In the third part of my S-Log series, I took the Sony F3 outside on a nice bright day to show what S-Log can do. Jeff Lee and I headed up to the roof of AbelCine and found some shade to stand in; this allowed us to show the most contrast possible. I shot the blog in S-Log, as well as in the standard video modes with the AbelRange profile that I created earlier this year. This way I could show just what S-Log enables in the camera. All the clips in the video, except the last one, were recorded to the Gemini 4:4:4 recorder, which records in uncompressed DPX stacks so nothing was lost along the way. The last shot was S-Log out to a PIX240 recorder in ProRes422 HQ, which was enabled with the new F3 1.31 firmware from Sony.

I graded all of the shots in DaVinci Resolve Lite. The difference between S-Log and my range profile can be slight at times, but pay close attention to highlight areas, especially the skin highlights. Thanks to Isaac Kiener from Sales for helping out. Stay tuned for the last part of the series on look up tables in the camera.

To learn more about the F3, S-Log and LUTs make sure to check out our F3 Training Class in NY and LA.

24p is ‘jumpy’ by its nature, but the look of 24p video in a digital video camera can still come across a little more ‘jumpy’ then you’d expect. There are several reasons for this, which I explain in article for HDVideoPro magazine called “Did I Judder.” As part of the article, I put together these two videos to show the difference between native 24p video and 24p converted to 60i. I did this because much of the 24p content we see has been converted to 60i already, which has a smoothing effect that is very noticeable on television.

Watch the above video to see true 24p and then the video below in 60i. The difference is subtle on a computer monitor, so it’s difficult to show online. However, the difference helps explain why so many of us are surprised by 24p judder. To learn the other reasons for 24p judder, check out the full article over at HDVideoPro.com and on magazine shelves now.

Let’s start by looking at two HD formats, each with a different color space and sampling. On the new PMW-F3 camera, you can record internally in the XDCAM EX format, but also output an S-Log RGB signal over Dual-Link or 3G SDI. More specifically in 1080 24p,

XDCAM EX =  8-bit YCbCr 4:2:0 1920 x 1080 23.98p recording (35 Mb/s recorded)

S-Log RGB = 10-bit RGB 4:4:4 1920 x 1080 23.98Psf (Uncompressed)

The text in red describes the color model, and the text in blue describes the color sub-sampling. Here is a short description of both.

Color Model

A color model is a system for representing color as a series of numbers, three numbers in the case of a RGB or YCbCr. Combine a color model with a color gamut, and you get what is called the color space. Gamut and color space are topics for future blogs, but it’s good to understand the relationship.

RGB – In the RGB color model, each pixel of the image has three chroma values (red, green and blue) and each of those color values has luminance (brightness) information as well. So in the RGB model, each frame is made up of three different color images. This becomes very useful in color correction, because each frame is essentially three different images put together.

YCbCr (YPbPr) – YCbCr describes a digital signal and YPbPr describes an analog signal, but they share the same basic model. Each pixel of the image contains luminance Y, blue-difference chroma Cb, and red-difference chroma information Cr. In the YCbCr model, each frame is made up of a luminance only (black & white) image, and two chroma components. Green information comes from a combination of the chroma and luminance data. YCbCr was created in order to reduce the bandwidth of the video signal by removing the redundant luminance signals in RGB (each having their own luminance value). This also enables the reduction of color information without effecting the actual resolution of the image. This process is called Color Sub-sampling.

Color Sub-Sampling

Color Sub-sampling is a method of reducing color resolution without effecting the overall resolution of the image. By reducing color information, the bit rate of the video stream can also be reduced. In the YCbCr color model, this is easily achieved because the luminance (Y) part of the image is separate from the Chroma (C) parts of the image. Looking at the diagram above you can see that the luminance signal is essentially a black and white image; our eyes are better at seeing brightness, so the Y signal is generally kept at full resolution. However, if we reduce the resolution of the color images (Cb, Cr), we have a hard time telling the difference.

We often see those numbers associated with a compression like ’4:2:2′ or ’4:2:0′. They refer to the method and amount of Color Sub-sampling. Just for this blog, we will call them a sampling triplet. The numbers break down quite simply: just imagine a simple 4 pixel wide by 2 pixel tall box at the center of a 4:4:4 sampled image:

1) The first value in the sampling triplet represents the luminance (Y) sampling in the image across both lines. As you can see in the image above, we have Y values for all four pixels on both lines. Therefore, we have full luminance resolution.

2) The second value of the sampling triplet represents the amount of chroma information in the first line of the box. In the image, above we have four chroma samples (C), so we have full chroma resolution on that line.

2) The third value of the sampling triplet represents the amount of chroma information in the second line of the box. In the image above, we have four chroma samples (C), so we have full chroma resolution on that line.

Thus, the image above could be called 4:4:4, with full color and luminance information.

When we start to reduce that information, things get a bit more tricky. Let’s look at 4:2:2 -

As you can see, we still have full luminance information (the first ’4′), but we have less chroma information, two on the first line and two on the second. So we have effectively cut our color resolution in half on both lines. This sounds bad, but 4:2:2 does very well for almost all situations. The chroma information that is removed has been averaged into the other value to keep the image accurate.

What about 4:1:1 -

Again we still have full luminance information (the first ’4′), but we have less chroma information, one on the first line and one on the second. This means that each chroma sampling is an average of 4 horizontal samples, which does not always give good results. With four horizontal samples averaged together, colors on the edges of your image can have a tendency to bleed. To fight this problem, most new compressions use 4:2:0 -

The luminance information is still kept at full resolution, but now we have two chroma samples on the first line and none on the second. Sounds bad, but really the chroma samples are an average of two horizontal pixels and two vertical pixels. This uses the same bit rate as a 4:1:1 compression, but without the same color bleeding issues.

It is possible to have a 4:4:4 YCbCr image, but RGB is preferred for color correction. Color Sub-sampling cannot be done in the RGB color model, because each color channel also contains luminance information. We often see the term RGB 4:4:4, but it is actually a redundant use of the term. I guess that Sony and other manufactures figured it was better to be redundant then to have people confused.

Most professional HD video cameras today have an HD-SDI (High Definition – Serial Digital Interface) output, so this will be a good place to start the discussion. An example of a typical HD-SDI video output would be a 1080 60i Uncompressed 10-bit signal. If you are scratching your head already, let’s break it down a bit with a couple definitions.

Bit Rate:

Sometimes also referred to as Data Rate, this is a term used to describe the amount of digital information (bits) that is conveyed or recorded per unit of time. In our world, this is typically expressed as an amount of bits per second (bit/s) that make up the digital video signal or recording. The higher the bit rate, the greater the amount of information being transmitted, and generally speaking the higher quality the video signal. When talking about a compressed video format such as DVCPRO HD or XDCAM, the bit rate refers to the amount of data recorded in a second. As the bit rate of a given compression increases, so does the amount of data recorded in a second.

For example: the XDCAM EX format in its highest quality mode has a data rate of 35 Mbit/s (35 Million bits per second), which translates to about 50 minutes of footage per 16 GB SxS card.

Bit Depth:

Though the word ‘bit’ is also used in this term, Bit Depth actually describes something completely different. Bit Depth, aka Color Depth, describes the amount of information stored in each pixel of data. As you increase bit depth, you also increase the number of colors that can be represented. In the case of an 8-bit RGB image, each pixel has 8-bits of data per color (RGB), so for each color channel the pixel has 2⁸ = 256  possible variations. In the case of a 10-bit RGB image, each color channel would have 2¹⁰ = 1024 variations.

That wasn’t so bad, so now let’s go back to the original example of the 1080 60i Uncompressed 10-bit video stream:

1080 60i – This is the video format, which has a resolution of 1920 x 1080 and a frame rate of 60 interlaced fields per second (59.94 to be exact).

Uncompressed - This refers to the fact that there is no compression being applied to the video signal, but it also refers to the Bit Rate of the video signal. An uncompressed video signal has a bit rate of 1.485 Gbit/s, which is a whole lot of data. In fact, it would fill up a 256 GB hard drive in just under 22 minutes.

10-bit – This is the bit depth of the signal, so each pixel has 2¹⁰ = 1024 variation per channel.

Another example would be the XDCAM EX codec, which in 24p we would describe as a 1080 24p 35Mbit/s 8-bit compression.

1080 24p – This is the video format, which has a resolution of 1920 x 1080 and a frame rate of 24 (23.98) progressive images a second.

35 Mbit/s – This is the bit rate of the video format, which is greatly compressed from the uncompressed video signal. The compression makes the video much smaller to store on a hard drive; on a 256 gb hard drive I could store up to 800 minutes of footage in XDCAM EX format. Many people believe that increasing bit rate also means increasing quality, however the quality of the image is greatly dependent on the compression type, which can often deliver excellent results at much lower data rates. This is the case with XDCAM EX or the AVC HD / AVCCAM formats available in many cameras today.

8-bit – This is the bit depth of the signal, so each pixel has 2⁸ = 256 variations per channel.  While an 8-bit type of compression reduces the number of color variations, XDCAM EX cameras, as well as other camera’s with 8-bit compression, are designed to minimize any effects that are associated with an 8-bit color space. The good news is that many cameras with 8-bit recording formats also output a 10-bit HD-SDI signal. Combined with a camera mounted recorder, you can greatly increase the overall quality of your recording.

I hope this guide has helped clear up any mysteries about bit rate and bit depth. In the future we will also cover chroma sampling and color space.

A few weeks ago, Jesse Rosen wrote an article explaining the technical definition of LOG-encoded video and how it differs from a viewing gamma such as REC709. To better understand how you might use LOG capture in production, I shot this video illustrating the differences.

REC709 is a monitoring standard for a final image; it is the contrast and color that generally replicates what you see to eye. But, it doesn’t necessarily show everything the camera’s sensor is capable of capturing. With the latest wave of Digital Cinema cameras, such as the ARRI ALEXA, the Dynamic Range detectable by our shooting devices has expanded dramatically, approaching that of film and pushing past what REC709 is capable of showing. To capture the imaging capabilities of these latest generation sensors, a LOG gamma can be used.

There are many situations where you do not have full control over lighting conditions, such as a bright window in the background or a city street at night. Even with full control, it is still common to color correct a project once it is edited together. If the image is captured in REC709, the ability to adjust the final image is limited by the contrast range already imposed on the image. LOG is a flatter contrast, allowing far greater image manipulation. Take a look at the video to get a sense of the difference.

Our eyes perceive brightness differently than an electronic sensor does. Video monitors are locked into a standard for color and contrast, but now Digital Cinema camera systems are being used to capture images that are used and displayed in media that can see well beyond those limitations. Brightness is captured on a Linear scale (LIN), and usually stored with video gamma (REC709) or with a more film-like Logarithmic encoding (LOG). What does this mean, how does one use it, and what capabilities does it represent?

In the first of a series of Technical Resource articles and CineTechnica posts, I present an introduction to the concept of LOG, its origins and capabilities. Click here for the article.

It seems that every few days, we hear about a new camera or a new lens. While it’s exciting to learn about new gear, it can also be a bit confusing. This is especially true when it comes to cameras with different sensor sizes. For a long time, it was only 2/3″, 16mm and 35mm in the professional world, but now we have a huge variety of sensor sizes.

Mitch Gross recently published his sensor size comparison chart, which is an excellent reference that has been very popular with our viewers. Following his lead, I developed this free online tool which allows you to compare different sensors at different focal lengths. We call it the Focal Length Comparator. Funny name I know, but it allows you to compare the fields of view of different sensors’ sizes at different focal lengths.

For instance, it can be used to visualize the field of view difference between a 50mm lens on a 2/3″ sensor and the same lens on an ARRI ALEXA sensor. I also threw in a bunch of extra information like crop factor, angle of view and specific sensor sizes. Check out this new online tool by clicking on the image above, or learn more about this tool at our online resource.

As a standard we started with the Super-35 film frame. For clarity, an aspect ratio of 16×9 was selected for all of the cameras, which is both the native aspect ratio for most of the cameras as well as the HD framing standard. For cameras that can shoot various aspect ratios (like the REDs), the largest available 16×9 motion picture shooting mode was selected.

On every sensor, the Super-35 image area has been indicated in Kodak yellow, with the individual camera’s image area indicated by a full color image. So on sensors larger than Super-35, that sensor’s extra image area has been indicated in full color surrounding the yellow Super-35 area. On sensors smaller than Super-35, the smaller image area has been indicated in full color, surrounded by the yellow Super-35 area. One will notice that on several of the sensors the difference is almost invisible, and to additionally help understand the difference in size there’s also a number indicating the percentage difference compared to Super-35 (a sensor a little bigger than Super-35 may be 103% while one much smaller might be 85%).

Underneath each image, the Shooting Format indicates the setting selected on the camera, with the physical size of the selected Shooting Format noted alongside in millimeters. The required Image Circle (projection size of light from the lens) has also been indicated in millimeters.

The chart is not meant to be a comparison of resolution or image quality in any way, but rather a specific comparison of relative frame sizes and their resulting fields of view.

Click here to download a larger .pdf version of the 35mm Digital Sensor Comparison Chart.

(Click image for bigger)

You may already know that MTF is an abbreviation for Modulation Transfer Function. MTF provides a way to objectively measure image sharpness in a practical and repeatable way that also correlates well with subjective perceived image sharpness.

The smallest practical unit for measuring image detail is a line pair, which is a black line and a white line, side by side. The measurement is line pairs per millimeter, and not lines per millimeter. It is very important to note that we are always talking about a black and a white line pair, and we are counting them as one line pair (1 lp/mm). In the digital realm, we need at least one pixel to represent the black point, and one to represent the white point. So we need a minimum of 2 pixels to image 1 line pair.

Each line pair is equivalent to a cycle and can be measured in either the horizontal or vertical direction. MTF is always measured in cycles per millimeter, which is the same as line pairs per millimeter. We use the term spatial frequency to describe coarse image details. In the old method of measuring resolution, low spatial frequencies would correspond to the larger black and white bars on traditional lens test charts. For example, 5 line pairs per millimeter (lp/mm) and 10 lp/mm are considered low spatial frequencies. High spatial frequencies are usually 40 lp/mm and above, and can go as high as 200 lp/mm, or even more for some highly resolving lenses.

A low resolution lens (low contrast at high spatial frequencies), but with good contrast at low spatial frequencies, can be perceived to be as “sharp” as a lens with much higher resolution (higher contrast at high spatial frequencies) but does not exceed the contrast of the low resolution lens at low spatial frequencies.

Looking at the two lenses on a lens test projector, it would be easy to see that one lens has higher resolution. The higher resolving lens might show the individual black and white lines on the 200 cycles per millimeter target, while the other would show just a gray square in the same place on the target (0% MTF = zero contrast between black and white lines = gray). Yet the lower resolving lens can look quite good through the camera’s viewfinder. This is because, at smaller viewed image sizes, the human eye is more sensitive to the contrast between coarse image details than it is to finer image details.

So how do we quantify the difference between those two lenses in a practical way? If we had the MTF graphs from both lenses, it would be possible to look at the size of the area “under the curve,” and with some experience interpreting the graphs, quickly estimate that the low resolving lens can still be quite good for most applications where the final output does not require the higher resolution (web video, SDTV and, in many cases, HDTV).

But if the intended use was for 2K digital cinema projection on a large screen, or a 2K digital intermediate, the higher resolution lens might give a better result, especially if its coarse image detail contrast is as good as, or close to, the low resolution lens.

MTF measurement is valuable because it provides a method of measuring image resolution that is completely objective. The traditional method of shooting test charts (or projecting test reticles) with resolution test patterns (usually squares consisting of black and white line pairs of a particular spatial frequency) is too subjective. It depends to greatly on the observer’s subjective interpretation of whether or not the target is resolved. Also, MTF is the only practical sharpness measurement method that would allow you to choose the right lens in our hypothetical example above.

Breaking down the term MTF, modulation in this use means contrast. Transfer function in this use essentially means “the ability to reproduce contrast.” In other words, what percentage of the subject contrast that is “fed into” that part of the imaging system (for the sake of simplicity we will consider lenses only) will be “output” by it onto the image plane.

For example, a lens with 50% MTF at 40 cycles per millimeter (c/mm) at f2 (T2.1) will produce an image where a 40 c/mm target will have half the contrast of the target (test chart, reticle or slit). So 50% is the transfer function at 40 c/mm and f2.

In the example photos below (shot with a digital camera from the image projected by a lens test projector), the older lens on the left (Fig. 1) is resolving 100 lp/mm and possibly 140 lp/mm, at least in the X (horizontal) axis. But it is doing so with very low contrast. Some observers would argue that even the 100 lp/mm target is not well resolved.

FIg. 1

The new, high optical performance lens on the left (Fig. 2) clearly resolves 200 lp/mm. But notice that the white bars at 200 lp/mm are slightly gray, not pure white as the lower spatial frequency targets from 10 lp/mm up to about 50 lp/mm. What you are seeing is the natural MTF response fall off of even the best performing lenses. Contrast at 200 lp/mm cannot be as high as contrast at 50 lp/mm. Looking at Fig. 1, our old lens again, notice that even the low spatial frequency targets at 10 and 12 lp/mm are still gray, not white. The lens tested in Fig. 1 has very low MTF, even at low spatial frequencies.

Fig. 2

So how much MTF do you need and at what spatial frequency? In the digital cine world, a good rough guide can be the pixel pitch of the camera’s sensor. No sensor can record details at higher spatial frequencies than the pixel pitch (which must be divided by two since, as mentioned above, you need two pixels per cycle or line pair).

The Phantom HD Gold camera, for example, has 12.5 µm (micron) photosites (pixels), giving a pixel pitch of 80 pixels per millimeter. This corresponds to 40 cycles per millimeter. 50% MTF at 40 c/mm is good optical performance.

MTF can also be used to measure and calibrate back focus (incorrectly but commonly referred to as “collimation”, but a lens cannot be collimated, only light can be collimated). The advantage of using MTF for back focus is that it will give very consistent results with a large group of test operators. Many lens manufacturers prefer it for this purpose.

MTF is a very complex subject, and this blog entry only scratches the surface. I will return to this subject in the future to explore it in more depth.