Monday, January 9, 2012

Is this thing on?


Haven't posted to this blog in years..... about to build a new website and I need to brush up on how all this works. I think I may have picked the wrong horse. Blogger and Wordpress are the two competing templates I'm looking into.

Adding a picture here just to keep it interesting.

Friday, October 30, 2009

Anamorphic lenses with the vDSLRs


I've been shooting tests with various anamorphic lenses lately and am finding out a lot about how they work and how to use them best with a vDSLR camera like the canon 5D mk2. Anamorphic lenses use curved elements, and sometimes prisms to stretch an image in one dimension while keeping another the same. In the case of moving pictures, they have been used to shoot widescreen/panoramic films using regular 35mm film with a common aspect ratio. I won't go into all the tech and history here, but suffice to say starting in the late 1950s and on into the present, there are many "cinemascope" ratio movies that have had a certain look that translates to our eye as "cinematic," due to many of the unusual properties anamorphic lenses create, in addition to the obvious, which is that they make a picture much wider than it is tall.

Cinemascope, the dominant trade name and format most often was presented as a ratio of 1:2.35, which is close to the shape of a US dollar bill. Composing images in this format leads to a different "language" cinematically, since a single medium close up often leaves the frame feeling too center weighted or lonely in the sense of balancing what the eye finds pleasing. It's been fun seeing images in this new way - painting on a wider canvas.

Getting that look with a vDSLR means using anamorphic adapters, or else adapting cine lenses to the vDSLR, a project that some are experimenting with nowadays. It's a new field and a group effort to figure out all the details and possibilities, and user forums such as DVXuser.com, Cinema 5D and cinematography.com are abuzz with discussions of how to master this approach.

There were many predecessors to this work, however besides of course the original cinematographers and engineers who pioneered this work in the 1950s and 60s. After the widescreen look caught on, amateur film makers adapted the technology to 16mm and Super eight film making, both of which use a somewhat square format to begin with. The non-theatrical 16mm film distribution network quickly expanded to include means to present anamorphically squeezed film prints using simple, two-element anamorphic lenses that screwed into the regular projector lenses, and these are the most common lenses on the market - people started using them with 16mm film camera taking lenses - some were designed for this double duty, in fact and marketed to amateurs as a kit solution to making widescreen home movies. The most common brands were made in Japan and marketed under many different names with slight variations on the size of the rear element for use with different brands of projector lenses but the name Sankor or Singer seems to be the home company that produced them, in model names like 16F, 16C, 16D, etc that designate rear sizes. Fitting these lenses to SLR taking lenses is problematic but the 16D models have a size that can be adapted with screw in stet-up rings. Other need a special mount of some sort to be made. These lenses are small and vignette at focal lengths wider than around 90mm.

The favored model is the Iscorama, made in Germany by Isco and sold originally for the Practika camera which uses an M42 screw in mount. It was designed with SLR lenses in mind, and sold to still camera users as a means to shoot widescreen images, mostly for slide presentation I suppose where the user would then mount his lens onto his slide projector to unsqueeze for proper presentation. On the Iscorama lenses, the anamorphic element fits onto the front of a spherical "A" lens like an adapter. The model I have, the Isco 42 has a 1.5x horizontal squeeze factor. These lenses were made decades ago and are no longer available except used and sometimes NOS in dusty camera stores.

Isco made several versions of anamorphic adapters and the most common seem to be the Isco 36, the Isco 42 and the largest, the Isco 54. These numbers seem to refer to the size of the rear element's threaded coupling, and smaller version seems to vignette at wide focal lengths in comparison to the larger models, but the 54 is slightly unwieldy, users say because of it's large size and weight. The model I bought had a built-in A lens that seemed to be a 50mm focal length and it screws in quite tight to the anamorphic, having a very recessed and small front element. It is not a sharp or fast lens, and it also has a rear element that protrudes too far back into the camera body and interferes with the SLR mirror.

I bought some 52mm cheap UV filters and removed the glass in order to make a threaded spacer that allows me to fit the Isco 42 front anamorphic unit onto a range of my manual focus Nikkor lenses, such as a pancake 50mm f1.8, my 85mm f1.8, a 105mm f 2.0, etc. Since the anamorphic's job is to stretch the horizontal field of view but to leave the vertical the same, there is a range of primes that starts with the 50mm or so ( a zoom might get you a bit wider) but has the field of view (width that is) of a lens 1.5 times wider than that in practical terms. In other words, you have a horizontal view that is more like a 35mm lens shoots, but then a height that is still like a 50mm lens. This becomes your "wide" and the 85mm and 105mm are more like the normal lenses. My cheap 200mm zoom lens makes a slight telephoto, akin to a 135mm when shooting normal stills. It's a good enough basic set of lenses to shoot a short film with, and I'm starting to use it in my everyday work but the learning curve is still shifting as I figure out all the peculiarities, of which there are many.

The Iscorama 42 was sold in various generations and sometimes is found by itself, without the dedicated M42 50mm "A" lens. The 54 was often used with a beaulieu super eight zoom lens and can be found in some camera packages that are hitting the used market as Super eight continues it's noble but sad decline.

The Sankor lenses are harder to focus, and give a less sharp image but have their good sides too, chief of which is that they are inexpensive and easy to find. If all you had was a sankor, it would be advisable to cheat and shoot wide shots by simply cropping a spherical lens in post and hoping the audience doesn't notice the difference in the overall look, but you wouldn't fool everyone. Mixing the two might be worse than simply shooting a whole film project by cropping the top and bottom of your shots.

This post is just an introduction to the subject, and I'll write more as I continue to learna nd experiment and have more images to share.


Saturday, May 9, 2009

We have a winner.... fridays are for celebrating...

here's a tune to hum as you read this....


Looks like we have a winning solution, and the basis of a good workflow in place....

The missing piece was a third party program called Remaster, part of Neo Scene HD, put out by Cineform.   

Here's a step by step description of what we did.  

We need to test more clips and examine the frame blending very carefully, but for now the basics seem to be in place.  Now we need to test and refine to be sure it's really all working.  But it is certainly a relief compared to where we were before... thanks to everyone who has helped and offered help.  Each one of you has contributed, even if it was just psychic good vibes..  most did a lot more than that, patiently explaining terms and theories.  

-----------------------------------------------------------------

workflow to take 30p files from camera into 24p and maintain audio sync and good color (no crushed blacks)



--------------------------------------------------------------------------------------------------

import clips to Hard Drive (make backups!)


open the desired .mov file with Remaster (from Neo HD, by Cineform) 


select the clip in Remaster


go to parameters, the little triangle at the bottom (before you convert it) 


select tab for output > choose Pro Res HQ


there you can choose the quality, including Film 1 and Film 2 (what are these?  likely you need 4:4:2 material for these to be active, canon d5 is 4:2:2) 


create a new file, don't overwrite the old file


convert one item or the batch of files you load up


select convert, choose where to put the file


name the file filename-prores30.mov (so that you know that this is the 30fps file) - then make a duplicate of that file, and call it filename-prores24.mov, open THAT ONE in cinema tools


(this is where you set the naming convention)


then you open cinema tools, it comes up with a dialog box, "Create new data base"  - continue 


next dialog box, just cancel


went to file> open clip


chose your filename-prores24.mov clip (this is the one we're converting to 24 fps), and it opens the clip


click the bottom button to conform


opens a dialog box that gives you option of changing the frame rate to 23.98


hit conform, and it is done instantly


close cinema tools


then open compressor, and bring in the prores24clip


apply Pro res Apple 4:2:2 setting to the clip


this is where it gets tricky


select that setting, look in the inspector window  - (read Denver's method and go from there) 



you have open



create a new time line in FC change it to 23,98

drag the 30 frame clip into the timeline

say NO when it asks you to change it to 30fps, 

go to the last frame of audio and get that TC value

now go back to compressor

click on the settings that you have applied to the clip

select the settings that you have applied to the clip and go to the inspector window

select the frame controls tab

click the star next to frame controls and turn frame controls on

set rate conversion to "best" 

"set duration" to the length of the clip you got in final frame of your audio - the duration of the clip you put into the 24p time line

the hour setting should be 00 not 01


look for a value something like 79.13%  - it wont be the same each time


hit submit and wait to render

aprox one hour per minute




Friday, May 8, 2009

side issue for FCP people, making canon D5 m2 generate color bars

One challenge I have not been able to master is calibrating a monitor from the camera.  The camera doesn't generate color bars, but in theory one could get a video of bars from a DVD or some such source and place it onto the Compact flash card from a computer/ card reader, and then play back that file thru the camera to the monitor.  Here is a discussion about some of that topic....

http://www.cinema5d.com/viewtopic.php?f=24&t=2180

I'm not even sure if you can put a jpeg back onto a CF card and get that to play, much less a .mov clip but I don't know enough to know why it's hard.  I should have paid more attention in video school.....

Tuesday, May 5, 2009

canon 5D Mk2 testing continues

-----------------------------------------------------

this is a temp blog post to address some global issues I'm working on regarding testing the canon d5 mk2 camera.  We've shot some tests and have seen a film-out to 35mm and are encouraged to continue learning the best way to get the great images this camera is capable of making from the native media files and on to 35mm film.  

I've written a lot of the same information various ways so don't be offended it I repeat myself.  SOme of this is just to keep it straight in my own head.   There are quite a few interested parties willing to chime in and help us solve this problem, or at least refine the shape of the stumbling blocks.....  any input is appreciated, but PLEASE read the links below before making any suggestions.  

Ideally, we will gather a few good folks together and each work on one facet of this puzzle and make headway together, and share what we learn and compare notes without having to deal with a bunch of newbie stuff or duplicating efforts.  


----------------------------------------------------

OUR GOALS FOR POST  TESTS

Searching for a workflow that begins with canon D5 Mk II (w manual iris Nikkor glass) and ends with 24p HD with no crushed blacks.  Hopefully includes Final Cut Pro and works on a mac but willing to explore options.  Opinions appreciated, actual hands-on experience preferred!  

This post at Neo Scene Insider seems to be a great place to start, and if you haven't already read it,  or at least glanced at it to know what the issues are please don't make suggestions.  The camera records info in the shadow detail areas that are usually clipped by most all Non linear editors - Avid and FCP and others.  Retaining this shadow detail is essential, and any workflow solution has to begin with keeping this dynamic range visible.  


relevant points to consider, terms that have been floated, etc. (ie, stuff I wish I understood better and need to test)

different ways to home-convert 30p to 24p - cinetools, compressor, etc
when to do these conversions in a long form project's edit workflow
when and how to color correct
how to get best media for possible high end color work in a full service digital suite
how to do color correction and dynamic exposure corrections ourselves
possible use of Apple's "Color" portion of FCP and its compatibility with neoscene (or other programs that help save blacks)
Apple Color and Quicktime 7.6 vs neo scene
CoreAVC
ffMpeg
Neo Scene HD
Cineform's reputation and tech support in general
bringing Canon D5 mkII clips into ten bit color space
YUV vrs RGB color
broadcast color standards and how they do and don't apply to us

To restate goals a bit:  the basic idea is to shoot a long form project, either doc or narrative feature with canon 5D mk II and do as much editing and rendering and color correction ourselves in a home-built edit studio, with the goal of having a festival-ready project in 24p HD that is very close to ready for x-fer to 35mm film out.  One possible path would be to do no color correction or frame rate conversion at all, but simply work to retain as much info as possible along the path in order to bring the media to a full service digital post facility prior to film-out - while creating a parallel workflow path that builds a festival audience ready HD version that has our own home studio 30p to 24p conversions and basic color correction for shot-to-shot matching, rough color grading etc.  

I am not a post production person, just a motivated creative type who is searching for solutions and beginning to research options on my own.  I wish I could say I understood all this better already, but that is why I am seeking advice.  

thanks in advance.

-----------------------------------------------------------------------

SHORT VERSION

we want to make a sundance-wannabe film with this d5 mk2 DSLR camera.  We've already shot some tests and seen the footage converted to 24p and transferred out to film.  It looks great and the camera works well, after a fashion.  

The issues we can improve on are two

1 - we need to retain the color quality in our workflow - camera very easily creates files that get the blacks crushed and highlights clipped in non linear editors.  We want to stop that from happening, and bring the best color and contrast into final clor correction prior to 35mm film out

there is some discussion of bringing the media into "color" portion of FCP in order to work with the 8 bit color in a 10 bit space, and how this helps even though it doesn't create any new information - is simply leaves room for it to be manipulated and gives the best look but it only works with certain types of files

2 - we will have to convert 30 p to 24p, and it seems like the best method is to use FCP and Cinema Tools to interpolate the data using a method detailed below by Denver Riddle.  but we dont knwo the limits of this, or if it is truly the best, and when to do it in the day to day workflow of a feature production on location.  

-----------------------------------------------------------------------------

great thread about 30p to 24p conversions

listen to this guy DENVER, aka denverkr   he knows what he is talking about

http://www.cinema5d.com/viewtopic.php?f=13&t=1201

THIS SEEMS TO BE THE BEST METHOD

can anyone improve on this?

----------------------------------------------------------------------------



---------------------------------------------------------

color space notes

here's the best intel we're read about color space, and what the camera is putting out:


read the link he starts out with, which is here


but follow the updates....  he corrects his info twice.  

In any case, there is a lot to learn here about what the camera is spitting out as media.  

-------------------------------------------------------

frame rate 30/ 29.97 NOTES


There is also a new note about how the camera SAYS it is 30p but is in reality 29.97, which also seems to be confirmed. 


-----------------------------------------------------------------

how to get to our test clips, and where to post your edited clips or examples


contact me via email to get info on how to reach our server for the first time, and log in to get an applet file that will get you to my server where you can download and upload clips

 clarkleewalker1(at)mac.com

Once on the server, then look for the little red cone icon that says "logon_ to_project" and copy that file to your hard drive somewhere, like on your desktop

Use that method to log on next time, as the password will change soon.  

Click on it and the mac should connect to the server and appear on your desktop.  There are a few clips there already, more to come. 

Thanks again for your help.  


----------------------

end

Sunday, May 3, 2009

CANON D5 to 35mm test shoot - 27-APR 09 - 30 APR 09


CANON D5 to 35mm test shoot  - 27-APR 09 - 30 APR 09


I worked all this last week shooting tests in Los Angeles for an experienced feature film director with the Canon D5 Mk2 DSLR shooting HD video for transfer to 35mm film.  The director wants to shoot part or all of an upcoming indie feature using the camera, for various reasons, both economic and aesthetic.  Yesterday we saw the footage at a lab in Hollywood, and learned a lot about the advantages and disadvantages to using this camera for feature film production and long-length documentary where the final product is film.  The package is owned by a close friend, a director who has made many features, and it's been exciting to see the delight in his eyes when we look at the footage we are getting.  He understands CINEMA like Rodin understood bronze, and he is practically giggling like a schoolgirl every time we set up a shot with this extremely minimal package.  


some stills and video can be seen here


http://www.flickr.com/photos/jeandodge/sets/72157617523029177/show/


The DoP is J.C, from the EU,  who has shot dozens and dozens of international corporate ads, and plenty of indie features, European TV MOWs, etc and like me comes from a 35mm and 16mm film background but JC also has experience with other HD camera platforms, including the F900, aka the CineAlta, better known as "that thing they shot the new Star Wars with."  He has also shot cell phone commercials and other regional advertising spots already with the Canon D5 Mk2, but only with canon lenses in various auto-iris modes in 30p for broadcast in Europe.  


None of us are camcorder/DVX style film makers, but of course as professionals we've seen and worked with the DVX style equipment as mentors to students, or by having an EPK crew buttonhole us, etc. and occasionally shooting side projects of our own with hastily borrowed equipment.  Nor are we experts in Final Cut Pro, Avid, color correction suite work with Pango etc although once again we have all worked many times with this stuff in the capacity of director/DP/producer etc.  Sometimes I wish I had more experience with "video," but I've never liked it.  Digital cinematography however, is finally getting interesting.  


(More about our specific workflow below, but first I'll try to get to the good news/bad news first impressions stuff. ) 


The bottom line conclusions we reached are not surprising to those who have been following the tech blogs and user forums.  The camera makes nice pictures very easily but is not user-friendly for film makers hoping to finish on 35mm.  However, the low-light quality and easy form factor advantages  (plus minimal cost when compared to say, super 16mm or Red One production)  outweigh the temptation to completely write the thing off and wait for a professional version to emerge.  The camera is nearly fatally crippled by not having the ability to acquire media in 24p - it only shoots 30p and therefore prior to film out the footage must be compressed by some means into 24p.  This step lowers the quality - more on how and why and what to do about that later, or elsewhere - it just does, adding contrast and compromising color rendition, and remains the single most major hurdle to clear.  But what we learned  after our simple tests, in the screening room looking at film is that there is plenty of room to account for that crippling factor as is, and that additionally, by going to film many of the original HD related problems like aliasing and solarization are masked effectively enough for the jury.  


What am I saying?  I'm saying you can make a feature film with this camera and to a vast majority of viewers, even trained eyes it will look like it was shot on 35mm film if you do it right and are careful at each stage of the process, and don't mind standing on your head and reciting the greek alphabet backwards before starting to shoot each shot.  It won't be as good, and it won't win any awards probably either but you never know.  I'd give it a B-plus when I thought it was going to be a C-minus going in.  Hopefully that B-plus can be raised to an A minus soon, with advances on post workflow. 


Here's a bit about our workflow:


Camera

Canon D5 Mk2 DSLR


support / follow focus

Zacuto rods, baseplates follow focus, hand held grip


on board monitor

7" HD monitor  - 720 lines


lenses

Nikkor still camera lenses, manual focus Auto Indexed, circa 1968.  


20mm Nikkor Auto 1:3.5   No. 476940

24mm  Nikkor Auto 1:2.8  No.337571

28mm Nikkor  Auto 1:2  No.317324

35mm NIKKOR -N Auto 1:1.4 35mm Nippon Kogaku Japan No. 350150

50mm Nikkor Auto 1:1.2  No. 390746

85mm Nikkor Auto 1.1.4  No.  202947

105mm Nikkor Auto 1:1.2.5  No.  525188

135mm NIKKOR -Q Auto 1:2.8 135mm Nippon Kogaku Japan No.192131

180mm Nikkor Auto 1:2.8  No.  330918


filters

ND 3, ND 8


lighting package  (essentially none) 

daylight

household tungsten bulbs

christmas lights

mixed street lighting

car headlights


grip package

french flag

car window clamp with a tiny ball head

no bounce/ fill, etc


THE SHOOT


We pretty much shot what we saw, and went with the lighting we encountered in the real world as a baseline for what to transfer to film.  We only had a test budget to look at about 200 feet of film and decided on a philosophy of "let's see what it looks like straight out of the box, with no help along the way."  Helping us to make this decision was the fact that shortly after arriving from London on sunday night, the DP got a call for a paying gig he couldn't refuse - assignment in North Africa - Casablanca to be exact.  This moved up our original lab deadline by two days, and meant we had to shoot all of our tests on Day One, edit on Day Two and turn in the film by morning of Day Three.  Goodbye prep, so much for scouting, forget about the charts and MTF tables I brought.  Rev it up and Go.  


The bulk of the test shot was conducted in director's kitchen/ living room less than four paces from where the cardboard shipping containers were piled.  Both Josep and I arrived in Los Angeles less than 24 hours before shooting began.  I had read up on the Canon but never held one in my hands.  The airlines had lost Josep's luggage sunday night.  I had come in on Monday morning, which was my birthday, and stopped in for a quick beer at Che Jay on the beach in Dogtown as is my habit.  We were tired, we stank, and the gear was still covered in styrofoam peanuts and we'd had zero discussion about what would be best to film first, and how.  So naturally we started shooting right away.  


SCENE 101 DAY INT:  LIGHTING W PRACTICALS, BEAUTY SHOTS of ACTORS and MATCHING  


We shot a short dramatic scene in the kitchen with whomever was visiting the house at the moment.   Two young men talking, as one gets ready to cook some eggs.  Master shot, reverse shot, a crossing/ OTS shot then on to a couple of nice looking long lens close ups and a cutaway we didn't end up using.  The idea was to look at something with an unprejudiced eye that seemed like it was a piece of a movie, and to leave out all worries about what the tech involved.  Do we "buy it" as cinema or not?  


CONCLUSIONS


Here the results were obvious, once we screened the footage - many issues, but it looks "good enough to pass," and it can look a whole lot better if we continue to make progress with preparations and research, consultation, testing, etc.   Shot to shot matching, we can cut from take one of a close up to a reverse and then back to take two of a close up and the shots match, etc.  No one was 100% sure going in that even this simple part would work, so it was a great relief even to see the basics work out okay.  The camera can make a consistent look if you lock the shutter/ISO combo and control the iris manually.  (This is Chapter One of the "stand on your head and recite the greek alphabet" factor I mentioned earlier, as the camera has no manual controls and has to be tricked into shooting in a consistent fashion from shot to shot.)


So, yes, the footage matches, and the transfer seems fine as far as HD to film look is concerned.  Bad news is that the contrast added by going to film is not pleasing on top of the contrast that was added both by the camera's tendency to crush blacks and the added degradation inserted by the 24p-to-30p compression.  It looked like poorly processed negative  - greys went to black, no shadow detail in zones 1 and 2.  But we kind of knew that going in.  Hopefully, that can be dealt with soon as we experiment with different ways to get in and out of Final Cut Pro.  When playing back directly from the camera into a plasma HD screen thru the HDMI cable, it's much better, of course.  If the goal is to keep that look thru film out, we've got a long ways to go and may never get there.  But we have to get closer... and we should be able to.  This is just a first test for us.  


Our workflow with 30p-to-24p was good at all the motion stuff but not so good at the color rendition part.  This remains the biggest challenge.  Shooting 24p with this camera is going to be a dream of many...  it looks like with the right gear we can at least recover the crushed blacks prior to the 24p conversion, so that's half of that battle.  Keep praying for a firmware hack, film fans.  


SCENE 201 DAY FOR NIGHT, ALIASING and MOON LIGHT


Next, we drove out to a lake at sunset and shot some day-for-night shots of sunlight rippling across the water in a backlit east-facing-west setup, something that would affect the scheduling of the director's intended project.  It too looked good but there was aliasing (zebra stripes on what should be diagonal white lines, etc) visible on the handrail of a boat dock which we didn't like.  We were hoping to shoot the sunset shot as day-for-night and then eat dinner while it got fully dark and try some moonlit exteriors but sadly the phase of the moon was not anywhere near full.  Instead we ate Mexican food and played with the camera in the restaurant marveling at how well it shoots in typical lighting environments like this.  We rolled camera on some much gorgeous B-Roll everywhere we went that eventually we got tired of even reviewing it later.  Most of the time the camera makes great, great 30p HD video. (The challenges are elsewhere, and some are still unknown, like infared pollution issues and workflows in post, mostly relating to the crushed blacks and 30p-24p paths. more on that below)   


CONCLUSIONS


The day-for-night scene was a delight. The footage looked alright since it already had crushed blacks and the film grain smoothed out the aliasing on the handrail.  Encouraging.  Both the camera and the lens seems to have fairly smooth highlights that carried through all the compression stages just fine.  



SC 301   LOW LIGHT, BLACKS and extreme DoF issues - SHOOTING IN THE DARK


Having dumped the moon we headed to an edit bay to try some workflows we'd researched regarding changing 30p to 24, in preparation for transfer to film.   Along the way I shot a few shots in the car of the director lit only by the tail lights of passing cars, at 3200 ISO using the 50mm at f 1.2 - aka "WFO" - the lowest possible light level where the camera records an image with the fastest lens in the kit. What's amazing about the Canon D5 is that it doesn't add "gain" to shots like this.  Blacks are black, and nothing is grainy or poor quality.  If there is enough light to expose an image, the image looks good.  In this case we were picking the worst of the worst - red tungsten lighting that probably metered two stops under even the level we were shooting at, in a moving car, handheld with a lens at a focal length you wouldn't want to shoot handheld with in 35mm.  (I wish I had tried the same shot with the f 1.4 85mm, without the support rods and FF - I think I could pull it off.  The camera handles better in some situations without the HH &FF support, just as an SLR is easier to take on a roller coaster than an Arri SR would be.  


CONCLUSIONS


In 30p HD video, the shot begins with focus on the leading car, a huge semi with a lot of puppy lights and then pans while it racks to the side de of the passenger's face.  When the camera got to the passenger, I was still fishing for the proper focus - shooting without focus marks and operating myself - and so there was a moment when the only thing visible in frame was a blurred face-skip.  The camera did some very digital looking things as it struggled to make sense of the shot, but soon righted the look when it had sharp focus on the face.  This was a noticeable video artifact and seemingly rendered the shot unusable, interesting as it was.  30p to 24p conversions actually smoothed some of this out, but it was still an obvious digital shot.  


But, low and behold, the film grain smoothed out the 10-12 problem frames and made it look more like a varite' shot would look...  just a slow operator bringing a shot into focus.  As troublesome as the crushed blacks are in the D5 Mk2, they worked in our favor here again.  The shot looked like something out of Alien 3 instead.  Blacks STAY black in the camera and there was almost no gain visible.  The process passed with flying colors on this one...


Maybe not THIS extreme, but at the extreme low light end of shooting this camera is the world class leader at the moment.  This camera does things no other camera can do, period when it comes to low light cinema look shooting.  Amazing stuff and to really take advantage of this camera one would set the story entirely in a coal mine or some such...  Again, I think that there will be six feature docs at sundance next january that shoot using this camera.  


I'd hate to be a festival programmer in 2010.  There are going to be a lot of fairly good-looking films made by inexperienced storytellers.  


POST PRODUCTION, TAKE ONE


Hoping to save time, we had actually stopped by a college where the director moonlights as a lecturer on the way back from our dusk-day-for-night shot at the lake.   Unfortunately the school's equipment wasn't really up to snuff and so we had to wait until the morning to try the 30p-to-24p workflow we had researched.  We were hoping to use "Color" a part of Final Cut Pro to try to recover the crushed blacks everyone is struggling with, and still have time to shoot some additional tests in the morning before turning in media to the lab.  



scene 401  PANNING, DAY INTERIOR, SCREEN DOOR EFFECT AND DYNAMIC EXPOSURE CHANGES IN COLOR CORRECTION STAGE


In the morning we did a few day interior shots at noonish light to test both dynamic color correction and panning issues.  Just very slow panning shots of the living room facing large french doors with a 35mm f2.0lens, with the camera's ISO/shutter combination feature deliberately unlocked.  (For the test, this 35mm was our wide prime, since the wider ones were still in the shop for CLA and some milling related to rear area clearance issues. See below for more about wide lenses.)   The scene is static, a 90 degree pan of the room that is dark but backlit from bright ambient outside the wall of glass french doors.  The old f2-f22 split shot, a classic case of horrible lighting. 


We were looking for "judder" issues, instances where we have seen and heard that the compression of the camera causes a panning image to stutter and jump somewhat.  If you don't see the "judder" on the monitor as you are shooting, it doesn't seem to be there in playback... but this is not the end of this line of testing.  But FWIW, it looked okay in that department. Instead, what we found in our intentionally under lit and still harshly backlit-from blown out windows pan shot was that the camera doesn't like dull walls and neutral areas of the frame, especially underlit areas.  This is where the camera's native codec decides to hide all the compression.  Here we found solarization, and also the dreaded "Screen Door" effect in evidence.  Large blank areas take on the form of a slowly moving, dancing/floaty paint-by-numbers painting - that's what I mean by solarization.  (And by "dancing/floaty" I don't mean jello-cam - another issue altogether not a factor here) Screen Door effect has to do with the Bayer pattern of the CMOS sensor, and it's not pretty when it arrives.  It looks like hundreds of vertical bars of rust colored shit that doesn't move with the image.  Imagine you were looking thru the camera and the ground glass was streaked with rust drippings....  yuk.  It looks like gain, grain and a pain.  This was some of the worst footage we saw going in to color correction - it was rated unusable by the director, the DP, the editor and myself.   We did less than one hour of color correction in a full service suite and spent maybe ten minutes creating a moving dynamic change to cover the exposure change the camera performed when we finally panned off all the french windows and were shooting the interior wall.  The colorist also did what he could to hide the screen door effect by  applying some selective color filtration, similar to notching out a high pitched whine in a sound track.  


CONCLUSIONS


Interestingly enough, after going out to film both the screen door and the solarization were effectively masked by the film grain.  The shot was still an ugly shot, but it had a lot less problems that made it look digital.  Instead, it looked like a bad film shot, which was a great victory.  The dynamic (means it happens in shot) exposure change worked well, too and wasn't as noticeable as an iris pull would have been.  The Canon D5 is incapable of changing shutter speed manually, which is how one would correct a shot like this with a Panaflex camera if re-lighting was not possible for whatever reason. It does however have an auto function for doing exactly this, and if you disable the camera's ability to auto iris, it does the riding of exposure by changing a combination of both ISO and shutter speed, with obvious shifts visible in shot.  


WIDE LENSES  :  a note about wide lenses, older nikons and cinema


The widest we had at the time, since our 20mm, 24mm and 28mm Nikkors turned out to have a rear element protecting flange that needed to be milled off in order to clear not the mirror, but a small part of the internal housing surrounding the mirror.  These flanges also protect against stray light bouncing across the rear element in nikon bodies so we'll have to test those thoroughly when we get them back from the shop.  It's likely that these wider lenses of this generation of Nikkors will have issues when used for cine - they are not as good as the newer rectilinear lenses at correcting for barrel distortion, and limit your ability to pan without drawing attention to the flaw.  As still lenses, they are fine and can hide this effect, but as cine they are definitely sub-par.  We may end up shooting "twisted" canon glass at the wide end of our package, or possibly Leicas.  The testing continues....  


SCENE 501 POGO CAM, SHUTTER RATE and DAY EXTERIOR


Day two was set for  a meeting with the lab, some basic editing of clips, and getting the 30p material into 24p somehow.  While the director made breakfast and the DoP changed his flight back to London, I put the body only onto a still camera tripod, raised the post and spread the shortened legs slightly and made an impromptu "pogo cam" just to see what the weight was like.  I know people are using steadycam junior and other gimble-only stabilization gizmos with this camera and wanted to give it a try.  I'm curious if the weight could be kept to a minimum and still achieve wireless follow focus capability, so you could shoot without a spring-arm but still have the full focus capability.  But time was short as we were off to the lab right after breakfast.  


Feeling brave, and only having one victim handy I put a 50mm lens up to see if I could keep a moving actor in a pleasing single MCU on a shot that began on the upstairs deck and when down the staircase to the street.  Tough enough shot for even an experienced crew, but a simple test would tell me what I wanted to know.  I did one take, with predictable results but was at least able to wing it somewhat.  I had hoped to be able to rig the Zacuto monitor magic arm to the bottom of my monopod to use as both a counterweight and a monitor, making an impromptu sled but didn't have a standard "magic arm" handy.  Zacuto's arm is sized for support rods, not 58' baby spuds.  I figured that with a mafer and a baby spud I could have rigged a monopod into a facsimile steadycam sled easily.  For a gimbel one simply uses your hand.  Instead had to just set the frame with the camera's  on-board screen at the start of the shot and try to eyeball it by feel and experience instead.  For exposure I deliberately let the camera overcrank the shutter - it was probably around 1/100th a sec or 125th a sec from the looks of it later, maybe even 160th.  My actor walked towards camera and we kept a five foot interval as I descended the stairs forward, shooting essentially behind my back as a steadycam operator would - half turned at the waist but feet oriented forward for safety on the steps.   Once on the street level, I let him pass in profile as I bananna'd out to allow the old " cowboy walks off into the sunset" ending.   It was a one take wonder - as in, I wonder why we only did one take.... but again I was just playing around, and we had to leave for editing immediately after.  It proves the point to me that you can shoot with the rig without a spring arm if you want to get the right rig and rehearse it at least once before you shoot!  In hindsight I didn't do so bad, once I got halfway down the steps - keeping the actor in frame and not giving "max headroom."   I wanted to try a second take but there was no time, because right after that we all ran to the lab for a preliminary meeting.  In the car, we were talking about steadycam systems and such and so I showed the little clip I had just done to the DP and director as a way to illustrate an example, making apologies and excuses all the while for the shoddy test shot.  Imagine my horror when the DP and director both said - put that shot into the test folder!   In reality, it served as our day exterior shot and also as a test to see what the 30p to 24p would do with fast shutter speeds.  The steadycam part was superfluous, but now I knew I had to see it 40 feet tall at the Lab along with assorted producers, colorists and the like.  But we were in a hurry and under the gun to get some material to the lab, so c'est la guerre.   


CONCLUSIONS


Again, the film smoothed out some of the digital-ness but certainly not so much that it didn't look like Saving Private Ryan. My pogo cam skills looked good enough for the civilians to assume we knew what we were doing... so I didn't have to hide under the seat.  This camera will "fly'" but it remains to be seem what the best rig will be.  I'm hoping to do some more tests sans spring arm soon.  Of course the vest-and-arm gives you a lot more rest and stability, but like with so many other things this camera is showing us, less is often more.  Odd rigs like the ones that resemble steering wheels, etc need to be tested as well.  


more on the FOCUS ISSUES


I'd been shooting at f2.8 for most everything we tested, since that's what we like in film usually but the camera's full frame sensor is larger than a super 35mm negative.  Actually an f4 and maybe even a 4-5.6 split seems to be giving us the look we are used to in cine at f2.8.  That's good news for the focus puller.  The nikkors are a pain in the arse to pull focus on.  I think we'll end up with a set of focus rings calibrated one for each lens and trust those instead of trying to look at witness marks atop the camera.  Of course still camera lenses have witness marks on top, while cine lenses have the marks at 3 o'clock where the focus puller can see them easily.  


This is a huge area where camera prep has to be explored, refined and tested. Our 48hr scramble barely scratched the surface of this.  Shooting with Nikkors is not like shooting with Primos... but it is somewhat like shooting with Zeiss super speeds - the ergonomics are less than ideal, but the results ought to be good enough in skilled hands.  Only time will tell.  


SCENE 601   - CAR SHOT HOLDING DOF


After the lab consult the director also wanted to do one last test, which was a couple of simple car shots holding two actors in focus that would give us an idea of what deep focus looks like, and also a sense of the motion blur in a moving car - something that always varies from camera system to camera system even in film, due to the different shutter angles you encounter with different cameras.   It seems the Canon D5 camera shoots two frame rates we film makers can deal with - one is 1/50th of a second at 30p, and the other, more common is 1/30 a sec at 30p, which the camera defaults to at anything above 200 ISO.   Ideally, if it shot 1/60th at 30p that would give us a look of a 180 degree shutter,  but what we get, both 1/50th and 1/30 is not too much of a problem.  Motion blur is not excessively cheap looking in low light.  


In real world conditions either of those rates usually looks good, very much like we are used to seeing in film.  Again, we only had a 35mm lens to use, where a 20mm or a 24mm would probably have been better but what we shot looked good.  We did two "hostess tray" style shots looking across the actors from the window - one looking across the passenger and the other looking across the driver, each holding focus on both guys.  With the 35mm, we had to stop down all the way to f8 and the deep background was going to be in focus, not the desired effect for a shot like this.  


Just from playing with the thing in the car, we knew we were going to like the look but had some speculation regarding the overall look.  Video is notoriously easy to spot by the unwanted depth of field, where film almost always is employed in a manner to direct your eye towards only that which the director wants you to see - which will be what is best in focus, while the background is soft focus.  Here we had two compromises - one being not holding the f stop down, and two not having the exact shutter rate we thought would be best.  Would this be a situation where the camera was not right?  Inquiring minds need to know.  


We also threw in some tight checkered shirts and some silky scarves to check for Moire patterns or solarization that often appear with "problem" wardrobe.  


CONCLUSIONS


The good news is that when the car is moving, the action outside the window is blurred due to the speed of the car.  The camera is so light and small it's much easier to rig and less obstructive to place slightly outboard - no more, say, than an extended side view mirror.  This means for indie film productions we'll skip the process trailer and the hostess tray and the speedrail rigs and stick with small limpet mounts (suction cups) and magic arms, etc.  


The looks we get from shooting inside the car have all been great - you can drive at freeway speeds and it seems very natural. On film often 35mph looks like 65mph, esp hostess tray stuff (looking across profile on driver)  With this camera 40mph is 40mph, and 60 is 60.   If that's a good thing for dramatic scenes, maybe it's a loss for stunt driving scenes, but I don't  see that as a bad trade off, and there are ways to make it seem faster, I'd guess - long lenses, tight cuts, wide angles close to the tarmac, etc.  


Even the slightly slower shutter seemed to help mask the depth.  When the camera comes to a stop the background is sharp at f8 but it's likely you would cut around that by selecting another angle.  


On film it looked like it does in HD.  No worries there.    Also the wardrobe seemed fine.  In HD, we could see some solarization on the shadowed headrests behind the actors.  Film smoothed that out, but we made a note that these solarization issues come up in bland areas and are usually most visible in large blank spaces.  If I were to build a living room set, I'd choose wallpaper over paint...  



POST PRODUCTION, TAKE TWO


too tired to blog all this post tech stuff - not my area of expertise anyway - but the short version is this


On an apple computer running OSX,  with Final Cut Pro we edited down the clips and batch converted the footage from 30p to 24p and did not do any color correction in FCP.  We sued a method that uses cinetools, not compressor.  The conversion adds contrast and in addition, FCP is not getting all the color the camera is recording.  It looks like there are separate solutions in the works for both of those problems but we were not able to employ them in the time allotted, nor has anyone solved both color space and 30p-to-24p issues in combination.  


While the mac was rendering (slowly, it was old and borrowed) the director made cocktails for everyone and I used the camera in handheld mode to shoot some verite style footage in the kitchen with a 50mm lens and follow focus.  It's pretty easy to get cool stuff that way, and I had a blast shooting it.  Again, this camera wants to make documentaries.   I already prefer it over an Aaton super 16 in many ways.  


THE SCREENING


The lab is located on land that was once Gene Autrey's ranch - the first one, before he moved to Newhall next to William S. Hart.  The film building is now dwarfed by the adjoining digital building... but it was great to see a real live Kodak truck pull up to the loading dock and deliver pallets of 35mm film stock.  Our God's not dead, they say... but He was coughing up blood last night.  


As I mentioned above, there was a short and simple one hour color correction session done at the lab's digital suite prior to the film-out.  The files were then loaded into an Arri Laser recorder and burned onto Kodak Vision negative film stock.  The lab retains the negative for their files of test shots, but we took home the print in case we want to look at it again somewhere else.  I'm unsure what the negative emulsion numbers were, I need to check but the print stock is Kodak Vision 2383 Color Print Film, and we sat down to watch all 289 feet of it.  The colorist in the film lab gave it a one-light print treatment, and pulled some green out based on his own judgement.  


Then, the lights went dark, and the capitol pictures logo came up.... and we saw the first film out from the D5, as far as I know about anyway....  


It was so good we sat through it twice.  


Actually, the truth of the matter is that we pleased to see the process worked in any way, shape or form and most of it was not beautiful in any regard but neither was it shot that way. We shot to FIND problems and amplify them, in many regards.  We did learn something from each separate test.  Clearly the work that remains to be done has to be in the area of recovering the crushed blacks that are there in the camera - you can see them prior to bringing the footage into ANY non linear editing program.  Then, we need the best possible 30p to 24p conversion possible.  The motion part seems to be working, but the added contrast and resulting image degradation is the worst part.  


As for the rest, it seems like the problems are simply a new set of limitations that almost any system presents, and the advantages can easily outweigh the liabilities given the right project.  All cameras are a box with a hole in them, and all camera systems are a tool in a cinematographer's tool box.  CHoose the right tool for the job, use it well and you will like the results.  










Sunday, April 12, 2009

BANNED by Jim! Hilarious...


So, dear readers I've just been banned from posting to RedUser.com by none other than the CEO Jim Jannard himself for either "excessive truthiness" or for touching a raw nerve somehow. Granted, it's his sandbox and he can do whatever he wants, so I'm neither shocked nor dismayed, just amused, and a bit surprised given his usual willingness to withstand criticism. The discussion revolved around two recent studio features CHE and KNOWING, that were both shot with his gizmo, the Red One camera. 

Look for the discussion to continue over at cinematography.com and for a longer explanation of what may be behind all this here in time, if I learn more. Right now I'm too busy daydreaming about how this might be a PR coup for my blog..... ha ha.

I was trying to be as kind and deferential as I could but the subject at hand was an important one - two big budget studio features had just been released, and both directors had chosen to shoot with a new and radical camera system that has a goal of emulating, and rivaling or bettering the imagery made by 35mm film.  Personally, I think his camera is on the verge of delivering the goods on that promise, and I say so often.  But it hasn't crossed the finish line just yet, and that's pretty much the consensus opinion across the industry at this point.   But I guess when it's your fortune on the line, you can get sensitive about such things.  

Here's the thread or at least a link to the end of it, as currently displayed. I better get a screen shot just in case.

What won't be evident as the discussion continues is that he simply barred me from further posts, and didn't bother to announce that fact to others. Clever way to win an argument... or at least get the last word.