Thread: MINILF-MONSTRO thoughts ;)

Reply to Thread
Page 9 of 13 FirstFirst ... 5678910111213 LastLast
Results 81 to 90 of 122
  1. #81  
    Quote Originally Posted by Björn Benckert View Post
    What I wrote/mean is that you can add a ND .3 lets say to ”slide the DR” of a camera. And yes that works as a DR slider,
    You do nothing to camera DR, you reduce the incoming light intensity hitting the sensor. : )

    Camera DR is fixed. Sensor sees X light intensity as the smallest and clips at Y intensity.
    You cannot do anything with the camera to change its DR handling.

    You can present it differently by a different transformation which makes it seem as DR is distributed differently - "ISO" digital gain-equivalent feature. Nothing changes with capture properties.

    When you put an ND or close the iris you don't "slide the DR", it just seems that way visually on a chart out of a wider context.
    You feed the sensor with reduced light intensity, which burries the image down in luminance. It is not the same image as the one exposed brighter. It is not just a noise difference. The image is much thinner. You sample less light.
    Shadow reach and cleaner shadows are something else, which doesn't change that or compensate the loss of image depth.

    Quote Originally Posted by Björn Benckert View Post
    The dubble Iso change I also mentiond is just to compensate for that light loss so in the end picture looks as bright as before adding the ND.
    Yes, with reduced density and boosted imperfections.

    Because you stretch the data from lower recorded values.
    Analog > Apollo wooden handgrip http://omeneo.com
    Digital > Primers - professional image transformation tools http://omeneo.com/primers

    imdb


    "Como delfines en el fondo del oceano
    volamos por el universo incentivados por la esperanza"

    "L'esperanza", Sven Väth
    "It's a poor sort of memory that only works backwards"
    Jung/ Carol
    Reply With Quote  
     

  2. #82  
    All that is understood hrvje.
    Im not saying using NDs change the dr of the sensor. But it moves the dr of your aray/setup. For example all Monstro setups does not have the same DR depending of what olpf is in use the DR slide left for lowlight olpf and right for skintone olpf.

    So this is not true ”Camera DR is fixed. Sensor sees X light intensity as the smallest and clips at Y intensity.
    You cannot do anything with the camera to change its DR handling.”

    Or atleast not if you dont count olpfs as part of a camera. Very few shoot without them.


    To me its interesting to see the difference between the cameras when the clipping points are aligned with shutter or ND. As when doing so and developing the captures to look identical I can see what noise levels I get on the cameras when leveled. To my findings even when monstro is hit with less light in this way to match alexa, I still get a cleaner image at 4k mastering. And yes then I have still more than one stop more to give on monstro.

    If comparing images from cameeas where one is clipping is kind of like using mesurements that are of the charts. Sure you csn shoot multiple tests with different exposures etc but I find it good to do it this way as it narrows things down and keeps other settings equal sure NDs contaminate colors but when doing comparison tests with leveled clippingpoints and doing it first with ND and then without ND and instead using a shorter shutter its quite easy to see what the nd contamination is, normally not much if such low as ND.3 or .6
    Björn Benckert
    Creative Lead & Founder Syndicate Entertainment AB
    +46855524900 www.syndicate.se/axis
    VFX / Flame / Motion capture / Monstro
    Reply With Quote  
     

  3. #83  
    Member
    Join Date
    Apr 2007
    Location
    Miami
    Posts
    90
    Quote Originally Posted by Jacobo Martinez View Post
    Thanks for this! Would it be possible to see the same clip with only the arri standard rec 709 LUT?
    Here is the test with Arri Rec709. Shot 800, rendered at 1600 to H265 UHD resolution. 50mm Arri Master Prime, available light. No denoising, very light color grade applied. I'd say the camera performs very well at 1600 ISO and noise is quite acceptable and usable for my taste. By the way, footage was recorded RAW and stored to Arri Raw HDE (40% lossless compression) which Davinci read easily from an SSD drive.

    https://vimeo.com/382500794
    Reply With Quote  
     

  4. #84  
    Quote Originally Posted by Björn Benckert View Post
    All that is understood hrvje.
    Im not saying using NDs change the dr of the sensor.
    You said "camera". : )

    Quote Originally Posted by Björn Benckert View Post
    But it moves the dr of your aray/setup.
    It doesn't "move the DR" literally, it reduces the incoming light intensity. DR is a ratio between min and max. "Moving DR" is a symbolical notion.

    You capture lower intensity light, bury the image in light capture, relying on sensor shadow reach and then dig it out when quantized in data - "stretch it" back in the image development relying on lattitude.

    This is not a free ride.
    You don't have the same negative as if exposing higher.

    Quote Originally Posted by Björn Benckert View Post
    For example all Monstro setups does not have the same DR depending of what olpf is in use the DR slide left for lowlight olpf and right for skintone olpf.
    OLPF can leave more of the IR range open. That is not the same as overall sensitivity nor does it compensate for the light loss from lower exposure.

    Quote Originally Posted by Björn Benckert View Post
    So this is not true

    ”Camera DR is fixed. Sensor sees X light intensity as the smallest and clips at Y intensity.
    You cannot do anything with the camera to change its DR handling.”
    You are mistaken.

    : )


    Quote Originally Posted by Björn Benckert View Post
    To me its interesting to see the difference between the cameras when the clipping points are aligned with shutter or ND. As when doing so and developing the captures to look identical I can see what noise levels I get on the cameras when leveled. To my findings even when monstro is hit with less light in this way to match alexa, I still get a cleaner image at 4k mastering.
    You get a cleaner image because of Monstro's lower noise floor and higher resolution.
    You also get a thinner image that way.

    You can ride that latitude range due to high bit depth and great sensitivity, night scene photographers do it all the time, but that comes at a cost of trading that image depth and contrast range. For a quality image you need the contrast range for healthy curve with proper toe and shoulder and image depth for tonal definition and subtleties.

    The loss matters less with starry skies, city scapes, corporate events etc and still images. In a grading suite with imagery requiring a special aesthetic tunable in subtleties and presented on big screens it matters more.

    How much it matters, is valorization which depends on subjective.
    The fact that there is a loss, doesn't.
    Analog > Apollo wooden handgrip http://omeneo.com
    Digital > Primers - professional image transformation tools http://omeneo.com/primers

    imdb


    "Como delfines en el fondo del oceano
    volamos por el universo incentivados por la esperanza"

    "L'esperanza", Sven Väth
    "It's a poor sort of memory that only works backwards"
    Jung/ Carol
    Reply With Quote  
     

  5. #85  
    I give up. If you expose alexa right under clip and expose monstro the same way you will have cliping in the monsteo image. Thats all Im saying. And when clipping you are not using the full dr of the camera so that makes it difficult to compare. Using different Olpfs on monstro will make more or less clipping but it will still clip. Exposing to cloose to the clipping point will also give a thin capture so its ofcourse a balance act to eak out as much as possible out of any sensor. But again to my findings and I worked quite a bit with shooting alexa and monstro used in the same scenery even for composite shots, ND 3 infront of monstro or close down a stop compared to tge alexa is a good starting point. And then yes they need to be balanced in grade a good starting point for that is dubble iso on monstro and soft highlight setting in ipp2.

    Monstros finer noise structure then covers up for the dual gain and things looks quite equal. But still then with the stop loss of light monstro still gives slightly cleaner blacks so one need to grain match the two. ie add some alexa mimicing noise onto the red footage.
    Björn Benckert
    Creative Lead & Founder Syndicate Entertainment AB
    +46855524900 www.syndicate.se/axis
    VFX / Flame / Motion capture / Monstro
    Reply With Quote  
     

  6. #86  
    Junior Member
    Join Date
    Aug 2008
    Posts
    1
    Hi,

    I was wondering if someone could explain to me why the Monstro looks behind the Alexa LF in these dynamic range tests. These seem to be some pretty thorough tests.

    https://www.agdok.de/de_DE/kameratest2019-en
    https://vimeo.com/333200960

    I’m confused with the 1600 ISO trick as I was under the impression that with red raw you aren’t changing anything on the sensor level and correct me if I’m wrong if you under expose 800 and push it up you effectively create 1600 and so it would display the same dynamic range?

    Thanks Michael
    Reply With Quote  
     

  7. #87  
    Senior Member Jacobo Martinez's Avatar
    Join Date
    Jul 2013
    Location
    Nicaragua
    Posts
    314
    Quote Originally Posted by Hernan Herrera View Post
    Here is the test with Arri Rec709. Shot 800, rendered at 1600 to H265 UHD resolution. 50mm Arri Master Prime, available light. No denoising, very light color grade applied. I'd say the camera performs very well at 1600 ISO and noise is quite acceptable and usable for my taste. By the way, footage was recorded RAW and stored to Arri Raw HDE (40% lossless compression) which Davinci read easily from an SSD drive.

    https://vimeo.com/382500794
    Thank you ! Amazing camera, amazing lens .
    www.jakobfilms.com
    contact@jakobfilms.com
    Instagram: insta_jakob_films
    Reply With Quote  
     

  8. #88  
    Senior Member Jacobo Martinez's Avatar
    Join Date
    Jul 2013
    Location
    Nicaragua
    Posts
    314
    Quote Originally Posted by Hrvoje Simic View Post
    You do nothing to camera DR, you reduce the incoming light intensity hitting the sensor. : )

    Camera DR is fixed. Sensor sees X light intensity as the smallest and clips at Y intensity.
    You cannot do anything with the camera to change its DR handling.

    You can present it differently by a different transformation which makes it seem as DR is distributed differently - "ISO" digital gain-equivalent feature. Nothing changes with capture properties.

    When you put an ND or close the iris you don't "slide the DR", it just seems that way visually on a chart out of a wider context.
    You feed the sensor with reduced light intensity, which burries the image down in luminance. It is not the same image as the one exposed brighter. It is not just a noise difference. The image is much thinner. You sample less light.
    Shadow reach and cleaner shadows are something else, which doesn't change that or compensate the loss of image depth.



    Yes, with reduced density and boosted imperfections.

    Because you stretch the data from lower recorded values.
    Yes!! Exactly! It works the same way with any A/D converter, think of audio. If you capture low audio signal, you will have capture less DR, noise floor will be higher if you push the level up in post, the audio will be thinner. I agree with you 100% what you are saying . DR is always fixed.
    www.jakobfilms.com
    contact@jakobfilms.com
    Instagram: insta_jakob_films
    Reply With Quote  
     

  9. #89  
    Quote Originally Posted by hmmm View Post
    Hi,

    I was wondering if someone could explain to me why the Monstro looks behind the Alexa LF in these dynamic range tests. These seem to be some pretty thorough tests.

    https://www.agdok.de/de_DE/kameratest2019-en
    https://vimeo.com/333200960

    I’m confused with the 1600 ISO trick as I was under the impression that with red raw you aren’t changing anything on the sensor level and correct me if I’m wrong if you under expose 800 and push it up you effectively create 1600 and so it would display the same dynamic range?

    Thanks Michael

    In that vimeo test monstro -6 looks cleaner than alexa -6 but then monstro clips at +6....
    so now if you change nothing but add nd .3 infront of monstro thrn monstro +6 will not clip and -6 will still look fine as monstro got more to give in the low end. and yes to make them look the same you would ofcourse need to compensate with double iso for monstro.

    Arri and red sugest different mapping of the dr. Red sugest to expose the sensor more. Now for example panavision rate monstro 1600 which distribute the DR more as arri does. It all kind of comes down to what you see as acceptable noise. If you have no problem with noise then you can rate your camera higher and get more headroom and more colorfidelity in your highlights, if you priorities a more clean image then you might have to sacrifice the highlights and even clip them.
    Björn Benckert
    Creative Lead & Founder Syndicate Entertainment AB
    +46855524900 www.syndicate.se/axis
    VFX / Flame / Motion capture / Monstro
    Reply With Quote  
     

  10. #90  
    Senior Member luigivaltulini's Avatar
    Join Date
    Oct 2009
    Location
    Switzerland- LUGANO
    Posts
    2,492
    Quote Originally Posted by Björn Benckert View Post
    In that vimeo test monstro -6 looks cleaner than alexa -6 but then monstro clips at +6....
    so now if you change nothing but add nd .3 infront of monstro thrn monstro +6 will not clip and -6 will still look fine as monstro got more to give in the low end. and yes to make them look the same you would ofcourse need to compensate with double iso for monstro.

    Arri and red sugest different mapping of the dr. Red sugest to expose the sensor more. Now for example panavision rate monstro 1600 which distribute the DR more as arri does. It all kind of comes down to what you see as acceptable noise. If you have no problem with noise then you can rate your camera higher and get more headroom and more colorfidelity in your highlights, if you priorities a more clean image then you might have to sacrifice the highlights and even clip them.
    +1
    I would add to download the original files.
    Vimeo is certainly not very good at understanding the noise you have on the clip.
    Grading Suite Davinci Resolve, Rental RED Digital camera.
    MONSTRO 8k CF #831 "BLACK PEARL" , EPIC-W GEMINI # "DAFNE".
    ARRI ALEXA EV , ALEXA MINI LF "ROZI"


    SITE
    www.cine5k.ch
    Reply With Quote  
     

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