Although there were many new cameras showcased at NAB 2014, I still think the BlackMagic Pocket Cinema Camera stands strong capable of shooting RAW video direct to SD Cards for under $1K. A camera I feel i’ll still be working with quite a bit this year. It was especially great to see Redrock Micro alleviate some of the ergonomic hurdles with their new RetroFlex LCD ViewFinder + Cage + RunStop Video Remote Handle.
With the optional top shoe clamp, you can add 15mm rails. With the optional HDMI lock added to the cage you can prevent accidental disconnections or damage to ports on the BMPCC.
This kit is definitely at the top of my list once it becomes available. So far the RetroFlex Rig is the only BMPCC cage on the market with seamless integration with an LCD Viewfinder (something I feel is a necessity with the BMPCC). If Redrock Micro finds a way to port this retro stylish viewfinder + cage + handle kit over to other camera bodies (like GH4 / Sony A7), this could be an entirely new line for them.
The answer to that is quite simple. You shoot composites to achieve physical impossibilities and overcome aesthetic difficulties. For example, most of the time you couldn't easily bill a client to take a model, crew and gear to a remote jungle for a shoot.
Aesthetically, if you want an HDR-style background with lots of visceral light that isn't there on vacation, it's quite difficult to cut around your model in order to separately style them and the background, and if you get grip in the shot, you'll have to clone or crop it out if possible. If not, tough luck.
On the other hand, you probably could bill that client for the use of jungle shots as background plates that you took on vacation a few years ago. And if you shoot on green screen or white seamless, it's easy to cut out your model and remove any grip in the sides of the shot. Plus, your background will be perfect since it'll be shot and processed like a landscape photo.
In essence, we composite because it's the easiest, most practical and cost-effective route to get the final image we envision.
Shooting for Compositing
There are some considerations that must be made when shooting for compositing. There are also extra steps you need to take when you're not doing the post-work, but you're simply sending in the RAW files to a separate designer or digital artist. The most important things are lighting and posing.
The lighting must be able to work in a variety of scenarios, wherever the subject may be composited into. If you're lucky, you'll know beforehand and will be able to work with that knowledge to set your lighting.
In this case, and likely the more usual scenario, I have no idea where my model is going to be put so I have to ensure that there's enough data to allow a reasonable amount of manipulation and re-lighting of both him and the background in order to blend them together. I'll look at exactly what I did next.
Because you don't know what format, market, or media the image will end up in, you can't be sure that one pose may work where another won't. So you have to shoot several varying poses in order to accommodate the variety of end-uses the image could see. I'll look at my posing shortly.
1. Tearing Off Skin?
What's this about tearing off skin? Well, that's the main aspect of what we're working on for this shoot. The idea is to have the model tearing his skin off, revealing his "true colours" or "true nature" underneath.
How are we supposed to do this? It's actually quite simple. We just need to have another material doing the actual tearing, that we can integrate into Brad's skin seamlessly. The obvious choice for torso-tearing like this is simply a skin-coloured t-shirt.
We went with that over a neutral-coloured shirt (a mid-light grey) that can be used to differentiate in Photoshop between the shirt to merge and the shirt to stylise (if it's branded apparel) or remove (if it's some kind of VFX shot).
So, in order to get the skin, we do a shot with no shirt. This provides skin colour, texture, specularity and muscle contours. Then you have to do an almost identical shot (certainly identical around the manipulated area, anyway) with the grey shirt under the skin-toned shirt, where the outer shirt is being torn in whatever way you want.
Thus the contours of the tearing shirt can be blended with the texture and contours of the skin, which should be relatively easy due to it being a similar tone.
2. Versatile Lighting
I didn't have time to keep changing lighting and shooting every pose under every lighting scenario I can think of that Brad could be comped into. So I have to come up with a single versatile lighting scenario that should cover most possibilities, particularly the more likely, dramatic options.
My solution is simple. Three-point edge lighting, close to white seamless. A five-foot octa from front-top replicates diffusive lighting from the sky for outdoors, while being burnable down to a dimmer non-specific indoor bounce light source or large window. This covers everything from mid-afternoon outside to a futuristic gym to a stone-walled dungeon corridor.
Then the edge lighting provides a reasonable rim, not too hard, which can be blended with outdoor backlighting or toned down a little for interior work as necessary. One of my rim lights is a little iffy on the recycle time and is unpredictable whether it'll fire or not.
Knowing I would be turning Brad to camera left in any asymmetrical poses, I set this light on the right so it doesn't matter too much whether it fires or not on any one shot, since the result would still be naturally blendable in a dramatic graduated background (say, sunlight on the left, storm clouds on the right). These rimlights were angled inward so that they would light both Brad and the background, so I wouldn't have to worry about extra background lights taking up unnecessary performance space.
He was positioned quite close to a nine-foot white seamless so that he could also catch some bounce from the rim off of that surface at a slightly different angle, making him appear more enveloped and naturally lit. No grids here, natural light is chaotic.
I wanted to keep the poses relatively simple, bold and strong to suit the sports-fitness theme itself, but relatively easy to create in duplicate. I ended up going with just four poses with simple codenames: "stomach," "chest," "neck," "back," referring simply to where the tearing was taking place.
Brad was fantastically understanding and outgoing, so we nailed each pose with just one shot, usually with a second "for safety."
First was the shirt-off, this was where I had Brad do the full act, muscle tension and expression and all. I gave him something to pull against to make the arm muscles actually appear to be doing work. In hindsight, a couple of inches of 1/4" steel rod or 1/2" hardwood dowel would have been much better, as it could have been contained entirely within his hand and wouldn't need to be separately removed in Photoshop.
Next was the harder part of recreating the above four images with the shirts on and torn.
4. Reposing for compositing
Here we added the shirts, and I mostly "tore" them with a Stanley knife while Brad was wearing them (there's some trust!) to maintain control over its path. I did a bit at a time, in the same order we shot the previous images. The stomach tear was elongated to make the chest tear, which produced a little excess fabric but this could be gathered up in Brad's fist no problem.
I found that verbally guiding his extremities into place was difficult, with him being a mirror image of the camera's view, and I couldn't verbalise precisely enough. Brad was having a hard time gauging exact distances without a mirror or anything to work by.
This was where using a short focal length close to the subject came in handy. I could flip around the LCD panel, and stand in front of the camera, observing the LCD closely and then precisely replicating it by physically manipulating Brad's limbs. This was much quicker and easier, and since we were just replicating arms and torso we had all four shots done in well under ten minutes.
Pace and Tempo
Keeping up the pace of these kinds of shoots is important, you've got to keep it quick and light. Because there's no obvious interaction between subject and environment due to the fact that they'll be composited in later, they have to use their imagination to try to act on what they think is going on.
If you're unclear as to what the purpose of the shoot is and the strange posing or behaviour makes no sense to them, they'll start getting uncertain and bored and you'll lose them. Better to get the poses not-quite-perfect and keep the energy up.
Keep them involved all the way through. Show them any pre-production work you've done, any work of others that's similar technically or conceptually, and generally try to build a solid image in their head as to what is going on, both technically and creatively.
Since I wasn't sure what's going on creatively, I showed Brad a similar example of the end results (easy to find on sites like 500px and Behance) so that at least he could conceptualise the process and his part in it. The actual shoot was only about 25 minutes long, and I only took 17 shots all-told, including the one lighting test shot which didn't really require much interaction from Brad. Since this was the first time we were working together, I used this brief time to get to know him more and let him acclimatise to the studio environment.
To wrap up, here are the poses as side-by-sides, showing how close I got. The closer you get, the easier it is in post! None of these are quite perfect, but they're all close enough that post work isn't too difficult.
Well, that's all for now on the shooting aspects, I hope you've enjoyed the ride so far. Be sure to check out Thiago Storino's tutorial covering the compositing and manipulation of my images.
Lately, we have been seeing a lot of artwork depicting models pulling away their skin to reveal something underneath. The team from Ars Thanea first developed, and popularized this effect back in 2013 in a series of images they did for the Sao Paulo Football Club. Since then, the style has become very popular with artists, and we frequently receive tutorial requests for it.
In this tutorial, we will not only show how to create the effect in Photoshop, but we will also show how to light and shoot the photography, as well. In this part, we will explain how to extract the model from the background, remove his tattoos, as well as how to pull away his skin, and insert some mechanical objects underneath. Let's get started!
Speed Art Video
Full Video (No Audio)
For those of you that prefer video tutorials, we also included the full 1 hour and 30 minute, real-time video for you to refer to. While this video does not include any audio instruction, it should be a great reference if you get stuck on any of the steps below.
Before you begin, see this tutorial from Rob Taylor that explains how the key photos for this tutorial were taken.
To complete this tutorial, you will need the following assets. Please download them before you begin. If they are not available, or if you do not wish to purchase them, you will need to find alternatives.
Open _MG_6590.DNG and with Pen Tool (P), separate the model from the background. Create a path around the model, press Command/Ctrl-Enter to make a selection and then use Feather (Shift-F6) around 0.5 to make the selection softer.
In the next few steps we will need to retouch the model just a bit. Create a new layer, and with the Clone Stamp Tool (S) remove the imperfections and tattoos from the skin, as well as the objects that the model is holding.
Now, with the Pen Tool (P), create a selection of the teeth and use Hue/Saturation to decrease the yellow tone of the model's teeth. Select Yellows and set the Saturation to -80.
To accentuate the lighting on the shorts, make a selection, and with the Curves adjustment, increase the highlights.
Then, mask following the direction of light.
To reduce the eye's shadow, use the Curves adjustment again, and mask the area between the eyes and nose.
Then, mask following the direction of light.
Now use the Dodge and Burn tools to accentuate the highlights and shadows. First, press Alt key and click on the Create a new layer icon at the bottom of the Layers panel.
In the dialogue box that opens, change the Mode to Overlay and mark the Fill with Overlay-neutral color 50% gray option at the bottom.
Then, with the Dodge tool, paint as shown below.
After that, use the Burn tool, to accentuate the shadows using the settings below.
2. Insert the Background
Open the old factory image and position it behind the model.
To isolate the model, you will need to blur the background with the Lens Blur filter, (Filter > Blur > Lens Blur) using the settings shown below.
3. Pull Away the Skin
Open _MG_6598.DNG, rotate and position the ripped shirt like below.
Then, with a soft brush, mask to show only the rips and folds. Make sure not to mask over the hands. Make a path around the fingers leaving the thumb out, you will not use it.
Now, you'll adjust color and contrast. First, make a merged copy of the model, duplicate all his layers and merge with Command/Ctrl-E, then link to the rip layer and set the Blend Mode to Color and Opacity to 75%.
After that, create a Gradient Map layer and set like below.
Now, adjust the contrast with Curves.
The next step will be adding more folds on the chest. Open _MG_6595.DNG and cut out some of the folds.
Then, with a soft brush, mask some parts to better integrate.
After that, use the same techniques used before on the ripped clothing. Once again, duplicate the model's layers and with a soft brush, mask it to reveal the skin texture where the shirt texture is very evident.
Using Curves, create the shadow of the hands on his chest.
To finalize this part, use Dodge and Burn, as was done previously.
4. Add Mechanical Parts
Open the machine image, and with Hue/Saturation, totally remove the color, then rotate and position like below.
Then, position the gear image and desaturate it, as well.
To equalize the contrast, use Curves as shown in the following image.
With a soft brush, mask to unify the images.
Using Curves, mask to create the shadow between the skin and mechanical parts.
Now, add some wires on the chest, you can place them as you wish. Open the wires image and with Pen Tool, cut out some wires pieces.
Scale the wires to a compatible size, then rotate and position them where you want them to be placed.
Desaturate with Hue/Saturarion set to -100 and mask to the wires look like some are passing behind some of the mechanical parts.
Using Curves, make shadows on the wires.
Group all the layers in a folder using Command/Ctrl-G, and on the folder, create a mask with path for the rip.
To finish this part, use Color Balance add a blue tone.
5. Environment and Mood
Create a Hue/Saturation layer and set Saturation to -25.
For contrast, create a new Black & White adjustment layer and set the Blend Mode to Soft Light and 30% opacity.
Use Color Balance to add warm tones. To conclude this tutorial, add two Color Lookups adjustments. First, use the Crisp Warm option with 30% opacity. With the second one, use Film Stock 50 and set the opacity to 20%.
In this tutorial, we showed you how to combine a series of photos that we shot in a tutorial in our Photography Category with some stock photos to create a photo manipulation inspired by Ars Thanea's Sao Paulo Football Club Series. In the process, we explained several techniques that can be applied to your projects including how to pull back the skin, how to extract a subject from its background, as well as how to combine stock photos to create mechanical objects. We hope that you learned something from this tutorial and can use the techniques used to create something exciting of your own.
There have been a few affordably priced 4k monitors on the market, but all of them were limited to 30 hz refresh rate. While 30 hz is fine for video, it can be a little painful for normal computing tasks. The Samsung 28 inch 4k monitor is one of the first somewhat affordable 4k panels running at 60 hz to hit the market. At a price of $599 you get a 3840 x 2160 (Ultra HD) workspace in a 28 inch desktop friendly package.
Originally the estimated price on the Samsung 28 inch 4k monitor was $699, but amazon dropped the price to $599 this morning so I ended up pulling the trigger. This will be replacing my much loved Catleap 2560×1440 panel that I’ve been using over the last few years.
Apparently one of the benefits of the Samsung 28 inch 4k monitor is that you can run the monitor off of a single displayport cable at 60 hz making it much more attractive than some of the dual DVI 4k options. I’ll let you know how it turns out when it finally shows up. Might have to upgrade my graphics card, not sure if my HD 7970 GHz Edition or my GTX 680 4GB will be enough to handle a 4k panel. I’ll only be using this monitor for editing so I might end up with enough GPU power to get by.
On a side note, if you are running Nvidia cards in SLI there is currently no support for this monitor. So if that’s your setup you might want to hold off until Nvidia finally releases an updated driver.
In a new episode of our talk format ON THE COUCH, Greg Crosby from G-Technology gives us a run-down of their new storage products targeted specifically to the increasing storage needs of professional filmmakers and creatives.
Their new Studio series was designed with the new Mac Pro in mind, both in looks and functionality: it supports Thunderbolt 2, uses server-grade 7200 3.5″ HDDs in a 2-bay (G-Raid Studio) and a 4-bay (G-Speed Studio) version with up to a whopping 24TB. The G-Raid Studio supports RAID-0, RAID-1 and JBOD, while the G-Speed Studio is a hardware RAID that supports RAID-0, -1, -5, -6 and -10 configurations. The Helium technology allows individual hard drives to store up to an amazing 6TB now.
Greg also talks about the their new SSD version of the G-Drive ev, made for the versatile G-Dock, which is a very convenient Thunderbolt solution with swappable hard drives that can be used individually via USB-3. This has become my go-to on-set no-brainer storage workflow solution and it’s perfect when the editor works with a G-Dock as well.
We chatted about the increasing needs for storage with 4K and RAW footage taking over the filmmaking world, and what that means for on-set backup strategies.
G-Speed Studio, with the top opened to remove the individual drives.
While cinema5D is not sponsored by G-Technology, I personally am one of their brand ambassadors and occassionally speak on their behalf about my work and how I use their products in my workflows.
The AX100 is of a species we thought extinct in the filmmaking world… the consumer camcorder. I had a brief hands on with the camera to find out how much of it is high resolution point and shoot and how much cinema camera.
Every Sony camera I have ever owned has been cutting edge inside, Mr Bean on the outside. Thankfully Sony’s ergonomics did take a step forwards with the RX series and those ergonomics were semi-successfully transplanted into the newest Alpha cameras like the A7 and A6000.
Despite the presence of an X in the model number, unfortunately AX100 has no such luck! What is necessary here is a total rethink of Sony’s camera ergonomics. I really hope the technology goes into the RX10 M2 because in the Handycam form factor you have to grapple with the world’s worst touch screen, all the important physical controls on one dial, horrible button placement and the same confusing ‘computer-like’ approach to overall design which effected the NEX-VG E-mount camcorders and even the semi-pro Sony FS100.
Technologically though this is an interesting camera. The specs are exciting, but Sony have created some confusion over the megapixel count. To video people they say it is 14MP. When talking about the stills mode of the camera they say 20MP. This had me thinking for a moment that maybe it’s a new 14MP sensor, the lower megapixel count making for better 4K and low light performance. Dig deeper down the specs sheet though and Sony say it is a 20.9MP effective pixel count. The RX10′s sensor which is similar has 20.2MP count. Sony use a 14MP crop of it in the AX100 for 4K video, which begs the question… how? The 20MP sensor is 5.4k wide. Taking a 3.8k crop from that is going result in quite a dramatic reduction in sensor size. In stills mode on the RX10 when set to 16:9 the output is 17MP.
When I had my hands on the AX100 I wasn’t specifically looking for a crop in field of view or really expecting it. All I noticed was a dramatic punch-in effect when active image stabilisation was turned on instead of standard optical stabilisation. I’ll have to take another look at this when I get a look at the camera again so follow EOSHD on Twitter with 14,200 others for updates!
The sensor is 13.2 x 8.8mm so a 2.7x crop over full frame at best. Obviously with a F2.8-F4 lens you’re not going to get the look of a GH4 at F1.4. Neither will you have a full frame feel to images with a shallow depth of field permeating even wider fields of view, though at the longer end of the zoom you will get a shallow depth of field. The lens is 9-111mm F2.8-4. Pretty good! Overall you can expect a Super 16mm aesthetic from this camera in terms of the sensor and lens.
I’d have much rather seen that long end sacrificed for the inclusion of the lens from the RX10. That is an 8.8-73mm but has a constant F2.8 aperture and very ergonomic stepless aperture ring. Here Sony thought the longer zoom and more powerful stabiliser would be what consumers wanted. They also felt it would be more convenient to change exposure by hitting a button to put it into auto-mode. Manually changing the aperture is done with a press of the iris button for manual control and a swipe of the jog wheel, then another swipe, and another swipe, and another swipe, and another swipe, then you may get it from F5.6 to F2.8 by the time sunset begins. Best light at magic hour! I don’t disagree with Sony when it comes ‘what consumers want’ but it make little sense creatively. Consumers need to learn to shut up and face the wall in focus groups… Or they don’t get their cookie afterwards. It’s a real shame to lose that constant aperture and stepless control where perhaps you need it most – for video!! On a video camera!
Live 4K HDMI output
I tried the camera’s live HDMI output to a 4K Sony display. I wanted to check if this could be used to bypass the internal XAVC codec and record uncompressed 4K to an Atomos Shogun in the future. Pretty amazingly, for a consumer camcorder, the image was clean of icons and noise grain was very fine, with an uncompressed and unprocessed feel. Resolution is indeed full 4K (2160p in the HDMI menu). Pretty exciting bonus feature, but there’s some lag between moving the camera and the image updating over HDMI which may be a deal breaker for some.
I was unable to tell from the TV if it was a 8bit or 10bit output. Same goes for colour sampling. I expect it is 8bit 4:2:2 like the Sony A7S. This makes the AX100 a whole lot more attractive in terms of potential image quality. Unfortunately mounting the recorder is going to be tricky. There’s only a hotshoe and this is recessed so only Sony’s own accessories attach cleanly. Others will need some kind of third party adapter.
The camera menus over HDMI by the way are hideously pixilated. Time to get some 4K sprites Sony!
Noise grain was fine and uncompressed on the 4K TV, a sign that the signal bypasses the compression engine of the camera.
Slow-mo is accessed through the menus in two ways. The 720/120fps is labeled smooth slow mo and there’s also a golf shot mode with a mysterious unknown frame rate. Image quality is better in the smooth slow-mo mode. I wouldn’t put it up there with the GH4 and certainly not the FS700 but it’s not the worst I’ve seen. The golf shot mode is a curious feature. How many of Sony’s customers outside Japan go to a driving range every weekend? Does it change to Cricket Shot Mode when you put it in PAL mode!?
First impressions and opinion
This could be a good documentary camera or behind the scenes camera. It’s a good consumer and travel camera. Rich tourists and oligarchs will enjoy it for times on the beach when their tripod is replaced by a girlfriend.
For filmmakers the main shortcomings of the camera when it comes to the 4K image is the codec and ergonomics. The image profiles are uniformly terrible and the codec can be quite muddy. That said this mud is greatly reduced when scaling the 4K output in post to 2K. Colour is pretty nice, if rather high contrast. Scenes which don’t demand a lot of dynamic range look more filmic than the average camcorder that’s for sure.
For scenes where there’s a lot of contrast, roll off to the highlights is harsh and they blow easily, indeed the image looks like video in such circumstances. With the RX10 you could dial video very flat using the Alpha picture profiles. On the AX100 we have a Soft High Key effect which lifts the shadows but messes with colour and some toy camera filters. CinemaTone whether on or off doesn’t really offer any perceivable benefits to image quality. I have prior experience of CinemaTone on the NEX VG E-mount cameras and I ended up turning it off because it looked worse!
XAVC-S on the AX100 uses a bitrate of 60Mbit/s for 4K. It is a shame there’s no higher quality mode for those who don’t mind slightly larger file sizes. XAVC is an important new proprietary Sony standard. Why such a tepid implementation? Mass produced consumer devices like the A7S and AX100 are exactly what Sony need to make XAVC a standard for video recording. The AX100 makes XAVC look low-quality due to the bitrate and the A7S doesn’t bother with XAVC-S 4K at all! Very odd! The GH4 shows what is possible in terms of consumer codecs and that is cheaper than both the Sony XAVC-S cameras.
1080p in the AX100 is delivered in XAVC-S and AVCHD formats.
I am really on the fence whether or not to buy this camera. Some shots require the speed of operation and zero-rigging such a camera as the AX100 can offer. Unfortunately direct manual control over the image is painfully slow and fiddly. In once sense it is convenient to carry around a camera with a very high quality sensor more in-line with a DSLR than a camcorder, 4K, XAVC, small file sizes, built in ND filter, 9-111mm zoom, very effective stabilisation and 20MP stills mode all in one. Stick it all in auto and just go and shoot. There is appeal in this but the ergonomics of the AX100 spoil the fun. The touch screen is truly diabolical. You can’t even swipe up and down on menus like on a phone. It’s almost in need of a stylus, so small are some of the buttons.
If the RX10 successor gets 4K, I may consider adopting this technology. Until then, I feel 4K alone is not enough to revive the Handycam for anything more than consumer point & shoot.
Here’s a rumour of the latest GoPro specs. Source is not known to me so take it with a pinch of salt until it goes official!
The GoPro 4 is expected to shoot 4K and sure enough judging from this rumour it will, at up to 30fps. But perhaps of more use to GoPro users will be higher quality 1080p at 120fps for slow-mo and 720p at 240fps.
The following is from the supposably leaked press release:
GoPro Hero 4 is really capable of recording video in 4K resolution at a speed of 30 frames per second, 1080p at 120fps and 720p at 240fps, which became possible with the advent of the new SoC Ambarella A9 chip (dual A9 ARM Core ® CortexTM-A9 processor 1 GHz and FPU acceleration ) for advanced applications, wireless connection, and streaming video in social media.
GoPro Hero 4 will also come with a 13-megapixel photosensor and a completely new lens for shooting in dark. Multi-Exposure and HDR WDR tone mapping, electronic image stabilization, also improved MCTF. Dual sensor interfaces 12-lane SLVDS/HiSPiTM/subLVDS, 4-lane, remote viewfinder playback which is extremely easy to use. You shouldn’t to have special technical knowledge or skills to do it. GoPro Hero cameras are amazing because they have a waterproof housing with a flat lens for shooting under water (up to 60m). Built-in Wi-Fi module, USB & HDMI – the only thing that brings it closer to homeliness. The camera has the ability to capture still images during video recording and perform encoding H.264/ BP / MP / HP Level 5.1 and MJPEG.
GoPro Hero 4 will hit shelves in the summer of 2014. Action Cameras of GoPro have become extremely popular as an excellent way to capture high-quality video at high speeds relative to the price that usually is $340.
Varavon had some very interesting prototypes at their booth this year at NAB. They had one of the sexest looking Panasonic GH3/GH4 cages I came across on the floor. Machined from a single block of aluminum, the reps said this was actually one of the first prototypes from the factory floor. The final version will be anodized black.
The Varavon Panasonic camera cage has a nice lip at the front of the unit to prevent any kind of camera twisting and built in cold shoe adapter at the top of the unit. It’s less of a protective cage and more of a all in one mounting solution.
There are cut outs around pretty much every control you’d need to reach on the camera with plenty of 1/4 20 mounting points. They’ve also included a nice leather hand strap for hand held use. It’s hard to get an idea of the actual size of the cage from these photos, but the camera cage is slightly smaller than a Canon 6d body.
Varavon also included a locking mic port as well as a cut out for the LCD screen that allows for full rotation. The reps said there would also be an optional HDMI locking port available upon release. Estimated street price is $200 for the cage and leather handle, but they weren’t sure if the HDMI lock would be included in that price or if it would be sold separately.
There are cages like the PNC Honu Panasonic cage for around $99 but the bolted side and top plates tend to make them flimsy. I haven’t seen a whole lot of other options in this price range that look as well polished. This will be on my buy list once it’s released, should be a perfect fit for the Panasonic GH4.