F55 exposure using zebras

Posted on April 8th, 2013

Now that the F5 & F55 are trickling out in more quantities and Sony has yet to release a professional exposure tool in-camera there are growing questions on how best to set exposure for when working with the cameras. Obviously with an external monitor or even a VF by Zacuto or Alphatron you can use the tools included, but when those add-ons are not practical or simply you don’t own them here are a few tips to setting exposure when shooting in slog2 mode on the F55:

In slog2 mode 90% reflectance (like a white card) should read 59% IRE. (Compare that to slog for the F3 folks which is 68% IRE for 90% reflectance.)

You can try setting your zebras to 60%-70% and not let anything very white or reflective-white show stripes.

Personally, I’ve been setting my stripes for 70% and making sure that nothing stripes unless its a super shiny surface, or say a “hot source”. I will probably move down to 60% but so far been pretty comfortable with 70%.

Posted in Filmmaking

F55 slog2 s-gamut mode test

Posted on March 9th, 2013

This frame grab was from my first test shooting the F55 using the 4k XAVC format in slog2 gamma mode and s-gamut color space. In post-production I applied Sony’s recently released slog2 to Rec709 LUT which was actually meant for the F65 but Sony says it should also work on the F55. Special thanks to Nate Weaver who discovered the LUT in conversation with Sony over on their boards and then shared it with others.

The video clip below has two parts: part-one is the full clip played normally and part-two of the clip features a 200% blow-up and a post digital zoom with pan and scan.

I found that Sony’s slog2 to Rec709 LUT crushes the blacks far too much so I created a slightly modified version of the LUT which worked out better for my scene. You can download it below if you want to give it a try, I also included another modified LUT which gives a nice golden hue to the highlights for a warmer look.

ORIGINAL SLOG2:

LOW CON 01 LUT:

GOLDEN TEXAS LUT:

 

DOWNLOADS:

Sony’s slog2 to REc709 LUT for the F65/F65: download

Original ungraded original slog2 + s-gamut 4k XAVC file: download

slog2sg_to_rec709_STARCentral.001 – LOW CON 01 LUT: ZIP | RAR

slog2sg_to_rec709_STARCentral.002 – GOLDEN TEXAS LUT: ZIP | RAR

Further Reading: What is a LUT?

Posted in Filmmaking, PWM-F55 | Tagged as: , , , , , ,

F3 SLOG LUTs

Posted on March 1st, 2013

If you work with slog on the Sony F3 in-camera using the picture profile slog version or work with slog from the F3 in post, at some point you will wish there was an easy way to add a simple LUT to de-SLOG either in-camera or in post to help get an idea of what your image is going to look like.

Well fortunately I’ve created a picture profile for the F3 as well as an additional LUT that can be used in post to do just that.

Downloads:
F3 slog Picture profile (adds LUT in-camera but does not push middle grey from 38%)

F3 slog to Rec709 LUT for post (matches es slog middle grey 38% to 50% IRE)

sonyf3sloglut

Enjoy!

Posted in Filmmaking, PMW-F3 | Tagged as: , ,

Sony F55 trumps RED

Posted on February 25th, 2013

f55r5cinemastyle-10-30-12-01

Now that the Sony F5 and F55 are starting to ship it will only be a matter of time before we get to see some great (and probably some not so great footage too) from these cameras. Sony has really upped the ante with these new F cameras offering not only a RAW format with 16 times more precision than RED, but 14 stops dynamic range, global shutter on the F55, and other features both cameras share like simultaneous RAW recording and SXS Pro+ cards.

Check out the FD Times 96 page report for more information and details behind the promo shorts shot with the F5/F55; Manout, Dig,  and The Contract.

Some of the biggest differences between the F5/F55 and RED Cameras:

  • The F5/F55 can record HD or 2k internally using a variety of formats from MPEG, XAVC and SR Codec. (The F55 can record 4k internally as well).
  • F5/F55 provides dual recording in two ways: two formats to the same SXS Pro+ card (for example XAVC and MPG at the same time to the same card), or simultaneous RAW recording and  HD/MPG/XAVC/QFHD to SXS Pro+ card. This feature makes it ideal for editing smaller and manageable proxy files and later linking up your RAW files when you’re ready for color grading and final output.
  • The bit-depth of the Sony F5/F55 RAW format is 16-bit offering 65,000 levels of data quantization which is 16 times more data precision than RED’s Redcode Raw format.
  • RED shoots 2k in a windowed mode (cropped sensor) while Sony uses the full sensor for lower than 4k rez formats.
  • The F5 is rated 800 ISO in standard mode, and 1600 and 2000 ISO in SLOG and SLOG2 modes. The F55 is rated 500 ISO in standard mode, and 1000 and 1250 in SLOG and SLOG2 modes. (The F55 is rated lower due to global shutter)
  • RED is rated somewhere between 400 and 800 ISO, and pushing to 1600 is also possible depending on the scene.
  • The F5/F55 has built in ND filters of 3 and 6 stops.
  • The F55 has a global shutter.

RED definitely seems to have their work cut out for them on their next “sensor upgrade”.

Posted in Filmmaking, PWM-F55 | Tagged as: , , , ,

Canon C-LOG exposed, literally

Posted on January 27th, 2013

canonEOS

There are many views on how to properly expose images on digital video cameras from use of waveform monitors, histograms and zebras, but anyone who’s familiar with my approach to lighting for digital acquisition knows I like exposing digital similar to how I would expose film in the sense of not being able to see the image until it was later developed. ie. I use a light meter to expose 18% middle grey at 50%IRE 99.99% of the time. I might shift my middle grey value however depending on the specific scene or if I’m using any special gamma curves to boost specific tonal information in the scene. (On the F3 cinegamma 4 is my all time favourite for boosting the mid-tones)

Having spent the last year to improve my exposure techniques shooting on the digital format using cameras like RED (with a linear and raw based format) or the Sony F3 (with standard or S-LOG gammas) I thought it might be good to delve into the Canon C-LOG to see if I could help clear up any confusion or misnomers on how to properly expose for it.

Whether it’s S-LOG, C-LOG, LOG-C or any other flavour of the LOG gamma curve they all have one common goal and that is to preserve as much data information over the entire dynamic range of the camera sensor as possible. Shooting linear gamma (as RED does) or LOG gamma (as does ARRI, Sony F3 and Canon C100/300/500) are at quite opposite ends of the spectrum in terms of what is happening to the spread of data information used to represent the signal and to some degree the ”lattitude” that can be gained depending on how you decide to expose either format. Refer to my article “When Raw is not Raw” to help understand the key differences between linear gamma and log/curve gamma formats.

When you start working with C-LOG and learning how to expose it properly what’s critical to know is how reflectance values have been re-mapped in the camera giving you “new” exposure levels. Because a LOG image will look extremely flat and lack any contrast it can be very difficult to expose it properly by using your eye, histogram or even a waveform monitor. Your best option is to use a grey card, white piece of paper, and a LCD monitor that can show you the IRE value in a digital readout display. You can also use a lightmeter but since most people don’t use one properly I strongly advise you perform an excercise to determing the effective ASA of your camera and lenses and get familiar with the concept (also your meter might be wrong so this is a worthwhile exercise). You can read an entry I wrote on “how to use a lightmeter“ with digital which should help cover the basics.

The chart below that I put together is based on information provided from Canon’s whitepaper on C-LOG. From the chart you can see how image brightness values are remapped to new values. For example 18% middle grey which is normally 50%IRE moves down to around 32-33%. 90%IRE (white) will appear on your waveform monitor at only 62-63%. It’s really important to use these new values (even if you don’t understand them) when setting your exposure in C-LOG mode since it will maximize the dynamic range your sensor is capable of and preserve as much of the scene information as possible. This “maximum” amount of information is needed later when you de-LOG your footage and begin grading and color correcting in post.

Also observe how changing your ISO on the Canon will impact your exposure lattitude below and above your middle grey point. In essence this allows you to chose an ISO that will let you retain specific image detail for a particular scene. Perhaps if it is low-key, flat, or high contrast. For any DP this an extremely powerful feature because it allows you to select exactly what you want to protect for without having to cheat and shoot using different middle grey values and then push/pull it in post to bring it back to normal.

Remember that anytime you push or pull your image in post you are altering its quality by introducing the possibility of posterization depending on how far you re-adjust exposure, color correct or grade images. Pushing/pulling in post effectively increases (or descreases) the signal-to-noise ratio so the more accurate and consistent you expose your images in-camera when shooting C-LOG it will pay dividends in post and this is even more true if you are using a low bit-depth to record your images.

Posted in Filmmaking | Tagged as: , , , , ,

When RAW is not RAW

Posted on January 10th, 2013

Since RED first introduced the RED One 35mm cinema camera in 2006 eventually we learned about its powerful 12-bit RAW format. Today more and more cameras including; BMC, FS700, F5/F55/F65, C500, ikonshkop, and D16 (in development) are capable of recording data directly off the sensor and storing the RAW data.

“RAW” gets a lot of credibility for its ability to contain “more information” but it has much more to do with the fact that when shooting RAW photos with DSLR’s or RAW video with camera like RED, the signals are stored using a linear gamma. i.e. data directly off the sensor which is linear based. Later in post processing once gamma correction is applied to match how we perceive light (more logarithmically) the data which represents the entire tonal range becomes skewed and biased towards the bright tones. So “more information” in this case is really only available in the bright tones and not so much the middle or dark tones.

Probably everyone has heard someone say when using a histogram keep the hill to the right. With linear based encoding there is good reason for that since you will use more of the data bits available to represent the overall image. Example:

Original scene:

What the sensor sees and how a linear signal looks using 5-bit representation (32-steps):

Gamma correction applied to same linear signal so we can view it:

(note the increased levels in the bright tones)

Not all RAWs are equal either. ARRIRAW shoots RAW but uses LOG to encode the signal. BMC is also doing something similar to LOG, and the new F5/F55 is not very clear to me yet but judging from Ben Allen’s latest article which indicated log is being applied in post processing, it leads me to believe Sony RAW is also perhaps using linear gamma to encode the information and later in post curves are applied.

So what is RAW really? A format that contains unaltered Bayer sensor information – but how it is encoded (ie. log or linear gamma) makes all the difference and becomes especially important for knowing how to handle exposure not only on different cameras but the overall and complete format itself being used.

Posted in Filmmaking | Tagged as: , , , , , , , ,

Will Dragon bid farewell to 12-bit linear?

Posted on January 2nd, 2013

With RED’s latest announcement that the new Dragon sensor may offer up to 20 stops of dynamic range, it has many wondering what’s in store for the insufficient 12-bit RAW based linear file format we’ve all known and grown to love since 2007. Will RED completely gut out it’s EPICs for the Dragon update – or will there merely be a new shell with new ADC boards, upgraded SSD modules, fans and power systems?

1357064610

My speculation is that if RED want’s to keep its 12-bit file format for the new 20 stop capable sensor they will need to move over to a Log encoded recording format similar to Sony, Canon, and ARRI. That or seriously jack up the bit-rates used in their current 12-bit based REDCODE format. Currently RED’s REDCODE format uses linear encoding which means the linear signal from the sensor is digitized and stored that way. Log encoding however applies a logarithmic curve before it is encoded and stored.

I could be wrong to imply that Log would allow MORE dynamic range to be captured within the same bit-depth size, but even so Log has several other benefits over linear such as a more even distribution of data levels per stop throughout the entire tonal range. And less bit-depth means smaller files and less file space to store “more” information. A good example of this already in place is with the Alexa using a 10-bit Log signal to carry almost as much information as a 12-bit linear signal, or a 12-bit Log signal to carry almost as much info as a 14bit linear signal.

Anyway, it will be interesting to see what becomes of this.

 

Posted in Filmmaking | Tagged as: , , , , , ,

Movember Episode #4 DOP Notes

Posted on January 1st, 2013

Welcome to my final blog entry on the Movember web-series and as promised in this final instalment we’ll take a look at how I did some advanced grading on the web-series in Adobe Premiere without the use of any third party plug-ins or add-on filters.

Below is a file I created in Photoshop which features six distinct layers each containing a unique gradient which can be independently selected or combined with others once inside Adobe Premiere. Once you bring a layer in Premiere you can rotate it, stretch it and adjust it’s opacity to meet a wide range of use on your video.

To download a copy of this PSD file and try it on your own projects simply click on the image below! By the way you don’t need Photoshop, the PSD file will directly import to Adobe Premiere and most other NLE software. When you import it however be sure to import all the layers individually, and not merged.

psd-grade-file

So below is an original frame from the series without any color correction or grading added to it. This is a pivotal moment in the episode where the evil boss is being confronted by the leader of the League of Extraordinary Moustaches. To really draw the audience into the center of the frame where the action is taking place it would help to increase the contrast betweet the center and some of the brighter areas of the frame. In particular on the left we have the actor in the foreground who’s shirt is a little bright and the sky in the background is also quite distracting. This scene is a perfect candidate to have some grads applied to it… in post.

movember_ep4_a

To darken the shirt on the actor in the foreground I drag a horizontal left edged grad from my PSD file to my Adobe Premiere timeline. By setting the anchor point to the left edge it allows me to rotate the grad slightly to match the angle of the shirt edge, and also stretch the grad depending on how much of the image I want to cover. Here’s what it looks like against a white background:

movember_ep4_b

Below is what it looks like layered overtop of the video. At 100% opacity it’s too dark and needs to be lightened using the opacity setting.

movember_ep4_c

After lowering the opacity to 20% the highlights from the shirt are reduced and already the center of the image is becoming more predominant.

movember_ep4_d

Two more grads are applied and their opacity adjusted: a flat edged grad on the top of the frame and a rounded corner grad on the bottom.

movember_ep4_e

After the final color correction and adjustments to the RGB curves are made at last we have our final look:

movember_ep4_f

There are indeed several ways to achieve the same effect of applying post production grads to images and this method is neither the best or worst. For this specific project the benefit of this method was that all color correction and grading were completely self-contained within a single project file which made it easy to send the project file back to the producer without having to include a new set of separate corrected video files.

Movember

That wraps up the series of blog entries on the 2012 Movember web-series videos and I hope you enjoyed the web-series and insights behind the scenes of the production. I’d love to hear your comments or questions so please feel free to post.

 

Posted in Filmmaking | Tagged as: , , , , , , , ,

Movember Episode #3 DOP Notes

Posted on November 21st, 2012

I hope you’ve been enjoying all the Movember episodes so far and be sure to catch the final episode #4 during the last week of November. Remember you can watch all the episodes by visiting the LOEM Movember YouTube channel.

In my previous Movember blog entries I’ve mainly been discussing DP and production aspects of the Movember web-series, so in my next two entries I’ll get more into the specifics of post production and color grading techniques that went into achieving the “final look” for the series.

Scene from Episode #3

In my episode #1 notes I mentioned the strict use of my light meter and working with a contrast ratio between 4:1 to 5:1. I used various types of lights to ensure consistent exposure of my skin levels and that I was working within my desired scene contrast ratio. Establishing good levels of tones on-set and in-camera goes a long way to being able to further grade the image in post and achieve different “looks”.

For color correcting and grading I used a stock copy of Adobe Premiere CS6 and no other programs, plug-ins or add-on filters. The reason for this was so I could do all the color correcting and grading within a self-contained Adobe project file which made it easy to simply email the project file to the producer for feedback and eventually final rendering and output.

Color Correcting and Grading in Adobe Premiere

I primarily used two filters built right into Premiere: Fast Color Corrector and RGB Curves and depending on the clip I sometimes adjust contrast using the Brightness/Contrast filter.

Fast Color Corrector Filter

Fast Color Corrector is a great way to move your overall scene hue balance and color angle. I find this useful to get a little nudge or shift in color temperature. For example if the shot is too cool or too warm you can move the “balance angle” in the direction of the color you are trying to increase. As seen in the screen shot below I moved the BALANCE MASTER slightly towards  RED to add some warmth to my overall shot and the BALANCE GAIN to 20 which is the max I recommend using.

The INPUT and OUTPUT tools (also found under the Fast Color Corrector) are very useful for adjusting/shifting your black and white points on the IRE scale. It’s possible to use this to re-expose your image by adjusting the black, gamma and white sliders. In this specific shot I lowered the white from 255 to 240. This was to compress and bring down the highs slightly. The flo lights are clipping as you can see (reading 110IRE) but I don’t care in this case since they are real world light sources.

Screen Shot Color/Grading Episode #1

RGB Curves Filter

The RGB Curves filter allows you to adjust very specific areas of the image tonal range but additionally lets you control independent RGB color values as well as MASTER levels. As seen above, I gave a very slight boost to the MID-TONES on the MASTER level and also slightly crushed the low-end (blacks) to create some nice contrast in the scene. On the red channel I made a small MID-TONE decrease to bring the reds back down leaving the low and high red tones intact.

 

Grading

Stay tuned to my next and final blog entry on the Movember web-series for a look at how I performed advanced grading in Adobe Premiere – again without the use of any third party plug-ins or add-on filters. The technique I used will blow you away – I guarantee it!

Posted in Filmmaking | Tagged as: , , , , , , , , ,

Movember Episode #2 DOP Notes

Posted on November 12th, 2012

The LOEM (League of Extraordinary Moustaches) Team hopes you enjoyed the second episode and incase you missed the first one just head over to the LOEM Movember YouTube channel to watch it. Please share. You can donate to the League here: http://ca.movember.com/team/845545

As mentioned in my previous blog entry the web-series was filmed on the Sony PMW-F3 and in this entry I’ll talk more specifically about the camera settings, setup, and lenses used.

Scene from Episode #2

Sound:

The F3 has dual XLR inputs and even phantom power if you like, but thankfully for me sound was recorded separately using a combination of lavs and an NTG-3 shotgun mic directly to a MacBook. It helped keep the camera free from wires and also from having to worry about monitoring sound on top of “visual” duties. Keeping the camera “free” like this is essential for any highly paced production where you are moving fast between camera and lighting setups, ie.. reverse shots.

Support:

The F3 was equipped with Zacuto Universal Base Plate v3, a set of 20″ carbon 15mm rods, and a set of RedRock 15mm hand grips and shoulder mount that made it easy for switching from tripod to shoulder mount configuration. The majority of the web series was shot with my Manfrotto 510 head and Manfrotto 3190 tripod.

Scene from Episode #2

Glass:

We shot the web-series “film style” using only one camera and shooting variations of master shots, medium shots, and then moving in for close-ups or insert shots. I had access to a set of high speed primes for the shoot but instead of them decided to stick with only two constant zoom lenses – the Tokina f2.8 16-50mm and Tokina f2.8 28-70mm. Coupled with the MTF Nikon to F3 adapter I was able to adjust the aperture of both lenses (normally electronically adjusted) with the outer ring on the MTF if I needed to stop down beyond f2.8.

“There is a time for primes – this just wasn’t one of them.”

Using zooms allowed for very quick reframing and adjustments of shots without having to move the whole camera setup on tripod and re-level the head each time. There is a time for primes – this just wasn’t one of them. I also knew that for the majority of shots I wanted f5.6-f8 for wide shots and f2.8-f4 for closeups and a zoom allowed me to stay further away from the talent and zoom in for medium and closeups adding even more background blur.

For this project the matte box and 5 x 5.65mm ND filters stayed at home. Instead I used a set of high quality Schneider screw on ND filters which allowed me to stack them directly on the lens depending on what strength I needed. I mostly used .3 (1 stop), .6 (2 stops) and sometimes stacked them for 3 stops total – of course that was in addition to the built-in ND filters on the Sony F3 (3 stops and 6 stops).

Settings:

All clips were shot internally to Sony SXS cards in 8-bit 420 @ 1920x1080p, 24fps, 0db gain, 800ISO and I used a clean picture profile which only had the Gamma changed from REC709 to cinegamma4. If you’ve seen the cinegamma4 on a waveform monitor shooting a grey card you’d know that the mid-grey level is placed a little high around 50%. What I really like about this gamma curve is it gives a very nice lift to the mid-tones so later in post I don’t have to adjust them and risk deteriorating the quality of the image.

Scene from Episode #2

I hope everyone enjoyed all this info and please stay tuned for my next blog entry on the web-series where I’ll go into more detail on color and grading in post production – all performed within Adobe Premiere CS6 without 3rd party plugins or filters!

 

Posted in Filmmaking | Tagged as: , , , , , , ,