EDIT: HD-DSLRs Vs FILM…this time the gloves are off…

actually best if we put them back on as it’s taking place in Chicago and it’s ‘effin freezing there, especially after coming back from Dubai the other day!! Brrrrr

Organised by Zacuto, I am going over to their HQ where we will be pitting the Canon 1dmkiv, 5dmkII, 7D, Nikon D300s, Nikon D3s, Panasonic GH1 against….35mm Film.

Film is beautiful and I would love to shoot on it sometime, but most of us just can’t afford it so just how good do these cheap pretenders do at giving us that coveted “film look” without actually, like, being film!? Of  course we already know these cameras are not going to be able to match film in a number of areas, but can it beat film in any? How good can they look?

Shooting identical scenes with different conditions then taking them into a professional grade and then projecting them onto a 40 foot screen.

Should be a lot of FUN.

Despite my advancing years, I am one of the new breed of DPs who have only ever shot digital but have always craved the the look of film. We are getting so close now with the RED and other camera….it’s these crazy cheap HD-DSLRs that have taken us all by surprise and it these that ill be taking charge of.

I also really want to get the best out of the Nikons, so if anyone can recommend the optimum picture profile settings for them would be very appreciative. I hear the low light of the D3s is incredible…

Watch this blog for news and do come to the meet up on Monday if you can!

 

EDIT: All test are done and the results have been gobsmacking. Stay tuned for further information. I will share info as soon as I get permission to share 😉

76 comments

  1. This looks really interesting, I’m glad you’re going to the trouble of such a broad comparison. And particularly pleased you’re testing the GH1, too, as this is likely going to be my next camera for a variety of reasons.

  2. Interesting challenge!

    Guess a lot will depend on how you record; do you expose to the left (under) or to the right (over)?

    Going after the film-look would suggest expose to the left with a flat profile, but whenever you do color grading you tend to get noise easily in the darker parts of the image. Even with slight denoising, that don’t look good on a big screen…

    Read this great post on ProLost: http://prolost.com/blog/2010/1/26/color-correcting-canon-7d-footage.html In this post, Stu Maschwitz suggests to expose to the right as much as you can, without blowing anything out.

    1. >Stu Maschwitz suggests to expose to the right as much as you can, without blowing anything out.

      >without blowing anything out.

      This is the tricky part. If you blow important details out the shoot is lost. If you underexpose you get noise but the shoot maybe still usable. With RAW still images there is quite a latitude to recover highlights. Not with JPEGs and movies.

      Too bad that most HDSLRs don’t help much with good exposure.

      1. I know this is the tricky part, I am still trying to work it out myself.

        HDSLRs aren’t really good for anything with lots of repetitive detail (thankfully, faces are the exception), so I am guessing that going for a deep focus shot (F11 and down) really is risky in terms of noticeable aliasing patterns. I especially hate the unexpected moving moirés! On the other hand, if you use it for that really shallow DoF (F2.0 and up), you can achieve things regular videocams can only dream of.

        Anyone here with solutions to either avoid those moirés (e.g. diffusion filters?) or fix them in post (software)? I am open to suggestions…

      1. I presumed that there would be a 35mm film camera that would be used as the benchmark. I was wondering if this was going to be a full Hollywood job needing 4 people to operate it and costing a gazillion dollars or something more modest?

  3. Love the idea. Did something close to this last week? Truly amazed at the quality. Would love to see all these cameras side by side, along with the REDone. Wonder if the average person can even tell the difference on the big screen.

  4. I wish you were doing some fil-MING comparisons.
    Shooting stills will reveal latitude, etc. (Please use the same lenses on the film * digital cameras) but I’d really like to see a few hundred feet of movies shot with the vDSLRs and with a film camera.

    I mean, isn’t that the REAL test- to recreate the look of 35mm motion pictures (not stills)?

    I certainly hope you test more than close up face shots.

    Try and duplicate some of the shots on here:
    http://www.dvxuser.com/articles/article.php/20

    like slow zooms in and out on houses with shingles and bricks (as if you were to follow someone leaving or going toward the house). How do each of the cameras handle moire?

    Or waves on the beach, surf, as shown in the CMOS examples there.

    Or fluorescent lighting here:
    http://www.dvxuser.com/jason/CMOS-CCD

    Show sports action, where the subject moves quickly or the camera does. (rolling shutter) Strobe action (entertainment news coverage).

    Don’t just do the namby-pamby shots that we know all the cameras can easily accomplish. Do something CHALLENGING.

    Create scenarios with 8 stops difference, 11 stops difference, 15 stops or more. Inside a dark room and outside with a bright sunlit scene. Etc.

    I look forward to reading your results.

    1. I think the the point of the test is to show how well an HD DSLRS footage stacks up to film on the ol big screen and to help us achieve similar results, not so much to illustrate the well known shortcomings of the HD DSLR (which at this point everyone knows by now, unless you live under a rock).
      With that being said, please explain to me logically why a professional would take your advice and knowingly shoot crap footage to illustrate the very well known shortcomings, when Barry Green and others has already done it for us?

  5. This is gona be really interesting, i’m sure you have planned to shoot some cool studio set-up but it would also be great to see the differences outside under natural light.

  6. I wonder what kind of stock you guys have in mind using? I would def. go with a neutral one as in vision 3 250d and the new vision 3 500t, do a 1 light process. Expose on the dot, I’ve done this test with s16mm and I’m sad to say 5D/7D fans 16 blew it away. The latitude and color rendition is truly no match. What I did and would recommend doing, is underexpose 1 stop over expose a few stops plus shot in very hard lighting conditions. Brutalize the stock, see how far it can handle it.

    1. But one thing i have to ask about your test is, what was the cost difference between the two? how much for a 16mm camera compared to a dslr? I think for the money a dslr does exactly what i need it to do.

  7. Not to devalue this test but instead of HD-DSLRs vs. FILM a more appropriate title may be Apples vs. Oranges. These are two entirely different beasts. It will be interesting to see how this plays out and what opinions are formed.

  8. As an old cinematographer, I would be very interested in seeing the results. Did you ever consider super 16 mm? That would probably be cheaper as well as being the medium closest to HD.

    Are you testing on an Arri or Panaflex?

    What film stock are you using?

    More importantly, do you need help? (chuckling)

  9. Hey Philip…

    I’m a current excited filmmaker (and 7D owner) and have waited a long time to see the amazing revolution of digital cinematography develop to this point!.. (started the adventure with Quicktime 1.0 and a Quadra 660AV… first Mac or any computer with Audio and Video In and OUT!…)

    Enjoying your passion, your wonderful eye, and your discoveries!

    and also shot many miles of film…

    the DSLRs have much less latitude than modern negative film. No doubt about that!… but they are actually better than working with older film stocks or reversal film or Tri-X… the newest film stocks are pretty amazing…

    and due to compression the DSLRs will be blown away by any 100 – 500 ISO 35mm film…

    This is the gap the new RED Mysterium and Scarlet/Epic cameras will fill.

    This all being said… it all depends on DISTRIBUTION…

    for HD TV and WEB the DSLRs are fantastic… IMHO…

    I’ve been introducing clients to the whole thing and more and more they have heard about it…

    In the shoot-out how will the film be processed and projected?

    Given the new trend of shooting (and developing) neutral, will you show the neutral versions?… and how will you judge the grade-ability of the footage?…

    Cost, ease of use etc…

    Best Wishes!

  10. Having done a similar test between the Canon 5dmkII, Canon 7D, and film some months ago after which the film-maker chose to shoot his entire feature on the Canon 7D at 24fps, I’m familiar at how these tests go. Very rarely in favour of digital.

    I’ve even created a workflow of setting up TC for Canon clips, and editing on Avid (don’t ask!) and grading in Color and Resolve, and film out.

    But I’m very interested in your results. If you need help doing a filmout of your Canon files, do let me know.

  11. Movement in the frame, especially in low light and high velocity action, that’s where film should be better, it would be nice to see DSLR footage (talking heads) intercut with S16 film, with good grain reduction in post.
    Big question: would it be acceptable for HD broadcast?
    Lots of broadcasters shy away from S16 now because of compression problems in HD streaming…
    I shoot 12 min S16 film for about 200,- GBP, and next I will be going for a GH1, many thanks for all the info on blogs like these, but I am not ready to give up film yet…..

    1. I agree, any test needs to cover camera movement with deep focus. I also want to see how the GH1 and D3s compare to the 5D2 on a moving camera shot with lots of detail.

  12. I’ve recently started directing spots for advertising agencies with very good budgets. For an upcoming shoot, the agency producer suggested that I shoot on 35 and gave me the budget for it. I’m opting out, instead choosing a mixture of the 5D and the Red. Film blows digital out of the water in a lot of ways – latitude, color information, richness, and not to mention the aesthetic of film grain. These small digital cameras are not only great to shoot with because of their mobility, instant playback, 35mm lenses, real-time picture profiles etc. – Even if you have the budget to shoot film, if these cameras are allowing you to achieve the desired “look” (something very close to film), AND they are cheaper to rent, then you’re taking the money that would have gone to film stock and development and you’re taking it home as profit. Maybe that means I’m not a purist. I love these cameras for lots of reasons, not just there cheaper operating cost. Let’s face it, film is out of here.

  13. Hi Philip, shoould be a good test. Can’t help but wonder why you havent chosen to include at least a Sony EX3 with Letus 35mm adapter, or are your vidcams now redundant?

  14. I’m really looking forward the test results. Sure video – especially the DSLR’s – still is miles away from real film stocks. But nevertheless i like the kind of filmlook, you get with a DSLR for a much lower budget. There are 35mm cameras sold for such high prices that you could buy a house for it. And compared to the fistful of limitations you have to accept or work around in videofilming, the price for a Consumer/Prosumer HD-Cam or DSLR is absolutly fair. If a project’s budget doesn’t matters – i would work on 35mm or Super16 film and an Arri camera, but in every other case – and there are many of them – i would give video and DSLR’s a chance.

      1. Fascinating, can’t wait for all to be revealed. I agonised over the 5DMKII versus GH1 for a long time, but found I shoot better footage on the GH1. Neither are perfect, both great but find the GH1 aids creativity more. Lens choice is good too.

        I can even use a Zacuto Z-Finder as a lens on the GH1!!

  15. I have a 7D and I take video using an mRAW setting instead of RAW. It seems to me that I can record as long as I like that way. I have gotten files of over 3 GB. Is there any information on the effect on the quality of the video? I am having a difficult time finding any information on the effects of using sRAW or mRAW with my 7D

  16. How did the D3s do on dynamic range? I thank it has the largest photosites of the bunch… I saw a test on the old D3 where it was getting around 13 stops (whatever that means ?) when shooting RAW stills… dose any of that range carry over to video?

  17. Hi!

    Thanks for a truly helpful website. Looking forward to the Nikon tests. I have shoot with Nikon for 15 years now. But I couldn´t wait for the 1080p so now I bought the Canon 5D Mk II, amazing camera, but doesn´t have the feel and ergonomics of the Nikon. i also saw that Nikon coolpix now hade 1080 p so I guess it´s a matter of time before Nikon release Full HD DSLR. Hopefully.

    Best regards
    Anders

  18. Hi Phil,

    I look forward to your results. I was fortunate to do a workshop recently where they were comparing the RED to an ARRI film camera – VERY informative.
    I’m having a lot of fun with my 5d2s. Once punters realise the “look” of film has a LOT to do with the glass that is used, their “look” will improve. I’ve been experimenting with LEICA, HASSELBLAD [Zeiss T], Schneiders, etc, and the difference between GLASS is critical. It’s the GLASS that fleshes out the actor. Here is a test I filmed last night with 3 actors using the Meisner Repetition, and you can clearly see the Canon 5d2 is absolutely breathtaking with the Zeiss lens.
    And the EASE of working with it makes it, in my mind, takes it to another level of producing “ART” that isn’t possible with film:

    http://www.reel2reel.tv/3s6k/Meisner_FedSquare_Feb_2010.wmv

    Bestest.
    Kostas Metaxas
    http://www.3s6k.com

    1. Results will be announced on the 15th March. Sorry for the wait. If it was up to me would tell you all now! Lot of editing to be done and more shooting. We are taking the stuff to Hollywood and Skywalker Ranch for screenings.

      1. The introduction of the Rebel T2i will make the March 15 test results rather outdated –or at least incomplete. This new Canon seems (almost, or) EXACTLY in the same range of the 7D video-wise (not so in still photo performance)but at HALF the price!

  19. Just stumbled across this – sounds very interesting as I shoot on lots of formats. Well done for doing it/being involved.

    DSLRS are certainly great but pretty terrible when they are handheld with mid to long lenses. I have yet to see ONE film clip that is handheld that doesn’t suffer from at least some serious form of rolling shutter. I tried a 35mm film (Ridley Scott) style look on the D90 when it first came out but I had to shoot on a tripod and be SO careful (you can still see some rolling shutter and distorted faces + I had to do a big whip pans on a friends old sony video camera as it looked totally useless on a D90): http://www.benjamintodd.com/Cinematography/benjamin_todd_Nikon_D90.html

    If I were on a budget I think I’d rather shoot on an similar priced HVX and cheat the focus than shoot on a DSLR (at least for anything with any movement, i.e, most things) – very shallow d.o.f and jelly pictures equal low production values in my opinion.

    Well done for doing the test but PLEASE, can it not be locked off or with slow tracks and slow pans etc, who realistically shoots like that all the time unless it’s for a personal project? If you take the DSLRS off the tripods it will prove why people set 35mm as a bench mark for more reasons than just it’s latitude.

    Anyway, enough waffle – well done! – looking forward to seeing the results!!

    1. On the verge of purchasing a 5D or 7D (or possibly waiting to see if Nikon release anything that can compete) after having regularly rented EX1 + Letus DoF adapter + Nikon lenses for small unit documentaries (using Red on larger crew TVCs and MVs), but am wondering if Ben has a point.

      Philip, how do these HD-DSLRs handle hand-held or being balanced on a cinesaddle?

      And, as a fan of high-speed (for slow-mo) shouldn’t I wait for the Scarlet?

  20. Hello Phillip,
    Could it be the Nikon blow-up was superior and the Canon kids are retooling to advantage their sales, thus the delay? A great story, we cheered for Rocky, could this be a parallel universe?
    Norman

  21. Ok so where is this???
    Please clarify, why is it not here?
    Is it under “Film shot on 7D wins SxSW juried “best film” prize”???

    Thanks a lot…

  22. Please state where we can find the update to this blog?

    I cannot find it on your site. Can you advise with specifics? Title of the Blog, URL, etc?

    It’s very unclear at present where we can see the follow-up to this blog! Thanks.

  23. Cool – I got it now I think! I still would like to know who “won” it, cannot figure out which one to get – 550D, 7D, or 5 Mk II…

    Keep up the good work, cheers

Leave a Reply