It was about 1.5 safe on the 2.5K, but on the Pocket Camera it seems like 1 stop above that, which isn't massive.

I've never actually checked Ultrascope, but sounds like it lines up. The important thing for me is saturating all three channels without clipping one too quickly on anything important.

IR Pollution, ND color casts, etc., all of those play a part.

I'm fairly sure I'm switching over to Schneider Platinum from Tiffen WW. The former's much more neutral to begin with and I think they may actually suit the camera(s) better. Still have to test... Platinum's expensive.

Another review of the Pocket. It compares the BMPCC favorably with the 5D Mk III and even the Red Epic.

http://www.learningdslrvideo.com/black-magic-pocket-review/

Let me start off by saying I do not have enough experience yet to do a proper review of this camera. Why? Because I am not a professional colorist that works on log images on a daily basis. However, I can give you a good run down if you are just starting out or at a level like me. The Pocket camera has taught me a lot in the past 30 days, in fact the images that I tried to grade a month ago I have since re-graded with much better results. I have asked several colorists and others many questions over the past month to help me get better with this thing. Guys I really think this is where things are heading so if you like me you need to start investing in yourself by either taking some advanced grading courses or experiment a lot with a camera like this, and at this price point this looks to me THE camera to learn on.

About half the time I am frustrated with the pocket camera because I can’t get the results similar to my 5D, and the other half when I do get the results I want I think it’s a wonderful image. Why is this? It’s not the camera’s fault, it’s because I am not an advanced colorist. The Canon 5D has spoiled me because I never had to work the image that much, but with the pocket camera in film mode you have to work each image a great deal. I understand the basics of color correction using ProLost settings and CineStyle, but correcting a log image is a different animal and a different workflow is needed, such was which node to place the LUT in to.

Is this guy accurate in saying he thinks the 5Dlll RAW has more dynamic range than the Pocket? Then again he's comparing RAW (5Dlll) vs ProRes (Pocket). Another thing that struck me was his comment about the 5Dlll RAW having the same resolution as the Pocket's ProRes. Makes me wonder how much of a boost we expect to get with RAW applied to the Pocket.

@Ian_T

To be fair in either case, neither cameras have been scientifically measured as far as DR, and I'm not even sure the 2.5K has been. Which is strange, because that one has been out for quite some time.

What's to be noted, though, is that Dave still doesn't know how to expose the camera, and two, the camera's DR is mapped differently than the 5D3, the latter having most of its DR down bottom and the former certainly up top.

I can only say from experience that the 2.5K certainly has a solid two to three tops up top over the MK3, and around a stop and a half or less down bottom (noise floor), which puts it about a stop under the 2.5K.

The Pocket is certainly right in line with the 2.5K in ProRes, and likely right on in RAW.

@matt_gh2 I believe for the majority of his tests he used H.264 in the 5Dlll vs the Pocket. But if you go a little bit more than half way down his page you’ll see he ports the 5Dlll RAW vs 5D111 H.264 vs BMPCC with a side by side picture of his face. Right below it this is what he says:

“Speaking of Magic Lantern, how does the Pocket camera compare to my 5D3 in raw? I didn’t really want to spend too much time on this because I will review this camera again when it shoots raw, but in terms of resolution the pocket camera seems to have the same amount of detail as 5D3 raw, but in terms of dynamic range I would give the win to the 5D3 raw.”

@Kholi yeah you’re right…I have not seen any proper testing of the BMCC when it comes to this. I’m not sure but I think Frank over at DVXUser has a blog where he tests it against a Red, Alexa and F5. But against the ML 5Dlll I’m curious about his statement because to me, or at least through my naked eyes, the Canon doesn’t even come close to the BMCC (and like you suggested the BMPCC will be in the same territory). We all know what it says on paper but to see actual side by side test results would be telling.

The first date with my BMPCC was quite weird... I felt like a newbee, even I had GH1/2/3 for 3 years... I just passed trough menu, prepare the camera and pressed the Rec. button, hopping that after grading I will be very pleased. Unfortunately I choose to slect film mode for viewing (very dificult with that flat image... today I will set the display to video mode), but I used zebra and some time I use iris (point a spot wich I want to be exposed properly, and when I pressed iris button, the camera set the aperture of the lens. Anyway, it apeares to me a little bit overexposed... I will try 90% zebra... I have several SD cards, but only Sandisk 32 GB, 45MBs work fine (did not worked with GH3 format, but exFAT 128b allocation unit). I will buy the faster and biger one for future raw (it is just aver the corner, see J Brawley blog!!!!). Is there something over 95MBs? Since now I am very hapy and confident. I am not afraid anymore... But I wish desperately they do something for sleeping the camera (I use a 7Ah/12V lead acid battery in a purse...) and asigned buttons for ISO and WB, like arows up/down for exposure. The only big problem for me is that I find not a way to export 10 bit 422 (PPro CC) for final use (usualy I will use part of my production for future projects, so I need a good export... Somebody help, please!... Thanks!

Guys, somewhere else on the net somebody asked a question I'd also like to know the answer to. While shooting is it possible on BMPCC to save to the card in 'film' mode (e.g. maximum DR) but to view on display and/or external monitor as 'video' (e.g. clipped DR)? The idea is to judge exposure better - eye, zebras, false colors and/or to give the external monitor more contrast so that focus peaking works better on it.

And on a separate note - is there some problem with bmpcc "forgetting" the settings you make?

Tux,

BMC v1 2.5k RAW is (apparently) 5MB/frame. It shoots 1.6x more pixels than the Pocket's 1080p, so: Pocket uncompressed RAW: ~3.1 MB Pocket compressed RAW (~1.5:1): ~2.1 MB @ 24fps = 50 MB/s (compressed) @ 30fps = 62.5 MB/s (compressed)

jbrawley says 2.0MB-2.4MB per frame, matches my numbers early in this thread. So you're looking at around 20mins @ 23.976fps on a 64GB card.

@atagunov, yeah if you read a few posts up, you can do that.

@_gl

I'm averaging 10-10.5 minutes to each 32GB Sandisk 95MB/s card. So, that's about right.

@29.97/30 I'm around 8 minutes or so.

I was definitely able to provoke a 2.4mb stream during stress testing.

Thx, also is there an issue with camera "forgetting" your settings as alleged on the internet?

I remember reading that aperture isn't remembered on active lenses.

John Brawley's First RAW images frm BMPCC:

https://www.copy.com/s/s36D39T6q7oa/Pocket%20DNG

Not impressed with those still.