Creative Storytelling in web, iOS, and Film

JVC GY-LS300 Camera — Is it a Solid 4K Solution for Run and Gun?

By on Aug 3, 2015 in Blog, Film Studio |

We’re making the migration to 4K. HD video is terrific, and we’ve had great success with it for many years now. Today, 4K is not a fad. It’s not up and coming. It’s the future and it’s present. Today.

So, we’ve been testing, shooting, reading, and doing the things most people do when they’re looking for a new bit of kit. Thanks to the terrific people at EVS in Burbank (and the very eager to please folks at JVC), I was given the opportunity to spend a week with the new JVC GY-LS300 camera. As I was traveling up the coast for meetings related to our iOS Studio, I figured I’d take it with me, shoot some stuff, and let EVS and JVC know what I thought.

If we spin the throw-back-lifetime wheel, I used to do a lot of product reviews. NAB, NAMM, CES, and multiple auto shows were the core of my collaborations, and it was great fun to check out a product before it reached the marketplace. For the last decade, testing gear had fallen off my radar – too busy with productions, deliverables, and clients. But now that we’re seeking real world answers to help us create better looking video for our clients, testing is once again hugely important.

The first thing I should share is that we don’t test like many reviewers do. There are plenty of reviews that go over the specs, the accuracy of those specs, etc. What most of those reviews don’t do is create real projects. Our testing is based on using the product (whatever it is) in the manner in which we’re seeking to solve a problem. In this case, it’s a run and gun 4K video camera.

After reading other people’s reviews, it seemed as if JVC’s latest jump into the indie pro-camera market would be a huge hit. The specifications as noted are amazing. Really way beyond what one might have anticipated just a few months ago for a camera with a list price under $4,000.00.

Some of those key specs include:

  • Full size video camera body – quite different from a DSLR
  • Micro 4/3 lens design, with interchangeable lenses
  • Full s35mm sensor – and software that matches it up to any lens manufacturer, so no conversion math
  • SD cards, so standard storage
  • SDI and HDMI out (both active while shooting)
  • .MOV files, so easy to edit with

Get some of the other specs here and here.

I picked the camera up on a Friday afternoon, checked it out, and put in the car and hit the road. I used it on multiple occasions over the week that followed, and while doing so, wrote several email notes to one of our most trusted DPs, Casey Goode. I was frank with him. No, I was blunt. In fact, I was furious. Casey (don’t blame him) suggested I put my email notes into a blog post. I’ll ass some bits here, but want to be fair – and everyone should note that there are many types of video camera users out there, and many of them will love this camera.

Why so serious?

I decided that to test the camera, I would shoot a run and gun documentary style piece. I chose the very cool Railway near Santa Cruz, the Roaring Camp lines. They have an old Shay locomotive that chugs up into the redwoods. There’s a lot of light changes, elevation changes, movement, detail, and all the things that make shooting with a camera important to storytelling.

I thought things were gonna be pretty cool when I got there. As I was purchasing my ticket, the young man behind the counter asked, “hey, is that the new JVC LS 300 camera?” What? Here? If this guy knew about it, that might be a good sign. But, reality set in when he followed up with, “is it any good? I heard it sucks.” I told him I was here to test it out, and hoped it would be fabulous.

What happened next was hugely frustrating. This camera does 85% of what it does really well. The other 15%… From my first email to Casey:

Been working with the JVC…

From a spec point of view, it’s pretty amazing.

Feels great to hold and at first glance is really comfortable – it’s a video camera design, not a DSLR!

But…

If you attempt to use this as a tool – you know, like a CAMERA, the thing falls on its fucking face. I mean faceplant right in front of the principal.

In auto focus, the camera “hunts” and cannot pick any image to settle on. So, nothing is ever in focus unless you’re shooting the side of a barn.

Any movement results in Disney Mr. Toad’s Wild Ride art effects. Note how crisp the coal is, but how soft and “painted” the flames look. Crap!!!!
The focus tools make images seem to be in focus via the multiple viewfinder/displays, but they aren’t – they’re soft. 
The above is supposed to be in focus. It’s soft. CRAP!!!! It sure looked to be in focus when I shot it!
The viewfinder is total crap.
The flip-out display is extra total crap. If it cost $2 bucks, I’d be surprised. The buyer at JVC got a bowl of soup when he made that purchase.
The “SHOTGUN” mic they give you is actually a lame lemon squeezer 22 cap gun mic. Sound quality is total Legoland, but.. um, not so awesome.
Oh, and if you try to capture anything with motion, as in panning, um… this sensor runs home to mama. 
See the distortion in the image above? That’s with a Vedra 16mm lens at 2.2 and a ND 1/64. Um, yeah… sucks.
Here is another example. Check out the sign and post. That’s as clear as it gets. Meanwhile, look at the images in the background – sharp as a tack. Whiskey Tango Foxtrot! And, we were moving at about 5mph. Sweet Baby Jesus this is a lame camera…
How could JVC release this total K-Mart Return Bin piece of junk???? I mean, it is so close to being a home run, except it isn’t. It’s a fly out. I mean… it’s really unusable. 
Stunning…
Oh, and the night stuff I shot yesterday…. LOL!!!! Every clip is corrupted. Unusable. No idea why. Bad dates, Indy?
I’m gonna shoot some stuff again tomorrow and see if anything changes, but I kinda doubt it. I will try shooting in HD at a higher frame rate, but if you want to shoot 4K… 
So, if you are gonna point this at someone for a portrait interview, the images should rock the house. Same for a wedding. But… you can use a DSLR for half the cost (GH-4). I don’t think either of those uses require SDI out or a s35mm sensor. 
Note that all of these images were shot in 4K/24fps, but downconverted by QT, as it doesn’t display 4K.
Why so serious?
When I first looked at the initial shots I grabbed, it was clear the images were fantastic. Unfortunately, as I started to rely on the camera to help me grab images that are essential to telling a story, the camera didn’t cooperate. My email to Casey explained it.
I took it out the next day and tried a different routine. I shot various images with it – and then shot the same material using one of my trusted Panasonic GH4 DSLR cameras. It’s also a micro 4/3 camera, so I was able to use the same glass on both cameras, leveling the playing field a bit.
My next email to Casey wasn’t any less serious…
As noted, once I get back into town, I’ll organize a proper write up on the experience of being used by the little bastard.
For now, here are a few additional off the cuff comments for your interim consideration. As you know, my humor is a reflection of 40 + years doing this work, and being on the cutting edge of technology for so many years with Apple, Radius, SuperMac, Macromedia, Adobe and others.
Today, the weather didn’t really cooperate for the types of shots I wanted to get.
In the afternoon, I did drive around a bit and this time, I brought along my Panny GH4.
I took some shots with the same setup both on each camera and location wise – the tripod wasn’t moved, the lens in use on both camera bodies was the same, speed, fps, exposure, etc.
Note that I didn’t do this as a scientific test, nor in fact did I do this as a “review” in the traditional sense. I did this from the perspective of a camera operator setting up the camera and getting the shot – as per my pervious email note – using this as a tool to get work done.
I’ll leave all of the snarky comments about functionality, etc. to my overall follow up once back in SoCal, but a few things jump out that I’m going to share here:
Check out this shot of a tractor – Lumix 300mm lens. THIS is what a s35mm sensor can do!
Here is the same shot with a GH4 (same lens, settings, etc.):
HUGE difference. The top image looks SO much better (no color correction either). Just one thing… the color in the bottom image is more accurate. Yup. The day was totally lousy with haze and June gloom, but that s35 sensor really does a job of grabbing detail.
Now, check out the flowers below – blowing in the wind. The focus assist told me this was in focus. So, what you’re seeing is either interframe junk or the sensor not processing fast enough based on 24fps (or some other possibility that only 10 year old genius engineers can explain):
Now, check out the GH4. Not nearly the image overall, but more accurate color, and less noise (arguable) in the movement. I can tell you that the image plays back better from the GH4 data.
The next shot is of water – lots of white and motion. Below is the JVC example:
Now the GH4 example:
The sensor is a really big part of this puzzle, and these experiences made me even more frustrated.
My overall opinion hasn’t changed much, but it’s clear (no pun) that the sensor in the JVC is capable of some fantastic resolution. It’s also BLUE as you can see. Makes a Sony imager look absolutely warm by comparison…
I’ll share more detail once I get back to the studio, but I think the bottom line for me is that it looks and feels like the JVC was developed be a collaboration of marketing and engineering types. While those elements are also used by other manufacturers, I would be amazed if JVC had INVESTED in getting “in-use” input from filmmakers, documentarians, or journalists.
The JVC 300 is without question impossible to live with as a daily A camera. It could potentially be a specialty camera (pending a sample cut in with color correction). The GH4 doesn’t capture the same image, but when viewed on its own, its an acceptable image, and the functionality of the camera (even considering the drawback of being a DSLR body) blows the JVC away in most examples.
And let me tell you, I SO wanted to like the JVC. We’re about to embark on a TV series for a built-in audience of nearly 700,000 people and it would have been awesome to have picked up several of these. I’ll share more in my follow up next week, but as noted, I have yet to find the magic I was hoping for.
I shared my impressions with EVS and JVC as well. I attempted to be balanced. The camera produces remarkable images under certain conditions. As noted early on, if the 15% of things that aren’t great can be remedied, this would be a remarkable camera for a number of different applications. In fact, if I was shooting a wedding, a portrait interview, or a cinematic locked off shot, it’s probably pretty great right now (as long as you have an external monitor).
As you might guess, in its current release, this isn’t a camera we can use. It is a camera that others may love. And, JVC is serious about making this little beast a hit. Perhaps if they update the firmware, replace the viewfinder, upgrade the quality of some of the plastics, and partner with a third party for an external monitor, the camera would be a valuable production tool – as long as the total price stays under $4,000.00.
Bottom line: I wanted to love it. It came so close to being amazing. But shooting run and gun is unforgiving, and as such, my issue with this camera was my use of it. The camera is bitchin’ for what it does. With a little more effort, it could be a game changer.
UPDATE! JVC is to be commended for listening to their customers. Since this review was written, the company has taken multiple steps to improve the visual capture, bugs noted, and overall performance of the camera. As such, this is now a camera that merits serious consideration. Great price, improved functionality, and the ability to help camera operators capture gorgeous images. Well done!