Digital Camera Video Basics
Update 26th May 2013
It is not that long ago, in 2005, that I were more than happy with my then Canon 350D. Prior to my 350D I owned the Konica Minolta A2. With the Minolta one could record basic movie files at something like 480 pixels by 360 pixels. With the later 350D, it was not possible to record movies. From that time onward the digital camera basically "exploded" in terms of absolute development. With each new model, processors were more powerful, the internal "operating" memory capacity increased, the reading / writing times were faster and today we at the point that digital cameras are capable of managing large amounts of digital data at lightning fast speeds, a condition perfectly OK for video. The mirror-less cameras plus live view options, were only another two important developments that prepared the "home" movie maker for serious DSLR and compact systems camera, video producing capabilities.
With this article I like to crunch literally books and pages of theory into one short article. It will therefor not be possible to go into any specific details with this article. The aim is to summarize the most important information and concepts into one short article and to thereby help you tie all those movie related loose ends. This article will also help you start your own research from a better and more clear basis.
From I started my own research on digital video recording, my own improved video knowledge cleared many faulty concepts I adapted when reading forums and reviews. In fact having gained a better perspective on how to plan and execute digital video recording projects, completely changed the way I approach my own video projects. Another consequence, was me adding the Panasonic GH2 to my camera bag;-) Amongst video enthusiasts as well as many professionals the GH2 is regarded as a movie making masterpiece. Personally, I enjoy every moment discovering the GH2 stills and video capabilities.
At the same time I have grown to better appreciate my humble E-M5 and its movie capability. With its amazing IBIS stabilizer, the E-M5 has turned out to become a home movie makers, dream tool. The E-M5 has less video "containers" & "codec" formats than for example the GH2 or the 7D. On the other hand, apply the correct video settings on the E-M5 and it will deliver excellent video quality. With IBIS active, it is as if the E-M5 are mounted on rails!! For example, the GH2 stabilization is also active during video recording, the challenge however are, lens based stabilization systems are mechanically restricted in its corrective movements. No other camera I tested could even remotely match the E-M5 video stabilization in movie mode.
Traditional cinema played a significant role in the creation of "acceptable" modern movie/video files. Modern Video-graphers generally like to "mimic" traditional movie recordings by applying at least three important "analog" concepts when recording digital "cinema" movies files. These "analog cinema" concepts again creates the typical theatre film viewing aspects like, shallow depth of field, 24 fps (frames per second) and the well known 180 degree shutter technique. Said frame rate combined with the180 degree shutter technique creates the movie flicker and motion blur patterns moviegoers grown accustomed to.
While doing my research for this article I saw that many modern video-graphers strictly use these older "cinema rules" and another newer breed video enthusiasts allow themselves more freedom while experimenting with the increased capability and options available with modern CSC and DSLR digital video recordings.
It is well worth researching these older analog recording techniques in more detail plus to learn more about older style cinema movie recorders and how and why film makers created the well known and used 180 degree shutter angle. This will help you understand why many older movie making traditions still exist today with "modern" digital camera video recordings. Fact is the modern DSLR and CSC (compact system cameras) like the 60D, 5D, GH1, GH2 and GH3 all enabled home movie makers the ability to create professionally looking home "cinema" movies at a fraction of the traditional film making costs.
In its basic form, the DSLR or CSC cameras are far from ideal when recording and producing movies. Within a very short time we witnessed a completely new world of digital camera "video" options becoming available on the digital video/film market. In its basic form, things like camera form factor, limited audio capabilities, different or inadequate video compression options and HD outputs all became limiting or plus factors when shooting movies with CSC and DSLR cameras.
For example when shooting video with high end cinema equipment, one will typically have a crew handling all the different aspect of the recording. One such crew member will for example manage only the focussing aspects of the recording. These crew members typically are referred to as the "Focus Puller".
In the above image you can see the focussing adjustment knob that are linked to the lens focussing ring. This method of focussing is more often found with DSLR cameras because the DSLR cannot focus while recording movies. CSC cameras has the ability to apply Contrast Focus while recording video. Cameras like the GH2 and the newer GH3 has increasingly better and more smooth focusing options while in the "video" recording phase. I found the E-M5 auto focussing to be capable while recording video, it is however important to regularly practice with the E-M5 and to know which of the different auto focus options are best for your style.
Reading reviews one do get the impression sensor size is important for doing high quality video recordings. When evaluating the different DSLR and CSC cameras one realize, different to still imaging that physical sensor size plus Mega Pixel counts are no key factors, determining video quality. The reason is full HD video only requires 1920 x 1080 pixels (2MP) files. What really makes the difference with DSLR or CSC cameras are the volume digital data these cameras can manage, continues imaging throughput plus finally things like high resolution camera screens or EVF's that are fitted to these "high end" cameras. Not to mention the large range of standard and specialized lenses or optics available for these cameras.
Lets take a moment to better understand the large amount of data that is created during a HD video recording. Each HD movie frame consist of a 2MP image. Each image holds 22'073'600 (2MP) bits of digital data. HD movies are recorded at a rate of 24 or 30 frames per second (fps). In many cases photographers prefer to record at even higher frame rates like 50 or 60 fps. Just think of the absolute disk space required to store the recorded video data second after second onto the disk. I think you will agree that without some type of compression any disk will quickly run out of memory space.
Another aspect to consider is the speed at which we generate the digital movie data. When we record at full HD (1920 x 1080), 30 frames per second, we generate 60MB of RAW (un-compressed) digital data each second. That means every second another 60MB of raw data is ready to be saved. One therefore need a memory card that can save 60MB of data every second. Having a quick look at my favorite 16GB, SD card and I saw it can only manage 20 MB each second. Fortunately modern video standards accommodate this challenge by applying different types of compression techniques to the RAW video files, therefore reducing the size of the saved video data. You will see your camera typically specify that it requires a memory card, class 6 or higher, cards that can handle 17 mbps or in extreme cases 24mbps (MB per Second). The new GH3 record up to 75MB per second.
Digital video compression is a complex subject and many different compression techniques are used today. I only briefly list and discuss the most important compression types. Please go search for more information on each if you think you need more information, it surely is an interesting subject. To put things into perspective I think it is important to understand that during the recording process the first type compression already takes place. The first type video compression takes place during the recording phase is referred to, as the first "scanning" phase. Two types scanning techniques are used, Interlaced or Progressive scanning. Next the data are saved in so-called containers and inside each container the data is saved in a specific codex format. Each if the three stages, scanning, container and codex, compress the video data and therefore potentially influence video quality and file size.
In terms of progressive and the interlaced scanning techniques, you will see videos are always specified as 720p or for example 1080i. The scanning type 720p stands for progressive scanning and the 1080i for interlaced scanning. A different way of expressing this are 720p50 or 1080i60. Translated 720p50 means 1920 x 720 progressive scan using a frame rate of 50 fps and 1080i60 means 1920 x 1080 interlaced scan using a frame rate of 60 fps (Frames per Second). I do recommend you to read more about the differences between these two formats, progressive and interlaced because it will help you understand the quality impact each has on the recorded and displayed video. Personally I tend to prefer using progressive scanning for all my work.
What is Frame Rate? With video we are challenged with things like flickr, 50 or 60Hz electrical energy, computer screen technologies and TV screen technologies. To produce steady movie files that also looks great on PC or TV screens, development engineers developed different recording plus display compression standards. One of these video quality "smoothing techniques" will display each frame twice, effectively doubling the recorded fps from 24fps to a new frame rate of 48 fps. Instead of using 48, the industry list it as 50 fps. Some newer cameras can manage the higher recorded data rates of 50 or 60fps and does not require the doubling up of frames to create 50 or 60fps.
The E-M5 specifies that it can record video at 30fps and thereafter save the recording as 60fps. The E-M5 therefor double up to create the 60fps frame rate. The new GH3 can record at a true 50 or 60 frames per second.
As said in the above paragraphs, another interesting aspects using DSLR or CSC cameras for video, is the focussing technique. Will video-graphers prefer auto or manual focus while recording video with their DSLR or CSC camera in the future. Both manual as well as auto focus options requires a slower and smoother action or movement, it should not hunt, stay out of focus for long periods or "hyper" react. During video recording, the automatic focus (AF) option on CSC cameras can tend to oscillate between in-focus and out-focus conditions.
The lens on the right was specifically designed for video. One of the unique aspects of this lens is the special "geared" type focus ring. Serious DSLR users will typically use a lens like this fitted with a unique focussing option to do the focussing manually during recording. (See above image)
Frame Rate & Shutter Speed
The next point I like to discuss are Frame Rate and Shutter Speed. You will see the E-M5 and other cameras like the GH2 will not let you adjust the shutter speed lower than 1/30th when recording at 30fps or a shutter speed of 1/24th when recording at 24fps. The reason simply is, while recording at 30 or 24 fps one cannot keep the shutter open for more than a 1/30th or 1/24th second. It is possible to increase the shutter speed and video-graphers generally recommended users to use a 1/60th second when recording at 30fps or a 1/50th second when recording at 24fps.
Again, one does not have to strictly apply these rules. You can also go ahead and record at 1/200th shutter speed if you have enough daylight exposure to do so. Just like with stills, the higher the shutter speed, the less risk one has for motion blur and the sharper the final image or video will be. In poor light a 30th second shutter speed could result in under exposure. My advice is to do more reading on this aspect of video work. We do not generally think of movie as we do with imaging, people typically expect video to be different. That is a mistake, All the rules we learned with imaging also applies to video recording. Depth of field, motion blur, framing and all the other rules apply with video just as it did with stills imaging.
If you do decide to record video at a fixed 1/60th shutter speed with your E-M5, keep in mind it could happen that you over or under expose the video recording, just as it would do in stills imaging. For this reason experienced video-graphers always keep ND filters ready. Reason is in bright daylight one will require a ND filter when recording at low shutter speeds and wide open apertures. At night you will have to again up the ISO to correctly expose your "video".
One would think that video are different to the typical stills imaging theories and that video expose differently. This is completely false. One of the best advantages using a DSLR or a CSC type camera is the freedom it provides in terms of depth of field, shutter and aperture effects. Master these aspects and you will shoot amazing results with your CSC or DSLR camera.
Container & Codec
Next I like you to take a look at how the different cameras prepare video material and in what format video files are saved. Video data is saved in a two step process, first step professionals referred to as Codec and in the final step professionals referred to it as placing the data in the AV Container. Codec really stands for "encoder/decoder" and is used to encode or compress the "RAW" recorded video data. The different "codec" formats typically are referred to as MPEG-4, Motion JPEG (MJPEG), H.264 or AVC. I read H.264 are the most widely used format today.
The next step is to insert the prepared codec into the so-called AV Container. There are two major AV container formats, Audio Video Interleave (AVI) and QuicktimeMovie (MOV). The E-M5 offer both types AV container formats. In the E-M5 section I will do a specific article that will discuss all this in more detail.
On the right you see one of the next most significant handicaps the DSLR owner has to manage when doing video. First the DSLR has to flip the mirror out the way and secondly the optical view finder becomes meaningless once you in video mode. The only way to then properly monitor recorded video are the camera screen. In direct sunlight this can be a real challenge. A special accessory like the one to the right helps the video-grapher to better "see" or view the external camera screen.
CSC cameras has a huge advantage in this respect with their high resolution EVF's plus in many cases the free moving camera screens. Between the EVF and the high resolution camera screens, the CSC video-grapher has virtually no reason not to be successful. I personally use my 7D only in static situations with the 7D fitted to a tri-pod.
On the other hand, when doing manual focus I find it often is close to impossible to correctly focus on the scene using the camera display or screen only. Using the EVF is much more effective making cameras like the E-M5 or the GH2 a real pleasure to work with.
Another important aspect of video work is audio or sound. Typically the producer will plan his or hers work well in advance before starting with the physical recording. This requires a recording scrip, describing all the major steps, speech, plus the main chapters or parts of the production. There are many sites with great information on how to plan and develop scripts.
The next step is audio. Cameras like the Panasonic GH2 or 3 offer several audio options and so does the E-M5. There are no reason not to record top quality audio during a video recording using any of these cameras today.
The 7D upgraded to version 2 firmware will enabled the operator several more audio options. Must say, older digital camera models like the 550D, the 7D or the GH2 are great for video plus they are much better priced on the 2nd hand market than newer models.
In special cases one can record video and the audio separately. During the editing phase one can then combine the audio and the video.
The final steps I like to discuss are the PC plus the Software you will use with your video recordings. What software one should use and how one should manage video recordings. Personally I use a MAC. Compared to my previous Window PC's, the MAC is just in a different class, working on video files. Personally I am an Adobe Creative Cloud subscriber, according to adobe the future way of working with graphical software. That means I have access to Lightroom, CS6 and Premiere Pro and all the other Adobe options in the creative suite. Both Lightroom 4 and CS6 are great for imaging and video editing. All I learned with digital imaging I can now use with video in CS6. I think it is really awesome to work with layers, apply curves and filters and whatever other CS commands when working with video files.
That said I think one should not under estimate the different MAC software options, iMovie and Final Cut Pro. These are very powerful software options and they deliver great results. I also found that Final Pro X is extremely powerful, offering many advantages over packages like CS Premier. In fact I am personally more inclined specializing with Final Cut Pro and not with the Adobe video tools in the future.
One often think it is only the specific camera model that will allow the user to record in slow motion or any of the other special video or movie effects. As you learn more you will see that this is not the case. Something like the E-M5 will deliver high quality video files at 720p30 and that combined with something like Final Cut Pro X and the sky is the limit, allowing the user endless options when cutting and building the final movie.
One can continue for pages more on this interesting subject. I hope with this short article you could fill the gaps and that you now know what to look for should you like to better your own video recording knowledge.