Thread: The Pocket 4K has half a stop more DR than the old Pocket

Page 2 of 6 FirstFirst 1234 ... LastLast
Results 11 to 20 of 52
  1. #11  
    Senior Member
    Join Date
    Aug 2014
    Posts
    179
    Quote Originally Posted by shijan View Post
    Just check values with Color picker.
    P4K patches left to right: 255, 255, 226
    Sony patches: 180 168 147
    They just cut 2 stops of dynamic range from Sony for unknown reason.

    This is some kind of strange dirty marketing dynamic range measure. From P4K footage never feels like outperforms older cameras in dynamic range. It looks sharper and with less moire, but always have that fake DSLRish "slightly not enough dynamic range to look like real film" feel.

    This is like samples from other thread, where two frames shoot with P4K and BMPCC, but P4K overexposed almost 2 stops compare to BMPCC and calm that p4K shadows are cleaner. The funny thing that even with 2 stops underexposed BMPCC with recovered shadows in post looks almost same as P4K already exposed 2 stops higher and downscaled from 4K to HD.
    It doesn't matter what your color picker says, Cinema5D clearly stated their methodology. It's not a conspiracy.

    The answer to your paradox is in the post above yours. The Pocket sensor is probably capable of capturing more total stops, but the P4K has a better signal to noise ratio, so slightly better useable DR according to their program.
    Reply With Quote  
     

  2. #12  
    What's also interesting is that they measured the DR in ProRes. It should be a little better in RAW.
    Pocket 4K "Chaplin".
    Reply With Quote  
     

  3. #13  
    Senior Member Taikonaut's Avatar
    Join Date
    May 2012
    Posts
    2,922
    Looks like some of the BRaw benefit has been added to the P4k Prores, though CineDNG I'm not 100% clear?
    Reply With Quote  
     

  4. #14  
    Senior Member
    Join Date
    Mar 2017
    Location
    Melbourne, Australia
    Posts
    708
    Quote Originally Posted by Stefan Antonescu View Post
    What's also interesting is that they measured the DR in ProRes. It should be a little better in RAW.

    Quote Originally Posted by Taikonaut View Post
    Looks like some of the BRaw benefit has been added to the P4k Prores, though CineDNG I'm not 100% clear?
    Prores is being denoised to help turn over which I don't really agree with since a big draw of BMD cameras was always the untouched image. As it's being denoised it scored slightly better than Raw in DR on cinema5d but once you denoise RAW the result are similar and once you add in highlight recovery and further manipulation Raw holds more DR.

    I think the NR should only activate if you are using the extended video mode and not the film mode.

    It was noted in someones recent review that the prores seemed too clean, too perfect and they were spot on it seems since it's being denoised.

    edit - it was papiskokuji's review here http://www.bmcuser.com/showthread.ph...f-the-BMPCC-4K
    Reply With Quote  
     

  5. #15  
    Senior Member Taikonaut's Avatar
    Join Date
    May 2012
    Posts
    2,922
    Seems that de noise might also be a feature of BRaw as well.
    Its look way too clean and you can see how much cleaner it is compared to RED Gemini. I agree de-noise should be LUT only and leave the noise well alone.

    De-noise is something you find in lower level consumer grade cameras, is that the way BMD is going?
    Reply With Quote  
     

  6. #16  
    Junior Member AndreeOnline's Avatar
    Join Date
    Aug 2012
    Location
    Zürich
    Posts
    29
    Quote Originally Posted by shijan View Post
    Those are very strange measurements.
    No need to sweat it. I think it's fair to say that the guys/gals behind Cinema5D are very transparent about how they measure. I also like the fact that they don't use the absolute number of patches that ImaTest can spit out (typically how RED measures), but that they are looking more at usable range,

    Don't jump to conclusions based on screenshots of frame grabs with DR patches. The cameras are set to expose the second patch to just within range with the help of waveforms and/or similar tools. There are log curves from more than one manufacturer that maxes out well before IRE 100. Normally I'd expect the second patch to be very close in brightness to the first one, but C5D are currently displaying tests overlapping two different procedures. I'm hoping we'll see an updated chart soon, with many cameras tested exactly the same way.

    Apart from that, since we're not looking at linear light here—but rather through a proprietary log curve—some cameras will show more even steps between patches while some will have more of a 'perceived curve' behind the patches. But at the end, the ImaTest software will spit out an objective score based on criteria x, y or z.

    And lastly, in the comments under the YouTube video there is a post by Cinema5D that addresses the ProRes vs raw noise levels and the answer they got from BMD. It's all very transparent and clear.
    Last edited by AndreeOnline; 10-30-2018 at 05:16 AM.
    Reply With Quote  
     

  7. #17  
    Senior Member
    Join Date
    Sep 2012
    Posts
    1,954
    255 IRE? They are transparent, I believe the issue is that the words don’t match the pictures.

    What lacks transparency is C5Ds relationship with BM but with no less than a dozen links to B&H in the body of the article I feel I have a pretty good idea of where they stand with both companies. $$$$
    Reply With Quote  
     

  8. #18  
    Junior Member AndreeOnline's Avatar
    Join Date
    Aug 2012
    Location
    Zürich
    Posts
    29
    Quote Originally Posted by Howie Roll View Post
    255 IRE?
    Had written something else initially (regarding sampling screenshots and RGB), but edited it and missed that. Thanks.
    Reply With Quote  
     

  9. #19  
    Member
    Join Date
    Jan 2013
    Location
    Sydney
    Posts
    86
    Quote Originally Posted by Howie Roll View Post
    What lacks transparency is C5Ds relationship with BM but with no less than a dozen links to B&H in the body of the article I feel I have a pretty good idea of where they stand with both companies. $$$$
    They're practically Satan.
    http://www.gamma-movie.com The IndieGathering 2016 best Sci-Fi feature screenplay.
    Reply With Quote  
     

  10. #20  
    Senior Member
    Join Date
    Sep 2012
    Posts
    1,954
    Quote Originally Posted by squig View Post
    They're practically Satan.
    I think an ethics statement is pretty common these days. Maybe you’re right though, nothing holy ever came from black magic.
    Reply With Quote  
     

Similar Threads

  1. Replies: 3
    Last Post: 10-26-2018, 02:22 PM
  2. Replies: 10
    Last Post: 12-07-2017, 05:34 PM
  3. [FOR SALE] Pocket Cam (LNIB) | Pocket Movcam Cage (Brand New)
    By UTR-Labs in forum For Sale / Want to Buy
    Replies: 4
    Last Post: 04-23-2016, 12:29 PM
  4. Replies: 0
    Last Post: 01-20-2016, 02:42 PM
  5. Replies: 8
    Last Post: 01-11-2015, 07:55 AM
Bookmarks
Bookmarks
Posting Permissions
  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •