Cinematic looking footage is the holy grail for most videographers at the moment.

What is it?

Depends on who you ask.

Generally, shallow depth of field, high sharpness veiled in overall “bloomy” softness with good dynamic range, especially in highlights with a little added colour magic seems to be the core it.

How do we get it?

The sharpness thing is fun. Most videographers complain about the over-sharp video look, so they soften it with a variety of antique lenses, some introduced haze or filtering.

Odd really. We chased sharpness for ages, now it has become the enemy for some. I guess the ideal is “natural clarity” rather than “clinical perfection”.

It seems you need to achieve it first, then you need to take the edge off it. Unsharp is just poor and gritty not acceptably mushy. Clean and sharp first, then smooth and soft combined is the business.

Depth of field is a double edged thing. You do not always need shallow depth, but you need a feeling of compression and grandeur, even at smaller apertures. This is often why full frame cameras and anamorphic lenses are used, because they add this easily, but M43 can also. Super-35 film, the Hollywood format of choice for decades is actually as close to M43 as full frame.

Dynamic range and colour science are intertwined also, because one without the other is less than half the story. Hard to explain, but you know it when you see it, there is an edge provided by some cameras, even with smaller sensors, when compared to many larger sensor cameras, often down to capture profiles, colour gamut and processing options. The BMPCC4k for example has that “something special” even though it is dated, less sharp than newer cameras and uses a smaller sensor.

Lighting. This is nothing to do with the camera, so anyone can achieve this.

*

So, where does that leave digital video, or more to the point using the bulk of cameras used as they were designed?

I have recently discovered a few things.

I love the Standard profile from the G9 with full sharpness (i.e. left at “0”, maybe reduced to -2 or -3, still testing), but with contrast dropped to -5 for some headroom. When I first researched the G9 and video, everyone** pushed Natural at -5/-5/x/x. I did as I was told, but it has never been a perfect fit. Natural tends to make highlights a little milky and I keep trying to bring the contrast back up, often too harshly. Standard profile gives me crisp whites, reducing the need to push contrast.

This breaks many of the above rules.

It is still contrasty*, but has a tiny bit of head room in post. The trick is, get it right in camera. I am competing with iPhones on auto here, which are not to be sneezed at, but I have the advantage of sensor size.

It has limited dynamic range*, probably around 8-9 stops, but that just means I have to pick my fights (locations), live with dramatically crushed shadows or blown out highlight or preferably supply some lighting. Often if I do not have much control over the location it is because TV are dictating a setup, which actually helps me no end.

If I keep the histogram centred or to the left, I am usually ok. I have a custom setting with Cine-D loaded just in case and might get the Vlog-L key.

It is too sharp by the cinematic school of thinking*. So why do I love how sharp it looks? A very mild Black Mist is available or some post. We will see.

I get basically straight from the camera, what I process towards. Just like shooting JPEG’s. If I nail this process, my footage does go beyond the basic video on the run look.

*

Because shooting both stills and video is tough, I need to get some things sorted. The above helps, but more has to be addressed.

Leaving a Variable ND filter or a BM, even a weak one on my main shooting lens for the paper does not make a lot of sense and changing mid stream is also not viable, so I am going to probably carry a dedicated “cinema” rigged out lens, maybe the legacy 25 Pen lens, or the 17 with the 15 as my low light stills shooter.

In good light the 17 is fine, but in low light, the 9 or 25 could cover me.

My filter kit is a mess, with multiple sizes and redundancies, but this is where poor choices and hoarding come to the rescue. I can basically kit a single 46mm lens out for video just for the work kit and have a full set of 62mm filters for the main video kit on the 12-40 zoom and 67’s for the S5 rig.

Using an Olympus lens on both G9’s keeps manual focus consistency and feels more natural to me with the advantage of the tactile manual snap back (the switches on the Lumix lenses are not). The Sigma 30 goes the Pana way, but will be used more for static interviews as would the S5’s lenses. If I get an S5 mk2 I would have serviceable AF available. This is for run-and-gun shooting so these two set-ups need to be intuitive.

So, back to the original question. Cinematic-soft or snappy-modern video? I actually like the snappier video look with maybe the slightest softness.

*If I add a 1/8 Black Mist filter I get some pleasant highlight softening and slightly increased dynamic range.

**Rich photo was the rare exception and I have to say, I liked his super sharp, super saturated footage.