by Sebastian Wöber | 19th December 2016
Many people were thoroughly impressed when the new DJI Inspire 2 drone was announced. And rightfully so, as it features a 5.2K RAW camera, Apple ProRes, obstacle avoidance, redundancy systems and double the flying power of its predecessor – all at a fairly reasonable price. I spent a week testing the drone in the field and I’ve seen what the camera is capable of after testing it in our lab. In this DJI Inspire 2 review I will show you how the drone performs and what kind of image quality you can expect. Stay tuned for part II of our DJI Inspire 2 review. If you are interested in our DJI Inspire 2 LUT and Inspire 2 RAW to LOG converter, please see this article. Please note: I abided by all laws to make this video. Due to concerns of some viewers I would like to point out that flight safety is very important. You should at all times see your drone. My night shots were filmed at dawn and I did not fly above people, buildings or streets for them. I flew slightly next to the highway, not above it, the church shots were filmed from above the yard and not above people and the shot inside the fog has an invisible cut, so I myself was positioned above and below fog for a clear sight of the drone at all times. Determine all risks carefully. Achieving many of these shots is much more work than it might appear. DJI Inspire 2 Review – The Next Generation This video review was my most extensive and time-consuming ever, but the more I worked with the DJI Inspire 2, the more I found that there is no way around creating an in-depth hands-on piece which I hope you will enjoy and find useful. As the successor to the Inspire 1, the DJI Inspire 2 really marks the next generation of drone flying for me. I was already impressed with the Inspire 1 RAW, but the 2 is a much more serious cinematography tool as the images it can produce are truly remarkable. Please note that the YouTube/Vimeo compression, even in 4K is very strong. If you want to see how the actual image quality is in 4K, make sure you download the source file and watch it on a 4K screen before giving your final judgement to the image. Photos shot by Gavin Fürst I don’t want to repeat myself too much, as I mention most of what I found important in my hands-on video. In this written DJI Inspire 2 review, I want first of all to simply give you “the list”: PROs The X5S camera quality 5.2K clean image quality – depth and resolution is amazing Apple ProRes Codec integration is very handy The overall ergonomics (case, battery, setup time, charger, etc…) The extended flying time of up to 25 minutes (I think I managed to get 20) Improved flying stability for smoother shots Improved flying speed, ideal for aerial cinematography SSD workflow Safety features (redundant batteries and IMU) Wider viewing angle Acceptable low-light capabilities CONs ProRes color always defaults to “none” and burns in a “bad look” when not changed manually Obstacle avoidance triggers too quickly and ruins my “flyby” shots The DJI Go App is too crowded, some buttons are too small and popups are unorganised The image transmission often broke up too quickly, at a distance of around 800m Every second shot had a bad horizon, although this can be fixed in post easily Focusing is still horrible. You tap to auto-focus but never know if you’re in focus or not until after the shoot. It is clear that DJI had been working on further improving the ergonomics of the drone, which is very nice to see. Small things like the remote control automatically charging your phone or tablet, the inclusion of an automatic landing gear, self-heating batteries, dual charging – all of these all make your life easier, especially as a one-man operator. It’s easier and safer to fly the DJI Inspire 2 than any other drone before. I’m sure DJI will be working on future updates to address some of the issues I encountered. One should keep in mind that the first DJI Inspire 2 units are only just starting to ship to a few testers, and that the software for this complex machine is still in its early stages. In my correspondence with DJI, however, I did feel that the company took my feedback seriously. DJI Go 4 App. The Color of ProRes recordings right now s defaults back to “none” whenever you change format. The Camera – Zenmuse X5S At the time of this review, I still couldn’t get the more affordable Zenmuse X4S camera to work, so I was only looking at the higher-priced Zenmuse X5S. The perfomance of the Zenmuse X4S will be, among other things, one of the points we will look at in Part II of our DJI Inspire 2 Review. DJI Zenmuse X5S Camera – Gives you RAW and ProRes recording options. As mentioned in the video, the results I got from the Zenmuse X5S are truly remarkable. Not only could I match it pretty closely to the the colors of the ARRI Alexa, but the 5.2K resolution for me is rather mind-blowing. This kind of frame size is very useful, especially on aerial shots. That said, the dynamic range is not quite up there with the Alexa, and the colors of the Alexa are a tad smoother and more accurate out of the box. You should also know that the Zenmuse X5S has a micro four-thirds sensor, while the Alexa shines in super35. During my skintone test with our model, I used the 45mm Olympus lens, which I believe would be a 60mm equivalent on a super35 size. I used the Zeiss 50mm Cp2 macro lens on the Alexa. I don’t intend to suggest that you should shoot your whole movie with a drone, even though DJI tried that in their promo video, but I do think the Inspire 2 can be used in high-budget movie productions mixed with ARRI Alexa footage. It is nice to see how powerful the DJI Inspire 2 RAW files really are, coming out of such a small and ergonomic single-operator drone. DJI Inspire 2 – 5.2K RAW file, converted to LOG (right-click open for full size) And then there is the Apple ProRes codec, although please note that it is limited to 4K UHD. In my tests, I found that 4K ProRes gives you nicer results than using 4K RAW, because the 4K RAW files have some aliasing distortion. Instead, you should shoot 5.2K RAW and convert to 4K later for the best possible quality in 4K. I took night shots for the beginning of my film by intention, in order to see how the camera performs in lowlight. The compressed YouTube film looks rather noise, but the actual footage had less grain and was clean, so shooting a lit city at night is very much possble with the Inspire 2 Zenmuse X5S camera. I shot at around ISO 400, though it is hard to determine exactly as the source was RAW. We will do further tests and present the results in my DJI Inspire 2 Review Part II that is coming soon. Inspire RAW to LOG Conversion During my analysis, I found that the best workflow for .cdng RAW from both the Inspire 1 and Inspire 2 is a little complicated, so I made a separate article and video that describes the process in detail. There you will also find a download to conversion presets and LUTs that can be used to improve and optimize the post production workflow of Inspire 2 Raw and ProRes D-Log files. In particular I tried to come close to Arri Alexa Log C gamma and colors. In the test lab, the colors coming from the X5S and X5R (Inspire) cameras appeared to be off by quite a bit in comparison to an Arri ALEXA. The idea behind the workflow and presets/LUTs was to convert Inspire 2 Raw (Or Inspire 1 for that matter) to a well-preserving video format in order to edit and colorize later (without having to go back to Raw). And furthermore to match and get a neutral Gamma on all my video files, ideally right while transcoding, so I can apply LUTs and grade easily. Go HERE for the workflow tutorial DJI Inspire 2 Package – What else do I need? If you’re going for the high end DJI Inspire 2 Premium model with the Zenmuse X5S camera, here’s what I’d recommend you get: The X5S package comes with most of what you need to fly professionally. Some kits come without the 15mm lens. You can either get that or any other of the recommended MFT lenses. The drone comes with 2 batteries. These will let you fly for about 20-25 minutes. If you want more fly-time, get more batteries. I’d get 3 more sets, for a total of 8. This means I can spend about 2-3 hours in the field. I’d also recommend you get another charging hub and power adaptor, because the included one only charges 2 of the batteries at once, which takes about 90 minutes. If you want to pilot and have someone else operate the drone, this can be achieved with a second remote. And another set of propellers might come in handy in case one of them breaks for some reason, although it is better to stay safe and never let that happen. Ah yes, the Inspire 2 bundle will ship without SSDs, so if you plan to record anything you’ll have to get those as well. If you want to save money, you could go for two 120GB SSDs and only record ProRes 422 HQ in 4K UHD. If you want to do 5.2K, then two 480GB SSDs would be useful. In any case, I’d recommend to get two, because you can’t delete individual clips and transferring files takes ages, but also because if you lose or break one it would mean you can’t film anymore. Don’t forget the SSD Reader. The minimum complete package is a little bit more expensive than expected, costing about $8,000. Considering the package DJI promotes costs $6,000, that’s a large extra and reminds me of the experience of getting a RED camera. But considering what you get, $8,000 is a well-deserved price, and in comparison to any other drone + camera that achieves this kind of quality, you could even consider this affordable. Conclusion There is no question about it, the DJI Inspire 2 is an impressive drone and a testament to the irritatingly fast pace at which DJI is advancing drone technology way beyond the competition. If you are a single operator and you are looking for a drone with great ergonomics and flying power, as well as the best possible cinema-worthy image quality in a small package, then look no further. There is nothing comparable out there, as any large drone will only outperform this if it is equipped with an ARRI Alexa. Even then, the resolution of the DJI Inspire 2 has no match. Maybe the RED Helium 8K could compete, but we hear the color science is not quite ready yet. 5.2K on a 4K screen is really, really impressive, the ProRes integration is a big improvement and the intelligent features this drone brings to the table will be a great help on any shoot. I hope DJI can work out some of the issues I have pointed out, especially the default color on ProRes and the horrible focusing limitations, but overall, the Inspire 2 is a great step up from the Inspire 1. Now all we need is this camera on a shoulder mount, but I have a feeling this is exactly what DJI will have in store for us soon. What option would this leave the rest of the camera manufacturers? I have no clue. We hope you liked our DJI Inspire 2 Review and camera analysis. If you have any questions or thoughts please let us know in the comments. Special Thanks To musicbed.com for providing the song. Stray Theories – “We Never Left” www.straytheoriesmusic.com Skintone model – Ieva Pocytė Big thanks to Gavin Fürst for his help flying the drone, photos and navigating the city!!!Read more
by Sebastian Wöber | 19th October 2016
The Blackmagic URSA Mini 4.6K is a truly affordable cinema camera with impressive specs that houses Blackmagic’s newest sensor. When it was announced last year, Blackmagic Design once again won many filmmakers over. Now that the camera has started shipping, there are many positive, but also some negative voices. Let’s take a look at the guts of the 4K vs 4.6K Blackmagic URSA Mini cinema camera in our lab. Comparison: Blackmagic URSA Mini 4K vs 4.6K On the outside these two cameras look identical. Inside the body they probably also share most of the same innards. What really differentiates one from the other is mostly the sensor as far as we can tell. The 4K sensor on the Blackmagic URSA Mini 4K is the same that was used on the large URSA camera and on the Blackmagic Production Camera. The new 4.6K sensor is 15% larger and similar in size to the ARRI Alexa and Canon C300 mark II sensors. We will focus on testing several aspects of sensor performance and evaluate the image quality. To make this review fair, we will also throw the popular Sony FS7 into the mix as an additional reference camera. A look at the Specs Specs-wise these cameras are virtually identical. Both shoot up to 60fps in 4K. The main difference lies in their maximum resolution and sensor size. The fact that these cameras shoot in all flavours of the Apple ProRes codec, as well cinemaDNG RAW is their big plus. It is, in fact, an aspect where all Blackmagic cameras have an advantage over most other low-cost cinema cameras on the market. Blackmagic URSA Mini 4K Max Resolution: 4K (4000 x 2160) Max Framerate 4K: 60fps Max Framerate HD: 60fps Log Gamma: Film Log Sensor: Aps-C (21.12 x 11.88 mm) Mount: Canon EF or PL Codec Bitrate 4K: up to ProRes 444 XQ – 312.5 MB/s Price: About $3000 Blackmagic URSA Mini 4.6K Max Resolution: 4.6K (4608 x 2592) Max Framerate 4K: 60fps Max Framerate HD: 120fps (windowed) Log Gamma: Film Log Sensor: Super35 (25.34 x 14.25 mm) Mount: Canon EF or PL Codec Bitrate 4K: up to ProRes 444 XQ – 312.5 MB/s Price: About $5000 Dynamic Range A good dynamic range rating allows us to capture a larger range of shadows and highlights in high-contrast scenes, an important property when it comes to comparing the URSA Mini 4K vs 4.6K and one where a main difference will become apparent. We’re testing with a DSC labs XYLA-21 transmissive test chart. For our dynamic range tests we use the Zeiss 50mm Cp2 macro lens (more on how we test HERE). Our software measured about 12 stops of usable dynamic range on the Blackmagic URSA Mini 4.6K (RAW). This is very similar to the rating of the a7S II and C300 mark II. [Update:] How we tested: We measured dynamic range using uncompressed RAW with an ISO of both 800 and 1600. We decoded the files in DaVinci Resolve 12.5 with BMD Film 4.6K Gamma applied. We also tested dynamic range with Apple ProRes 422 HQ and had the same results. Here’s a screenshot of the dynamic range of a few popular cameras compared. In comparison to the URSA Mini 4.6K, our software measured about 8.5 stops of usable dynamic range on the Blackmagic URSA Mini 4K (RAW). The Sony FS7 reaches 12.5 stops. For each camera there are different reasons why the dynamic range is limited. The Sony FS7 image seems to become unstable in the lower stops due to processing. There is noise reduction which cancels out noise, but it’s also apparent that we quickly loose detail in the darker areas. The URSA Mini 4K simply doesn’t capture as much dynamic range as the other cameras. The URSA Mini 4.6K would have potential for more stops of range, but noise becomes stronger in the dark areas. Unfortunately there is a lot of pattern noise there, more than on the 4K, which makes darker areas of the image less usable. Here is a shot of the darker steps (11, 12, 13 and 14). Step 13 and 14 were not counted as valid range by our software, because there is too much noise. See the same image with raised gamma for better viewing below. Blackmagic URSA Mini 4.6K – Pattern Noise in Dynamic Range Step 11, 12, 13 and 14. Blackmagic URSA Mini 4.6K – Pattern Noise with raised Gamma ISO? The Blackmagic URSA Mini cameras are not strong when it comes to low light performance. In comparison, the FS7 has a greater range in terms of ISO. On the URSA Mini 4K there are only three settings: 200, 400, and 800. The URSA Mini 4.6K goes up to ISO 1600. In our tests we found that there is little difference in image quality when comparing all of the ISO speeds available on a single camera. It might seem so at first because the image gets brighter with higher ISOs, but in reality the lower ISO speeds merely cut off the image video range. In other words, the same image only gets coded differently at different ISO speeds, seemingly without any difference in processing whatsoever. That’s why we recommend to use the full video range in ProRes, in order to get the best color gradations. This can be achieved by using ISO 400 or 800 on the URSA Mini 4K and ISO 800 or 1600 on the URSA Mini 4.6K. When exposing your image, though, make sure that you do not underexpose the image, as it will look brighter with the higher ISO setting (800 on 4K and 1600 on 4.6K). If you want the best quality, overexpose your image so you don’t get the noise from the darker areas into your shot, but be careful about highlight clipping. The lower ISO speeds (200 and 400 on the 4.6K) should be avoided. [Update:] In order to get the most range out of our footage, Blackmagic recommends rating the cameras at their native ISO, which is 400 on the 4K, and 800 on the 4.6K, and then processing the RAW files using the latest version of DaVinci Resolve with BMD Film 4K and BMD Film 4.6K Gamma applied to the RAW decode. In RAW mode, ISO can be selected during the decoding process. We found that a setting of ISO 800 gives you the best starting point to grade. Image Quality In terms of image quality, the Blackmagic URSA Mini 4K and 4.6K are highly regarded due to their codecs. The following images were taken from URSA Mini RAW files, the Sony FS7 with its native codec and the Fuji XT-2 mirrorless camera with an external recorder: 100% crops (except the 4.6K downscaled to 4K) What we can clearly see here is that the Sony FS7 and Fujifilm X-T2 have a cleaner image when it comes to fine details. The URSA Mini 4.6K and 4K, on the other hand, show a little bit of a moire pattern on fine lines due to aliasing. The resolution of the Sony FS7 UHD image seems similar to that of the URSA 4.6K RAW image downscaled to 4K in terms of how much detail they resolve. When we compare the Blackmagic URSA Mini 4K vs 4.6K we see that the URSA Mini 4.6K resolves more detail than the 4K. The same is true when we compare a recording in 4K resolution on the URSA Mini 4.6K, to a 4.6K image on the same camera. But image quality is not a thing that is black and white. Here you can see how the different cameras treat a natural object, as opposed to test chart stars: Contrast on all images above has been adjustedfor a rough match. Interestingly the image coming from the Sony FS7 seem much softer. Add some sharpening to the FS7 image, though, and you will find that the image gets much closer to the way the URSA Mini images look. In conclusion, I would say that the codec of the FS7 is its weakest point, but the image looks cleaner than the one from the URSA Mini 4K and 4.6K. All in all, the URSA Mini 4K and 4.6K have a similar looking image, though the 4K wanders off into a slight green tint while the 4.6K looks more magenta. There is a certain amount of noise in the shadow areas on both cameras, and the image is slightly sharpened in-camera. But the look is very natural and colours are quite neutral. [UPDATE:] Here is a version of only the Sony FS7 image, graded and sharpened to match the URSA Mini processed RAW images above. Here you can see that the detail is very similar, but also how the codec easily falls apart on some portions of the image. The image is more stable and ready to grade on the URSA Mini’s: Sony FS7 image graded and post sharpened to match the URSA Mini 4.6K Rolling Shutter Some cameras, like the Sony a7S II, suffer from a severe rolling shutter effect, a phenomenon also referred to as “jello”. Unfortunately, the rolling shutter that we see on most CMOS sensor cameras is also present on the Blackmagic URSA Mini 4.6K and Sony FS7. As we can see, the rolling shutter on the Blackmagic URSA Mini 4.6K is identical to the one we found on the Sony FS7. 11ms is an OK rating when it comes to rolling shutter. On most mirrorless cameras the effect is more severe. The Blackmagic URSA Mini 4K however gets the best rating, as it has a global shutter sensor that does not suffer from the rolling shutter effect at all. RAW vs ProRes When comparing the codecs on the Blackmagic URSA Mini 4K vs 4.6K we found that ProRes generally gives us exceptional results. RAW is a very nice option and should in theory extends bit depth of your files to give you finer gradations. We did not test this. Dynamic range is not increased when using RAW, however. [Update:] We have compared gradations on a RAW and ProRes file and we can confirm that RAW increases the bit depth. So if you want the best filmic look for heavy color grading, we recommend to use the RAW option on this camera. Conclusion It was truly interesting to take a closer look at the Blackmagic URSA Mini 4.6K in comparison to the URSA Mini 4K camera. We saw that the dynamic range of the URSA Mini 4.6K is similar to the FS7. It is a vast improvement over the URSA Mini 4K, which really lacked behind on this point. On the other hand, the URSA Mini 4K has a global shutter sensor and thus doesn’t suffer from rolling shutter effect at all. In terms of image quality, the URSA Mini 4.6K delivers a really nice image with balanced colors and a natural look. The URSA Mini 4K clearly comes from the same family of sensors, but has a slight green tint. There is slight aliasing on both the 4K and 4.6K when we compare it to the FS7, though, and noise kicks in quickly if you are not careful. Both cameras are no lowlight wonders, but there is an improvement on the 4.6K. [UPDATE:] Also, the 4.6K can shoot up to 120fps in windowed HD. The most striking argument for the Blackmagic URSA Mini 4.6K is clearly its dynamic range. It also has a sensor 20% larger in size and an extra of 0.6K in resolution, which most will deem marginal in a world of 4K, UHD or HD delivery. If dynamic range is important to you, then the URSA Mini 4K probably does not have what you want. But besides this point, both these cameras are very similar. In our opinion, for those shooting in studios, there is no good reason to upgrade to the 4.6K at this time. Everyone else will probably welcome the extra filmic quality the URSA Mini 4.6K can achieve. Would I consider shooting on the Blackmagic URSA Mini 4.6K? Absolutely yes. With its high bit depth and natural looking image it will deliver high quality 4K with a high codec quality. There are other good and comparable cameras, but when it comes to film aesthetics and if you put quality control issues aside, then there is not much that will take you this far at the low pricepoint of the 4.6K. What is your experience with the Blackmagic URSA Mini 4K vs 4.6K? Let us know your opinion in the comments.Read more
We only send updates about our most relevant articles. No spam, guaranteed! And if you don't like our newsletter, you can unsubscribe with a single click. Read our full opt-out policy here.