ISO Range and Gamma curves

[bill totolo]

Not a scientific test at all:

ISO 400 with Andy Shipsides' AB_Norm Picture Profile
47049


1600: Andy Shipsides' AB_Range picture profile
47048

Sony PL Mount lenses to SxS-1 memory cards. No color grading.

[TimurCivan]

Bill,

By 400ISO, are you going by what the camera says?

Try using "dB" instead. the ISO #, is innacurate by a stop and a half. 0dB on NON log is 800 ISO. 0dB in LOG is 1600ISO with a meter.

By going -dB, you are just cutting d/r off the top of your signal. Swish. Gone. One stop.

[bill totolo]

Shot one is 0db.
Shot two is 12db.

By this math does that mean shot two was really 3200ISO?

Thanks, Timur.

[TimurCivan]

Yes sir

[bill totolo]

Thanks Timur, I was disappointed with the noise level but now I know why.
When shooting to internal cards do you avoid going above +6db of gain (aka, ISO 1600)?

Thanks.

[TimurCivan]

honestly dude, I've shot 18dB with no problems....

The F3 is so grainless. even at 18dB so long as there are no saturated colors, youre good.

Make sure you black balance.


http://vimeo.com/20725393

[bill totolo]

honestly dude, I've shot 18dB with no problems....

The F3 is so grainless. even at 18dB so long as there are no saturated colors, youre good.

Make sure you black balance.


http://vimeo.com/20725393

A little too much noise for my taste. I can't stand it when the fall off gets too much macro-blocking and the subsampling in the blacks gets noticeable.

What picture profile did you use for the above video?

[bill totolo]

Simple exposure and gain test:


http://vimeo.com/38257827

[Douglas Villalba]

I keep hearing that the ISO rating by Sony is wrong, but any old school photographer would disagree. What is wrong is the way it is being measured.
The way a negative and a positive film are measured is different. Most people measure the 18% gray card at the middle between highlights and shadow. The correct way to measure a negative film is were you want details on the shadow. For positive (movie straight to projector like your old 8mm or photography slides) and video you measure the brightest area that you want details. Done this way you get the most out of your image. I personally don't trust a reflective reading, I mush prefer an incident reading to a flat disc. The half sphere is only good for average reading of multiple light sources.
Video is a little more complicated and different Gamma curves would change how you see it. Mush like crutching or stretching in post. A good starting point is using a regular setting (709) and reading the highlights at the ISO the camera said according to the db and adjust when you change the gammas.

[bill totolo]

Great input Douglas, and exactly why I posted the link. Testing S-Log and LUTs today on the F3.
Here's Sonys take: http://pro.sony.com/bbsc/ssr/show-highend/resource.solutions.bbsccms-assets-show-highend-pmwf3.shtml#/sLogGamma

[bill totolo]

Straight out of camera. SLog, 709 (800%) lut to SxS cards (Media export from Premiere as TIFF, then converted to jpeg via iPhoto):

49793

Frame on left is straight out of camera, frame on right is with minor color grade:

49819

[Douglas Villalba]

I think that a client/director will be happy looking at that picture on the monitor.

[bill totolo]

Thanks Douglas. This was just an experiment for an upcoming shoot.

Today I had a two camera setup (Varicam 2700) at proper angles. Then I just squeezed the F3 off to the right to see how the Rec709 (800%) LUT looked and if I was understanding exposure.

BTW, this is Marlen Esparza Flyweight women's boxing champ and TeamUSA Olympic boxer.

[TimurCivan]

Have you updated to the latest firmware? It enables slog out the A port while applying a lut to Video out. This way you can do a 709 lut out to client and you can be safe knowing slog has your back.

[bill totolo]

Thanks Timur, I haven't decided on an off board recording device yet.
Are you still happy with the Ki-Pro?