Don Kotlos

​That would be idiotic to do since they would now have 2 cameras that don't have good quality over 1. They went into the trouble improving so many things in video, it makes no sense messing the video quality on purpose. There can be other explanations too: 1. It is not as a big problem in real life shooting situations. 2. You got a lemon, although I cannot imaging how can that happen 3. software stabilization introduces more artifacts. That would be easy for you to test. You said you are working on it, so I am sure you will also grab the test scene that you have with all the other cameras. If you see the same sh*tty result, I would suggest sending Olympus an email with your findings. I am sure you can influence the development of the next firmware. ​Please expand on this point in your review. If you have the shogun it would be interesting to see how the clean HDMI looks. See if it is the codec or the processing before that point. Thanks

​Can you please provide us with some information on how much of an improvement there is with the Ninja star? I am thinking of getting it too but I am not sure if it really improves the video quality. Maybe a record a static scene with both the internal codec and the external ninja and then post the images? Few things that are interesting to test: 1. Resolution 2. Blocking/ compression artifacts in daylight 3. Low light performance. Thanks!

GTX line is fine for home video editing. Better spend you money elsewhere... Here are few informative links: http://www.pugetsystems.com/labs/articles/Adobe-Premiere-Pro-CS6-GPU-Acceleration-162/ http://www.pugetsystems.com/labs/articles/Adobe-Premiere-Pro-CC-Professional-GPU-Acceleration-502/ http://www.studio1productions.com/blog/?p=302

In general you should invest in a "balanced system", which means improving your weak parts first. There are several sites like this that you can find more details. From what you describe an SSD will be helpful, but it really depends on the compression of the 4k file... You can find out what exactly is slowing you down and produces the stuttered playback. For example if during playback a single core is maxing out but all the other ones are not being used (or the % use is close to the % of each core in windows task manager) then that shows your file is being decoded slowly. If there is minimal cpu usage and your hard drive is at its maxed speed, then that should tell you that you need an ssd. If all of your cores are maxed out, then you are CPU limited, but at least in premiere you can reduce the playback resolution. Your GPU should handle most things just fine, but you can see if it is maxing out too. Sometimes of course it is not possible to see what is limiting you (bus speed etc.) so you just improve few things at a time and see how it works out. Most of the times, you can change the way that you edit (transcode, reduced resolution playback, etc) but a modern system can provide some advantages too. CPU wise a 5960x has double the processing power (multicpu tasks) and the ddr4 at least double the bandwidth with what you currently have. A system like that with a asus x99 deluxe that supports thunderbold and 32gb ram will be ~$2000. Adding a couple of SSD for media/exporting will be important too. If you work with davinci resolve and your GPU is limiting you, a gtx970 might do the trick. It is important ot find out what is limiting you before you spend any money...

​I thought about that too but then there is a third option further down: 1280x720. So I was confused again. Also it supports UHSII!

​Don't get me wrong, the nebula is a great product but it is not as convenient as IBIS. Even though nebula s smaller than a shoulder rig, it still is something that you will have to carry around in a large bag and assemble it with the camera before you shoot. That is a great way to miss the moment with children. And as the video illustrates there is no way to focus it while in use so you are limited to a large DoF and only if the light/camera allows it.

​Well if you want usable video without a rig, the new E-M5ii will be the choice. I don't shoot professionally and a rig is not something I carry with me every day. Moreover, a rig does not make it easy to follow children around. Don't underestimate a perfectly usable video in a small run&gun camera. If you are a professional, think of it as a secondary camera which will get you shots that your main cam sitting on a rig will never be able to, most probably because it will not be with you all the time.

I just hope E-M1 gets the video improvements because I would hate to sell it. Coupled with the 25/T0.95 is a joy to use.

ETTR with 8 bit is asking for trouble.

​Yeah I get that, but since the camera supports downsampling and writting 80mbits/s to the card, why not allow the maximum bitrate for 24p? That would possibly result in a much more gradable image. It looks like they are limited to a maximum ~0.32bits/pixel for some reason.

​I haven't seen any videos taken with the new firmware, but since they went though the trouble improving all these things for video, was is so hard to implement a very high bitrate (80Mbit/sec) FHD?

​I come across this opinion quite often but I fail to understand it. Take the following example: Lets say a camera company decides to release an updated model every month. Case 1. If I buy a camera that fit my needs, why a new model a month later will affect the usefulness of the camera that I already have? Case 2. If I buy a camera that does not fit my needs, wouldn't the shorter release cycle allow me to get the camera that I need sooner than later? In general I believe that shorter release cycles are more beneficial for the consumer. That and open source firmware.

​Yes this is how the manual "mechanical" focus is implemented in some of the olympus and fuji lenses. Still fly by wire, but they have set markings on the focus ring. Nothing with gears.. So while this is still useful for street/still photography it is not as good for video since the focusing is far from smooth (unless you focus really slowly). Still better than purely fly by wire since no matter the focusing speed the location of the ring determines the focus. What they could do with a firmware update is during video mode the transition between focus points to be very smooth. It is very easy to do and I am sure if they think about it, they will do it. Any olympus engineers around?

​Unless they get rid of the x-trans or do a full sensor readout during video, I wouldn't expect any dramatic improvement for video.

It also has XAVC-S and S-LOG2! http://www.sony.jp/ichigan/products/ILCE-7M2/speclist.html I guess this is the EM1 update that I was waiting for.

How did you convert it after all? Detail is fine with minimal noise. What was the ISO? There is some nasty blocking in the shadows but I guess that would be visible only when pushing by quite a bit. Edit: Can it be due to the conversion? Are there any flatter settings? Thanks

The guys at imaging resource have posted some 4k NX1 videos for the brave.

Does it really matter if they offer 24p and manual controls if the quality is so bad? Unless they started reading the whole sensor, line skipping on x-trans creates terrible aliasing artifacts. They didn't mention anything on that end so I doubt they managed to fix it. They are going the right direction though, and I would really like a Fuji with very good video quality.

You can't walk as you do normally since your shoulders change their axial position as you do. You ll have to bend your knees and keep you shoulders at the same level.

It happens more frequently with the blue, but it appears in the other channels too. Even the new FS7 suffers from it :

Yeah it looks bad, but the exposure is correct and the main reason people buy this camera is to use it at these high iso values. Especially in a setting like this one you need quite a bit of DoF.

Ok like Andrew many other people report no difference in video mode, so there is something wrong either with this firmware, or the video comparison that they published or the guys doing the hack.

They have uploaded a comparison and the difference should be there. I wonder why you couldn't see a difference. They have two versions. I ll give it a try later today too. If this ends up being the ML alternative for the EM1, then Olympus might get the wrong impression for their 2.0 firmware when the sales suddenly increase and keep on improving their art filters. I don't have a problem with hacked firmware but at least I expect companies not to lie about the technical limitations of their cameras.

Transcoding should be preferred in order to reduce the CPU load during editing. "Once converted to ProRes 422 HQ the footage weighed in at 26GB." Andrew how good do you think will be the transcoded file for grading?

Andrew how does the quality compare to the EM1? It's a pity that they didn't enable the sensor stabilization in the GX7.