Thread: Sony PMW-F3

Page 22 of 26 FirstFirst ... 12181920212223242526 LastLast
Results 211 to 220 of 253
  1. Collapse Details
    Senior Member
    Join Date
    Mar 2005
    Location
    toronto, Canada
    Posts
    839
    Default
    Also, don't you sample diagonally anyway in the red ? ( sorta like Q67) ? Just curious.

    That's probably proprietary right ? :-)

    edit : I've always respected red's engineering team and you Graeme.

    Last edited by D_and_G; 11-12-2010 at 02:19 AM.


    Reply With Quote
     

  2. Collapse Details
    Red Team Graeme_Nattress's Avatar
    Join Date
    Apr 2005
    Location
    Ottawa, Canada
    Posts
    1,135
    Default
    Quote Originally Posted by D_and_G View Post
    Also, don't you sample diagonally anyway in the red ? ( sorta like Q67) ? Just curious.

    That's probably proprietary right ? :-)
    Nope, standard Bayer pattern CFA.

    BTW, here's a close up of a zone plate on F35 to show the chroma moire:


    If you think about it, it's pretty obvious that it must happen at high detail due to the size and spacing of the pixels, and how horizontally, there's a 2 pixel gap before another of the same colour. If you think on how you'd have to design an OLPF to avoid that.....

    The F35 photosites are smaller than RED's I think - they fit 3*1920*2*1080 = 12441600 in about the same area we fit 4520x2540 = 1148080. Ours are 5.4 microns square. If you have precise F35 sensor dimensions you can do the math.

    Graeme
    www.nattress.com - Film Effects and Standards Conversion for FCP
    www.red.com - RED - 4k Digital Cinema Camera


    Reply With Quote
     

  3. Collapse Details
    Senior Member
    Join Date
    Apr 2009
    Posts
    1,176
    Default
    That's kind of mesmerizing...


    Reply With Quote
     

  4. Collapse Details
    Red Team Graeme_Nattress's Avatar
    Join Date
    Apr 2005
    Location
    Ottawa, Canada
    Posts
    1,135
    Default
    It is rather!

    D&G - I've got to admit I was pretty much anti-bayer-pattern myself until I began working with them. Once I'd written a few demosaic algorithms and crunched some numbers on them, I began to appreciate both their benefits and their drawbacks. Fortunately, the major drawbacks are in the dealing with them in terms of compression and demosaic, which are rather tricky to get right, and that's the bit I do, so it's my own mis-fortune so to speak.

    Graeme
    www.nattress.com - Film Effects and Standards Conversion for FCP
    www.red.com - RED - 4k Digital Cinema Camera


    Reply With Quote
     

  5. Collapse Details
    Senior Member
    Join Date
    Mar 2005
    Location
    toronto, Canada
    Posts
    839
    Default
    Quote Originally Posted by Graeme_Nattress View Post
    If you think about it, it's pretty obvious that it must happen at high detail due to the size and spacing of the pixels, and how horizontally, there's a 2 pixel gap before another of the same colour. If you think on how you'd have to design an OLPF to avoid that.....

    The F35 photosites are smaller than RED's I think - they fit 3*1920*2*1080 = 12441600 in about the same area we fit 4520x2540 = 1148080. Ours are 5.4 microns square. If you have precise F35 sensor dimensions you can do the math.

    Graeme
    I assumed they had a 4 perf Full aperture size sensor. Using the above, it appears closer to a 3 perf Full aperture. Never shot with the cam, so I just assumed its size.
    I don't know why they didn't go 3 chip ccd, though. You can fine tune each olpf for the color associated, and avoid the compromises made with bayer chips.

    edit : I had questions here about the F35 test (S-log, lenses etc...) mostly out of curiosity - not questioning their results, just curious what the setup was.

    Aside from these tests; what's optimal for the F35 ?


    Last edited by D_and_G; 11-12-2010 at 02:21 AM.


    Reply With Quote
     

  6. Collapse Details
    Senior Member
    Join Date
    Apr 2009
    Posts
    1,176
    Default
    Graeme,

    Do you happen to have a larger version of that crop? I'd like to see what part of the chart this is...feels like I'm standing about 3" from a Monet.


    Reply With Quote
     

  7. Collapse Details
    Resident Preditor mcgeedigital's Avatar
    Join Date
    Apr 2005
    Location
    Potomac Falls, VA
    Posts
    7,926
    Default
    Quote Originally Posted by Graeme_Nattress View Post
    It is rather!

    D&G - I've got to admit I was pretty much anti-bayer-pattern myself until I began working with them. Once I'd written a few demosaic algorithms and crunched some numbers on them, I began to appreciate both their benefits and their drawbacks. Fortunately, the major drawbacks are in the dealing with them in terms of compression and demosaic, which are rather tricky to get right, and that's the bit I do, so it's my own mis-fortune so to speak.

    Graeme
    Well, you are certainly a wizard at it.
    Matt Gottshalk - Director/ Dp/ and Emmy Award Winning Editor
    Producer, Digital Creative for the United States Postal Service


    Reply With Quote
     

  8. Collapse Details
    Red Team Graeme_Nattress's Avatar
    Join Date
    Apr 2005
    Location
    Ottawa, Canada
    Posts
    1,135
    Default
    Quote Originally Posted by D_and_G View Post
    I assumed they had a 4 perf Full aperture size sensor. Using the above, it appears closer to a 3 perf Full aperture. Never shot with the cam, so I just assumed its size.
    I don't know why they didn't go 3 chip ccd, though. You can fine tune each olpf for the color associated, and avoid the compromises made with bayer chips.

    edit : I had questions here about the F35 test (S-log, lenses etc...) mostly out of curiosity - not questioning their results, just curious what the setup was.

    Aside from these tests; what's optimal for the F35 ?


    The capture looks s-log like to me - lens, probably one of the arri primes.

    You can use three chip + prism, but it's not without it's compromises either. There are prism artifacts, and it would mean on a 35mm sensor that you'd not be able to use standard PL mount lenses as the optical path would be wrong.

    Graeme
    Last edited by Graeme_Nattress; 11-12-2010 at 07:04 AM.
    www.nattress.com - Film Effects and Standards Conversion for FCP
    www.red.com - RED - 4k Digital Cinema Camera


    Reply With Quote
     

  9. Collapse Details
    Senior Member Duke M.'s Avatar
    Join Date
    Aug 2008
    Location
    CA & FL
    Posts
    2,736
    Default
    Quote Originally Posted by Michael Olsen View Post
    Pretty much nothing, aside from the fact that it is an S35mm sensor camera with the ability to use interchangeable lenses (as are the Alexa, D-21, RED ONE, Epic, the F35, the Genesis, the 9000PL, etc.). It's in a very different market segment.

    Oh, and there doesn't seem to be a proper home for it on the forum, so this thread ended up here.
    There was a proper thread in the SONY XDCAM section since it is a SONY XDCAM and started at nearly an identical time, but for some reason they closed that one and left the threads in the DVX and AF100 sections. Why? Go figure.


    Reply With Quote
     

  10. Collapse Details
    Senior Member Duke M.'s Avatar
    Join Date
    Aug 2008
    Location
    CA & FL
    Posts
    2,736
    Default
    Quote Originally Posted by D_and_G View Post
    [COLOR="#f5deb3"]
    I don't know why they didn't go 3 chip ccd, though. You can fine tune each olpf for the color associated, and avoid the compromises made with bayer chips.
    Three chip CCD in S35 size would be hugely expensive and cause even more over heating problems. That's why everyone is going to CMOS. The problem with CMOS is the early versions were/are slow.


    Reply With Quote
     

Page 22 of 26 FirstFirst ... 12181920212223242526 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •