Category Archives: Tests

A Tale of Two Lenses

Canon TV C-16 21.6mm-129mm F2.7 Zoom

Canon TV C-16 21.6mm-129mm F2.7 Zoom

Since my Panasonic GH series cameras can be adapted to use just about any lens, I am always keeping an eye out at thrift stores for vintage 35mm still camera and16mm motion picture and TV lenses. Usually when you find a lens in a thrift store it is usually either overpriced, not very special, or is in poor condition. I was very pleasantly surprised when my wife found not one, but two interesting lenses in the same week at the same thrift store. Continue reading

It’s Not a DSLR, It’s Not a Camcorder, It’s Black Magic!

Black Magic Cinema Camera

I’ve been shooting with the Black Magic Cinema Camera for the last month and what follows are my first impressions. In part one of this series on the Black Magic Cinema Camera, I will explain why this camera is different than DSLRs or even pro camcorders

Everybody knows that the Black Magic Cinema Camera was a revolutionary product when it was unveiled at the 2012 NAB show. Why then do they try to compare it to any other camera, and especially the Canon 5DMK2 DSLR? Continue reading

Panasonic Lumix GH2 Mid and wide angle lens tests

I was very excited to get my Panasonic Lumix GH2 DSLR for a couple reasons, the most important being the the extended tele converter zoom mode which adds a 2.4X zoom to the normal 4/3s sensor field of view by actually using 1920X 1080 pixels from the sensor without interpolating the resolution. This is also good in that it allows the use of traditional C-mount 16mm film lenses with the GH2 without using the digital zoom. So here are a series of test using various lenses with the GH2 in both the Extended Tele Converter zoom mode and in the normal zoom mode.
There are a couple things to note, while in the the Extended Tele Converter mode the camera does not output  an HDMI signal, the new HDMI output does help  with focusing this camera. All of  these tests were shot at extremely low light levels (about 26 Lux) The camera is set to ISO 3200 and the shutter speed is set to 1/50 sec. and all the footage was shot  23.97 FPS at 1080P.  I then edited all the footage on a 720P timeline in Final Cut Pro with minimal grading (only white matching). The C-mount lenses all were using the same Rainbow Imaging M4/3 to C-mount adapter. The 2/3′ broadcast lens used my custom  fabricated M4/3 to B3 lens adapter. Also note ththis test was to test lenses with a less than 50mm full frame FOV.
I’ll lead off with the most pleasant surprise. The first lens is a no name Chinese C-mount 25mm F1.4 CCTV surveillance lens that I picked up on Ebay for approx. $20. This is a retro-focus lens, meaning that the focus adjustment is behind the iris. The shot was shot at F1.4 and is surprisingly clean over the entire FOV and I didn’t have to engage the ETC mode to achieve full sensor coverage. As long as you don’t have to pull focus during a shot, this is an excellent and cheap lens for low light 50mm equivalent shots.
The next lens is my sentimental favorite, this is a Angénieux retrofocus 9.5mm f2.2 lens. If you want that  60s  Cinéma vérité look, this is your lens. Using the GH2’s ETC mode this lens is another  50mm  equivalent lens, while slower than the Chinese lens, it has a nice clean center with a slight soft vignetting around the edges. I also shot with this lens without the ETC mode, which I’ll show in the wider angle shots.
The next lens is also a surprise in that I had written it of as a bad lens at one point because of it weird chromatic aberrations when zoomed in . (More on that later) but when set to wide open iris and zoom, the lens actually looked pretty good. This is another C-mount CCTV lens, but it is a Canon and it is both fast (f1.4) and wide (11-70mm zoom). Again this is another lens that ends up being a 50mm equivalent.
This next lens was a huge disappointment for me. Having used ENG cameras for many years, I was very excited when I adapted  a 2/3″ Fujinon 9.5-133mm 12X broadcasts lens from my old Ikegami HL-95 camera to a M4/3 mount. For all the effort, I am heartbroken at the results. While the lens is really old, I can’t see any fungus when inspecing the lens, and after several careful back focuses, I have to admit, this lens looks terrible. Given the potential for this lens I just have to say I am disappointed.
So, there is no shortage of  fast lens that will give you  a 50mm full equivalent  FOV. The hard part is finding an inexpensive lens that will give you a wider FOV.
The next test is where I took the 1080P vignetted footage from the Angénieux retrofocus 9.5mm f2.2 lens and dropped it into a 720P timeline at 100%. This is my favorite shot. The Angénieux has nice soft edges and a wide enough FOV (28mm equivalent FOV) for most wide angle close-ups. It’s a great “love scene” lens. If you can stand shooting a vignetted shot, knowing that you will crop it in post, this is a great lens.
Next up were the two common kit lens that come with the Lumix GH1 and GH2. Don’t misunderstand that these are inferior lenses,because they are not, Both are fine lenses in their own right. The only downside is that they are relatively slow. f4.0 for the 14-140mm and f3.5 for the 14-47mm. Both these lenses give you a approximately 28mm equivalent FOV and are quite clean from edge to edge. If you don’t need a long lens or  Optical Image Stabilization that the 14-140mm offers, then the 14-47 is a good lightweight lens 9and cheap).
The final set of test are just to show the natural vignetting of the the C-mount and 2/3″ lens. They are quite interesting to note the edge differences.

DIY GoPro Jaybilizer Jib Arm Test

This is a test of a DIY jib arm design for small, light cameras like the GoPro Hero. the boom is a 12′ paint extension pole. I need to work on the trapeze so that it has less friction, but overall I can see this as a useful tool. I want to try my Samsung HZ10W next, as the GoPro is not the best video quality and the fisheye lens is so wide that you cannot avoid lens flares. Shot at 720 60p conformed to 24Fps. ungraded, edited as Prores in FCP, compressed with the Elgato Turbo H264HD.

Slomo: Conform vs. Optical flow

This was a test comparing 24fps conform to optical flow.The 60 fps to 24 fps conform using Cinema Tools is about the limit before the footage looks to jerky. With Optical Flow in Apple Motion, shot duration (RAM size?) seems to be the limiting factor. If I go beyond 30%, I get stuttering and artifacting after about 400 frames. I know that you can get better results using simpler back grond and foreground shots. Here is a great tutorial from CrumplePop on using Optical flow: 1000fps for free – using Motion Optical Flow instead of Twixtor
Shot with Lumix GH1 with 28Mbs settings, post in FCP6 and Motion, compressed to 5Mbs with Elgato Turbo h.264HD. No grading.

GoPro on a Mic Boom

This is a test mounting the GoPro Hero HD on a microphone boom, with some care and a better boom operator, this might be a good solution for quick and dirty crane shots. Below zero temps outdoors made this a an indoor test. This was shot at 720P 60fps conformed to 24fps. Edited in FCP with no grading.
Note for animal lovers: All the trophy mounts were inherited from my Father-in-law.. I only shoot wildlife with cameras.

Why the (hacked) Panasonic GH1 is a better video camera than the Canon5DMk2

If you are a Canon 5Dmk2 fanboy you may want to shut your eyes. The lowly Panasonic GH1 just got supercharged with a firmware hack that makes a great HDSLR video camera even better. Here is my list of things that make the GH1 a better video camera than the 5D. First let’s take a look at the 800-pound gorilla in the room, The Canon has a full frame 35mm sensor, therefore, there is no crop factor when using 35mm lenses. Canon’s full size sensor is 4 times the size of the 4/3rds sensor in the GH1, which also means that the Canon must compress 4 time the data for the same bandwidth, all thing being equal (which they are not.) Supposedly the 5D is writing data at 35Mbs, compared to the (stock) 17Mbs of the GH1, which means that it’s still throwing away half again as much data as the GH1 (4 time the sensor and only twice as much data as the GH1).

GH1 Camera rig

My GH1 Camera Rig

But the GH1 has a secret, with the Ptools firmware hack from Tester13, the GH1 can shoot true 1080P 24fps with no pulldown at 50Mbs. And since both cameras use h.264 compression, with the software hack, the GH1 is compressing the data from its sensor far less.

But there’s more, the GH1 hack also enables the 1080P mode using MJPEG compression at 100Mbs for short shots. While this setting is kind of like using nitrous and it’s really pushing the limits of the camera, it is a real boon if you want to get the highest quality video possible. The advantage that motion JPEG has over h.264 is that it only uses spatial compression instead of the temporal and spatial compression of h.264. This means prettier pictures. His holiness, Phillip Bloom posted a great blog post on what the firmware hack brings to the GH1 which you can read on his blog.
I’ll admit that image quality is close enough between these two cameras that the 5D might win if that were the only issue. But the GH1 is a far more usable camera than the 5D, First off, for shooting video you can’t beat the fold out LCD screen on the GH1. It’s just better than the Live View screen on the 5D, period, even with a Z-finder. Second, the GH1 has continuous auto focus when using the kit lens. With 5D in live mode, you are stuck with whatever focus you had when you started recording. Thirdly, even un-hacked the GH1 can shoot 30fps 1/2 speed slow motion. You just shoot at 60fps 720p and play back the footage at 30fps, no frame blending or doubling, just plain good old fashioned over-cranking.
Okay, Let’s talk about glass. The 5D still uses an optical viewfinder, which means it has a pentaprism and a mirror, which it must flip up to take video, which, of course, makes the optical viewfinder unusable. This arrangement also means that the Canon has a deep flange distance. the flange is the measurement from the back of the lens to the surface of the sensor. Traditional film and video lenses have a short flange distance and don’t work well with 35mm SLR bodies. The GH1 has no optical viewfinder and therefore has a very short flange distance, which means it can use practically any 35mm SLR, video or film lens that someone is willing to make a adapter for. There is even a PL mount adapter that cost as much as the GH1, so that you can use cine lenses that cost 100 times as much the GH1. I use all my old Canon FD SLR lens as well as some Minolta, Olympus, and16mm film lenses and even cheap CCTV lens with my GH1.
Finally, let’s take a look at the form factor. The 5D is big and relatively heavy DSLR. Great if you’re trying to impress other photographers or the bridesmaids at a wedding shoot, not so good if you’re actually trying to shoot steady handheld video. The GH1 is lighter, smaller, more balanced and just plain easier to handle.
I won’t argue that the Canon 5DMK2 is a great still camera, and better than the GH1. But when it comes to real world video shooting and video quality, 5D, you’ve been served!

Timelapse test GH1

A time lapse test with the Panasonic GH1 with kit lens shooting 1/1000sec shutter in shutter priority. F stop varied between 8 and 4.1. Some shutter “wow”. I’m thinking I should have shot f4.5 in manual. any advice is welcome. Next time , I want to compare using apeture priority. as long as the shutter speed doesn’t exceed the the interval time I should be okay.
Location: Golden Colorado