By Keith Mullin
After some unexpected delays in shipping, due to a pre-release update to the side grip, we were excited to receive our rental EVA1 at the tail end of November. What followed was a mad rush to shoot some sample footage alongside the Sony FS7 and FS5 cameras for an upcoming event at our shop. It’s taken a few weeks but I’ve finally had a chance to fully analyze and write up an initial review and comparison of the EVA1
Initial Impressions
If you are not familiar with the EVA1, it is Panasonic’s answer to the aforementioned FS7 and FS5 cameras, as well as the C200 and C300 from Canon. It is an attempt to bring the features and capabilities of Panasonic’s high end Varicam and Varicam LT cinema cameras into the sub $10,000 range. With a price tag of $7,400, it sits squarely in the middle of the Sony FS5 and FS7M2 cameras. The original FS7, which is still a current product from Sony, is only $100 more than the EVA1.
The big question is how does it stack up in features and capabilities?
Unboxing
Taking the EVA1 out of the box for the first time I was struck by how much of a clone of the FS5 the camera is visually. Almost every feature looks eerily similar. Here are just a few examples.
The camera feels cheaply made. Everything is plastic and feels like it. Holding the camera I can’t help but think about how well the individual components are fitted together. The finish on the plastic is also a bit on the slippery side. These are all things one would not really expect from a $7400 camera.
The body itself is a bit bigger than the FS5 and around the same weight.
The hand grip, which has already seen one major problem being fixed by a full replacement of the handle, has a couple of other flaws. First of all, adjusting the hand grip is not as convenient as it is on the FS5. To rotate the handle you have to move your thumb way out of natural positioning, to the point that you can’t really do it without a firm grasp on the camera with your supporting hand. This will make for more difficult adjustments on the fly.
Secondly, and maybe it’s a bit nitpicky, the camera will not sit flat unless the hand grip is in it’s fully upright position. The FS5 handle was very specifically designed to allow the camera to sit flat no matter the orientation of the grip. It might seem like a small thing, but it’s the attention to the little details that make for a really great user experience.
Other Features
A definite positive feature of the EVA1 is the inclusion of a timecode I/O port. This is a feature that is sorely lacking on the Sony cameras, the FS7 in particular since it sees a lot more work on higher end multi-camera shoots where timecode is a key tool. At least with the FS7 you can add timecode capability with the optional XDCA, but on the FS5 you are just out of luck.
The LCD viewfinder on the EVA1 has a built in sun hood, which is a great feature. The flip side is that the LCD on the EVA1 is highly reflective and is difficult to see if there is any extraneous light falling on it.
The EF lens mount is limiting. It is a bit baffling as to why Panasonic would license the mount from Canon when they have a perfectly functional, and highly adaptable, Micro 4/3 mount available to them. Perhaps this is because native Micro 4/3 lenses will not work with the larger Super35 sensor, but with the proper Metabones adapter you could have almost as much flexibility as the native E mount of the FS5 and FS7. In the end I picked lenses for my tests on whether or not they would work on the EVA1 since I knew I could adapt anything I needed to the FS5 and FS7 cameras.
Menus
Once the camera powered on I found the menu’s more or less logically laid out. I definitely didn’t struggle too much to find the functions I was looking for and was pretty familiar with the core functions of the camera after 20-30 minutes of poking around. That isn’t too bad for picking up an unfamiliar camera system. I have a high degree of doubt that anyone unfamiliar with the FS5 and FS7 would be in the same position after a half an hour.
One thing that jumped out at me while going through the menus was the recording codecs currently available. I knew that the camera was going to be limited to 30p in 4k until the firmware update, but I was not expecting that all of the recordings were going to be LongGOP, even in HD.
The Tests
Physical characteristics and menu structure is all well and good. But how does the camera LOOK? To find out I did a battery of side by side test shots with the FS5 and FS7. They are not the most scientific tests, and they are not field tests. I don’t have time to take 3 cameras out into the field and put them through their paces. The tests are also by no means exhaustive, there are a mind boggling number of ways to shoot these cameras once you start trying to do apples to apples. I decided to go with settings that were “most similar” and stuck to them.
Methodology
Unless otherwise specified, the tests recordings were all made with the following settings.
EVA1:
- V-log/V-gamut
- S35 5.7k sensor mode
- Native only ISO (800 or 2500)
- 3840×2160
- 23.98
- 180 degree shutter
- MOV 422 LongGOP 150Mbps
- Middle Grey exposed at 42%
FS7:
- CineEI Mode
- Slog3/SGamut3.Cine
- Rated at 2000 ISO
- 3840×2160
- 23.98
- 180 degree shutter
- XAVC-I
- Middle Grey exposed at 41%
FS5:
- PP9
- Slog3/Sgamut3.Cine
- Rated at 2000 ISO
- 3840×2160
- 23.98
- 1/50 sec shutter
- XAVC-L 100Mbps
- Middle Grey exposed at 41%
If you know the Sony cameras you will note that these settings put the FS5 at a pretty substantial disadvantage, since the 4k codec on the FS5 is only an 8-bit 4:2:0, while the FS7 and EVA1 are recording 10-bit 4:2:2. You should also note that the EVA1 at this time is only capable of a 150Mbps LongGOP recording. In the future the plan is to update that to an intraframe codec with 400Mbps recording. A release date for the update is not known at the time of writing.
It should also be noted that these tests were performed with an FS72, although we would expect to see absolutely no difference with the original FS7 as the sensor and processing are identical.
Watch the video below to see the cameras in action and read on to learn a little bit more about the process and what I noticed working with the clips.
Color/Skin Tones
Yup, the EVA1 a color camera. Below you can see the ungraded and graded color charts. It was easy and quick to match all of the clips together using a waveform monitor and the less than stellar color controls found in FCPX (10.3.4). It was probably a 10 minute process to get the clips to matching as well as the do below. The match is not perfect, but its very close. If you were flipping directly between clips you would see some variance in a couple of the color chips, but they grey remains the same and the majority of the color chips dont show any noticeable shift. With a little more time and some better tools the results could be much better.
The FS5 sample is the one that stands out as not quite matching, and would likely take the most work to bring in line with the others. This is due in large part to the 8-bit recording on the FS5 having less overall color information than the 10-bit recordings from both the FS7 and EVA1.
But charts are one thing, how do the cameras handle skin tones?
The FS7 and EVA1 at 800 ISO definitely have the most nuance, with the FS7 tending a bit towards pink/red, at least as graded. The lighting ratio in the shot is a bit drastic on purpose. The EVA1 has a bit more of a pleasing shadow side of the face, but the FS7 retains more detail in the deep shadows. If you look over the subjects right shoulder you can make out a bit more detail in the objects on the counter behind.
The stained glass “window” above the chart is actually a Kinoflo Celeb with a custom made filter of stage lighting gels. The idea was to test the highlight retention capability of the cameras but I didn’t have the settings right, or the time to fiddle with it. If you watch the video you can see how much it has been brought back, but it was never really blown out. Based on the specifications for Slog3 and V-Log I would expect that the Sony cameras have more latitude in the highlights. V-log clips at 79 IRE while Slog3 clips at 93 IRE but they have nearly identical middle grey settings. Correction, V-log L clips at 79 IRE. V-log as it is in the EVA1 clips at around 95 IRE
Dual Native ISO
One of the big marketing points of the EVA1 is its dual native ISO feature. The feature is reminiscent of the dual ISO found on the larger Panasonic Varicam 35 and Varicam LT cameras. On the EVA1 the two base ISO ratings are 800 and 2500, or 1 2/3 stops, which is a full stop less of a difference in base ratings of the Varicam’s, which have base 800/5000. The premise behind this feature is that you can have two base ISO settings that yield similar noise profiles, one that can be used in normal shooting situations without needing a pile of ND filtration, and another for use in low light environments.
Having seen the Varicam perform this feat of magic first hand, I was looking forward to seeing what the EVA1 was able to do, understanding that it was a more budget friendly camera system. Unfortunately the EVA1 yielded disappointing results in this area. It should be kept in mind that the codec currently being used is not as robust as the one that is coming in future firmware updates, how much that will improve the EVA1’s performance is unclear.
As you can see in the above image there is quite a drastic difference in the noise profiles. Even more damning is how the EVA1 stacks up against the FS7.
In the above image it’s really a toss up as to which noise profile is preferable between the FS7 at 2000 ISO and the EVA1 at 800 ISO. The FS7 contains a touch more color noise, while the EVA1 contains a more pronounced grain, and at higher IRE. Given that the FS7 is 1 1/3 stops more sensitive, the FS7 would give you more flexibility to overexpose your footage to improve the signal to noise ratio in the shadows, or allow you to shoot in lower light than the EVA1 at 800 ISO.
But what about the EVA1 at 2500?
As you might have predicted, there is significantly more noise in the EVA1 footage at 2500 ISO than in the FS7 at 2000. It’s not even close. In fact the EVA1 at 2500 ISO has similar noise to the FS5 at 2000 ISO.
The bottom line is that the dual native ISO does not live up to the hype, and the over all sensitivity of the sensor leaves a lot to be desired. As stated above, these tests will be worth revisiting once the higher bit rate codec becomes available.
Low Light Performance
Pixel peeping aside we were interested in how the EVA1 performed in low light. For this test I bounced a blue filtered LED light off the ceiling of the room and then used a BIC lighter just below frame. Each camera was shot using a 50mm Leica-R lens using the iris to compensate for the 1/3 stop of sensitivity difference between the Sony cameras and the EVA1. Due to the extremely low lighting conditions it was not practical to shoot the EVA1 at base 800 ISO so only the 2500 ISO base was tested.
All of the cameras performed pretty well given the incredibly low light in the shot. Nothing in the scene, with the possible exception of some of the reflections off the glass, went above 25 IRE, so the images were definitely deep down in the shadows. If I had to pick, I think the FS7 retained just a touch more nuanced detail than the EVA1 or FS5. That detail is probably impossible to see in the still frame grabs, or on the uploaded video, due to compression. That is something that you should keep in mind when picking the camera for your next project. Where is it going to go when you finish it? What you see on your expensive 10-bit master monitor is not what everyone else is going to see.
Sharpness
Another of the selling features of the EVA1 is a 5.7k sensor super sampled down to 4k. In theory this will make for sharper images with better color accuracy than the 4k sensors in the Sony cameras. We’ve already taken a look at color, so lets look at sharpness.
The FS5 is definitely not as sharp as the FS7 or EVA1. Between the two more expensive cameras the EVA1 might squeak out just a tad sharper than the FS7 but it’s almost too close to call. I think the numbers and the edges of the lines are just a tiny bit clearer on the EVA1. It’s probably not enough of a difference to be noticeable in the majority of situations or to make for sharper post process zooms or extracts. As with the intraframe codec slated for a future there are plans for enabling RAW output of the full 5.7k sensor readout which could yield better results when down-scaled to 4k.
High Speed Recording
The EVA1, like the FS7 and FS5 has the capability of capturing high frame rates for slow motion playback. The EVA1 will record 120 fps and 240 fps continuously, with a few caveats. First caveat is that you have to change the sensor readout to be able to do high frame rate recording. For 120 fps you have to change the readout to 2.8k, and for 240 fps you are down to a 2.2k readout. It is likely that the Sony cameras are doing something similar behind the scenes, they just aren’t telling you. The second caveat with the EVA1 is that to reach 240 fps the camera does a Micro 4/3 crop of the sensor. The third caveat is that at 240 fps the ISO rating of the EVA1 is cut in half.
All of those caveats aside, the EVA1 does admirably well at high frame rates.
At 120 fps all of the cameras performed very well. There is some noise in the black on all of them, but nothing horrendous. The EVA1 has the least noise, but the FS7 has the sharpest image.
The EVA1 definitely had the least noise and artifacting of any of the cameras at 240 fps (180 fps on the FS7). However it did crop in and lose half of its sensitivity. Again the FS7 appears to be just a tad sharper than the EVA1. The FS5 was the clear loser in the 240 fps round with lots of noise and compression artifacts, particularly in the blacks and in the bowl of the glass.
The results of all of these high frame rate tests would probably be improved by not shooting in V-log or Slog3, which is already stretching what the codecs can do. The results might also be different had the FS7 been shot in center scan mode instead of full scan.
TL:DR
The EVA1 footage is very usable and easy to work with. Over all the EVA1 had an edge on the FS5, but didn’t quite stack up to the FS7. These results will be rendered moot (or at least subject to revision) once the EVA1 gets a firmware update.
As always, if you are interested in buying or renting any of these cameras, you can give Z Systems, Inc a call or shoot us an email.
