford every one of them…Minus the Epic 🙂
Philip,
I was told many times by Sony reps you can only record S-log with 4:4:4 recorder.
Please can you confirm this.
Thanks jiri
Not true. You can only record RGB 444 with a dual link recorder but you can record s-log as I have shown here even with just the SxS but ideally euthanasia 422 high quality recorder like a kipro mini
For me, seeing this at NAB and now with the additional commentary helps to point out the strengths and weaknesses of the various cameras out there. Knowing where the my DSLR will not do well gives me useful information on how I should light, what I need to think of in terms of production design and what sort of situations I should try to avoid. This knowledge also helps on the other end when trying to optimize the image.
It also helps me know what cameras I should consider based on my budget and production needs. For example, the F3 really blew me away at how it stood up against more expensive cameras. Too bad the Red Epic didn’t make it in time.
I completely agree with you that the Zacuto/SCCE team did a really fantastic job with this and that it proved to be very educational in terms of how the upper echelon of cinematographers think and technical areas that I should brush up on.
All in all, it’s still horses for courses and the best camera for expressing your creative vision is the one you have.
always when i see these shootouts, i think that i miss something (or let´s say, i seem to understand the smallest possible amount).
the things that makes it so hard, are these massive colorshifts. everyting looks like another planet with individual physics.
another thing ist, what would i get after color correction? the 2010 shootout showed some correctiontests, but i don´t get it. some of these images look so flat, that i dont like them. the alexaperformance seems so great, but i actually don´t like the look that much. i feelt the redstyle so much more pleasing(in the first part espacially). and it makes me think, that i miss something (the real look, that i would get in the end).
sensorsizes and iso-levels also seem to distract my ability to judge.
what a relief to be poor and actually a director/writer (with much interest in camarawork, but glad to give that hard part away, when i grow up)
(sorry for the writing – iam not english and flirtet with the girls in school, instead of learning enlish :))
I was impressed by the F3 in general. I must admit though that part of that may be a subconscious desire to convince myself that the FS100 is a great camera (and since the sensor’s the same, hopefully it is…). Obviously the really high-end cameras like the Alexa do even better, but it’s nice to see the capabilities of that sensor.
On the other hand, being as I am in the position of penniless amateur, I almost have a perversely reassured feeling about DSLRs having a somewhat muddy, low resolution look. It means the bottleneck is the camera, not the glass, so I can afford to buy cheap old manual primes and not worry too much about their sharpness! That’s not entirely true of course, and if I do someday upgrade to, say, the FS100, it might bite me in the arse later!
One thing I was puzzled about – I was under the impression that ISO was a standardised system (hence “ISO”!) where two cameras set to the same shutter, aperture, and ISO should be exposed (at least at the 50% grey point) identically. ISO ratings were massively different across the different cameras however; is this to do with the contrast curves and tone-mapping that goes on between the sensor and the output? If so… is ISO really useful at all for video?
I would have loved to have seen some of the short films made with one (or maybe feature all of the cameras) the cameras at the end. Much like the First “Great Camera Shootout”, That I thought was half of what made it so good.
Oh, well.
Maybe they’ll have one at the end of Part 3
the problem is you saw all DSLR stuff in the first one. In this one you need to watch “Game of Thrones” and things like that!!
i quote steve “This shootout was shot with all different kinds of cameras and nobody seems to care.”
guess this could be the last sentence of part 3 :). what would you have done, if you shoot a documentary while being surrounded by the best cameras in the world?
so the shootout itself could please youre desire? 🙂
Revealing parts of the test:
1) The 1D Mk4 under-performs compared to 5D and 7D big-time. Nocturne could have looked better on a 5D I reckon!
2) AF 100/101 Performed better than expected but shows in the Polka-dot dress that it is somehow not down-sampling perfectly. It is doing weird box patterns and loosing detail, wonder if the GH2 does the same.
3) Film Highlights rock! Notice the neon sign is perfectly preserved in film as there is no digital top in range on a physical medium. Beautiful and something for digital to still conquer.
4) Nikon D7000 has better overall video down-scaling than Canon, this was shown since early tests on the net also. Still not perfect and far from Panasonic’s solution.
5) Sony F3 continues to rock, the F35 looks much noisier. Its a wonderful piece of gear.
6) Red and Alexa can’t be beat for faithful reproduction of a scene, wonderful looking images that clearly replace film. No wonder most new productions are ditching celluloid for them.
7) In the grand scheme of things all the Cameras including the DSLRs offer a great image that will probably not hinder the story telling process. As evidenced by the latest iTunes 5D MkII shot movie. A good script and direction/acting trumps all the resolution tests in the world.
Great test overall Zacuto, hope you decide to do a 3rd shootout!
Hey Philip!
Thanks for putting this together!
My favorite on this scene shots & charts is the AF100 for the cost compared to image quality, sure the Red One is amazing in sharpness but for what price?
My choice for low budget projects would be the AF101 or the new FS100 and for big budget projects the Red One (maybe the Epic 🙂 ) judging only on a personal feeling and subjective opinion of course!
But for the moment I’m fooling around with my 7D and HFS10 😉
Instead of “unfair” i’d rather say : so little differences for such a huge difference of prices !
This Arri Alexa looks like a monster indeed. But look how good little dslr’s stand in front of those beasts. And they weren’t even made to compete at first !
When all of us will go to hollywood, we’ll shoot with top rank cameras ; but for now let us make amazing films that lead us to this point… 🙂
Watching the Zacuto Great Camera Shootout was so interesting, even though none of it applies to me & I understand half of it. Things like this work to reason with people who try to brush all cameras with 1 stroke & say all work equally well in all jobs. Some of the DSLR stuff made me cringe, like the Polka dot part, but is logical since things like compression & processing is just part of DSLRs as they aren’t made to be solely for video. I wonder what they would look like with RAW footage & better processing or the option to process out of camera.
hello, i think this episode is boring comparing to the first one. By testing with static scenes or charts, they are missing something important that is motion capture where compression artefact can be also seen.
THe chart used for resolution
They should film a short with all the cams , just to compare in “real life” .
you know what i think about charts…some people do love them though!
Phillip:
Thanks for all you do. Was the FS-100 excluded from this test? I read something that said that they may have tested it a bit. Any thoughts?
Mike
Hi Mike. The test was done at end of January. The FS100 wasn’t out at that point.
P
I’m sorry Phillip,
But I did not understand your comment about watching Game of Thrones.
please clarify.
In reference to seeing work shot on the cameras used in the shootout as was done in shootout one. Because of the high end nature of the cameras just watch things like Game of Thrones which was shot on ARRI Alexa
What was the lens on the AF 100/101? Was it the CP2., the Voigt. or the generic 14-140? I would be really curious to see the AF 100s performance with the new MFT lenses. Using the older 4/3 leica lenses and voigt for my dabbling has resulted in some really spectacular pics and videos on my GF1. The size factor is such an advantage, no-one notices me, everyone is posing for the guys with the huge canon dSLRs.
i think they all had identical lenses on as far as I know. They say so in the doc.
I just recently bought a 60D, got it for $900 on Amazon before the price shot back up. I can’t wait to start making films with it but I was a little surprised at the color reproduction. However, this seems to be a problem with the compression and not the camera itself. The RAW photos it takes are incredible and the sensor definitely rivals the more expensive cinema cameras, but it was simply not designed for video. I think fixing moire and aliasing, switching to a well-implemented AVCHD codec like the GH2, and allowing for outputting 4:2:2 to an external recorder would make the Canons incredible. They are still the best value for the money though and are capable of so much.
Also, the show Supernatural was shot on ARRI Alexa.
Would be interesting to see how the Ikonoskop measures up considering it’s price..
Anyone had any experience with it?
Philip, nice write up and review.
These test are great and it’s a fun documentary but it really shows how awful a picture you can get with all of these cameras when you make the conditions so extreme.
I believe this is a great tool for students or new comers into this field because it really does show you that just because you purchase a $60k camera does not guarantee you a great picture.
Lighting is key with all of these cameras and knowing the limitations of each gives you the ability to light it as perfect as you need. Film is the standard and everyone knows it won’t hold up in low lighting, but light it right and well, you know the rest.
Just for fun I wish they could light every scene just right for each camera and see if there is much difference.
I am happy that my Panasonic AF101 got those results at the second test. I have bought it recently after I saw the first test by Zacuto SCCE. My iso settings on camera are 200-320-640 and never had a problem indoors or in low light. I am using of course fast lenses. But Canon 5D and smaller cameras have their own fields to be on use. And it is true today in my fashion shoot I only brought my Canon 5d and Canon 550d was much easier to move and not so heavy tripods around.
As much as I adore the stuff Steve and Jens do (I believe “Film Fellas” will attain legendary status, eventually), I think these test are merely gear porn. no more and no less.
Yes, we have learned (great deal or a little bit – it depends on how literate you are) but that is just about it.
The notion of “the right tool for the right job” is FALSE, IMHO.
It is a fallacy. it cannot exist as we, as human beings, cannot “digest” so many fact to form intelligent, rational conclusion.
People shoot because it serves some internal agenda they have : be it methods they know and are good at; financial constraints; the will to belong to some “elite group”… and so on and so forth.
Just remember the following phrase someone coined :
****************
“Extended attention to details (in our case – factors in choosing a camera), leads to poor attention for the main goal (in our case – making motion picture)”.
****************
Just me 0.02
absolutely! nothing wrong with a bit of porn no? 🙂
Epic eats all of these cameras lunch on all tests… period!
how do you know? it wasn’t in the tests…pure speculation.
“EXTENDED ATTENTION TO DETAIL LEADS TO POOR ATTENTION OF THE MAIN GOAL”
I liked the review, Phil.
When making a decision in attaining a camera, there are FAR too many factors to consider … which leads one to the inevitable conclusion:
99% of the viewers would not be able to make out something of the data as shown.
“Right tool for the right job”?
Naaa… it is a FALLACY.
There is no such thing. The human brain cannot cope with so many factors AT ONCE in order to make a rational decision (aka “Right tool for the right job”)
It is a known fact (scientifically proven if I’m not mistaken – read Martin Lindstrom’s “Buy-o-logy”)
***********************
There will ALWAYS be improvements but the importance of a camera body is vastly overrated and the gulf between consumer and pro hardware is much smaller than it was in the earlier days of DV.
JUST SHOOT !
***********************
(Andrew Reid)
And I say: NUFF SAID!!
you quoting Andrew Reid!? Blimey! Mr track-suit wearing chav!! 🙂
Well, it took me a while to understand you are not talking about a Chevy 🙂 (being the non-English-speaking-layman that I am).
😀
One thing that I think bears pointing out in regards to comparing film with digital cameras is that there seems (in this test at least) to be the assumption that film grain and video noise are both equally undesirable, and that crystal-clear clarity and sharpness are always a good thing. For me at least, film grain on certain projects would be a great thing, and something I would go out of my way to have as a purely aesthetic choice, whereas most of the time the same cannot be said for video noise. It’s interesting to see film next to these other cameras, but ultimately it’s apples and oranges. I would love to see Kodak and Fuji keep making film stocks that are clear and sharp and less obviously film-looking, but at the same time I wish they’d also have some options that go the other way and have varying levels of grain, contrast, saturation, etc. How cool would it be to be able to shoot a film with TMax 3200 and just go nuts with grain? As some of the folks interviewed pointed out, ultimately these are all just great tools, and what really matters the most is the vision of the person behind the camera. A good cameraman can make compelling images with a cell phone, an Alexa, a wooden 8×10 field camera, or whatever is at hand.
This is kind of unrelated to the sensor and low light issues in the test. I just happened to come up on these lenses on the fourthirds site -especially the Horseman tilt shift adaptor. Has anyone tried this?
http://www.komamura.co.jp/e/TS-pro/index.html
http://www.four-thirds.org/en/microft/cine-sys.html
You know in the end a lot of this stuff boils down to the implementation of the codec being employed. A bad implementation of the code can fault the hardware…. I think the GH2 has shown that against the AF101. I guess not too many people have been able to dissemble the Sony/Red/Ari gear, but for what its worth the best performance & size per buck has to be the GH2.
In the end minaturasation will win. I’m truly gop smacked by what Vitaly is doing with the GH2 at the moment.
Love all this Zacuto “gear porn”: a great snapshot of this fast evolving state-of-art. This reminds of a time when a leading Japanese photographer had a magazine assignment to shoot a famous Hollywood celebrity. He chose to use a handheld “consumer point and shoot” camera at an outdoor cafe near Malibu rather than any of his expensive pro gear…It got the job done while enjoying and capturing a relaxed lunchtime atmosphere. His big pro gear would have caused a huge distraction…
It is such a fantastic concept when so many people can collaborate together on such an interesting subject.
I am an electronics Engineer and applaud all the people who took time/donated etc for this project.
Thank you everyone!!!