What would you like to see in a HD100/200 to 35mm transfer test?

[Jose L. Martinez]

Hi,

I'm about to dp a short feature with the HD100 / 200 with Sgpro adaptor and Nikon lenses.

The film will be transfered to 35mm anamorphic (2,40 to 1) at Technicolor, and I managed somehow to get some budget for a 2 mins tests.

So, any ideas what should I test to get a good knowledge of the limits of this camera?. I've completed a small list of things I would like to know:

- Emulsion test: Macbeth chart, gray card, human skin before a white / black background, two 1k's at 2 meter, 45º. Expose at key and bracketing 6 stops over and under. With fujinon stock lens alone and with the adaptor. I'll try to have an astro to check IRE values, but still learning the damn gizmo.

- Filage: Normal and fast moving subjet at 1/48, 1/50, 1/10, etc.

- Resolution chart.

- Flare: I need to know how the lens behave with sunlight and practicals, how ugly they are, etc.

- High fequency / banding / stair -stepping / moirè. Just to know what not to do.

- Some interior and exterior test shoot just for the look.

- Some 60p slow motion, just for the shake of it...

Anything I'm missing, anything senseless.

All opinions appreciated. Perhaps I could scan the negative and publish it somehow, just for info.

Cheers,

Jose

[Brian Drysdale]

I wouldn't use a 35mm adapter for a film out. It's OK for TV, but it's going to degrade your image. Even on the high end F900 cameras people tend to go for the highest quality HD lenses if they're going a film out rather than use a Pro35. On the JVC you're fighting for all the image quality you can manage for a film out, even more so if you're cropping to scope.

[Jose L. Martinez]

Hi Brian,

First, thanks for your reply.

That's why I'm doing the tests both with and without the adaptor. I'm loosing a full stop with it, and nikon lenses are not exactly the perfect solution, but the story really needs a cinematic dof more than a crispier image... and, at least in the monitor, I like what I see... ;-)

Jose

[Boris Missirkov]

One thing concerning corrections in the post before filmout: do never pull your blacks to 0, or pure black. Always leave some space on the left of the histogram. Otherwise, the shadows that look OK on the video go into the lower part of the sensitometry curve of the negative and get all sorts of bizzarre tints - as there's no negative that has identical R,G and B gamma in the toe of the curve. Try to keep all your image in the straight part.

I had to learn it the hard way - trying to colour-correct a 100-minute film whose DI was done in a studio that claimed to know what it was doing.

Good luck!

[Seung Han]

Hey Jose,

I am very interested in what you discover with your tests. Please keep us informed and if you could include some pics it would be great. Thanks!

[Jose L. Martinez]

Well, test was shoot and should be printed this wednesday. It's just a day before the shoot, so maybe I won't be able to post the result in a week or so.

http://www.contacontos.com/test.mov is the url to a H264 small version of the test. It's 2 minutes long, but I've chopped it down to fit in my server.

I basically tested the Sgpro r2 with nikon primes and the fujinon alone. I've used cinelike, tim dashwood' amelie (thanks!),'Angel' presets (thanks!)and paolo ciccone true color 3 (thanks!), bracketing exposures from 5,6 (at key) all the way up and down in 1/3 increments.

I've also included some shots from Xan, a feature I dped, and 3 shoots from Tim Dashwood hd200 tests (wide and cinevert).

In all cases, white balance was done before touching the matrix.

Rendered in uncompressed 8bit 1280 x 720p and delivered in uncompressed tiffs. Will have it xferred via crt and arrilaser at Technicolor and Iskra, cropping to 2,40 aspect ratio.

Will let you know!

Jose

[Jose L. Martinez]

Hi again,

Back from shooting (quite bruised and with a toe broken by the jib), grew quite confident with this camera.

A day before shooting went to see the tests at Deluxe, and really got impressed. The output was done via arrilaser directly from tiff sequences, a minute in 1,85, then the same images cropped to anamorphic 2,40.

I'm trying to get all my ideas ordered in a longer post, detailing all methods used and options, but my conclusion by the way is that this camera is a really serious option for filmlike production, even in anamorphic.

Let me know if any of you has any questions or whatever. Will post back!!.

Jose

P.S. Have a look at the weblog (galician / spanish only, sorry) 12007.blogspot.com

[Oscar Honorio Pantoja]

José, estaba siguiendo tus post y he visto tus pruebas preliminares. He quedado impresionado, pero no muy satisfecho.
Sería bueno si puedes exxplicar un poco el procedimiento mas a detalle. Los usuarios en español (somos pocos pero los habemos) te lo agradeceremos.
Me interesa en particular el tema de los negros, tu ajustaste la camara en modo filmout?? o sólo usaste el TC3 de Paolo??
saludos

[Jose L. Martinez]

Finally got some time to post the conclusions I got from the tests. I'll try to put them as clear as possible (I'm in a 8-day week, 5 hours of sleep per day--). If any of you has some specific question I could answer, please let me know, and I'll try to answer it here.

Oscar, si te parece continúo este post en inglés, o más bien spanglish, y respondo a tus cuestiones en castellano en el siguiente, o buscaré ese foro en castellano de la jvc en el que creo que estabas colaborando (me equivoco?).

OK, so:

0) Test details: We used both a HD201e and one of the first hd100E to arrive to Spain. Both with the stock lens (fujinon 5,5-14), no filters, and alternatively with the SgPro lens adaptor with a nikkor 50mm prime lens and f3.5 75 nikkor zoom (not the best glass, by the while).

Subject was placed at 2 meters from camera, and ligths (2 fresnels 1k) were placed at 1 meter from subject in a 45º angle from subject, both hor. and vert. Lights were a bit old and used (I was using a media school, so...) so if any of them was close to 3200k it was unintentional :-)

Background was a white poliestirene reflector and a black velvet piece. We also add some golden props and a red costume to be used in the film.

Camera was rated at 160 ASA by comparing the auto iris reading with an incident meter. No white balance was used, but the 3200ºk preset, to preserve the color matrix values.

We tested the following scene file recipes: Default cinelike 24p, Tim Dashwood Warm Green and Cinevert, and Paolo's TrueColor V3. Thanks!. We also tested the scene file created for the feature film.

Key was rated at 5,6 and we bracketed all the way down and up in 1 stops increments.

Some scenes from my short feature Xan, A despedida (xanadespedia.blogspot.com), shoot with the HD100 fujinon were also included just to see some real footage. I also had to include some of the m2t generouslly shared by Tim Dashwood with his hd200 / cinevert tests because of some problem with my original tape.

Footage was edited in a Final Cut Pro station, ingested via firewire, edited in a HDV 720 24p native project (no AIC), sequence duplicated and reconfigured to uncompressed 8 bit (I didn't use 10bit for this to avoid dithering, but really don't know if I could have gained something by doing so, opinions?). Output to uncompressed TIFFs without YUV processing.

Transfer was done on Arrilaser (according to tech, it did all the uprezzing and lin-log colorspace conversion), on what I think was 5201 50 ASA vision 2 negative, maybe intermediate, but I have to confirm.

Test was cropped first to 1:85 and then another copy was cropped and fitted in the 2:40 anamorphic image.

The mov I've shared here suffered from poor compression and a gamma shift (directly outputted from FCP) I was unaware until already uploaded it to our server (4 hours before taking a plane, so...) so it is not a reference of color, latitude, quality... it's just a demo of what I was trying to do, not of what the camera is capable of.

[Jose L. Martinez]

Please note I'm speaking here of my impressions and opinions after the tests, so I cannot offer more scientific data than how I think it look according to my tastes. I heavily encourage you all to conduct your own tests, if possible.

1) FILAGE AND GRAIN -

I've tested around 6-8 cameras for filmouts, and this is, along with the varicam, the closest to film I've found in 6 years. The filage is virtually indistingible from the motion blur produced by a 35mm camera with a standard 180º shutter. I've recently shoot some 35mm tests (my experience with this format is still 'on the works') and I really think that, for a 'average' scene,it would be relatively impossible to distinguish from 35mm originated material, with non educated viewers.

The grain is really nice when is well managed (eg to avoid banding) and much more organic than what can be found on other cameras (hvx200, f900), and the colors, particularly with the true color and cinelike settings, are really close to what I expect in a film print.

2) COLOR SPACE

This is a 8 bit lin and heavily compressed camera, so I would try to get the closest image possible to what I want in camera instead of trying to get as much latitude as possible (I have no balls to show a muddy image to clients, trying to convince them I will fix it in post, a 550 EUR an hour of lustre).

3) DETAIL

As always, the gurus in this board are rigth. Get the detail waaaayy below. I won't deactivate it, -7-8, never off. I was gladly surprised by some long shots that simply could have been shoot on film.

The image got softer with the anamorphic print, but less than I expected. I wouldn't use the sgpro adaptor in an anamorphic project, but after seeing the tests, I would definetively go for it for an 1.85 project, for me, the slightly softer image is an acceptable price for the shallower dof. But check back focus almost with each setup.

4) LATITUDE, ASA RATING AND LOW LIGHT

The 'scientific' tests I tried give me an asa rating of 160ASA. My experience while shooting almost doubles it. It is heavily dependent on the scene file choosen, with the true color v3 gaining almost a stop over the cinelike. Why?. I simply dunno.

The lesson to learn, way better to underexpose with this camera. Lowlights seems beatiful, highlights over 100 ire, as all of you know, ugly at best. Better get your big highlights in post, in a log environment.

5) LENS SWEET SPOT

I've found it, for my tastes, around 4-5,6 with the fujinon. I've experienced some diffraction (seaside, f16, full zoom) and the color aberrations we all know.

6) 35 ADAPTOR (SGPRO) FOR BLOWUP

You loose a stop, and with the fujinon, is bulky as hell, but I would use it for 1,85 film out. Not as soft as expected, and really nice images. Have a try!.

Jose

[Jose L. Martinez]

Just in case you had any curiosity on how the thing ended, there's another post with images and a short movie to check.

http://www.dvinfo.net/conf/showpost....&postcount=133

Cheers,

Jose

[Eric Gulbransen]

Jose that's a lot of work, sharing all that info. Thanks for the time and effort. Congrats on your project. Looks great.

[Marc Colemont]

Thanks Jose for your detailed explanation, and to share your experiences with the camera for filmout.

[Jaadgy Akanni]

Thanks for all the updates on your project. I've always been convinced that this camera gives you the most filmic look of all the cameras in its price range. Those grabs are just beautiful-very nice "film-stocky" grains.

Hey José, given that you lose some color when using the adaptor, having you ever raised the color gain on either Tim's or Paolo's scene recipes?

[Dan Parkes]

Cheers Jose.

I missed your earlier post with your impressions etc of the JVC201/SGPro setup and since we are using the same equipment I found it most interesting.

Your conclusion regarding the stock lens sweetspot being 4-5.6 confirms what I have been told -we were struggling for light and ended up opening the stock lens right up on a recent shoot, and had some issues with soft focus.

You also mentioning checking the backfocus on each setup. I am curious to know what you used for onset monitoring? Did you have a HD field monitor?

Dan