dahlfors

I'm not sure if people are noticing or not, that the BMC video has very narrow depth of field on almost every shot. So you're mainly looking at footage where the majority of the image isn't in focus. Additionally, being a 720p vimeo video. Not the video I'd pick for trying to judge "perceived sharpness"...

There's a lens comparison of that RMC Tokina with a few other Nikkors: http://www.kenrockwell.com/tech/comparisons/2010-08-10-17mm-sharpness/ That lens is very well worth 6 euro. Seems to be one of the better wide prime lenses for Nikon mount, although Nikon's wide zooms are better.

Googled. Found this image from an expired online auction of RMC Tokina 17mm Nikon mount: '> To use this lens with metering on a Nikon body, you need D7000/D7100 or higher. Without metering it will function on most Nikons in manual exposure mode.

As a Nikon shooter, I wouldn't actually rule out Nikon. Granted, it doesn't look like most of the Nikon F mount lenses I've used (which all vary, depending on AI-S, AI, AF-S etc), but it's not too far off from what an AI-S 105mm f/2.5 looks like: '> It might fit. The RMC Tokina 17mm was made for Nikon mount at least. DreamingBMCC makes a valid point though - the rear element circle seems very small. But perhaps that's a part of the lens design. EDIT: This seems to be the same lens, for 35mm Pentax, so the small rear image circle seems to be part of the lens design: http://www.pentaxforums.com/forums/36-sold-items/241488-sale-sold-tokina-17mm-rmc-f-3-5-lens.html - So, I'd try to see if it fits on a Nikon - without forcing it in...

"In addition Nikon have all the technology to build a high end cinema line for pros like Cinema EOS, something which is turning out to be massively profitable for Canon. The sensor, the image processor, the codec, the lenses, everything." The last few years Nikon has been on recovery after their most important factories got destroyed (Thailand and Japan). That makes a huge impact on a company which have cameras as their main business and no photocopiers to fall back on. I believe Nikon's reaction to this is to be cautious about investing too much into new product lines and try to recover with what they know they're good at. Considering that situation, I think they're on a good track in improving the video modes on their cameras although they can't keep up with Panasonic in terms of features or handling. They've at least made incremental improvements with sensors, moire, aliasing and added 1080p60 modes in their low-end range - none of which Canon has done. What I just can't understand with Nikon, is why that bloody Live view aperture situation exists with their lower-end cameras since that issue is not there on cameras like D800. If that is planned segmentation of products - it is very retarded. If it's something that requires some reengineering to fix it, it's really about time that they fix it. I hope that Nikon is seriously watching what Panasonic is doing, and I hope they'll figure out how to do 4k and how to implement 10-bit 4:2:2 in at least 1080p. This is what Nikon EU product manager has to say about Nikon and 4k: - “We are aware of the need for, and request for, 4K video recording. It is a bit tricky, it’s not something that we are purposefully excluding from our cameras; however we need to approach it carefully. There are high-end cameras that produce this but it just puts such a load on the equipment. For us, because consumers are demanding it we are aware of this and will be looking into it for the future." --- Source: http://nikonrumors.com/2014/01/25/nikon-sees-future-in-4k-video-and-advanced-compact-cameras.aspx/ Sony, Canon and Panasonic all had a headstart thanks to producing video cameras for a long time. In my opinion Nikon hasn't been doing all that bad coming from a still photography camera background only. I certainly hope they will realize what handling issues and features that still are missing and provide that - either in mirrorless, DSLR, hybrid or specifically video-tailored cameras. And my own review of the D5300 and D3300 would be: - Great cameras for the newbies who want to start learning about video while also shooting stills. Far better choice than any other DSLR considering sensor, stills & video performance together. Panasonic G6 and Nikon D5300/D5200/D3300 are my first choice of recommendations for video+photo first-timers on tight budgets. There's handling issues and features missing that are available in more expensive cameras - but that's to be expected in entry-level/budget range.

Yep!

I see you have a (Redstan?) clamp for your Baby Iscorama there on one photo, although it wasn't listed in the Ebay ad. I got a non EOSHD member friend who has a baby Iscorama for which he has built some own clamp solution. Last time I spoke with him he was interested in finding some better clamp solution. So I thought I'd ask you if you'd be interested in selling your clamp?

@Bryan Harvey: very interesting to hear your behind the scenes comments! ... Also, Panasonic sure seems to have a winner here. It really is time for the slumbering big ones like Sony, Nikon & Canon to wake up if they want to keep selling in this market segment...

Updated rumors on 43rumors: http://www.43rumors.com/first-panasonic-gh4-images/ "4K video at 4096×2160 pixels and max frame rate of 96fps" - impressive if true!

And it's approved by Kurt Russell too: http://www.43rumors.com/ft5-picture-of-kurt-russel-with-the-new-g4k-and-new-info-about-the-video-specs/

No, Nikon's stabilisation doesn't work with Metabones speedboosters. Vignetting I'm not sure about. If there is, there might be a slight bit on the wider end, around 18-20mm but probably not. Autofocus with Nikons, nope. Aperture is controlled on the speedbooster and you focus manually with the lens. For stabilisation in lens, for long zoom range for run and gun usage - I'd go with a m4/3 lens.

Think it might be my post you're thinking of: they used the MOVI stabilizer to film the Burton Snowboards factory for an ad: '?do=embed' frameborder='0' data-embedContent>>

Yes. Again, why do you think they chose lineskipping for the iPhone 5s video? You yourself said the iPhone 5s video was flawless. It is not. If they'd do a proper resampling from 1440p or higher down to 720p without line skipping and proper high quality video encoding, I can assure you - they would not be able to do it in realtime at 120 fps. (I don't know about the sensor in iPhone, but there are sensors that very well can output higher resolutions at up to 120 fps - but it's still very resource intensive if you want to process the full sensor data and resample it down to 720p or 1080p from a full readout. Line skipping might or might not be necessary from sensor standpoint in an iPhone.) Does 5dmk3 have the power to take 4k-10 bit raw and downsample that to 1080 in realtime 24fps or so? Nope. Outputting RAW is wholly another thing. That means reading data fast and outputting data fast. It's more about I/O than it is about a processor actually doing heavy processing. Downsampling high resolution at high quality at realtime framerates for video is much more difficult processing. So, why doesn't the Nikon V1 output 4k Prores, 4k h.264 or downsample 4k to 1080p to Prores / higher rate H.264? (No, it doesn't have the processing capabilities to do the proper DOWNSAMPLING and ENCODING in realtime). And, the downsampling is key if you really want the best possible resolution and accurate color for 1080p with RGGB grid array sensors. My whole argument in the original post was: from a manufacturer stand point it makes much more sense to go 4k and let the user downsample in post to achieve excellent 1080p, it is much more expensive to do if you want all the processing done in realtime in-camera.

For my own needs I pretty much agree. I'd be fully happy with 10-bit 1080p without aliasing/moire and options for 422 / 444, and preferrably compressed with some efficient codec - and perhaps possibility to choose Prores/RAW when needed. I'm really happy with the video I get out of my D800: the main limitation is 4:2:0 and 8-bit color. And when shooting in certain city scenarios - moire/aliasing can occur. Ironically enough, what I'd lack from the pocket camera is a more efficient codec (10-bit 422 H.265 in a few bitrates perhaps) for when I wouldn't need Prores/RAW. So, looking forward to what Panasonic is cooking for 2014 :)

Did you read my post? My whole post was about the difference of "being able to process" and "be able to process in proper high quality". Here, take a look at the aliasing and crap you'll find in an iPhone 5S video: vimeo.com/77162053 (removed the http 'cause I didn't see it necessary for this thread to have the embedded video here) Why is there aliasing? Because the iPhone isn't up to do the processing in high quality and they have to take shortcuts. The video quality of the iPhone 5s is great for a phone, but my post was about processing it properly for far better footage (like in a possible upcoming GH4). Processing it properly, to get rid of aliasing, moire and getting the most possible resolution out of the footage while keeping the color information - whilst outputting it to a suitable compressed format tailored for high quality output. If you do it with proper downsampling algorithms optimized for quality, you'll need beefy hardware if you want it done in realtime, especially when we're talking more than 24 fps, 10-bit. Also, I'm perfectly well aware of Qualcomm chips. Our company uses those for hardware encoding video on smart phones.

Yes. I think this is where Andrew is going as well with the article: If you want to manufacture a camera with a lot better performance in 1080p, you are already designing a camera that can output a decent 4k video mode, since you want at least four (RGGB) sensor pixels for every 1080p pixel output. If the new Panasonic will output 4k at 10-bit 422, that video downsampled (in-camera perhaps or in post on a computer) will be some awesome 1080p 10-bit 4:4:4 material. So, from a manufacturer's standpoint: - They can already get sensors that could do 4k at 24fps or more (which can be used for really nice downsampling to 1080p). - With the right codecs you could encode 4k in fairly nice bitrate even to SDXC if the camera doesn't have CF / SSD. - Another option would be providing HDMI/SDI output to recorders. What's the remaining component then? - Processing the data fast enough in realtime with high enough quality. Doing that is a challenge when you want the highest quality of the downsampling. Here in fact, it might be easier to output a faster 8-bit/10-bit 4k and let people downsample it in post, rather than to attempt to do the high quality processing & downsampling to 1080p in-camera. And to think about the processing restraints: What's the major issue on BM pocket camera? ...and how long does the battery last in movie mode on a Panasonic G6/GH3/xxx or a Nikon D800/D5200/xxx? - The main difference in power draw is not the display nor the sensor, and there's probably not too big of a difference in writing to the card either - it's the processing. And that's still on a camera without 4 pixels (RGGB) per one 1080p pixel output. So, imagine the power constraints for designing a camera like the BM pocket camera and multiply that datastream by 4... The power & processing issues can be solved of course. But it will require more processing, more cooling, more power = bigger, more expensive camera. Want it done on a budget - get a decent 4k 10-bit and do the processing on a computer.

No need to jump to conclusions. Just because people don't see any bias in the article it doesn't mean people are in love with Bolex... Personally I find it interesting to learn his impressions and comparisons of the cameras, but I wouldn't still know for sure which I like better until I'd use the cameras. And yes, having Prores is for sure an advantage because people might not want to shoot raw for every project - just as PB listed in the article. Hence the reason I took it as an example of him not writing a biased article. I've watched & read PB's reviews of the BM cameras and the Digital Bolex and I don't see much bias in any of those articles.

I have no idea about difference in these two, other than that on your photo it seems like the coating is different on the two lenses. If anyone would know more differences about them, I bet it'd be Tony Wilson.. :)

What is unfair in his review? It's quite impossible for him to review it from every person's own use and perspective - of bloody course he reviews it with his workflow in mind - just like he always states that he prefers Prores to raw and in general work with 1080p than 4k/5k. ...And that doesn't mean that a Red camera or whatever with raw isn't the right tool for those people who need it.

No, I didn't find it skewed at all. You just need to understand which parts of what he says is his personal opinion regarding what suits his kind of use for a camera, and what parts are more objective of pros & cons for a camera. Pretty much like any individual reviewer out there.

Don't you worry about the original taking lens. Get a Dog Schidt Optiks, Nikkor lens or some other nice old manual lens for it. I especially love my DSO and Nikkor 85mm AI-S f/2 on mine.

What kind of spam troll user is this user called "F35...........i like that."??? REALLY. Original post reads "I`ve been eyeballing the black magic pocket camera, GHXXXX series...anything sub 1000$ basically". Any comparison to a Sony F35 in that case is very irrelevant, even more so when writing about it post after post. As for the original poster: Some of the most interesting alternatives will be Nikon D5300 or Panasonic G6 at that budget. Another interesting thing for skate videos is that both support 50p/60p at 1080, which will be useful for slowmotion. As for ultra wide angle alternatives for m4/3, I'm not so sure about what alternatives there are for such a small sensor. For D5300 there's at least a 10-20mm lens and a 10.5mm fisheye lens that might be useful.

As far as I know, it's all because humans can perceive a lot more than 24 fps. In computer gaming with 3D games, it isn't a fully smooth experience until you reach somewhere around 60-100 fps. In general, computer 3D games don't add any motion blurring (at least not older ones like classic Quake), so it's just a sequence of images rendered as fast as possible. It can't be really said to have a shutter speed, but let's say it has a shutter speed far exceeding what a camera would do, like 1/100000. At this shutter speed most individuals perceive it as a smooth animation at somewhere around 60-100 fps, some ppl even claim to be able to spot changes in the smoothness up to 120 fps or so (true or not is debated). What we can say though, is that it doesn't look smooth or "right" for people at 24 fps, because their eyes and brains can process images much faster. So at 24 fps, what makes 180 degree shutter angle better - is that it smoothens it out for our perception that exceeds the 24 fps. Yet it retains a bit of the stuttering due to the 24 fps - which is what we're used to in films (and similarily TV as well - in 25p PAL countries). Personally, always seeing 25p on TV, and films in 24p, but never 30/50/60 fps material - I didn't think 24p and 50p would be that much different for me. Not until I shoot my own 50p and 60p footage and viewed it on a computer. It was then when I finally understood what people meant with 24 fps being an important part of the film look.

On wikipedia, there is a nice animation of how the shutter worked on film cameras, which also makes it very clear why it is called a 180 degree shutter: http://en.wikipedia.org/wiki/Rotary_disc_shutter

At least, this is how I've understood shutter angle: At 1/100 shutter speed at 50 fps, it means that for the duration of each frame (1/50th of a second), the shutter was open for half of the frames duration. When you take one second of 50 fps footage and conform it to 25p slomo, you are spreading out your 50 frames for a duration of two seconds at 25 fps, that's the only thing that changes. When you step between frames, each frame will have captured half of the frames duration and motion in real time. The shutter angle remains at 180 degrees - because for each frame's duration the shutter has still been open half of the duration in real time. If you mix slomo footage with 25p at 1/50 shutter speed (180 degree) - the same property applies here - the shutter has been open half of a frames' duration in real time. And it is this property of capturing half the motion & time of a given frame that makes 50p slomo footage at 1/100 shutter speed look the same when conformed. Hope that makes sense :)