P337

Julian's images: saving the 4K example at quality 6 creates DCT macroblock artifacts that don't show up in the 444 example at quality 10. All the images posted are 420: that's JPG. To compare the 1080p 444 example to the 4K 420 example: bicubic scale up the 1080p image to match exactly the same image region as the 4K image (examples posted are different regions and scale). The 1080p image will be slightly softer but should have less noise and artifacts. Combining both images as layers in a image editor then computing the difference (and scaling the brightness/gamma) up so the changes are clearly visible will help show exactly what has happened numerically; helpful if the differences aren't very obvious on visual inspection.
 
We agree that 420 4K scaled to 1080p 444 will look better than 1080p captured at 420 (need to shoot a scene with camera on tripod and compare A/B to really see benefits clearly). 444 has full color sampling per pixel vs 420 having 1/4 the color sampling (1/2 vertical and 1/2 horizontal). My point is that we're not really getting any significant color element bit depth improvement which allows significant post-grading latitude as provided by a native 10-bit capture (at best there's ~8.5-9-bits of information encoded after this process: will be hard to see much difference when viewed normally (vs. via analysis)). Another thing to keep in mind is that all > 8-bit (24-bit), e.g. 10-bit (30-bit) images, need a 10-bit graphics card and monitor to view. Very few folks have 10-bit systems (I have a 10-bit graphics card in one of my machines, but am using 8-bit displays). >8-bit systems images need to be dithered and/or tone mapped to 8-bit to take advantage of the >8-bit information. Everything currently viewable on the internet is 8-bit (24-bit) and almost all 420 (JPG and H.264).
 
re: H.264 being less that 8-bits- it's a effectively a lot less than 8-bits not only from initial DCT quantization and compression (for the macroblocks), but also from the motion vector estimation, motion compression, and macro block reconstruction (which includes fixing the macroblock edges on higher quality decoders).

I haven't followed the topic, but I was curious if I could simulate the theory.
On Fred Miranda I found this topic that explains how Photoshop saves jpgs in 4:2:0 or 4:4:4
I used a full resolution JPG from the GH2 and took the following steps:
 
Cropped the image to 3840x2160 and saved as JPG quality 6. The result is a 4K 4:2:0 still.

 
Then I resized this image to 1920x1080 and saved it as quality 10 to make a 4:4:4 image.

 
I also resized the 4K 4:2:0 still to 1920x1080 and saved it at quality 6 to make a 1080p 4:2:0 image.

 
200% crops of the above images in same order:
 
4K 4:2:0

 
1080p 4:4:4

 
1080p 4:2:0

 
Not sure if this test is correct, but to me it looks like you can gain color resolution from downsampling the 4K 4:2:0 file to 1080p. Correct me if I'm wrong!

I like the analogy of Bit depth being like rulers for this argument, I see a "10bit HD" source as one person having to produce a measurements in full 1/4 centimeter while "8bit 4k to HD" is like 4 different people measuring in full centimeters then averaging to the median of their measurements.
 
For example let's say they have to measure a subject that is 3.25 cm. 
 
The guy measuring in 1/4 centimeters easily and accurately produces the measurement of 3 1/4cm.
 
The four guys measuring in full centimeters would have to choose between 3cm or 4cm initially, if three of the four measured it as 3cm and one guy measures it as 4cm then their result would be accurate at 3 1/4cm.  But if they ended up with a different set of initial measurements they could result in an inaccurate measurement of 3 1/2cm or 3 3/4cm.
 
So it is possible to down sample 8bit 4k and gain more accurate color depth similar to 10bit HD but you will be introducing an opportunity to get slightly inaccurate results compared to an actual 10bit source, in fact each pixel has only a 1 in 4 chance of being accurately sampled to 10bit, this would be noticed on edges of color objects and in gradients.  I don't think it's worth it as a 10bit replacement but it's definitely worth it for 4:4:4 and hey it's better than 8bit so if you're stuck with it I would down sample it.  
 
Bit-rate compression is another story, as 100mbps is not enough for 4:4:4 HD.  Luckily we are bloating to 4:4:4 after recording so we are able to reset the bit-rate, allowing adequate room for the extra information to keep the same detail of the original, I would suggest 400mbps at least up to 1.6gbps if you really think it's 10bit. 
 

I think the problem is doing a high quality debayer on the raw. The compression to JPEG isn't the processor intensive bit. It is the debayering.
 
All the camera has to do for DNG is dump the 1s and 0s to a very fast 512MB buffer memory chip. Then write 512MB to the card (that takes longer).
 
Imagine with every DNG file having to process all that data - 5MB of it - like opening a raw file in Photoshop - but in-camera.
 
Clearly the camera is doing a very simplistic debayer of the 2K raw to provide an image for live view and the H.264 encoder.
 
But we need a much higher quality one for decent 1080p MJPEG, otherwise you may as well stick with the existing H.264.
 
That's my understanding of it.
 
DNG is the option with the least processing, least CPU intensive, but sadly the most data to shuttle off to the card.
 
Of course Blackmagic cameras debayer their 2K raw just fine to ProRes in-camera, even the Pocket Cinema Camera. So it is doable with today's technology. What Blackmagic are doing could be very specialised though. Whether Canon have that technology, I don't know.

I thought the buffer has a write speed limitation though, independent of the card?  I guess you'll find out if you keep sticking faster cards in there and performance plateaus.  

What needs to be understood is that these "2K" DNGs actually have about 1920x1080 pixels of "actual image", the rest are black bars.
 
So far each model seems to have a set width for the "actual image" but the image height can be changed by setting the camera's recording resolution in Canon's menus.  Also in the 5x and 10x mode the width can be increased but the height is about cut in half.  Further more 1080p and 720p modes seem to use the same "actual image size" of about 1920x1080 but the DNGs also mimic the frame-rate so in 720p mode you get a usable 1080p60.  However the memory buffer to your storage card is limited to about 30 frames, so that's about 1 second for most frame-rates and half a second for 60 fps.  
 
Here are a few examples of "usable image" sizes:
60D in 1080/720 = 1736 x 1154
60D in 480p = 1736 x 694 (almost 720p but needs a little up-resing in post)
60D in 5x = 2520 x 1764
60D in 10x = 2520 x 1080
 
5D3 in 1080 = 1931 x 1288
5D3 in 5x = 3592 x 1320 (a bit of a crop factor though, Andrew said about 1.3x)
5D2 in 1080 = 1880 x 1248
 
There have been very little advancements, 1% tried 720p 24fps on the 6D and only got 54 frames.  That is about 20 more frames and about 2 seconds and so far the longest 24fps recording on record.  They have also found that turning the camera to "JPEG" (Not RAW+JPEG or RAW) increases the frame buffer to about 50 also but won't compress the DNG.  The problem is that the buffer simply isn't fast enough and as I said before they will have to compress the DNG much like BlackMagic does.  
 
Alex just threw in a quick DNG converter to see if it would work, it does but it was designed by CHDK some time ago and based on a previous version that didn't support compressions, if they can implement the newer DNG then they might be able to add compression but even at a basic 3:1 compression ratio that's only 3-6 seconds before the buffer fills again.  

Thought maybe it was my setup (Nanoflash) but your footage also shows a green => magenta shift from HDMI to H.264.

To say no, it is not better, without at least a simple green screen test seems unfair. I can't see why people were expecting a drastic difference

rich and vibrant colours can be got by the colour profile you shoot on in the camera
and by grading in post
 
Panasonic and olympus make m43 lenses to use without adapter
but I would recomend you do use adapters and it will open up a whole world of great lenses you can use
Canon FD lenses are superb on a GH2 and cheap
Carl Zeiss are also superb and have great blacks and contrast
 
so have a look on ebay ....

I really hope you guys are right and all these technical explanations that ensure great results on paper end up making the G6 a true continuation of such great image quality, only improved.

Panasonic apparently pixel binds 2x2 pixels to reduce the sensor's full resolution to 1/4 before reading the sensor (approximately 12Mp to 3Mp for the GH3), then further reduces that when processing to "FullHD" (about 2Mp).
- http://www.imaging-resource.com/news/2012/09/28/qa-with-panasonic-the-story-behind-the-new-gh3-and-compact-system-tech
 
Also Panasonic claimed disabling the multi-aspect feature of the sensor helps them improve their sensor reads, enabling 1080p60 as well as a cleaner image.  This makes sense since the GH2 had to down res approximately 3.5Mp to 2Mp with an older processor while the G6 will be converting 3Mp to 2Mp with a newly developed processor designed to further increase signal to noise, preserve details in noise reductions and widen dynamic range.  Checkout the G5 since it also used the GH2 sensor with supposedly better results, and it did it without this new processor. (I'm still trying to track down where I read all this so take it with a grain of salt)
 
Can't help to think this could be Panasonic trying to show off what they can do with their own hardware technology, vs the Sony sensor in the GH3, justing on the specs and price Panasonic seems to be ok "cannibalizing" some of the GH3 sales to make this point. 

Panasonic apparently pixel binds 2x2 pixels to reduce the sensor's full resolution to 1/4 before reading the sensor (approximately 12Mp to 3Mp for the GH3), then further reduces that when processing to "FullHD" (about 2Mp).
- http://www.imaging-resource.com/news/2012/09/28/qa-with-panasonic-the-story-behind-the-new-gh3-and-compact-system-tech
 
Also Panasonic claimed disabling the multi-aspect feature of the sensor helps them improve their sensor reads, enabling 1080p60 as well as a cleaner image.  This makes sense since the GH2 had to down res approximately 3.5Mp to 2Mp with an older processor while the G6 will be converting 3Mp to 2Mp with a newly developed processor designed to further increase signal to noise, preserve details in noise reductions and widen dynamic range.  Checkout the G5 since it also used the GH2 sensor with supposedly better results, and it did it without this new processor. (I'm still trying to track down where I read all this so take it with a grain of salt)
 
Can't help to think this could be Panasonic trying to show off what they can do with their own hardware technology, vs the Sony sensor in the GH3, justing on the specs and price Panasonic seems to be ok "cannibalizing" some of the GH3 sales to make this point. 

Everyone seems to be pre-occupied bickering over youtube videos so I did some digging and one thing I found is that the GH3 incorporates 8x8 and 4x4 pixel macroblocks into all of their their Intra and Inter frame compressions where as the GH1 and GH2 only used 4x4 on everything.  
 
As I understand it, this helps the GH3 be more efficient with bandwidth and supposedly helps retain fine detail areas, so if this was carried over then the G6 should have the potential of slightly better details than the GH2.
 
 
 

You know an LCD screen is not the same as looking through a viewfinder...
Also, they've been using measuring tape to focus for over 100 years of film, maybe more than their eyes.  :)

One positive of this camera vs the GH2 that the video illustrates: big blue skies everywhere, no banding in sight.

I think you have to decide for yourself what is most important.
 
Image Quality: BMPCC probably wins big time with the dynamic range and the raw editing options - But the GH3 is still very good
Ease of use: GH3 - Flip-screen, oled viewfinder, lots of buttons, touch screen. Reasonable file sizes
Lens flexibility: Both pretty flexible, GH3 wins because finding good wide angles or bright standard lenses for the BMPCC is going to be more difficult (and expensive) with the 3x cropfactor.
 
The GH3 is a much more complete camera than the BMPCC. The Pocket is just the cheapest way to get raw footage.

you are watchin on youtube - mega compressed so beware...- lets not write it off before we have all downloaded files direct off the camera,
If it has Venus engine inside then its hackable....its been done before by Vitaliy and Driftwood etc

Eagerly awaiting your Pocket Cinema Camera? You already have a micro for thirds camera and some C-mount lenses? Want to know if they will cover the sensor of the Pocket? Lets find out! I hope you will add your results, so we can make this list growing.
 
I will only add lenses to the lists when you have proof, in other words: images.
 
How?

Because we know the active sensor area of the BMPCC measures 12,48 x 7,02 mm, it is fairly easy to check if our C-mount lenses will cover the full sensor. Calculate this by taking a picture with a lens on your micro four thirds camera, and crop out the image area of theBMPCC.
 


In Photoshop:
Open the image. Go to Edit > Image Size, uncheck resample image. Change Image width to 19 centimeters, press ok Go to Image > Canvas Size, change dimensions to 12,48 x 7,02cm, press OK to crop the image to BMPCC size. Resize to 1920x1080 pixels Post your results! Note: If you shoot on the GH3 or other MFT camera's, the sensor size is 17 x 13mm, so change the width in step 3 to 17 cm!

To lazy to do it yourself or you can't work it out? Upload the full resolution files and I'll do it.

List terms explained:

Yes = covers the full sensor of the Blackmagic Pocket Cinema Camera
No = doesn't cover the sensor
Needs modification = Doesn't fit on C-mount to M43-adapter without modifications
Equivalent = The focal length and depth equivalent on a fullframe camera (5D Mark III for example)

Blackmagic Pocket Cinema Camera Compatibility list

Primes
 
Apollo 25mm f/0.85 - Yes = 72mm f/2.4 equivalent [link to proof]
 
Angenieux 10mm f/1.8 Retrofocus (Fixed Focus) - Yes (dark corners) = 28,8mm f/5.2 equivalent [link to proof] [more info]
 
Carl Zeiss Jena Tevidon 10mm f/2 - Yes - Needs modification = 28,8mm f/5.8 equivalent [link to proof] [more info]
Carl Zeiss Jena Tevidon 35mm f/1.9 - Yes - Needs modification = 101mm f/5.6 equivalent [link to proof]
 
Century 9mm f/1.8 - YES (poor quality) [link to proof]
 
Computar 8mm f/1.3 - NO [link to proof]
Computar 16mm f/1.4 - NO [link to proof]
Computar TV Lens 25mm f/1.8 - YES = 72mm f/5,2 equivalent [link to proof]
 
Cosmicar 8,5mm f/1.5 - NO [link to proof]
Cosmicar 12.5mm f/1.8 - YES - Needs modification = 36mm f/5.2 equivalent [link to proof]
Cosmicar 25mm f/1.8 - YES - 72mm f/5.2 equivalent [link to proof]

Ernitec 6.5mm f/1.8 - YES (heavy distortion) [link to proof]
Ernitec/Navitar 17mm f/0.95 - YES (v. blurry corners & distortion) [link to proof]
 
Fujinon TV 12.5mm f/1.4 - Yes (blurry corners) - Mod.? (unknown) = 36mm f/4 equivalent [link to proof]
Fujinon TV 16mm f/1.4 - NO [link to proof]
Fujinon TV 35mm f/1.7 - YES - Needs modification = 101mm f/4.9 equivalent [link to proof]
 
Leitz Macro Cinegon 10mm f/1.8 - Yes (dark corners) = 28,8mm f/5.2 equivalent [link to proof]
 
Kern Switar 10mm f/1.6 - Yes (slight vignette & blurry corners) [link to proof]
 
Nikon Cine Nikkor 13mm f/1.8 - Yes =  37,5mm f/5.2 [link to proof]
Nikon Cine Nikkor 25mm f/1.8 - Yes = 72mm f/5.2 equivalent [link to proof]
 
Pentax 25mm f/1.4 - YES - 72mm f/4 equivalent [link to proof]
 
Schneider 10mm f/1.8 (silver version) - No (almost) [link to proof]
Schneider-Kreuznach Cinegon 11.5mm f/1.9 - No (almost) = 33mm f/5.6 equivalent [link to proof]
Schneider-Kreuznach Cine-Xenon 16mm f/2 - Yes = 46mm f/5.8 equivalent [link to proof] [link to proof (2)]
Schneider-Kreuznach Xenon 25mm f/0.95 - Yes = 72mm f/2.7 equivalent [link to proof]
Schneider Xenoplan 17mm f/1.7 - Yes (blurry corners) - [link to proof]
 
SLR Magic 11mm F1.4 - Yes - [link to proof] (added by EOSHD)
 
Tokina TV Lens  8mm f/1.3 - NO [link to proof]
Tokina TV Lens 16mm f/1.6 - NO [link to proof]
 
Taylor-Hobson Cooke Kinic 25mm f/1.3 - Yes = 72mm f/3.7 equivalent [link to proof]
Taylor-Hobson 25mm f/1.9 - Yes - 72mm f/5.6 equivalent [link to proof]
 
Wesley 25mm f/1.4 - YES = 72mm f/4 equivalent [link to proof]
 
Wollensak Cine Raptar 12.5mm f/1.5 - Yes = 36mm f/4.3 equivalent [link to proof]
Wollensak Cine Raptar 25mm f/1.9 - Yes = 72mm f/5.6 equivalent [link to proof]

$ 25 noname 25mm f/1.2 CCTV - YES = 72mm f/3.5 equivalent [link to proof]

Zooms
 
Ernitec 6-12mm f/1.4 - NO [link to proof]
 
Kowa TV Zoom 12.5-75mm f/1.8 - NO [link to proof]

@Rungunshoot
Before buying any of those (around $500 price range), I'd rather wait 'til they are produced by SLR magic or others (or abuse some of the cheaper or older lenses Andy listed). I expect the BM Pocket to be a bestseller, it will push MFT tremendously and people will demand wide angle lenses. Panasonic and the other MFT sellers say 'thanks BM'!
 
Until then, there are two existing MFT lenses that allow a run&gun style (equaling 36mm, the classic reporter's focus length) and are fast enough to provide some shallow DoF, this is the SLR magic 12mm (originally an observation lens as well) and the Olympus 12mm (with the automatic functions working, I am curious to see the auto exposure work with the 13 stops and high iso).
 
As for a standard lens, why not the Lumix pancake 14mm that's seen on the camera on the BM site? (EDIT: I know 50 mm is considered standard, but that's valid only for a 3:2 aspect ratio, it would be 47 mm for 16:9, 42 mm is a good compromise. You wouldn't shoot portrait close ups with it, but as far as portrait lenses are concerned, there are of course hundreds of options for MFT.)
 
With the BM cameras, we will see another game changing. It will no longer be about 'as shallow a DoF as possible'. There is nothing special about that any more. It will start to be about image quality. We are not yet fully aware of the avalanche that has been triggered. Ordinary hobbyists will buy the BM Pocket for under $1000 (in Germany, you can order it for 894,88 €, that's almost 200 € less than the GH3, body only). They will be able, at least, to use Resolve lite 10 for free, which has more editing features than #9 and works platform independant (a very shy love relationship between BM and Apple has to be acknowledged though). They will be told that RAW as well as ProRes 422 10-bit provide better quality than their pathetic '16,7 million colors' display can handle. They will be convinced by the industry that their old displays don't suffice anymore. They will buy 10-bit monitors (a few years ago 5-10.000 bucks, now starting at ~500, allegedly Dell has some good and affordable models. BTW: This is a serious limitation for all iMac users. Their displays are 8-bit. Probably there is/will be a solution from BM, I don't know, BM says 'thank you Apple'). They will learn to see the higher quality (frankly, it's easier to see than the difference between, say, 720p and 1080p). 
 
This is what I call a change.
 
EDIT: We can only speculate which lenses will be best for the Pocket. Sensor size, lens characteristics, post-debayering, we simply can't tell how these things influence each other. I really think it's wise to start with the most basic pancake and then collect experiences.

I don't think anything is going to give you that look on the BMPCC. I've been researching the same thing, and the focal length of lens you're seeking is 8mm.  While those lenses exist for s16mm, they would need impossibly large apertures to equal the wide-yet-shallow DOF in your sample images.
 
You might get something vaguely similar with this combo:
 
m43 speed booster (doesn't exist yet)
voigtlander 17.5mm f/.95
 
That would something be like a 35mm f/1.4 on full-frame.  
 
As far as I know, nothing is equally fast and wide in the m43 arena.

Don't think that is what Tim is referring to Caleb.
 
The spec flange for EF is 44mm and this is what Blackmagic went with.
 
The actual flange on EOS DSLRs is more like 43.5mm to allow for some tolerance for infinity and AF micro adjustments. This wasn't factored into the Blackmagic Cinema Camera as it would have knocked cinema glass off their marks.
 
Still think they should dump EF and go for E-mount.
 
I bet you anything there will be an E-mount version of the 4K camera come the end of the year.

By the time 4K actually becomes a mainstream delivery format it could be 5+ years from now. Even "HD" brodcast in the States is often 1080i or 720p.  I think everyone is jumping the gun on the 4K for future proofing.
 
Using 4K to re-frame, track, zoom, that makes sense. But it's more of a luxury, less a necessity at this point.

But i fear my eyes will see it, when i try to bring out shadow detail that has been thrown away!
 
There is a reason why .CR2 files (which use actual loss less compression) are so big for stills images...

Please correct me if I'm wrong.
 
If you record a format, then compress it, and it looks the exact same after playback, then it is "visually loss-less".
 
If you record a format, then compress it, and you lose no information whatsoever (including detail that can't be seen upon native playback) then you have a true loss-less format. 
 
It sounds like the compressed Raw is going to throw away some information.