Sunday, May 22, 2016

New Website

You can now find all my information and links to recent work at:
http://billtotolo.com

Tuesday, September 15, 2015

TotoloVision - A picture profile for the Sony a7s

TotoloVision:

Thank you for visiting. I hope this color profile is helpful, I offer it for free to the community.
Please note: this is not for any other sensors than the original a7s.

Created with a DSCLabs Chroma DuMonde Color Chart and the Vectorscope on the Convergent Design Odyssey 7Q.

This has only been balanced against daylight, I have not tested under Tungsten yet.
Care should be taken to make accurate exposures.

Black Level -4
Gamma: Cine 4
Black Gamma: Middle; Level -1
Knee: Auto>Auto Set Max Point 100%>Sensitivity Mid
Color Mode: ITU709 Matrix
Saturation -2
Color Phase -1
Color Depth:
R 0
G +1
B +2
C 0
M 0
Y +1
Detail -2
Adjust: Mode: Manual; V/H Balance 0; B/W Balance Type3; Limit 7; Crispning 0; Hi-Light Detail 0

Monday, July 13, 2015

IRND/Hot Mirror/Firecrest tests

The camera used is a Sony a7s in stills mode. Shooting 16x9, 10 Megapixels, Extra Fine jpeg, ISO 200, Cine3 Gamma in Pro Color Mode (no other adjustments), sRGB.
Filters shot at an f/ 2.8 in full frame mode. The idea was to shoot at a stop that neither helps, nor intentionally hurts the image (regarding vignetting, etc).

I used a 50mm Zeiss ZF for its color neutrality, and relative sharpness at f/ 2.8

The first frame is without filtration which can be used as a monitor reference. You should not see a magenta or cyan skew. This was shot under blue skies, mid-day in southern California. White Balance manually set to 5600K for reference. Zebras set to 90%, what you’re seeing is a 4x4 showcard.

This was shot at a deeper stop because I wanted as ideal a reference shot as possible. You should see a field of white with a page from a notebook in the center:




Performing a white balance yielded a result of 5800K. You can see a minor difference:


First filter up is the Tiffen IRND 2.1 at 5600K:



Of course you have to white balance an IRND. A manual white balance with the filter in front of the lens yielded 4000K:



Next is the Tiffen Hot Mirror IRND 1.2 pictured at 5600K:




And with a manual white balance again yielding 4000K:


Next is the Pancro IRND 0.9 at 5600K:



Now with a manual white balance of 4200K:



The Formatt Hi Tech Firecrest (IRND) at 5600K:



And finally, Formatt Firecrest manually white balanced to 5400K:




© 2015 Bill Totolo
All Rights Reserved

Monday, June 08, 2015

Sony a7s - shooting a feature with a pair of Sony a7s's

The project, which changed in scope on an ever-continuing basis, evolved into a
character drama centered around one location for 11 days, and various locations after
that. These included bar night-interiors for two days, bar night-exteriors for two nights,
and several days of day exterior locations.

G&E was very humble: We started out with a 1.2K magnetic ballast HMI, (4) 1x1 bi-
color Litepanels, (1) Mole 2K zip, (2) 750w Lekos, standard ARRI kit (1K open face,
650, 300, 150), a China ball with a 200w tungsten globe, a box of full spectrum CFL
household bulbs, some Christmas lights, and an assortment of low wattage household
globes and mr-16’s.

I used 4x4, and 2x4 foam core boards, a Westcott 6-in-1 reflector kit, white queen size
bed-sheet, skate-board dolly, (2) 1K dimmers, Letus Helix gimbal with 4th axis, and a
knock-off steadicam vest and arm purchased from ebay provided by our 1st ac.

My crew consisted of 1st ac, gaffer, and occasionally I had a best boy. My DIT was an
inexperienced data wrangler who learned on the job.

The budget for this feature film was in the $60K range, so initially I thought I would be
very limited in my camera selection, however this being LA, we were offered everything
from RED Dragons, Epics, and Scarlets, to every DSLR on the market as donations or
very close to free of charge.

A major factor driving the decision-making process was based on my confidence in the
post team, or lack thereof. In my experience, the success of production largely relies on a
good relationship with all departments including an experienced post team, and that relationship usually begins in prep. As it turns out, on this project, the producer and director decided they were going to handle post, and it was evident this was not a team with an extensive post-production
background and certain workflows would be burdensome on this team.

As a DP I often find myself in an advisory position. It was my opinion that to maintain
consistency over the work put in throughout prep and production that a streamlined
workflow was in order.

I have owned a Sony PMW-F3, Canon 5D MKII, and 5D MKIII, and have shot with all
the major cameras out there today (ARRI, RED, Sony F5/55, Canon C100/300/500, etc).

After selling off my cameras I decided to purchase the Sony a7s because I like owning
a small DSLR for personal projects, and for experimenting with lenses, etc. It turns out
the a7s has a very shallow flange focal distance which allows it to adapt
to many existing lens systems.

By now we’re aware of the a7s’s low-light capabilities and its extended
dynamic range. I set out to test the acceptable noise level of the camera and ended up
being disappointed. I wanted to use S-Log2 for its DR but with a nominal ISO of
3200 found the noise unacceptable. I could get great results with a Picture Profile but I
didn’t want to throttle all that dynamic range into a Rec709 container. I found that by
using Miguel de Olaso’s (aka Macgregor) recommendation of shooting with a PP where
the Gamma is set to Cine2, and the Black Level raised +15 that I could emulate the DR of
Slog while using much lower ISO’s. The Color Mode I liked most is “Pro”.

After my tests migrated from test charts to real world scenarios with people moving about
I ran into another obstacle. I found I could get a nice look using Miguel’s PP but as I’m
not as talented a colorist as he, I was spending a lot of time grading each image. And as
I’d volunteered to be the colorist on this picture (as opposed to the option of possibly no
colorist) I needed a reliable workflow that would allow me to color hundreds of shots
efficiently.

I spent a lot of time honing some kind of skill-set with Resolve and became comfortable
with curves, LGG, and motion tracking. And as I honed my skills the latest hot thing to
hit the market was look-up tables, commonly referred to as LUTS. I wasn’t looking for a
magic bullet but it was evident a lot of people were getting some great results with
popular LUT recipes.

The problem was that I wanted a very naturalistic look for this film and I really think
what Sony gives you in Slog can be very beautiful. It just takes the right lens choices, and
a subtle hand in grading. So I set out to find a way to make Slog work for me.

What I found is that by exposing Slog2 at Sony’s recommendation of 32IRE (347 10 bit
CV) for middle grey and 59IRE (582 10bitCV) for 90% white, or over exposing by no
more than 1 stop I could get very easy to grade results using the Sony 1D input LUT:
Sony Slog2 to Rec709. With a little adjustment to the knee and toe I had a
look I was very pleased with. And for instances where I wanted to shoot under extremely
low light conditions (5-10 foot candles) I purchased a plug-in called NeatVideo. With a
light touch this plug-in can transform the noise of Slog into something very acceptable. I
had some files projected on a 30 foot screen and the results were enough to make a
believer out of me.

I recorded data to SanDisk Extreme Pro 95MB/s SD cards and to Convergent Designs
Odyssey 7Q loaded with a pair of CD 256GB SSD’s. This allows me to have robust files
recorded in 8bit 422 ProResHQ files. (The Odyssey records in 10 bit, but the a7s only
sends out 8 bit).

ProRes transfers very reliably between Adobe Premiere Pro and Resolve, and any files
that were grabbed on the a7s without the O7Q were transcoded to ProResHQ from the
native XAVC S files via Rarevision’s 5DtoRGB app.

With a nominal ISO of 3200 and a DR in the neighborhood of 14 stops I noticed
artifacting in my lighting units I wouldn’t see if we were shooting in REC.709. I rented
gear that’s been well used so I’m not saying there are flaws inherent in any particular manufacturers
design but noticed flicker in our HMI and banding from our LED panels and CFL's.  I swapped out all my LED panels and CFL’s for good ol’ tungsten units, and that old magnetic ballast HMI had to go for a more reliable ARRI 1.2K w/ electronic ballast.

I don’t know if it’s particular to this camera but I found using fader ND’s sucked the life
out of the images, muting color, and creating artifacting in the bokeh. If I had them
in time I would have used my Formatt HiTech Firecrest ND’s. As it was, I used my B+W
pola to adjust exposure or I switched to Miguel’s PP with Cine2 and lowered the ISO.

We encountered a stuck pixel on one of the cameras but this was easily remedied by
remapping the pixel. The camera does this automatically every two months, so the trick is
to move the date in the menu ahead two months and restart the camera. As soon as I did
this, the pixel was remapped and the camera performed fine.

Camera A was on a Letus Helix, sometimes with the 4th axis on a vest and arm. It was in
a HotRodCameras cage with aluminum rods, CINEGEARS wireless FF, Teradeck Bolt
300 HD transmitter, Odyssey 7Q powered by Anton Bauer Dionic 90’s attached to a
Gold Mount plate fixed to a home made belt clip. All Lemo and custom cables were
made by Media Blackout. All adapters made by metabones.

Camera B was mounted on a Zacuto V3 baseplate, mounted in a Movcam cage, with
Zacuto Arm supporting a Convergent Design Odyssey 7Q+, RedRockMicro
microMounts with microSpuds, CINEGEARS wireless FF, Teradeck Bolt,
HotRodCameras hand grips, OConnor O-Box, D-Tap breakout box, Zacuto Zwiss plate,
Anton Bauer gold mount plate, and Dionic 90. Vinten 5LF head and corresponding sticks were used.

Lenses used:
35mm Zeiss ZF 1.4 cine modded by Duclos lenses.
50mm Zeiss ZF 1.4 cine modded by Duclos lenses.
85mm Zeiss ZF 1.4 cine modded by Duclos lenses.
100mm Nikon E series 2.8 cine modded by Duclos lenses.
135mm Nikon AI 2.8 cine modded by Duclos lenses.
Canon 24-70 MKII L series f/2.8
Dog Schidt Optiks Flare Factory 58 Mild Magenta/Normal Gold; Nikon mount.

B+W 77mm slim circular pol

What I found truly unique about the a7s was the ability to crop the frame, so every lens
effectively had two focal lengths. Also, to mitigate CMOS artifacting it is wise to
consider shooting in crop mode. If the camera is locked off or doing a slow dolly, jib, or
gimbal move shooting in full frame mode is fine.

The cameras were very reliable other than a stuck pixel, and occasionally, swapping a lens
on the metabones e-mount put the camera into factory reset, which can be hazardous if
not caught right away. Other than that, we didn’t have any corrupt files in camera, shut
downs, or other ghosts in the machine.

Keep in mind the full frame sensor occupies a lot of real estate so I advise extensive
observation of the CFA to check for dust particles. A blower bulb followed by an
inspection with a sensor scope, and a cleaning with a Delkin sensor cleaning kit does the
trick.

Oh, and those tiny Sony NP-FW50 batteries don’t last very long. We got by with
4/camera and a bunch of chargers, but barely.

In the end, I’m extremely pleased with the images we captured. I honestly don’t think
anyone is going to guess we shot on a DSLR. I’m glad we used the a7s under these circumstances.
I’d only consider selling the a7s if Sony releases a version with In Body Image Stabilization, and it would be nice to be able to control ISO in SLog.

As small as the camera is, it gets heavy quickly and using the gimbal all day is
exhausting. In the future I would seriously consider using a readyrig, or easyrig.

© 2015 Bill Totolo
All Rights Reserved

Saturday, April 25, 2015

Cinematography Reel 2015

A montage of material mostly shot 2014/2015 with a few classics in there. Trying to show more narrative work, but I still love lighting and shooting live music performance. A lot of stuff on there shot with Sony F55 and Canon C300 with Fuji Cabrio zooms, and some tilt/shift Canon lenses. Zeiss is also present.

Monday, January 19, 2015



My new cinematography demo reel showcasing work from 2014. Please share among friends. Thanks to all the crews involved, too many to list here.

Monday, November 24, 2014

Canon Cine-Servo 17-120

This blog has been neglected the better part of 2014, and I've been meaning to update it for a while. With the explosion of camera/tech blogs and podcasts it hasn't seemed as neccessary to post. I've been immersing myself in other people's experiences, and quite frankly, was left wondering what I had say.

All I can offer is the experience of one DP: the ups, the downs, the false starts, and frustrations. And of course, the satisfaction of a job well done.

Before I delve into projects from 2014, I want to share something relevant to Nov. 23, 2014- The new Canon Cine-Servo 17-120. A lot of camera-persons are set up as small businesses and are looking for end of year investments, and a lot have their eyes on the Canon. Somehow, I was lucky enough to get booked on a gig that subbed in one.

As an operator, my expereince was very positive. It was easy to adjust back-focus. The macro acts as you would expect a macro to work on an ENG/B4 style lens. There's minimal breathing, bokeh from the 11 blade diaphragm is easily obtained and quite nice (with a slight cat-eye effect on the outer perimeter, which I happen to be a fan of), the servo is incredibly smooth, it's about 6 1/2 pounds, and it appears to be very sharp across the aperture range. This last statement is just an observation by eye and not properly evaluated with a test chart.

Before returning the gear I had time to record some images and thought I'd share:

https://vimeo.com/112614454





Thanks for bearing with the absence, I'll try to keep the blog up when I have something to share. In the mean time...I don't know, I need a catch phrase. 
See you on set.

©Bill Totolo 2014, All Rights Reserved