RED One Camera Activate All the 4Ks

How to Calculate RED Camera Data Rates

You don't want to be caught shooting RED with no inclination of how much footage can fit on a drive. So, I'm going to share with you the formula I use to determine how much data is pumping through the camera.

To be able to accurately determine data rates on the RED became one of the main motivations behind the RED One Pocket Guide. Finding this information online was tough and I spent a great deal of time figuring it out.

Understanding How the Data Works

To accurately determine data rates on the RED One requires an understanding of how the data is created, compressed and stored.

The basic unit of computational data is the bit. In imaging terms, the most basic unit is the pixel. When working with digital cinema footage, bits and pixels are intrinsically related. The more of either bits or pixels in a file, the bigger the file will be.

To give you an idea of scale, a single frame of RED footage at 4K and a 16:9 aspect ratio takes up 9,437,184 pixels.

Basically, the RED camera is able to move vast amounts of data and that data is dependent on four major factors:

Frame Rate

Frame rate refers to the amount of frames per second (fps) that the camera is recording. The more frames recorded per second, the more data moved since there will be more images created. For most film productions, this will be 24 fps.

Aspect Ratio

The amount of the sensor being used to record an image will affect data rates. A 16:9 aspect ratio, for instance, is larger than a 2.35:1 aspect ratio. Because of that, the 16:9 frame will use more pixels and more pixels = more data.

Resolution

RED has made no secret that it is able to shoot 4k (now 4.5K) of resolution. This refers to the number of lines of resolution made up by pixels. More resolution will be using more pixels meaning more data is moved.

Compression

Part of the reason for RED’s success has been their proprietary compression algorithm referred to as REDCODE. In short, REDCODE works to make files smaller while retaining resolution, colors, dynamic range, and more to the best of its ability.

Almost every digital camera uses compression, even if it is labeled as a “RAW” format.

REDCODE is available on the RED One in three settings: REDCODE 28, 36, and 42. In this naming system, the bigger the number, the less the image is compressed. The less the image is compressed, the more data there will be.

Since the development of RED Epic (and now it’s release), RED has switched these numbers to represent compression ratios — i.e. 10:1 or 15:1.

REDCODE is also a variable codec which means it does not compress at the same rate all the time. REDCODE is affected directly by image content.

This is a crucial concept to understand. The way that REDCODE compression works is by sampling an image and compressing on the fly. The more “going on” in an image, the more data there will be.

That means wide landscapes with many trees, different colors and movement will record more data than simply a black wall.

The Formula

So now that you understand the basic premise of how data works within the RED camera we can dive into the formula used to calculate it:

(pixel width x pixel height x framerate x 12)/8/1024/1024 = uncompressed data for one second of footage in megabytes

In the above formula, the 8/1024/1024 is the conversion of bits per second into megabytes per second and the 12 is the bitrate (4 bits per channel; three channels). These stay constant.

But let’s say you want to know how much footage will fit on a certain card or hard drive?

How To Use It

Using this formula, I calculated approximate data rates and subsequently the time a certain amount of gigabytes would hold of footage. Remember, these are approximate estimates since REDCODE is a variable codec.

Now I will take you step-by-step of that process using an example calculation.

For my example, I will calculate the amount of storage on an 8 gigabyte (GB) compact flash card shooting 3K resolution, 2:1 aspect ratio, 24 frames-per-second, at REDCODE 42 compression.

Step 1 – Calculate data for one second of footage

Formula: (pixel width x pixel height x framerate per second x 12)/8/1024/1024 = uncompressed data for one second of footage in megabytes (MB)

Example: (3072 x 1536 x 24 x 12)/8/1024/1024 = 162 MB/s

Using the formula, plug in the numbers that are specific to your camera settings. If you can’t find these numbers, they are published in the RED support manuals and FAQ.

Step 2 – Divide by compression ratio

Formula: (Step 1 result)/(compression number) = data rate at determined REDCODE setting

Example: 162/7.5 = 21.6 MB/s (data rate at REDCODE 42)

Once you have completed step 1, you end up with the amount of uncompressed data for one second of footage. Find out which REDCODE you will be shooting at and divide the number by that. For the new RED Epics, these will already be ratios, for the RED One here are the comparable ratios:

  • REDCODE 28 ~ 10:1
  • REDCODE 36 ~ 9:1
  • REDCODE 42 ~ 7.5:1

Once you have done that, you now have a compressed data rate in MB/second. If all you are looking to calculate is a data rate, you can stop at this step.

Step 3 – Calculate data rate in minutes

Formula: (Step 2 result) x 60 = Data rate in MB/min

Example: 21.6 x 60 = 1,296 MB/min

To determine the data rate in minutes, simply multiply your number by 60.

Step 4 – Determine and convert the storage capacity

Formula: (Usable storage in GB) x 1000 = Usable storage in MB

Example: 7.5 x 1000 = 7,500 MB of usable storage

The next step is to determine what storage unit you will be shooting on and the capacity of it. For instance, are you shooting on the new RED Solid State Drives or are you shooting on an 8 gigabyte compact flash card?

It’s important to remember at this step that due to the nature of formatting storage, there is not as much free space as advertised. I am not computer savvy enough to explain it in full detail (read more here) but everytime you format a drive or card for use, it takes up space.

In the RED One Pocket Guide, my rule of thumb was that there would be about 7.5 gigabytes of free space for every 8 gigabytes the storage unit was advertised at.

You should also convert your gigabytes into megabytes since that is what the data rate is in. To continue with my example, we were shooting on an 8 GB compact flash card of which approximately 7.5 gigabytes will be usable.

Step 5 – Divide storage capacity by the data rate

Formula: (Storage determined in step 4) / (data rate determined in step 3) = Total minutes able to fit on that particular storage unit

Example: 7500/1,296 = 5.78 minutes or 5 minutes and 47 seconds

Once you have determined your usable storage capacity, you can then figure out how much footage will fit onto it. Simply divide your data rate into the storage capacity and you end up with the amount of minutes that will fit on it.

Dealing with Data

Once you have this information, you should be able to more accurately determine the amount of footage you’re able to shoot in one day. I was always frustrated with RED for only releasing select data rates. I wanted to be able to get as exact as possible so that I could relay reliable info to a cinematographer or director.

Data management requires great care on set. Make sure you are using reliable hard drives and best practices when transferring footage. If possible, have somebody who’s entire job is to dump and check footage.

As I wrote in my 10 Things Every Camera Assistant Should Know About the RED Camera Post, “When a director or DP asks you how much time is left on a card, they don’t want to be told 2 GB or 25% – those numbers don’t really mean anything. Knowing how much footage is left, in terms of minutes, is the camera assistant’s job.”

  • Pingback: Hidden Cost of RED Epic and Digital Cinema Cameras - The Black and Blue - Tips for Filmmakers and Camera Assistants - The Black and Blue

  • Graeme

    You’ve made a small mistake there. Each channel is 12 bits, and there are three of them. So you need to multiply the number by three.

    • http://www.theblackandblue.com/ Evan

      Hey Graeme — I appreciate the comment, but I don’t believe you’re right. 

      Each channel is 4 bits making for a 12 bit image overall. I did make a mistake in my article saying “12 bits per channel” though and I’ve corrected that — so thanks for making me aware of that!

      Using my formula, the data rates align up almost perfectly with charts RED themselves have provided. Use the 3K 2:1, REDCODE 42 @ 24fps example to compare to this chart:
      http://www.redgrabs.com/up/1293435891.jpg

      The data rates also align up with my own experience and expectations for shooting times.

      • Marcus2012

        No, each channel is 12 bits, 24 bit computer images AKA true color uses 8 bits per channel, 32 bit uses 8 bits per channel and adds an alpha channel (which is used for transparency) He is correct.

  • Pingback: Shooting with RED Epic #7: The Difference Between REDCODE Data Ratios | The Black and Blue

  • Brian McClatchy

    Hi Evan,

    thanks for the formula, your calculation matches up quite well with real world test I have run so is a big help to calculate run times a various frame rates. Just wondering about the 16 / 12 bit debate. Red states that the Epic and Scarlet are 16 bit, and in an article in Wikipaedia on Redcode it states that it records 4 channels, 1 for Red, 2 for Green, and 1 for Blue. I assume the 12 in your formula stands for 12 bit, can you clarify if Redcode records at 12 or 16 bit?

    • http://www.theblackandblue.com/ Evan

      Hm I haven’t heard RED say that, but I don’t doubt you. If they say it’s 16 bit, it would make sense that 4 channels x 4 bits per channel = 16 bit.

  • http://www.facebook.com/marconcio.vonmarconcio Marconcio Von Marconcio

    4 bits per channel would mean 2x2x2x2 = 16 colour variations per channel, like some 90′s GIF image…