maxotics

Raw is a different thing entirely to 10bit 4:2:2.
 
Raw is just the 1s and 0s from the sensor. Everything else like bit depth, sampling, debayering, compression, encoding, comes later.
 
So 10bit 4:2:2 does not have mailability close to a raw photo actually.
 
The advantages of recording to a high bitrate is that the image doesn't break up when the movement or scene gets too complex for the codec. Also a finer more film-like noise grain is maintained.
 
However the stuff you can do with raw like changing the white balance entirely in post as if doing it in camera with no side effects, you cannot do with even the best RGB / YUV codec in 10bit 4:2:2, like ProRes.

let's make this even simpler, and use a dynamic range to show why you can't always resurrect higher bit depth (even with error diffusion):
 
Assume there are two types of A/D conversion:
 
1 bit (taking on the value of 0 or 1) 2 choices 
2 bit (taking on the values of 0 1 2 3) 
 
Let's assume that analog values are:
 
(0.1) (1.2) (2.0) (2.1) (3.0) (4.1)
 
and that A/D conversion assigns the closes possible digital value. 
 
1 bit A/D conversion becomes:
 
(0) (1) (1) (1) (1) (1)
 
2 bit A/D conversion becomes
 
(0) (1) (2) (2) (3) (3)
 
at half resolution you get either:
 
(0) (2) (3) or (1) (2) (3)
 
either one represents 3 levels of light , which you cannot represent in just 1 bit. 
 
Is this a contrived example, yes. But the point is to try and show they are not mathematically equivalent. 

It's actually 100% the shooter, 100% the editor and 100% the camera :)

Absolutely. In fact, most of this article just re-covered the ground of Andrew's fairly recent "Dear Nikon" article. I was hoping for a more in-depth review of the actual camera. Not another generalised rant about why MFT cameras are great. I have a G6 and love it, and agree that mirrorless is the future, but my D5300 is getting more use ATM because I like the image so much. Often I don't find the Nikon quite sharp enough, but resolution is OK and the other qualities trump the G6 for me. The 5300 really is a pain in the neck to shoot with though.
 
There is a lot about this article that is right, but there is also a lot that is unsaid or unacknowledged. There are some things I'm very disappointed EOSHD didn't address and that I actually feel make this quite a misleading/lazy review.
 
My responses to the article are below, mainly adding up to my opinion that for everything the D5300 lacks compared to other similarly priced cameras, it makes up for in other areas - making it one of the best all round choices at the moment. Having said that, I do completely agree that in a year's time the level of detail on the 5300 will look slightly archaic, and the extremely frustrating interface design and feature set makes it extremely frustrating and probably a no-go for a lot of work.
 
So, addressed in article-order:
 
 
Quote
The D5300 comes into a world where video enthusiasts are shooting 4K on Panasonic consumer cameras
 
Er, not yet they're not. And the D5300 actually came out in 2013.
 
 
Quote
I just wanted to shoot nice 1080p, conveniently, for a low price with interchangeable lenses. The D5300 to some is $799 for a Super 35mm camera that shoots quite nice 1080p with no moire & aliasing problems, good in low light, great articulated screen – and free Nikon stills camera into the bargain.
 
Agreed. This is where I'm coming from.
 
 
Quote
I just cannot get over the…
Baby Photo Mode
 
This is all an important point, but not really a problem specific to the 5300. But agreed. Nikon are being stupid about their video mode, and so are Canon.
 
 
Quote
The LCD has almost invisibly faint transparent masking marks for 16:9
 
I don't find this a problem at all, once you're used to it. A small, small, tiny inconvenience.
 
 
Quote
Simply by repackaging the D5300 and redesigning the firmware, Nikon could make it 10x more useful for everyone in the world with an interest in shooting artistic video.
 
Again, agreed. Hugely frustrating.
 
 
Quote
The Blackmagic Pocket Cinema Camera is $999 with a 10bit 4:2:2 ProRes codec, crisp detail and 13 stops of dynamic range. The D5300 does have a better screen than the Pocket Camera, which is articulated and it does have an APS-C sensor, 1080/60p in addition to the film frame rate of 24p and 25p, so it’s not all bad…
 
The Pocket has a very different workflow that a lot of shooters just aren't going to want, and is even more frustrating to use than the 5300 in many ways.
 
Quote
Neither can quite match the GH3 for detail in 1080p (let alone the GH4). I believe the trick Panasonic are using to give us such crisp 1080p on their cameras is to down-sample the sensor to roughly 2.5K and then oversample 1080p from that higher resolution raw image. The D5300 looks like the 5D Mark III’s stock video mode for resolution – it’s a bit mushy. You notice this the most when shooting in daylight at focus points between infinity and roughly 5 meters. Sometimes you don’t notice the softness much at all, so it’s not the camera’s main problem.
 
Agreed. Interesting theory about Panasonic too.
 
Quote
Standard or Vivid picture profile ... Whatever method you use the results after grading look similar.
 
FFS ANDREW WHY HAVE YOU NEGLECTED TO TALK ABOUT FLAAT?!
 
 
Quote
Low light performance is very good, even with the focus assist zoomed in 2 levels you will find it difficult to see any noise on the LCD while out shooting in low light at ISO 800. The image maintains rich colour at high ISOs and on brighter areas of the image at high ISOs noise almost vanishes altogether. ISO 1600 and 3200 are perfectly usable in video mode and even 6400 and 12,800 are better than on many cameras at the same price, closer in fact to the Super 35mm sensor in the Sony FS100.
 
IMO this is the major strength of the camera. It is superior to the Pocket, even with Speed Booster, here - because the Pocket has very washed-out colours in low light that completely negate it's low-noise levels. I haven't had a chance to look at how the GM1/GX7 do in low light (and forthcoming GH4). I really hope the GH4 matches the 5300 here. That will make me really happy.
 
 
Quote
The punch-in focus assist is generally a bit slow to use and you can’t simply half-press the shutter button to come back out of it, instead you have to tediously reverse back out out with the ‘minus’ key.
 
Minor niggle and entirely wrong. Pressing 'OK' (centre button) brings you straight out.
 
 
Quote
The lens mount is way too restrictive.
 
Seriously, WTF? You go on about the graetness of the Speed Booster all the time. Pocket + SB = awesome. Nikon mount = grim.
But I agree that it's ridiculous the way Nikon cripple their low-end cameras for using their own glass.
 
 
Quote
with the Olympus OM-D E-M1 any lens you put on it automatically has class-leading stabilisation better than any VR lens in the entire Nikon range including their pro lenses.
 
This is the case with any camera other than Olympus - not just Nikon!!!
 
 
Quote
Last year’s D5200 has an identical image in video mode and costs just $400 used, which makes it hard to justify the D5300 if you don’t need 1080/60p.
 
On the surface this is true, but in actuality it is entirely, categorically false. The D5300 has none of the banding/fixed pattern noise of the 5200. This means far better low light performance and crucially, AMAZING DYNAMIC RANGE POTENTIAL USING FLAT PROFILES. I know that you know about this Andrew. WHY HAVE YOU NEGLECTED TO TALK ABOUT THIS?!?! A quote you made on our older D5300 thread:
 
 
Quote
EOSHD, on 16 Jan 2014 - 8:11 PM, said:
Today I bought a Nikon D5300 for review, and a bit of shooting, but mainly for the blog with intention of sending it back after.
Well, so far it is surprising me.
The dynamic range with the flat picture profile is really quite something. Head to toe with ProRes on the Blackmagic Pocket Camera. I'd put it at 12 stops. Very good colour and good shadows, and again good low light performance. The codec in 1080/60p seems ok so far too. The main drawback seems to be the cheap-mid-range Nikon ergonomics (not enough buttons and dials). Wish they had put this video mode (and articulated screen) in the D7100 instead.
But so far so good peeps!
 
Why didn't you address this in the article? You do know about this. DYNAMIC RANGE IS A MAJOR, MAJOR PRO OF THIS CAMERA. Far superior to any other low-bitrate camera in the price-bracket.
 
 
Quote
As of today, DSLR video is over. Dead. Kaput!
 
Er, I'd probably pick a GH4 over a 5D+RAW for convenience's sake and features, but I'm pretty sure I'll prefer the 5D image (?) and I think we'll be seeing just as much 5D RAW as ever, even after the GH4 is out. But yes, I hope in the long-run mirrorless is the direction we are headed in (and I believe it is too).
 

Quote
The lack of video features and 4K will put DSLRs at a very significant disadvantage on performance relative to the best mirrorless cameras this year.
 
Agreed.
 
 
Quote
To give you an idea of how antiquated the D5300 form factor is you still can’t change the aperture from the camera whilst live-view mode is engaged. Why on Earth not?
 
Agreed. Ridiculous.
 
 
Quote
But in the cold light of day for $1999 (maybe less by the time final pricing is announced), Panasonic offer us 4K video. At $799 Nikon offer us a Baby Photo Mode. Why bother with this crap any more? Really?
 
Ridiculous comment. $2000 is is over twice as much as $800. For some people that's a big difference.
 
 
Quote
Pros
OK video quality and very good low light performance
Better than Canon Rebels and 70D on image quality, for both video and stills
Pleasing colour straight out of the MOV files, richness of tone maintained in low light
1080/60p useful for slow-motion video when converted to more cinematic 24p frame rate
38Mbit codec avoids break-up in 1080/60p mode (only 24Mbit VBR in 24p and 25p mode though)
No significant moire or aliasing issues (though resolution falls short of being truly full HD)
PAL / NTSC switchable for wide variety of frame rates
Manual focus magnification has an ultra-detailed display mode (though painfully slow frame rate)
Quicktime MOV file format benefits (easy editing and access, thumbnail preview in Explorer and Finder)
Very nice smartphone standard 3.2″ articulated screen
The sensor produces immaculate stills quality for the price
$799
 
Agree with all of these, though it seems almost misleading for someone with the respect you have from your readers to neglect talking about dynamic range with the D5300. IMO DR is on-par with low-light as the greatest strength of this camera.
 
 
Quote
Cons
Charmless – looks cheap, bland shooting experience
Very poor ergonomics by Nikon standards
Extremely dated form factor
 
I really like the size, weight and shape of the camera. It is a pain in the backside to use though.
 
 
Quote
No real improvement in image quality over the D5200 in terms of video or raw stills
 
FFingF's sake!!! The banding is gone!!! That was THE major drawback of the D5200. You said so yourself when you reviewed that camera! At least bloody mention it!!!
 
 
Quote
Blackmagic Pocket Cinema Camera offers much better image for filmmakers (13 stop dynamic range, 10bit ProRes 422) for just $200 more
 
The Pocket workflow is a very, very different proposition for some people, and it requires a much greater outlay in terms of gear and time spent in post. The 5300 is also a much better all-rounder.
 
I agree wholeheartedly with all of your other cons.
 
 
As is becoming standard Andrew, I have been very critical of you. As always I still have the greatest respect for you and your site (in fact I may actually just change my signature to say this ;) ).

I too, was confused before working with the actual bits of RAW video data.  When the sensor is exposed to light a chip reads the values from each sensel (which is eventually combined into a pixel value).  The sensels are basically just voltage resistances.  The camera can read the values to as much precision as the electronics allow.  You can liken this to getting the voltage from you household electric.  A simple meter would show 220 volts in the U.K., but a fancy one might show 219.556345 volts. 
 
The very first sensel, top left, is usually a red filtered sensel (keep in mind the sensor are, at heart, monochromatic).  A Canon camera, for example, reads that value and stores it as a 14-bit number, 0 to 16,383.  That means it records 16,383 shades of red.  It does the same for the next pixel, a green one.  Below that green one, on the second row, will be a blue sensel, again the same for that.  Both Magic Lantern RAW and Blackmagic, save these values (the blackmagic with some lossless data compression, not to be confused with visual compression).
 
Each of these values is a number between 0 and 16,283.  From the three, if you construct a full-color pixel you'd end up with a value at around 4 trillion, say it was white.  And another pixel, black, was 0.  Unfortunately, our eyes can only see about 16 million variations in color so the 4 trillion dollar white would actually only look like a 16 million value.  For the most part, we can only see, and our display and printing produce, color values within 16 million shades.  So why do we want 4 trillion colors when we can only see 16 million?
 
EXPOSURE!
 
First, let's turn to the GH4 (or any H.264 tech).  In those cameras each sensels value is saved into a 256 (8-bit) value.  Those 3 values are then combined to create a 16 million (24 bit) full color value. Assuming you've exposed the scene perfectly, and the colors are what you want, you will happy as Andy with a gold-plated G6 :)  Also 4:2:2, and all similar nomenclature, are about color compression for motion; they degrade each still image.
 
But what if you didn't expose correctly?  What if you made a mistake and exposed +2.  Now your scene is washed out.  When you try to get lighter values at the bottom you can't get them.  The detail (shading) just isn't there.
 
But much of it WAS there in the RAW data from the sensor!  Let's say with the GH4 you exposed perfectly, and it happened to be the 16 million color values between 2 trillion and 2.16 trillion.  Then, you expose +2 and it saves the values from 2.74 to 2.9 trillion.  With the RAW data you can pick up those values and CONSTRUCT your 16 million color H.264 video data.  Naturally, the sensor works better when you expose in the middle of its range, so you can't fix the exposure perfectly in real life.
 
The only benefit of the I-frame data is that it doesn't try to compress the 24bit values from one pixel in one frame, to another, which can create artifacts.  It does not, in any way, solve the problem of available color depth, if you're comparing 8bit to 14bit.  
 
It sounds like the GH4 will provide fantastic resolution in low-dynamic range shooting conditions.  It will not do what the Blackmagic cameras do in high-dynamic range shooting conditions (save enough data for you to fix exposures in post, or recover details from shadows, etc)
 
Filmmakers who don't understand the difference between the cameras may end up having the wrong camera and that would be a shame.  They are both fantastic technologies.  You can no more have both than you can have a Range Rover and Jaguar in the same car ;)

I too, was confused before working with the actual bits of RAW video data.  When the sensor is exposed to light a chip reads the values from each sensel (which is eventually combined into a pixel value).  The sensels are basically just voltage resistances.  The camera can read the values to as much precision as the electronics allow.  You can liken this to getting the voltage from you household electric.  A simple meter would show 220 volts in the U.K., but a fancy one might show 219.556345 volts. 
 
The very first sensel, top left, is usually a red filtered sensel (keep in mind the sensor are, at heart, monochromatic).  A Canon camera, for example, reads that value and stores it as a 14-bit number, 0 to 16,383.  That means it records 16,383 shades of red.  It does the same for the next pixel, a green one.  Below that green one, on the second row, will be a blue sensel, again the same for that.  Both Magic Lantern RAW and Blackmagic, save these values (the blackmagic with some lossless data compression, not to be confused with visual compression).
 
Each of these values is a number between 0 and 16,283.  From the three, if you construct a full-color pixel you'd end up with a value at around 4 trillion, say it was white.  And another pixel, black, was 0.  Unfortunately, our eyes can only see about 16 million variations in color so the 4 trillion dollar white would actually only look like a 16 million value.  For the most part, we can only see, and our display and printing produce, color values within 16 million shades.  So why do we want 4 trillion colors when we can only see 16 million?
 
EXPOSURE!
 
First, let's turn to the GH4 (or any H.264 tech).  In those cameras each sensels value is saved into a 256 (8-bit) value.  Those 3 values are then combined to create a 16 million (24 bit) full color value. Assuming you've exposed the scene perfectly, and the colors are what you want, you will happy as Andy with a gold-plated G6 :)  Also 4:2:2, and all similar nomenclature, are about color compression for motion; they degrade each still image.
 
But what if you didn't expose correctly?  What if you made a mistake and exposed +2.  Now your scene is washed out.  When you try to get lighter values at the bottom you can't get them.  The detail (shading) just isn't there.
 
But much of it WAS there in the RAW data from the sensor!  Let's say with the GH4 you exposed perfectly, and it happened to be the 16 million color values between 2 trillion and 2.16 trillion.  Then, you expose +2 and it saves the values from 2.74 to 2.9 trillion.  With the RAW data you can pick up those values and CONSTRUCT your 16 million color H.264 video data.  Naturally, the sensor works better when you expose in the middle of its range, so you can't fix the exposure perfectly in real life.
 
The only benefit of the I-frame data is that it doesn't try to compress the 24bit values from one pixel in one frame, to another, which can create artifacts.  It does not, in any way, solve the problem of available color depth, if you're comparing 8bit to 14bit.  
 
It sounds like the GH4 will provide fantastic resolution in low-dynamic range shooting conditions.  It will not do what the Blackmagic cameras do in high-dynamic range shooting conditions (save enough data for you to fix exposures in post, or recover details from shadows, etc)
 
Filmmakers who don't understand the difference between the cameras may end up having the wrong camera and that would be a shame.  They are both fantastic technologies.  You can no more have both than you can have a Range Rover and Jaguar in the same car ;)

Hans Punk shoots my kind of camera porn ;)
 
Awesome stuff! 
 

 
Like Andrew meets Wim Wenders
 

Hans Punk shoots my kind of camera porn ;)
 
Awesome stuff! 
 

 
Like Andrew meets Wim Wenders
 

Loving my A7.   The Video really is crap, but it's amazing on stills.   Currently rocking  a 20mm Nikon f4, Super Tak 35, Konica 40 (damn, is that thing sharp) Minolta 45, Nikon pre-ai 50 1.4, and Nikon AiS 85 F2.   Still waiting on my FD adapter.
 
If you shoot stills, this thing is amazing.  If you want video, get a Pocket Cam.

Andrew I really like your blog and I don't want to come across like a d**k, but I've been wanting to ask you something for a while now and with this article I feel I have to. It's a genuine, respectful question - it's not intended to be confrontational:
 
Do you need to reevaluate what EOSHD is about?
 
From where I'm standing - admittedly the very bottom of the pile - some of the 'guiding principles' seem to be loosing prominence.
 
From 'About EOSHD':
 
Which leads me to a slightly more delicate question ...
 
From the EOSHD Legal Disclaimer:
If this seems out of order I apologise, but I come to this blog believing I know where you're coming from. Does the above still hold true? Associated is a pretty broad and loose term.
 
I'm not intending to criticise you - obviously it's natural that you move onward and upward. I'd just like to get a clearer picture of where you're coming from, or perhaps where you think EOSHD is going.
 
I think it's changing.

To me that sounds like it might be just an internal reflection rather than an actual bug software/hardware bug. 
 
A kind of a glare reflection which appears only during very specific circumstances; when bright enough a spot light source is shining straight into the sensor from a specific (low) angle, and then the shiny surface of the sensor gets reflected back to the lens, and then some of the lens surfaces or a filter surface (or some other reflective surface) will produce this ghost glare. I've seen something (kinda) similar happen before, under certain circumstances. The easiest way to minimise it was to use a lens hood.
 
I'd try reproducing that scenario with different bright lights straight into the lens from varying angles, trying to create various glares on purpose, and then experiment with different lenses, filters and perhaps even the Speed Booster, if possible. Chances are the stripes will only appear with certain lens and light angle (and colour) combinations.
 
It it really is what I believe it is, I believe it will make no difference if it's a still photo or video. The reflection may just be easier to spot and reproduce in a video clip.
 
Well, just a thought, might be worth investigating.

for fully resolving a pair of lines (one black, and one white), 2 rows of pixels should be used, right? Nope, 4 rows needed because of Bayer pattern: blue,green,blue,green. so in matter of resolution,  4k is not actually 4k, let alone 1080p. so 2 megapixel Sigma Foveon like sensor  (3 layered color filter), has potential to be equal to bayer 4k. BUT, its insane to make 2 megapixel Full Frame sensor, cause every single one of pixels will have a massive area, almost 17um x 17um. that will heavily suffer from electron overflowing, and you need a thick dark ND filter for your every outdoor shot. what we lack right now, is not resolution. we lack Acuity. colors are not correct, much of that is based on mathematical guess, thanks to demosaicing, edges are soft for exactly same reason. lots of moire and false data. 4k isnt going to solve all of these, but downscaling, a decent downscaling, can give us some of that acuity we are lacking right now. the problem is, there is no camera equipped with "Built-in Best-in-Class hardware-accelerated downscaler". all the processing is on your own (PC) shoulder. otherwise, why should I care if gazillion pixels has been read out to give me my sweet 1080p? 

It works if the lens has an external OIS switch, at least it does on my 14-45 OIS lens. 

Mirrorless camera sales in the US have slipped because in Japan they have a "smaller is smarter" mentality and love novelty, but in the US it is a "bigger is better" culture and people equate something like a Nikon D4 with quality, because of the heft and substantial size.
 
The biggest mistake the mirrorless market continues to make (even the Sony A7R) is that they are focussed around small size.
 
Mirrorless is the future direction of all cameras because digital technology is the successor to mechanical and optical technology. The mirror is not going to still be flapping up and down in 100 years time, we will have EVFs so good as they will resolve far more than an optical viewfinder and give us super-vision, better than our own eyes can see, especially in low light conditions.
 
Instead of driving home the fact that mirrorless is a POWERFUL technology and DSLRs are a past tech, the Japanese manufacturers gave us cute little GF cameras and dinky PENs. Ridiculous.
 
Only now are they on the right path going high end, but Sony are still trying to downsize their FE mount lenses at the expense of lovely fast apertures, and Olympus are still adding gimmicks on their high end cameras, and none of them have an answer to the workhorse pro cameras like the 1D X and D4. At least not for stills.
 
I have always found a use for mirrorless cameras because the video has often been better - both in terms of image quality and features. That hasn't changed.

I would want good 1080p first, like blackmagic prores 1080p and such for very low price. I think most of the first 4K consumer cameras will have some bullshit codec with low color depth, low bitrate and sharpening pushed to the max.

Things like bit depth really matter for post. I can approximate  higher bit depth by downsampling from higher resolution, but the individual precision of each pixel remains the same. So in certain cases, you won't pickup subtle changes in tone. 
 
And all else being equal, down-sampling does not afford you better dynamic range.  And 4K raw is still way too much data, so you will lose something by going to 4K. 
 
As for content delivery, almost eveyone is still stuck below 1080p (besides Blu-Ray). iTunes, Netflix, Cable, almost all of their HD looks like heavily compressed 720p or 1080i. 
 
The Canon c300 is a special case, because it uses its 4K sensor not just to down-sample intelligently, but to minimize motion artifacts.  Most of the 4K cameras that are going to come out won't be using the same process, and therefore won't get the same benefit from a 4K sensor.

I understand that 4k gives better resolution, dynamic range and colour than 1080 - and that these benefits are transferred when 4k is downscaled to 1080 in post. But I don't understand why this is an argument for 4K for people who don't need 4K output. Surely those people would benefit much more if the greater processing power and bitrate was put into a better 1080p codec (e.g. like the Pocket's prores). They would then get the benefit of the larger files in the form of grading latitude, rather than just chucking away information from their very full cards as soon as they got home. Wouldn't they? Personally I'd rather chuck that information away after I've done something useful with it, like a bit of colour correction.
 
I'm not hearing a lot of people complaining that the Pocket isn't 4K. I am however hearing a lot of people complaining that the Pocket is a pain in the a**e to use. Imagine if Panasonic put BMPCC-like tech inside a GH3. We'd all go completely wild. Why isn't that the immediate future? With 4K it just seems to me like we'll be starting the whole H264 journey again, just at a higher level. Why not make HD the best it can be before moving on to 4K? The whole thing smells a lot like the megapixel race to me and, to be honest, the ugly side of capitalism.
Anyway, this is my question: Leaving aside reframing options, why is compressed 4K better than high bitrate 1080 for those who don't need 4K output?
 
Just to be clear, this is a genuine question. I am genuinely hoping to learn something. I am not being pointlessly antagonistic in the hope of rubbing someone up the wrong way. That's just the card I was dealt at birth - to forever write forum posts that elicit the wrath of Hades.

I'm going with 4K too , I'm sticking with Panasonic as they make great cameras that just work  , plus most of my glass is optomised for m4/3
 
So roll on gh4 or what ever they call it

And not to forget a special thanks to Andrew for hosting this site.  All the best.  ;)

I had been a projectionist for a long time (now this profession has died), and from 2000 on I had also been a digital projectionist. Until 2011, when automation finally killed the job, I used to compare my own stuff to the DCPs, side by side on the big screens, in the last two years also as DCPs, when easyDCP and openDCP became available. The largest screen was 78 feet x 32 feet (that's for scope, for 16:9 the width then was 58 feet).
 
First thing I noticed is that resolution doesn't influence sharpness to the expected degree. And it also doesn't influence subjective quality very much. In fact, an upscaled SD DVD ( anamorph pixels with scope-crop, really the worst way to treat a video) could be shown to a big audience, and (back then) nobody complained, the class-A hardware scalers made it look good. I know this is hard to believe, but we once had a festival with student films, ranging from DVD, BD to genuine DCP (a Red!), and the one best looking was a masterfully graded HVX200 short, played from SD DVD.
 
On the other hand, there was a way to know instantaneously what was film and what was video: Colors. 

 
I know this comparison is only 8-bit, but I have to find a way to describe aesthetic subtleties here. With a camera like the GH2 ("Musgo"), for example, one would be well advised to fill the frame with detail, textures (resolution, that's the GH2s strength) and not with skies and other big areas of glorious colors.
 
Right now we grade for 8-bit, so 12-bit raw is *just* a bigger palette for grading. Color depth seems to add a new dimension to our video. It's fun to tear the, er, bloom off the images and to dive through the colors. Would it stand against an Alexa? I can't tell, really, but I'm convinced it would do better than many others.
 
I can't wait to see a DCP with the 12-bit preserved in my old cinema.

Very nice review as always.  I would add this:
 
o. Takes 24 to 30 1920x1080 RAW (DNG) photographs per second.  These images can be post-processed by almost every photo processing software known to man :)
o. Showing 24 fully-realized photos per second is what real film does.
o. If you were on a million dollar movie set and shot side-by-side with the production camera and showed it in a theater most people would not see much of a difference.  Am I wrong?  
o. If the BMPCC shot 720p it would be a fantastic achievement!  I'll argue the camera is already 2nd gen.
o. The BMPCC simply blows away camcorders from a few years ago.
o. Davinci Resolve Lite adds thousands to the value of the camera.
o. H.264 looks like complete crap in low-light.  The BMPCC has a natural grain/noise look.  Sensor size doesn't even come into play.
o. The 5D3 is it's closest competitor and the CF cards are very expensive and ML untrustworthy.  
 
The only negative about the camera for me is I no longer have ANY EXCUSE whatsoever to film anything creative.   Yes, it's easy to want this and that lens, and rig, and whatever.  Yet what you can do with the BMPCC and a cheap 14-45 IOS lens is, quite simply, STAGGERING.  Seriously, every day I don't use it I feel a bit ashamed.

Very nice review as always.  I would add this:
 
o. Takes 24 to 30 1920x1080 RAW (DNG) photographs per second.  These images can be post-processed by almost every photo processing software known to man :)
o. Showing 24 fully-realized photos per second is what real film does.
o. If you were on a million dollar movie set and shot side-by-side with the production camera and showed it in a theater most people would not see much of a difference.  Am I wrong?  
o. If the BMPCC shot 720p it would be a fantastic achievement!  I'll argue the camera is already 2nd gen.
o. The BMPCC simply blows away camcorders from a few years ago.
o. Davinci Resolve Lite adds thousands to the value of the camera.
o. H.264 looks like complete crap in low-light.  The BMPCC has a natural grain/noise look.  Sensor size doesn't even come into play.
o. The 5D3 is it's closest competitor and the CF cards are very expensive and ML untrustworthy.  
 
The only negative about the camera for me is I no longer have ANY EXCUSE whatsoever to film anything creative.   Yes, it's easy to want this and that lens, and rig, and whatever.  Yet what you can do with the BMPCC and a cheap 14-45 IOS lens is, quite simply, STAGGERING.  Seriously, every day I don't use it I feel a bit ashamed.

This will change it immediately with no waiting and no lag, HD on demand, no frigging around to select it even though you selected to always play HD in your Youtube preferences, but it doesn't fn work properly!! This will play HD 100%, if you use Chrome!! Enjoy.... :D
 
http://www.autohdforyoutube.com/
 
ps. In the tools/settings, you can set it to 720p or 1080p by clicking on options button.

Nothing planned for Blackmagic yet. What do you think guys? BMPCC guide next up?

I hope it's OK posting this under this topic:
 
 
Is there a BMPCC Shooter's Guide planed or (perhaps) already in progres?