Episode 2, “Sensors & Sensitivity” of the three part series continues with tests covering sensitivity, resolution, compression and the relationship between them.  These tests were designed and administered by Robert Primes ASC, director of the Single Chip Camera Evaluation (SCCE) and shown at 2K screenings around the world to indie filmmakers, event shooters, commercial DP’s, directors and corporate filmmakers alike.  Their opinions on the footage are invaluable when it comes to understanding what all this data means in real world shooting situations.

To measure the sensitivity, the SCCE team used the Signal to Noise ratio of each camera to determine the threshold of acceptable noise. “Michael Bravin shot an opto-electronic conversion function or OECF chart,” says Steve Weiss, director.  The chart uses twenty different grey patches that are analyzed by software to determine how the sensor converts the illumination into digital values.” A low light scene was lit by Stephen Lighthill, ASC to show how noise can affect a shot in the real world.
 
Matt Siegel was in charge of measuring the resolution of each camera by shooting a 3ft wide Siemen star chart.  This chart is used to find the Spatial Frequency Response (SFR) of the sensor which shows the smallest details a camera can capture.  The final test in episode two covers color compression and sub-sampling. Each camera was recorded, if possible, to an external recorder to capture the most uncompressed image possible, with the Wringer chart, the differences between on-board and off-board recordings are shown. “Some cameras can record in 4:4:4” explains producer Jens Bogehegn, “but other cameras compress the image by removing color data, this can be 4:2:2 or 4:2:0.” A Still Life scene was also shot by Steven Lighthill, ASC and Nancy Shreiber, ASC to show the real world implications when resolution and compression are pushed to their limits.

Come watch Episode 2: “Sensors & Sensitivity” and learn from some of the best in the business!  We promise you won’t be disappointed.  
 
The featured scenes in this episode are shot by Michael Bravin, Matt Siegel, Stephen Lighthill, ASC and Nancy Schreiber, ASC.  The still life scene was designed by Rhonda Rolston.
 
Commentary: Jay Lee, Jack Cummings, Dan Freene, Daniel DeMoulin, Daren Finner, Nino Leitner, Jonathan Bregel, Jon Carr, Jan Crittenden: Product Manager-Panasonic, Bruce Logan, ASC, Cinematographer “Tron,” James Kallemeyn, Sam Shinn, Johnson Liv, Chris Cooke, Mark Steel, Dave Kittredge, Richard Crook, Ken Glassing: Cinematographer, “CSI-Miami,” Gale Tattersall, ASC, Cinematographer, “House,” Michael Lewis, Sebastian Tr, Mathew Medeiros, Paul Ream, Michael Watson, Peter James, ACS, ASC, DP “Driving Miss Daisy” & “Meet The Parents,” Robert Primes, ASC, Ryan Koo, David Johnson, Rodney Charters, ASC, Cinematographer “24,” and Robert Haddad.

 

CAST & CREW
The web series documentary features two different independent crews. The SCCE Crew: Administrator Robert Primes, ASC; Station Chiefs: Michael BravinStephen Lighthill ASCNancy Schreiber ASCMatt Siegel and Mike Curtis; Line Producer Josh Siegel. The Shootout 2011 Crew: Editor Karen Abad, Graphic Designer Chris Voelz, Producers: Daniel SkubalScott LynchJens Bogehegn and Eric Kessler; Web Series Director Steve Weiss.

Additional SCCE Testing Methodologies
All of the manufactures were invited to be involved with the tests and provide a technician with their camera.  In cases where the manufacturer declined to send a representative, Bob Primes assigned a camera master who was an expert with that particular camera to manage it as it rotated though all of the series of tests. Each test also had a station chief who kept the tests consistent across cameras.

Leave a Reply

Your email address will not be published. Required fields are marked *

*

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

140 Responses to “The Great Camera Shootout 2011: SCCE ~ Episode Two”

  1. Bumki Cho on July 27th, 2011 9:40 am

    Loading the second episode as I type! So excited! Thank you, zacuto!
    By the way, your EVF is the best!!

  2. Zacuto on July 27th, 2011 10:08 am

    Thanks!  We appreciate your support and we hope you enjoy this episode, let us know what you think about it when your done. ~Dave

  3. Spawnd on July 27th, 2011 10:53 am

    HEY, Zacuto

    Is this episode still reencoding on vimeo or the orginal file had such high compression?
    All I can see are blockes of compression, please reupload with higher quality.

  4. Scott Lynch on July 27th, 2011 10:59 am

    Hi Spawnd,
    We are currently uploading a new version of the show, which should be less blocky. I’ll make an announcement when it finishes uploading to vimeo and is available here. 

    Basically the issue we ran into was the best way to compress our 64Gb ProRes444 master file into a file we could upload to vimeo. The compression we used for Episode 1, didn’t work as well with this episode, I think because the tones and grain are so subtle with some of the cameras.

    -Scott

  5. Spawnd on July 27th, 2011 11:04 am

    Thx for quick response.
    Have You tried modyfying the keyframes? This usaully helps in those kind of scenes.
    Waiting for the real quality video.
    BTW, the Episodes are great :)

  6. Scott Lynch on July 27th, 2011 11:34 am

    Spawnd,
    Yeah the new version has modified approach to setting the keyframes which seems to work a lot better. There will still be compression, but I think everyone will be happier with the new version.

    -Scott

  7. Darren on July 27th, 2011 1:24 pm

    Another great episode.  I am appreciating more and more the extra added white “iron man” circles to tell me what I should be looking at in the scene specifically instead of it just being explained verbally.

    Looking forward to Episode 3

  8. mattie b on July 27th, 2011 1:44 pm

    I look forward to the higher quality file. Right now, everything is blocky so the compression aspect of the test will be hard to differentiate.

    Thank you for all of your hard work.

  9. Bumki Cho on July 27th, 2011 1:57 pm

    Great episode! Perfect demonstration of the numbers don’t always translate into the actual footages and why different cameras are in different price range.

  10. Jon Sovey on July 27th, 2011 3:05 pm

    Great episode! It would be nice to have some uncompressed stills of the comparision shots. 

  11. Scott Lynch on July 27th, 2011 3:45 pm

    Ok! The new and improved compression is now live, you may need to clear your cache on your browser.

    -Scott

  12. eco_bach on July 27th, 2011 4:13 pm

    Hard to evaluate on the web..would be nice to be able to download a higher res, less compressed version.

  13. Spawnd on July 27th, 2011 4:27 pm

    thank you, watched it, way better the the previous version.

  14. Anonymous on July 27th, 2011 4:34 pm

    Zacuto Rocked yet again! Loved it! Veru surprised at how well the 7D is doing. 

  15. Scott Lynch on July 27th, 2011 5:29 pm

    I’m working on a full prores444 version of the tests that can be downloaded. I’ll be trying to get this out to you guys in a few days from now.

    -Scott

  16. Jon on July 27th, 2011 6:01 pm

    Hey Zacuto,

    Loving this series so much, what a wealth of info. Are you planning on doing three episodes again like last year, or will there be more?

    Thanks,
    Jon

  17. Christopher Marino on July 27th, 2011 7:18 pm

    Very well done to all involved. I share Ryan Koo’s comment on the numbers really don’t mean the end result but more a reference. You have to test, test, test for your particular scene or setup to find out for yourself. Again great work!

  18. Murray on July 27th, 2011 8:00 pm

    The Panasonic AF100 really beat the pants off the Canon 5D Mark 2. I’m going to take a serious look at that camera. Makes me wonder about the Sony FS100 too.

  19. Hans Steinert on July 27th, 2011 8:06 pm

    yeeeeessssssss

  20. Joel Mielle on July 27th, 2011 8:28 pm

    These camera shootouts are starting to if anything, confuse me.  I’m not crtisising the production here as it’s awesome and give them credit for undertaking such a task.  So thank you Zacuto and all the sponsors.  I just feel as film makers we are getting too technical.  What I would like to see is a regular audience with a some sort of pleasure meter that rates the feeling they obtain from different cameras.  I can tell you now, there would be no difference if you’re into a story.  We are comparing $2k cameras with $30k cameras yet the end result to our audience would be exactly the same.  My point being that I’d like to replace my 5D with a proper video camera but nothing makes sense. So I still have no idea what to buy. I’m just hanging for the 5D MIII. Am I wrong?

  21. Joel Mielle on July 27th, 2011 8:33 pm

    I was thinking the same thing, but when you get into it, it just doesn’t have the magic of the 5D, the depth etc.  The AF100 might be sharper with less noise but it has a very sterile TV look, not as cinematic.

  22. Steve Weiss on July 27th, 2011 10:34 pm

    Joel, take a breath…  these are just cameras.  focus on the story, that’s what’s going to make your film work or not.  Nobody is going to see these differences.  Stick with the camera you have and focus on how to master it strengths and weaknesses but focus on the story.  That’s what we do.  This series was shot with some HVX200, EX3, 5D, 7D whatever.  People don’t watch a movie and say boy if it only was 4:4:4 I would have like it soooo much better.

  23. Glenn Hanns on July 27th, 2011 11:47 pm

    Notice that the blowup at 20:08 has it listed as the F35 yet it clearly has grain and moving dirt in the image, im assuming they mixed it up with the 5213 footage?

  24. Glenn Hanns on July 28th, 2011 12:12 am

    Begs the question then “wheres the F35 blowup?”

  25. Martini Hong on July 28th, 2011 12:17 am

    Of course we all know here we are comparing the behavior or each cameras on different situations. The perfect example here would be girl’s red dress with white polka dots. If your “story” needs that detail to be included, for example: the other girl talks about that girls red dress with white polka dots. And the viewers are only seeing red dress, as shown on DSLR results, then that’s a bummer…

  26. Hans Steinert on July 28th, 2011 12:24 am

    Joel, what you say is true, but the primary difference in these cameras has to do with the flexibility of achieving the look you desire, which is part of story telling afterall.

  27. Dustin Uy on July 28th, 2011 2:28 am

    id disagree. cinematic is very subjective.

  28. Dustin Uy on July 28th, 2011 2:29 am

    lowlight and dynamic range can be controlled by lighting.
    you cant control soft images(low resolution) & aliasing/moire.

  29. Chris Medico on July 28th, 2011 6:18 am

    For some reason I can’t get it to play. If I click through to Vimeo it indicates the video is private. :’(

  30. Chris Medico on July 28th, 2011 6:18 am

    For some reason I can’t get it to play. If I click through to Vimeo it indicates the video is private. :’(

  31. Chris Medico on July 28th, 2011 6:19 am

    For some reason I can’t get it to play. If I click through to Vimeo it indicates the video is private. :’(

  32. Bumki Cho on July 28th, 2011 8:42 am

    Not surprised D7000 didn’t measure up nicely against big boys in this episode (resolution/compression/shooting charts). The same thing can be said about 1D, 5D, and 7D. When just comparing DSLRs, I thought D7000 image looked (subjectively) better than Canon DSLRs in low light scene. I also thought D7000 image looked sharper than Canons in flowers and spices scene (again it’s my personal taste). But moire in D7000 was bad. It was just weird. Funny thing is I didn’t notice it in the first episode and that to me is an important piece to keep in mind. People watch stories not pixels.

  33. Mandy Rogers on July 28th, 2011 9:10 am

    Hi Chris,
    Are you clicking play up above on the video? What browser are you using? 

  34. Mandy Rogers on July 28th, 2011 9:10 am

    Now your famous! ;)

  35. Mandy Rogers on July 28th, 2011 9:12 am

    Hi Jon- Glad you ar loving the series. We will have a third episode, like last year.. stay tuned. 

  36. Mandy Rogers on July 28th, 2011 9:12 am

    Thanks for the Z-Kudos! 

  37. Steve Weiss on July 28th, 2011 9:25 am

    Remember that what you are seeing with the polka dots is a huge blow up of the image.  

  38. Steve Weiss on July 28th, 2011 9:28 am

    It does not have as shallow DOF as the 5D but the 5D has more shallow DOF then any film camera ever made except IMAX or OMNIMAX

  39. Steve Weiss on July 28th, 2011 9:33 am

    Sure you can.  You can sharpen in post and or try not to shoot scenes that produce aliasing/morie.  I know that sounds harsh but in the 80 & 90′s red was a color that never really worked on video.  It just buzzed.  When Nancy Reagan wore that red dress everyone in the news pool freaked out.  We did what we could to get around the problem (lowered the chroma).  You do what you have to do to get around your cameras weaknesses and exploit it strengths.

  40. Chris Voelz on July 28th, 2011 10:26 am

    Hi Glenn,

    It is in fact the F35.  We noticed that the F35 had some weird clipping going on in the red channel.  If you look closely at the wide angle you will see the black dots only on the dress and red pillow. The blow up has noticeable noise because the camera has a maximum resolution of 1920×1080.  The image was blown up by 450%, so there isn’t a lot of information there to support that kind of zoom.  So this is why it looks so noisy.  Thanks for watching!

  41. AWS on July 28th, 2011 10:31 am

    Man, I’m loving the Alexa!
    I’ve always found the color reproduction on the Canon DSLRs to be lacking. Even still, I was surprised at how poorly they handled that red dress. I really hope that Canon puts out a DSLR that’s geared more towards professional video/cinema shooting in the near future. Use a better codec or maybe even raw. Now that they’ve “stumbled into it”, they should latch onto DSLRs as cinema cameras.  
    It’s a sad thing indeed that you guys couldn’t get a Red Epic for this test. I would love to see how it measures up against the Alexa.

  42. Eric Whitehead on July 28th, 2011 11:22 am

    Haven’t worked with the FS100 yet, but by all accounts it definitely seems to be a very viable choice. The chip is the same as the F3, so you can expect similar image quality. Check out Phillip Bloom’s comparisons here: http://vimeo.com/23294197. And with it only being slightly more expensive than the AF100 (Check B&H’s prices), the decision between the AF100 and FS100 has become very tough. They’re both great cameras, and from what we’ve seen so far in the shootout, the F3 and AF100 have outperformed the DSLRs IMHO. Although… We haven’t seen episode 3 yet. :)

    On a side note: Incredible work Steve and the entire shootout team. Such important and game changing work in the democratization of film. Cheers!

  43. Bumki Cho on July 28th, 2011 11:30 am

    Just wondering which manufacturers sent a tech person to set up their cameras. Can someone answer please?

  44. Scott Lynch on July 28th, 2011 11:43 am

    Panasonic, Canon, Phantom were there for Phase 1 and 2, Sony was there for Phase 2. Fujinon reps were also on hand to check the backfocus of the lenses.

    -Scott

  45. Bumki Cho on July 28th, 2011 12:23 pm

    Thank you, Scott. Too bad not all of the manufacturers saw the benefits of this extensive testing on their cameras.

  46. Bruno Perosa on July 28th, 2011 9:03 pm

    thanks Zacuto

  47. james wood on July 28th, 2011 10:13 pm

    I’ve been waiting for this since episode 1.

  48. Adam Harness on July 28th, 2011 10:15 pm

    Dang, as a person really excited to start using the 7D, I have to say it’s made me want to start even more now, seeing how well they compete with 100,000+$ rigs. @Kesslercrane Thanks for all the help in E.2

  49. James Vonk on July 29th, 2011 2:18 am

    Great test work, and kudos to all involved. I watch every episode in and admire your great work. Taking into consideration many details, including the Vimeo java player compression, I am interested why you did not attempt to use unmodified Canon/Nikon glass on the DSLR’s. In the Still Life scene, all the DSLR’s appeared to have chromattic shifting on the detail highlights. Could this be related to the modified PL mount bodies to fit the Fuginon lens? Are the PL lenses checked for sharpness vs the Canon glass? I would hate to modify my camera and lose sharpness. May I inquire who modified the DSLR bodies?

  50. Stunko on July 29th, 2011 2:37 am

    Boy, oh boy, this one was SO BAAAAAD. At least we learned what we already sort of knew, namely that DSLRs were designed to takle still pictures. And not video. Surprised to see how extremely poorly the pricier Canon clunkers — the 1D and 5D — fared. But here in Episode 2, even the 7D and the D7000 flunked it. Shouold have tested the 60D and D5100 instead, perhaps, as it looks like the cheapere DSLRs are outperforming the pricier ones. 

    Sumamry judgement: you need to spend the bucks to get any quality. 

  51. Steve Weiss on July 29th, 2011 3:33 am

    Wrong conclusion.  I have seen brilliant DSLR work.  These tests to me even made film, Alexa and RED look bad.  They are huge blow ups and stress tests.  Nobody would ever stress a shot like this in real life.  Don’t get yourself into such huge stressful situations.  Light for your DSLR, don’t have shots that are 5 stops over-exposed.  That is a total cop out.  Anyone of these cameras can look great if you are a great DP and know how to use your camera.  I did not intend for this test and documentary to make people feel bad about their camera.  It’s more to see some differences and see what camera shines best in each scenario.  Some tests the DSLR’s are better, some the bigger cameras and some film.  Interesting but irrelevant.   Even Bob Primes said in episode one, it’s all in your talent as to if any of these cameras are going to look good.  I’ve seen shit done with a RED and I’ve seen brilliance done with the same RED and that goes for every other camera including the DSLR’s.  With the quality we are at now getting these days in all of these cameras (which is 10 times better then the video cameras I’m accustomed to in the 80′s and 90′s, stop your bitching please) there is no excuse not to have amazing images once learn to master the strengths of each camera and the weaknesses as well.  

    Remember, content is kind and no matter how good your footage looks if your story sucks then it’s all shit.  Nobody said boy did that was a horrible film but it really looked great.  Story, acting sucks, whole film sucks.

  52. Stunko on July 29th, 2011 4:10 am

    All true, Steve. Biggest differentiation criteria in Episode 2 seems to be between cameras that can only record internally to flash cards using a super busy-body codec via the line-sklipping method, and those that can in fact bypass mediocre internal recording entirely and output via HDMI and/or HD-SDI to an external field recorder that can deliver and capture way better bitrates and algortithms. 

    With the Canons and Nikons, it is line-skipping, internal flash card recocrding, 4:2:0 chroma subsamljning, 8-bit color depth all the way. Too bad, really it is. 

    Panasonic AG-AF100 can output it signal bypassing the H.264 codec, but it is still doing pixel dowsampling, and its small, 4:3 aspect ratio sensor is a real off-beat novelty item in this group.

    That would sort of leave the Sony NEX-FS100 and PMW-F3 and above as the contenders to pick from, at least based on Episode 2′s rather shocking results. The HDSLRs pretty much failed this second round of testing. For the life of me, I have no clue why the pricey Canon 1D Mark IV was even chosen to run — its performance was pretty much atrocious across the board. 

    Biggest surprise to me was the Panny AF100′s results, I think this must have surprised a lot of skeptics. And the new Sony Super 35 sensor in the FS100 and F3 seems to be holding up pretty good. 

  53. bahman Kormi on July 29th, 2011 8:46 am

    Great testing! Big Result! I appreciate the bandwith of possible scenarios using different tools!
    Bahman Kormi
    EU-based DP
    mail@bahmankormi.de

  54. Steve Weiss on July 29th, 2011 9:29 am

    thanks bro, I’ll let the team know

  55. Scott Lynch on July 29th, 2011 11:54 am

    Hi James,
    The use of the PL glass was not the cause of the chromatic shifting that you’re seeing. The PL Lens mount conversions of the 1DM4 and 7D were done by Clairmont Camera and the 5D was converted by Hot Rod Cameras. Both of these companies do superb work. The Fujinon lenses used are also some of the best zoom lenses available, and there was a Fujinon tech on set to make sure the back focus was correct for each camera.

    I’ve attached an image below of the Canon 7D in the still life scene. There is a frame from the h.264 video and a RAW still. The shifting you are seeing is compression and line skipping when the camera creates the video file.

  56. Nuno Rocha on July 29th, 2011 12:04 pm

    Great Job everyone! I love this show! 

  57. Alan Lasky on July 29th, 2011 12:08 pm

    For anyone who lives in the world of motion imaging sensors the Zacuto guys are doing pretty thorough work on sensor testing. Everyone else will likely find it incomprehensible and boring. Hats off to these guys though, as Hollywood-centric camera tests usually end up in agenda and legacy based nonsense.

  58. Anonymous on July 29th, 2011 12:39 pm

    Video is not embeded on your site.

  59. Rob Imbs on July 29th, 2011 1:53 pm

    This is SOOOO well produced and edited. The 3 dimensional charts are wonderfully animated, excellent work! This series should be on PBS.

  60. Dave Gish on July 29th, 2011 1:58 pm

    The DSLR cameras all have HDMI ports that output video while recording.  My understanding is that this version of the video has no chroma sub-sampling or other compression artifacts.  There are little boxes that covert HDMI to HD/SDI.

    So why couldn’t the DSLRs use the higher quality recorder as well?

  61. Scott Lynch on July 29th, 2011 2:03 pm

    The DSLR’s do not output a clean signal, they put little white boxes and other graphical elements on the video which makes it unusable.

    There’s is no way to record a clean signal from the cameras we tested, through the HDMI, as of right now.

    -Scott

  62. Dave Gish on July 29th, 2011 2:34 pm

    Sounds like a good candidate for a firmware hack…

  63. Anonymous on July 31st, 2011 3:13 am

    unless you had installed magic lantern on a canon 550d :)  - The new hacked firmware for the canon 550d allows you to turn off all graphics from the hdmi output – but as the output drops to 480P on the hdmi port making it less useful then we’d like :(

  64. Felix on August 1st, 2011 9:00 am

    The GH2 would be interesting too!

  65. Perry on August 1st, 2011 10:03 am

    Why are you guys not shooting with Cinestyle picture profile on th canon DSLRs? I feel it is a strong tool that is accessible to anyone with the camera…

  66. Scott Lynch on August 1st, 2011 10:38 am

    Hi Perry,
    These tests were shot in February and March of 2011 and the Technicolor proflie was simply not available or announced at that time.
    -Scott

  67. blahberstein on August 1st, 2011 8:24 pm

    Is it possible to record from a DSLR to a Codex or S.Two?  Or would the HDMI transfer rates be less than the write speeds for the SD or CF card?  Are there other options for bypassing the 4:2:0 compression of DSLRs?

    My understanding is that the transfer speeds of HDMI were around 10.2 gbps, whereas SD Class 10 cards (for example) are around 80Mbps.  Any thoughts?  Why did you guys opt for the SD and CF cards in the camera?

  68. Josué-Joshua Owens on August 1st, 2011 10:50 pm

    What about Super 16mm?

  69. Scott Lynch on August 2nd, 2011 9:10 am

    The issue we had when recording the DSLRS is that the HDMI signal always has graphics on it. And there is not a way to record a “clean” signal from the camera. There was talk that a hacked firmware could remove the graphics, but there are a lot of other problems that could affect the camera if you used a hacked firmware. It was decided that we could only really use the camera as it was designed, and that was to record the video to the SD/CF cards.

    -Scott

  70. Scott Lynch on August 2nd, 2011 9:12 am

    We were only testing cameras capable of imaging areas similar to 35mm. S16 is too small of an imaging area, and every camera you add to a test makes it that much more difficult to compare.

    -Scott

  71. blahberstein on August 2nd, 2011 1:09 pm

    Awesome, Scott!  Thanks for that info.

  72. Anonymous on August 2nd, 2011 9:23 pm

    I understand. Thank you. I was just wondering about the types of things you were testing, like resolution, compression, and dynamic range.

  73. Josué-Joshua Owens on August 2nd, 2011 9:24 pm

    I would imagine that Super 16mm would have performed similiar to 35mm, but with more grain.

  74. Josué-Joshua Owens on August 2nd, 2011 9:27 pm

    I thought the DSLRs performed very well compared to the cameras that were much more expensive than they were.

  75. Brad Ohlund on August 3rd, 2011 9:34 pm

    Such valuable info. Like most artists, cinematographers care deeply about what their equipment/technology can do to help them make the directors vision a reality, but even more importantly, how their images affect their audiences on a visceral level. As one person commented, if you need the polka dots to show up to tell the story, pick the camera that will do that. But if you start with the highest quality camera that is suited to your production/budget/story/ format/venue, you then have a larger “palette” to work with. This information provides such a valuable service to cinematographers in the current world of technology.

  76. Mike Freze on August 3rd, 2011 11:56 pm

    I was lucky enough to see a rough cut of this at a private screening at NAB in Vegas with Rodney Charters (24, Shameless) Robert Primes ASC, Academy members and Sony execs.   This is the best explanation of the capabilities and comparisons of the newest HD cameras I have seen yet. 

  77. Anonymous on August 4th, 2011 5:46 am

    the af100 seems to be holding up pretty well. I really wished theyed used a hacked gh1 or gh2 to see how it would hold up

  78. Raydon Price on August 5th, 2011 1:51 am

    Well done!

    NOW EPISODE 3 hits us …when?

    Peace!

  79. Anonymous on August 5th, 2011 1:23 pm

    Thank you very much for doing these comparisons!

  80. Surfer Mike on August 6th, 2011 5:56 pm

    Another great webisode, again congrats to everyone involved. What this episode highlights for me is the drawbacks in recording anything to H264 format, I suspect if the DLSR’s could “run a bypass” there specs would have been higher. I thought i would never hear a DP say he was shocked at film grain, Lol. Looking forward to the final webisode. 

  81. Stunko on August 7th, 2011 11:11 am

    Of course they do, you just have to pick the right ones.

    http://atomos.com/ninja/dslr.xml

    For example, the Panasonic Lumix GH2 (and G3) output perfectly clean video, at least according to some folks who have tested this with the Panasonic HDSLR hooked into the Artomos Ninja portable video field recorder.

    Of course, if you must shoot with a Canon or Nikon DSLR instead, you are sort of out of luck when it comes to external recording from your DSLR.  

  82. Stunko on August 7th, 2011 11:14 am

    Of course they do, Scott, you just have to pick the right ones.  
     
    http://atomos.com/ninja/dslr.xml  
     
    For instance, the Panasonic Lumix GH2 (and G3) mirror-free HDSLRs  output perfectly clean video at 1080p and 1080i, at least according to some knowledgeable folks who have tested this fact with the Panasonic HDSLR hooked into the Artomos Ninja portable video field recorder via HDMI connectivity.  
     
    Of course, if you must shoot with a Canon or Nikon DSLR instead, you are sort of out of luck when it comes to external recording from your DSLR.  There is always hacking, of course. 

  83. Stunko on August 7th, 2011 11:35 am

    Wow, Scott, “Super 16 has too small of an imaging area,” eh? Compared to what, I wonder — Super 8 or Regular 8?

    Anyhow…. DP Vilmos Zsigmond, ASC lensed “Surrender, Dorothy,” a television movie featuring Diane Keaton, for director Charles MacDougall. Zsigmond has compiled some 70 narrative credits during his storied career, including Close Encounters of the Third KindThe Deer Hunter and Deliverance. Movie was shot with two side-by-side Panavision Elaine Super 16mm cameras mounted with 4:1 and 10:1 zoom lens used for variable focus. The shorter lens was usually used to cover master shots while the longer one was focused on close-ups that director Charles McDougall used for cutaways.
    After the film was edited offline, the conformed negative was scanned at 2K resolution at Burbank’s Modern VideoFilm, where Zsigmond timed the movies as well. “I was surprised by how much detail you can see on an HD monitor without a hint of grain,” he says. “I would use this format again on the right movie.”

    SO, I guess for some of the digital shooters Super 16mm motion picture film is unacceptable, I guess these old-school ASC cinematographers are hopelessly stuck in the past. 

  84. Stunko on August 7th, 2011 11:44 am

    Surfer Mike, I know thatere was a qualified cinematographer in charge of these tests, and that this was a 100% independent, honroable test. Now, that does not eman that everything was done with the films stock, processing, scanning, etc. like if another ASC member was running camera on a $100 million budget feature shootintg on 4-perf or 3-erf 35mm film. 

    If someone had made me a scan transfer this bad on my S16 film negs, I would be furious.

    Re. video-enabled DSLRs that can bypass the deplorable internal recording mayhem to flash memory cards, check out the latest Panasonic Lumix models. They can do what the Canons and Nikons cannot, apparently.

    http://atomos.com/ninja/dslr.xml

    Unfortunately, even when you record from a Panny Lumix GH2 to something like the Atomos Ninja in the Apple Pro Res HQ codec at 220Mb/sec bitrate, you need to keep in mind that these cheap HDSLRs will only give you a seriosuly crippled 4:2:0 chroma subsampling quality, coupled with lowly 8-bit color depth. Not a desirable combination, for sure.

    At least with soemthing like a Panny AG-AF100 or Sony NEX-FS100, you can record internally a 4:2:2 chromae, albeit also only at the lower 8-bit color fidelity.  

  85. Surfer Mike on August 9th, 2011 8:30 pm

    Hi Stunko Thanks for the explanation. Point taken about the post process on film but it’s still interesting to hear the statement in the 1st place. I remember a tape camera in the 90′s a Ikegami “something?” which it’s big claim to fame was it looks just like film. Which of course it never did!!. Shows just how far camera development has come since then & the differences between a digital format & film is shrinking. But there are still differences!    

  86. jkoenig on August 12th, 2011 2:44 pm

    what iPad app are they using in this episode?

  87. Mandy Rogers on August 12th, 2011 2:57 pm
  88. rizibo on August 15th, 2011 12:55 pm

    When will episode 3 be released?  We are dying to see the surprise ending.

  89. Anonymous on August 16th, 2011 10:04 pm

    Somebody may said that 720P from Panasonic GH2 better than 1080P from Canon 5DMKII

  90. Andrew Jakobs on August 18th, 2011 5:36 pm

    I’m just looking at Episode 2, and looking at the comparisons I’m wondering why all the different ISO’s between camera’s, how can you compare camera’s if you don’t have them all at the same ISO just like in the 2010 shootout.. But I guess I don’t really understand the test-methods used, I’m just getting into camera’s and trying to understand everything.. Personally what I’ve seen till now I would stay away from the weisscam HS-2.. …continueing watching episode 2…

  91. Doleep Mahmoud on August 20th, 2011 9:14 am

    When episode 3 will Air or stream! its 20th of August and no word about it yet?

  92. James on August 21st, 2011 6:00 pm

    true that mate. It gives the illusion of motion. People feel like the’re accomplishing something. ;)

  93. Jonathan Hunter on August 25th, 2011 12:48 am

    So the secret surprise in Episode 3 is the testing of the new Sony a77.  Wow!  

    Very nice addition and gratz on getting it for this shoot.  Wonder if it is able to top the Canon 7D and Nikon D7000.

    Thanks for all your hard work!

  94. Jerrod Nash on August 25th, 2011 3:18 am

    I wondered why you’ve emphasized that you shot on two film stocks, but you only show one of them in the comparisons. Did they yield results too similar to include both in the compairisons? Maybe you could include a blurb on why the Kodak 5319 was selected (also why it was selected vs the other Kodak stock that was shot). It may be obvious to those who have worked with MP film for years, but not as obvious to me, someone who has only worked with $10K and under cameras and still film photography.

    Thank you for doing these tests. I know it must have cost your company (and Kessler) a lot of money. Thanks to you both, but especially to Kessler since they are from my home state of Indiana (and North Central Indiana at that). I am happy to see a great company like theirs succeed as they have.

  95. Anonymous on August 26th, 2011 4:03 pm

    When will Episode 3 be released?

  96. Doleep Mahmoud on August 26th, 2011 10:13 pm

    I think we will be watching eposide 3 on the 31th of august ? zacuto where are you??

  97. James on August 30th, 2011 11:04 am

    Where is episode 3? I know I know, I’m an angrateful sob. Guess I’m a little cranky with all the product delays this year. Especially the pix 240. Delayed again! October….

  98. hisonni johnson on August 31st, 2011 11:49 am

    where is ep 3. please let us know

  99. Mandy on August 31st, 2011 12:51 pm

    We are still tweaking the edit, should be a couple weeks! ~ Mandy 

  100. Mandy on August 31st, 2011 12:52 pm

    We are still tweaking the edit, should be a couple weeks! 

  101. James on August 31st, 2011 6:56 pm

    ouch! Thanks.

  102. JimWil on September 2nd, 2011 8:37 pm

    I’d also like to know what the delay is with episode three.  Can we get word??

  103. Steve Weiss on September 2nd, 2011 9:55 pm

    It takes a long time to edit these episodes, there is tons of footage and it is an enormous job.  Karen is constantly working on these episodes for months.  It just doesn’t come together.  You need to create a story in every documentary and that’s what we are working on.

  104. Stunko on September 4th, 2011 6:53 pm

    Episode 3 has probably been canceled, due to lack of interest. No great loss, really, since Episode 2 was pretty awful already. Also, half of these cameras featured are getting to be on the ‘Endangered Species’list already. Time is a-marching on, guys. 

  105. Steve Weiss on September 4th, 2011 7:26 pm

    You know you want it Stunko, 3 weeks and it’t coming.

  106. Stunko on September 4th, 2011 7:28 pm

    You got me there, Steve. 

  107. Stunko on September 4th, 2011 7:32 pm

    I was wondering about that, too. All I ever shot on was on Agfa, ORWO, and Fujifilm 16mm stock, starting way, way back in 1970. But I guess if you really have the deeper pockets for the higher prices, Eastam stock will also do. 

  108. Robbin Groot on September 6th, 2011 2:24 am

    I noticed the lack of attention that the Nikon D7000 gets in these episodes. Or is it just me on this matter? Except for the resolution test the d7000 looks just sharper then the Canons (except for the 5d mkII) and much better shadow detail/low-light and Iso performance. Then again, i’m a bit dissapointed in the colorhandling of the D7000 and ugly moire. How does the lens tested on the nikon compare to that on the other camera’s? And what about the firmware updates? 

    I was shocked seeing the 5d mkII performing so bad in many tests, were i thought the 7D had the advantage over it’s older brother. Let us not speak about the 1D. I expect more from the upcoming 5D mkIII

  109. Steve Weiss on September 7th, 2011 9:02 am

    You need to watch all three episodes and all of your questions will be answered.  Episode three is about 2 weeks away.  Watch ep. 1, it answers many of your questions.

  110. John McAleese on September 11th, 2011 9:24 am

    This episode was fantastic!, good point for the SLR’s,. compression will always be a factor to keep in mind, remember compression like technology will also progress in time, so hang in there SLR users, however, Steve, we did see an umcompressed RAW still from one of the slr’s on the Siemen star chart, I would imagine this was a reduced resolution to match as near HD resolution as possible?

    I would have loved to have seen a FULL resolution RAW (5616 x 3744, ) still from the 5D in this comparison, ok, a pointless test for what the series is trying to achieve, however, in the animation industry it would be really nice to know what these cameras can deliver to us before the cut

  111. Dave Tally on September 12th, 2011 7:52 am

    First let me say, great job. I got so much out of last years series and was so glad to see it repeated this year.

    My question pertains to the 7D. What color style was used?

    I have had good results with the Technicolor CineStyle and S-curve LUT and color gradin using CineForm Neo and really would to find out what you used for the test.

  112. Scott Lynch on September 12th, 2011 8:33 am

    Hi John,
    The scene you are referring to where we show the uncompressed vs compressed. The uncompressed is full resolution RAW, so you are seeing the benifits of both the increased resolution and no compression. Both of the shots were blown up so you could see the differences better. But with anyone with a DSLR and a printer can recreate this type of test at home to see how your own camera performs in still vs video mode.

    Just print out a back focus chart, and then shoot it 8-10 feet away with both stills and video mode without moving the camera, then you can bring the footage in and compare the differences.

    -Scott

  113. Scott Lynch on September 12th, 2011 8:46 am

    Hi Robbin,
    The D7000 gets some more attention in EP3, but we noticed a lot of the people who saw the presentation were interested in talking about the larger cameras.

    I have a few theories about why the Nikon looks sharper, the D7000 records a less compressed codec then the Canon’s so it gets some help there. The other major difference, as you pointed out, is the lens. The Nikon was using a Zeiss ZF prime on it and the Zeiss lenses have a lot of contrast. This makes the image look sharper to us. The Canon’s had the Fujinon lens, which is capable of resolving more detail, but is not as contrasty and so it looks less sharp to us. This a trick of our perception of sharpness which we covered in this episode.

    With the firmware updates, the D7000 was using the 1.01 firmware. At the time of the testing in February and March, the 1.02 firmware updates had not been released.

    -Scott

  114. Scott Lynch on September 12th, 2011 10:11 am

    Hi Dave,
    We used the “Neutral” setting with Sharpness = 0, Contrast = -2, Saturation = 0, Color Tone = 0.   

    The Technicolor profile was not released at the time the tests were shot. 

  115. Dave Tally on September 12th, 2011 5:16 pm

    Do you have any comments on the technicolor profilem, adjustments, concerns, things to watch out for, etc.

  116. Anonymous on September 12th, 2011 6:00 pm

    I own a 5d MarkII for two years, i never got such unsharp images from this camera like you did.
    How did you record, which picture style, which lense for all the cameras? Did you crop in post?
    With the “Neutral” setting you cannot get the colour-style that you present.

    I’m really wondering!?

  117. Scott Lynch on September 12th, 2011 9:13 pm

    Hi Guest,
    Which part of the video are you referring to? Some of the scenes that we show are blow ups and some are not. We answer most of your questions in the episode itself, but here is a quick recap:

    - We recorded the 5DMII on the internal CF card
    - We use the Neutral setting with Sharpness = 0, Contrast = -2, Saturation = 0, Color Tone = 0 for all the scenes
    - The camera had a PL mount so we used Fujinon Premier Zoom lenses for most tests
    - Some scenes were blownup, but we mention in the show before we show you a blown up scene.

    -Scott

  118. Scott Lynch on September 12th, 2011 9:16 pm

    The only thing I would reccomend is to not shoot with the saturation turned down when using the Technicolor Profile. In my experience, I find that it is too difficult to bring back color with the h.264 compression, so don’t take too much of it out when you shoot. 

    The best rule of thumb when using a DSLR or a camera that only records in Rec709 is to shoot your footage as close as possible to how you want it to look in post. The more you can do in camera the better your footage will look.

    -Scott

  119. Anonymous on September 13th, 2011 6:44 am

    It’s the noise test from 7.24 around and at 9.45. The focus should be at the young woman and later the man in the car. The persons are sharp with many cameras but not the dslrs. Therefor the subjective impression of image-quality sucks at once.
    Especially the 5d has great low-light performance and a focused object IS sharp, noisy or not.
    DOF is smaller than with cameras with smaller sensors. You can see that the bush in the background filmed with the Alexa is sharp too. While i cannot determine any sharp point filmed with the 5d.

    Again, which lense? The PL mount lenses do not cover the sensor of the 5d, in consequence you can only use the center of the image which means a loss of pixels.

  120. Scott Lynch on September 13th, 2011 2:03 pm

    I think what you are are seeing is a combination the compression of the h.264 on the cameras and the re-compression of vimeo. I would suggest downloading the original file we uploaded to vimeo, from this page: http://vimeo.com/26772177

    But even if you look at the source footage when you compare a 5D to an Alexa, the 5D does not hold as much detail. These DSLR cameras do look softer because of line skipping and compression. 

    Again the lens we used in this paticular scene was the Fujinon 14.5-45 Zoom Lens (http://www.fujifilmusa.com/products/optical_devices/digital-film-cinema/pl-mount/145-45/)

    Some PL lenses are not capable covering the full sensor, but in this case this lens did cover the Full Frame sensor at the focal length we shot at. Also a loss of pixels is not really the best way to describe what happens, what you would see would be vinetting on the corners of the screen.

  121. Stunko on September 14th, 2011 9:38 pm

    Many of therse cameras showcased in the test are sort of obsolete already. I mean, why would someone use any of the pricey digital film cameras parede here, now that Sony had released their 8K resolution capable and 4K recording to flash card F65 CineAlta camera for only $65,000 with a color OLED VF? The ones Zacuto is featuring here will have no chance in hell in competing against the Sony F65, which will start shipping early next year.   

    Also, why would someone get an old school flipping-mirror type of a DSLR that has a 100% useless viewfinder in video mode, now that Sony now has two amazing Single Lens Translucent video enabled DSLRs out there, the SLT Alpha 65 and the ASLT Alphas 77? Maybe the hapless Canikons can battle it out amobngst themselves, but they have no chance to trump the new Sony SLT Alphas HDSLRs and their AVCHD 2.0 recording codec.

    These tests here are still useful, of course, just don’t rush out and buty one of these cameras in September 2011, is all I’m saying. There are some new boys in town, folks. 

  122. Steve Weiss on September 14th, 2011 9:55 pm

    And we are already working on a new shootout to include many of the new cameras not used in the 2011 shootout.

  123. Oli on September 15th, 2011 7:15 am

    You really love Sony, don’t you?

  124. Guest on September 15th, 2011 7:33 am

    How was the set ISO chosen for each camera?

  125. Stunko on September 15th, 2011 4:50 pm

    Not at all, Oli, in fact probably the opposite is true. However, it is nothing short of amazing to me how robust this company had gotten and the amazing cameras they are unleashing. Just within the past 12 months, you’ve got the NEX-FS100, PMW-F3, SRW-9000PL, and now the F65, all in the CineAlta category of D-film cameras. And having used the SLT Alpha 55 HDSLR from them, I have great hopes for the amazingly spec’d Alpha 65 and 77 models due out next month.

    I am not sure which other reputable, long estavblished camera company in the world could pull of such a feat at such amount of time, frankly. Remember, they also had to contend with the terrible earthquake’s aftermath. One should give credit where credity is due. 

  126. Stunko on September 15th, 2011 5:10 pm

    Re. the LENSES to be used on this new breed of high-end D-film camereas…. the biggest problem faced by those of us wanting the purchase or rent these newfangled Super 35mm sensor (app. 27mm diagonal) D-film cameras with PL mounts is the dire lens situation. There is precious few lens makers out there catering at this level, and what they have are routinely mundane glass at amazingly high prices. 

    Since the East German Jena Zeiss was gobbled up by the West German Carl Zeiss, the Zeiss lens prices have shot through the roof. No competition, see? If you want high contrast and super sharpness, Zeiss is your ticket, however Zeiss glass has to be softened quite a bit for facial close-ups. Very little by way of zooms from Zeiss. Cooke in the U.K. had gone lazy, they are doing primes only with a 6-month waiting list for orders, last I checked. But the famous “Cooke look” is really something else, admittedly. Angenieux of France is still delivering their famous zooms, but they have very few models and the spec’s are nothing to write home about.  I mean, this company had relativel light weight cinematic zoom lenses with 10x and 12x zoom ranges back in the late 1950s. But the 12x Angeniux zoom of today weights in at 11 kgs = 24 lbs and measures a whopping 0.44 meters in length, wow! Not every filmmakers’s cup opf tea, definitely.   

    So, that would basically leave the odd-ball Russian lens oddities, as well as the Fujinon and Canon cine lenses from Japan, which are basically the best values for the money in 35mm cine optics today. But altogether, with this avalanche of S35 CMOS sensor cameras coming out and becoming available (the latest add-ons being Red Epic and Sony F65 CineAlta), the major issue I can see is the specs, prices, and availability of the matching high-end cine optics from C. Zeiss, Cooke, Angenieux, Fujinon, and Canon. Compared to these ueber-priced cine-style lenses, the class-leading  35mm photo primes and zooms (Sigma, Tokina, Tamron, Rokinon, etc) begin to make a whole lot more sense, frankly. 

  127. Scott Lynch on September 16th, 2011 11:15 am

    The F65 will cost about $115k, and while I’m sure it is a great camera, I think it’s important to say that we have incredible tools that are available right now. If you understand how to use your camera properly and have a good story to tell you could use any cameras.

    The latest and greatest will always be better then whats out now, that obvious. What’s less obvious is that you don’t need the latest and greatest to make a good film, it has nothing to do with the camera and has everything to do with the people behind the camera.

    -Scott

  128. Scott Lynch on September 16th, 2011 12:10 pm

    The ISO was determined by the camera master, who choose the setting that gave the largest amount of high light and shadow dynamic range. For example in the scene with the candles and car, the ALEXA was shot at an ISO of 1280. At this setting the ALEXA records about 8 stops under and about 5 stops over 18% grey. 

    Because this was a dark scene, the camera master decided that they did not need as much highlight detail and choose an ISO to capture better shadow detail. ND Filters were then used to compensate for increased sensitivity. But the Alexa was an exception, as it had a very clean sensor that has very little noise.

    In that scene the girls face was lit to be properly exposed at an equivelent of ISO 640 at an f/2.0. Most of the cameras were shot at ISO 640 because any higher and they the noise levels would be too high.

    -Scott

  129. Tylermc on September 16th, 2011 2:36 pm

    what is the difference between .mov and AVCHD. and why is AVCHD an advantage?

  130. Stunko on September 16th, 2011 3:17 pm

    Scott, your infor is a bit dated, it seems. These are Band Pro’s prices:

    http://www.bandpro.com/products/new-products-and-specials/175-pre-order-your-sony-f65-camera-package-today.html

    $65,000 for the camera body, PL-mount, and ultrra-high rez OLED color vieewfinder. In other words, the Sony F65 CineAlta is a heckuva lot cheaper out of the box at only $65K than a similarly equipped Red Epic, not to mention most of the 2011 Shooput D-film cameras.

    If you buy two, that would come to $130,000, I suppose, unless you can negotiate a two-camera discount.

    F65 with OLED VF, Arri PL mount, field recorder, flash media, transfer station, etc. will still only cost $85.000. Click on the link and see.

    So, I have no clue where your $115,000 price popint is coming from — please, do tell. M<aybe Chicago pricing will be tens of thousands of bucks more than NYC and LA prices, I don’t rightly know. 

  131. Stunko on September 16th, 2011 3:28 pm

    Forgot to mention, if you do not need 4K resolution and 16-bit RAW recording immediately, you can get everything I mentioned for $65,000, and the attach a $1,000 to $2,500 video field recorder to the F65, and you will still be shooting glorious digital film with it. But the Sony SR-R4 digital recorder and the SRMemory cards are amazing, I handled the SRMemory card last year in NY, and it is about the size of a 2.5-inch SSD, but can store up to 1TB and write/read much faster than the fastest of today’s overprised solid state drives.

    Anyhow, the $20,000 extra for the field recorder, SRMemory card, and the data transfer station is very, very reasonably priced by Sony, I should think. But then again, you can start shooting with it for $65,000, plus whatever external video recorder you should attach to it. Amazing times we are living, definitely.  

  132. Ilona on September 16th, 2011 3:30 pm

    Fuji lens link:age Not Found
    Our apologies, the requested page was not found.

  133. Anonymous on September 17th, 2011 12:15 am

    Here is the Red Package for 58k, comes with quite a bit more than the 65k package for the sony. 
    EPIC-M Brain with titanium PL mountBomb EVF®5in Touchscreen LCDDSMC Side HandleDSMC Side SSD ModuleREDmote4x 128GB SSD cards4x REDVolt batteries and travel charger

  134. Stunko on September 17th, 2011 12:37 am

    I spotted these tiny pictures of somethings, I guess Red is still a tad short on written specs and whatnot.  I certainly would not be parting with US$58,000 looking at these tiny images. 

    Anyhow, you can comapre Sony as an audio-video-film-cinema  company to Panasonic, JVC, Arri, Panavision, etc.  And you can comapre Red as a company to Weisscam, Silicon Imaging, etc.  Just let us not compare Sony as a comonay to Red as a company.  That would really be an apple to orange comparison. 

  135. Stunko on September 17th, 2011 12:41 am

    Now with the Sony’s fantastic F65 Cine Alta camera announced and shipping in January, I don’t think that shooting digital and getting the coveted film look will require any great talent.  The camera will do that for you.  In fact, any contemporary digital film camera priced at $50,000 and above should be able to do a pretty decent knock-off of the film look. 

  136. tylermc on September 17th, 2011 10:40 am

    what is the advantage of recording in AVCHD compared to the current .mov my canon 60D uses?

  137. Scott Lynch on September 19th, 2011 5:03 pm

    Hi Tyler,
    Well “.mov” and AVCHD are really two different things. “.mov” is a file container while AVCHD is a codec. Codec’s are the algorithms that compresses the image, certain codecs like AVCHD are good at giving you small file sizes, but they throw away a lot of information to do that. There are other Codecs that save more information and give you larger files, but if you have to manipulate your images in post, you have a lot more control.

    File containers are sort of like boxes that generally hold many types of data, such as a video stream, audio stream and metadata. 

    -Scott

  138. Ian Mpherson on December 6th, 2011 3:50 pm

    They go head to head DSLR v Film.

  139. Kostas Metaxas on February 2nd, 2012 2:50 am

    Hi Z,
    Great comparison. You have to love the gloom-doom over the dslr vs pro cameras. Wake up guys. Two years ago there was NO CHOICE. Today, we have oodles. Prices are coming down. I’m looking forward to the BOMBSHELLS manufacturers will release this year. It’s a great time to be a cinematographer.
    Kostas Metaxas
    EXERO FILMS

  140. Adam Collier on March 8th, 2012 7:01 pm

    Great watch if you want to nerd out on tech