Zebra issue on my BMCC

The place for questions about shooting with Blackmagic Cameras.
  • Author
  • Message
Offline

Patrick Finnegan

  • Posts: 210
  • Joined: Wed Aug 29, 2012 11:43 pm
  • Location: Los Angeles, & Santa Fe, New Mexico

Zebra issue on my BMCC

PostSun Sep 08, 2013 5:03 pm

Hi everyone,
I received my BMCC MFT last week and have been shooting numerous tests in al kinds of lighting and exposure situations. I am wondering if anyone else has been experiencing a similar problem. I found a discussion about this on another forum:
http://96.30.50.45/showthread.php?4992-I-really-need-help-with-exposing-footage-correctly-with-black-magic-cinema-camera
but thought I would bring it up here in hopes of getting some feed back potentially from Black Magic staff.

When trying to expose for 100% zebras my screen (exposing ETTR) goes so white and bleached out long before I see any zebras appear, so much so it is extremely difficult to judge the framing on the scene.

I decided to shoot a series of test shots as in the other forum users were complaining that they were getting their zebras, or exposure aids appearing on their external monitors like a TV Logic long before the cameras screen. In the test I shot two sample clips at 100%, 95%, 90%, 85% and 80% zebras, one just as the faintest hints of zebras appears and one with no zebras at all. I then took the samples into Resolve and looked at the scopes and here are my results:

at 100%, 95%, all of my shots showed some amount of clipping both with Zebras appearing and just without.
At 90% with no zebras showing, I was clearly safe from clipping but so close I could see that I was at risk of specular clipping over time.
at 90% (with some slight amount of zebras just showing) I had a safe margin in post to lift the highlights a bit if I needed and plenty of room to pull down the mid tones and shadows without introducing extra noise.

at 85% I was losing headroom I could otherwise recover in camera which would introduce unwanted noise in post if I lifted it them

In the other forum some users suggested that the Black Magic zebra was giving different values than the external monitors because the BM zebras were a direct indication of the sensor values while the SDI stream feeding an external monitor was being translated into a Rec 709 space and thus not as reliable.

At this point I am using the 90% just showing zebras to judge exposure but I am unclear whether there is something in my camera which is out of calibration or not.
Thoughts?
Last edited by Patrick Finnegan on Sun Sep 08, 2013 9:02 pm, edited 1 time in total.
Finnegan's Wake Films
Director/ Cameraman
Offline
User avatar

adamroberts

  • Posts: 4513
  • Joined: Wed Aug 22, 2012 5:27 am
  • Location: England, UK

Re: Zebra issue on my BMCC

PostSun Sep 08, 2013 5:40 pm

The zebras in camera are based on the 16bit sensor data. The same data that is being written to the CinemaDNG files.

The zebras on the external monitor are based on the signal coming out of the HDSDI port. That data is 10bit 4:2:2 and depending on the format will be BMCLog or Rec.709. This is the same data written to the ProRes file.

So the zebras will differ.

Zebras in camera are also an average of the data and in some cases a single channel might be clipping before the average value of the 3 colours clips. All depends on your scene.

Ultimately exposing comes down to personal preference but you want to feed the camera as much light as possible because you'll be able to pull the exposure down in post and lower the noise floor. The only way to find the perfect exposure for your taste and workflow is to test and refine.
Offline
User avatar

Tom

  • Posts: 1626
  • Joined: Wed Aug 22, 2012 5:08 am
  • Location: Manchester, UK

Re: Zebra issue on my BMCC

PostSun Sep 08, 2013 5:50 pm

If filming raw, just lower the iso until you can see clearly on the screen. It will still record at 800.
Tom Majerski
http://tetragrade.com/
http://www.imdb.com/name/nm5157752/
Offline
User avatar

adamroberts

  • Posts: 4513
  • Joined: Wed Aug 22, 2012 5:27 am
  • Location: England, UK

Re: Zebra issue on my BMCC

PostSun Sep 08, 2013 7:23 pm

Tom wrote:If filming raw, just lower the iso until you can see clearly on the screen. It will still record at 800.


Yup. And the Zebras will also still be displaying sensor clipping to you get the best of both. :-)
Offline

Patrick Finnegan

  • Posts: 210
  • Joined: Wed Aug 29, 2012 11:43 pm
  • Location: Los Angeles, & Santa Fe, New Mexico

Re: Zebra issue on my BMCC

PostSun Sep 08, 2013 9:00 pm

adamroberts wrote:
Tom wrote:If filming raw, just lower the iso until you can see clearly on the screen. It will still record at 800.


Yup. And the Zebras will also still be displaying sensor clipping to you get the best of both. :-)


great info and feedback...

but what about the fact that I cannot even see Zebras on my screen until the image is completely bleached out and clipping, particularly if I am shooting exteriors? Only when I dial down to 90% zebras do I trust I am being warned about clipping with enough margin of confidence. Is that something the rest of you have noticed (zebras very hard to detect at 100% on a bleached out image) or am I possibly looking at some kind of calibration issue on my camera? Since I have never used anyone else's Black Magic Camera, I have no basis of comparison for what I am seeing.
thanks
Finnegan's Wake Films
Director/ Cameraman
Offline
User avatar

adamroberts

  • Posts: 4513
  • Joined: Wed Aug 22, 2012 5:27 am
  • Location: England, UK

Re: Zebra issue on my BMCC

PostSun Sep 08, 2013 11:16 pm

As Tom said if you drop the ISO down to 200 the screen will become less bleached out and you'll be able to see the zebras easier.
Offline

Darryl Gregory

  • Posts: 939
  • Joined: Sun Dec 16, 2012 2:49 am
  • Location: LA

Re: Zebra issue on my BMCC

PostSun Sep 08, 2013 11:42 pm

Also try the 3M anti reflective film for Ipads be sure it is the 3M Brand!
with no glue or adhesive since this may leave a residue.
http://www.amazon.com/3M-Anti-Glare-Pro ... +for+ipads
Offline
User avatar

Tom

  • Posts: 1626
  • Joined: Wed Aug 22, 2012 5:08 am
  • Location: Manchester, UK

Re: Zebra issue on my BMCC

PostMon Sep 09, 2013 10:54 am

Patrick Finnegan wrote:
adamroberts wrote:
Tom wrote:If filming raw, just lower the iso until you can see clearly on the screen. It will still record at 800.


Yup. And the Zebras will also still be displaying sensor clipping to you get the best of both. :-)


great info and feedback...

but what about the fact that I cannot even see Zebras on my screen until the image is completely bleached out and clipping, particularly if I am shooting exteriors? Only when I dial down to 90% zebras do I trust I am being warned about clipping with enough margin of confidence. Is that something the rest of you have noticed (zebras very hard to detect at 100% on a bleached out image) or am I possibly looking at some kind of calibration issue on my camera? Since I have never used anyone else's Black Magic Camera, I have no basis of comparison for what I am seeing.
thanks


Set them to 100%, and trust them - if you do not see them but the screen is solid white - it is not clipping. In raw, because it is still in iso 800, you can really just use the iso settings to help visibility on the screen. The settings wont move the clipping point but will help you to compose your shots.
Tom Majerski
http://tetragrade.com/
http://www.imdb.com/name/nm5157752/
Offline

Patrick Finnegan

  • Posts: 210
  • Joined: Wed Aug 29, 2012 11:43 pm
  • Location: Los Angeles, & Santa Fe, New Mexico

Re: Zebra issue on my BMCC

PostMon Sep 09, 2013 6:33 pm

Thanks everyone. All of this is good information, particularly when shooting raw. As for Pro Res, if I want to stay at 800 ASA to maximize dynamic range, at present my screen image is so washed out at what should be the correct exposure that I would have difficulty even framing a subject. I like the trick of dialing 200 ASA to check the image and then resetting to 800 ASA for shooting.

Unfortunately, I think my problem runs deeper than trusting the overexposed image to be right as long as the zebras don't show clipping. I am new to resolve and interpreting its scopes, so I may be misunderstanding what I am seeing...
But, I retested my camera again this morning with my zebras set at 95%. Shot in raw as well as pro res. For the raw, I took the dng files into DaVinci and went into camera raw mode and Black magic film setting and then compared each of the test shots on the scopes. As you can see from the frame grab below of the pro res clip at 95% with the zebras just barely dialed away, the scopes are showing clipping in the red and blue channels. Additionally, I would never be able to even see the zebras when shooting exterior. They were so faint even using a loupe, and viewing interior in a dim room, I could barley detect them. I was able to pull down the exposure on the raw files and gain back a tremendous amount of information but the clipping was still there for parts of the image. I suspect my camera may need calibration.

Thoughts?


[img]
Pro%20Res%20image%20at%2095%%20zebra.jpg
[/img]
Attachments
Pro Res image at 95% zebra.jpg
Pro Res image at 95% zebra.jpg (240.01 KiB) Viewed 4217 times
Finnegan's Wake Films
Director/ Cameraman
Offline
User avatar

Tom

  • Posts: 1626
  • Joined: Wed Aug 22, 2012 5:08 am
  • Location: Manchester, UK

Re: Zebra issue on my BMCC

PostMon Sep 09, 2013 7:32 pm

Please share a dng showing the problem.

I dont follow exactly what the problem is.
Tom Majerski
http://tetragrade.com/
http://www.imdb.com/name/nm5157752/
Offline

Doge

  • Posts: 14
  • Joined: Sun Apr 21, 2013 7:40 pm

Re: Zebra issue on my BMCC

PostMon Sep 09, 2013 8:43 pm

Tom wrote:If filming raw, just lower the iso until you can see clearly on the screen. It will still record at 800.


+1

Also, when shooting in RAW, you don't need to dial your ISO back up to 800. You can keep it at 200 or 400 while shooting.
Offline

CaptainHook

Blackmagic Design

  • Posts: 1603
  • Joined: Wed Aug 22, 2012 4:50 am
  • Location: Melbourne, Australia

Re: Zebra issue on my BMCC

PostMon Sep 09, 2013 9:51 pm

Tom wrote:If filming raw, just lower the iso until you can see clearly on the screen. It will still record at 800.


I'm waiting to hear back from BMD after IBC about this, but i'm not sure if it's that simple (if you 'develop' in Resolve). Do a test shooting the same controlled scene at 200ASA/400/800/1600 in RAW. Open each shot in Resolve and you will see even though the exposure slider is at "0" for each shot, they appear darker/lighter so resolve is respecting the ASA setting. They question i'm waiting on a response to is, is it just an exposure adjustment that resolve does 'in the background', or are the same curves that the camera applies at those settings also being applied? If it's the curves, your ASA setting in RAW IS important since a curve (you can't currently manually change) is being applied. If it's just exposure, you still have to consider the range of the 'exposure' slider in Resolve is changed since "0" reflects the ASA setting, "0" does not always equal 800ASA. If you shot and 'monitored' at 200ASA, then you'd need to push the exposure slider in Resolve up 2 stops to see 800ASA. And IF resolve is also applying a "200ASA curve" to that, it won't look the same as setting 800ASA in camera.
http://www.captainhook.co.nz/blackmagic-cinema-camera-lut/

**Any post by me prior to Aug 2014 was before i started working for Blackmagic**
Offline

Patrick Finnegan

  • Posts: 210
  • Joined: Wed Aug 29, 2012 11:43 pm
  • Location: Los Angeles, & Santa Fe, New Mexico

Re: Zebra issue on my BMCC

PostTue Sep 10, 2013 6:14 am

Tom wrote:Please share a dng showing the problem.

I dont follow exactly what the problem is.


i am sorry. I don't mean to be confusing. Everyone's input for shooting in raw and dialing back the iso more or less solves the issue for raw...

but the problem remains for pro res

It is simple, when using zebras at 100%:-

1) the screen is bleached so white I can only visually detect the zebras by using a loupe in a darkened room,

2) I can barely see anything of the subject I am shooting because the screen is bleached so white

3) using the zebras at anything above 90% as exposure guides results in clipping


Tom, I can supply a dng but the raw is really not the issue as I can use 200 asa to see my scene.
The screen shot I showed in my last post shows a pro res test image I shot at 95% Zebras (which should be safe exposure) clipping in the red and blue channels.

So, in essence, the only exposure guide I have with the camera (zebras) is giving me false readings and in order to even use them I can barely make out my subject in the bleached image on the screen.

i am trying to find out if this is normal as I don't have another BMCC to make a comparison with

i of course can compensate by using only 90% zebras but find it odd that everyone else seems to be able to use 100% zebras on their BMCC's just fine while mine clips.

thanks for all of your valuable input concerning shooting raw
Finnegan's Wake Films
Director/ Cameraman
Offline
User avatar

Tom

  • Posts: 1626
  • Joined: Wed Aug 22, 2012 5:08 am
  • Location: Manchester, UK

Re: Zebra issue on my BMCC

PostTue Sep 10, 2013 9:43 am

Patrick Finnegan wrote:
Tom wrote:Please share a dng showing the problem.

I dont follow exactly what the problem is.


i am sorry. I don't mean to be confusing. Everyone's input for shooting in raw and dialing back the iso more or less solves the issue for raw...

but the problem remains for pro res

It is simple, when using zebras at 100%:-

1) the screen is bleached so white I can only visually detect the zebras by using a loupe in a darkened room,

2) I can barely see anything of the subject I am shooting because the screen is bleached so white

3) using the zebras at anything above 90% as exposure guides results in clipping


Tom, I can supply a dng but the raw is really not the issue as I can use 200 asa to see my scene.
The screen shot I showed in my last post shows a pro res test image I shot at 95% Zebras (which should be safe exposure) clipping in the red and blue channels.

So, in essence, the only exposure guide I have with the camera (zebras) is giving me false readings and in order to even use them I can barely make out my subject in the bleached image on the screen.

i am trying to find out if this is normal as I don't have another BMCC to make a comparison with

i of course can compensate by using only 90% zebras but find it odd that everyone else seems to be able to use 100% zebras on their BMCC's just fine while mine clips.

thanks for all of your valuable input concerning shooting raw


The waveform might show clipping, but try recovering the data - I am fairly sure you will find that there is detail still there.
Tom Majerski
http://tetragrade.com/
http://www.imdb.com/name/nm5157752/
Offline
User avatar

Tom

  • Posts: 1626
  • Joined: Wed Aug 22, 2012 5:08 am
  • Location: Manchester, UK

Re: Zebra issue on my BMCC

PostTue Sep 10, 2013 9:45 am

CaptainHook wrote:
Tom wrote:If filming raw, just lower the iso until you can see clearly on the screen. It will still record at 800.


I'm waiting to hear back from BMD after IBC about this, but i'm not sure if it's that simple (if you 'develop' in Resolve). Do a test shooting the same controlled scene at 200ASA/400/800/1600 in RAW. Open each shot in Resolve and you will see even though the exposure slider is at "0" for each shot, they appear darker/lighter so resolve is respecting the ASA setting. They question i'm waiting on a response to is, is it just an exposure adjustment that resolve does 'in the background', or are the same curves that the camera applies at those settings also being applied? If it's the curves, your ASA setting in RAW IS important since a curve (you can't currently manually change) is being applied. If it's just exposure, you still have to consider the range of the 'exposure' slider in Resolve is changed since "0" reflects the ASA setting, "0" does not always equal 800ASA. If you shot and 'monitored' at 200ASA, then you'd need to push the exposure slider in Resolve up 2 stops to see 800ASA. And IF resolve is also applying a "200ASA curve" to that, it won't look the same as setting 800ASA in camera.


Hmm, might have time to do a little test today, but this is how I thought it was:

Between 200-800, the exposure is just digitally reduced from 800 as meta data, so on playback in resolve the 0 exposure point is simply moved mathematically. at 1600, during ADC the shadows or low signal was given a slight bias to help with noise, and then flagged as meta data to playback at a higher exposure.

If you do find out for sure, could you let me (us) know your findings please.
Tom Majerski
http://tetragrade.com/
http://www.imdb.com/name/nm5157752/
Offline
User avatar

Tom

  • Posts: 1626
  • Joined: Wed Aug 22, 2012 5:08 am
  • Location: Manchester, UK

Re: Zebra issue on my BMCC

PostTue Sep 10, 2013 10:58 am

@CH

Just did a quick test: filmed the same scene at 800, 400, 200

In resolve I dropped the 800 by -2 exposure, 400 by -1, left the 200 untouched.

As you can see in both examples, they result in an identical waveform and image.

Example 1:
http://tommajerski.com/publicimages/iso_example_1.jpg

Example 2:
http://tommajerski.com/publicimages/iso_example_2.jpg


To those not following what I am trying to demonstrate;
I am showing that when filming in Raw mode, if the image on screen is too bright to see and not clipping, you can lower recording ISO and it will not change the dynamic range response or clipping point, but will make composing much easier. In these examples, I filmed a scene in ISO 800, 400 and 200, in Resolve I dropped the ISO 800 shot to -2 (to match with the iso 200 shot) and the iso 400 shot by -1 (to match the iso 200 shot). As you can see, this results in the same image, both visually and according to the waveform readout.
Tom Majerski
http://tetragrade.com/
http://www.imdb.com/name/nm5157752/
Offline

Patrick Finnegan

  • Posts: 210
  • Joined: Wed Aug 29, 2012 11:43 pm
  • Location: Los Angeles, & Santa Fe, New Mexico

Re: Zebra issue on my BMCC

PostTue Sep 10, 2013 5:22 pm

Thanks Tom for the results of the test. I will use 200 ISO when shooting raw for framing purposes as it will make it much easier, now confident that the sensor data will be exposed at 800 ISO.

I will go back and play with grading the (apparently) clipped images and see if I can full recover the highlights. I appreciate all of your feedback. This is an example of this forum at its best, when used to share knowledge and discuss possibilities rather than bemoaning disappointments!

As for shooting in Pro Res, I guess I will just have to get used to looking at a very bleached out screen if I want to maximize dynamic range.

As a side note, I own a Sony F3 and at a certain point Sony provided a firmware update which allowed users of the camera to dial an exposure index (EI) on the viewing image which did not affect the recorded image.

Don't know if the folks from Black Magic are reading this, but it would be great to have something similar for the Black Magic Cameras where you could dial down the (apparent) exposure of the screen so you could discern more contrast and detail in framing while being confident that what was being recorded to the SSD drive was maximizing dynamic range.

One can hope...
Last edited by Patrick Finnegan on Tue Sep 10, 2013 10:08 pm, edited 1 time in total.
Finnegan's Wake Films
Director/ Cameraman
Offline
User avatar

adamroberts

  • Posts: 4513
  • Joined: Wed Aug 22, 2012 5:27 am
  • Location: England, UK

Re: Zebra issue on my BMCC

PostTue Sep 10, 2013 5:56 pm

Just been thinking about this and I was playing with the camera earlier...

What mode are you viewing in? Video or Film?

Also have you tried adjusting the brightness of the display?

I generally don't worry about the display being washed out as I use an EVF for framing, focusing, etc.
Offline

Patrick Finnegan

  • Posts: 210
  • Joined: Wed Aug 29, 2012 11:43 pm
  • Location: Los Angeles, & Santa Fe, New Mexico

Re: Zebra issue on my BMCC

PostTue Sep 10, 2013 6:36 pm

Hi Adam
Generally, I display in video mode for more contrast in the scene and better focus assist and I have tried various brightness adjustments as well as antireflective film (which in the end I discarded as it made things even more difficult to see).
I own a Zacuto eve which unfortunately is HDMI in so until I get some kind of converter I can't use it. At present, I am using the waveform on a TV Logic monitor to double check my exposure tests.

I have found that a Hoodman loupe with their HOOD Riser nearly covers the entire frame of the Black Magic display and has a diopter adjustment as well. I am presently working on a way to rig that to easily come on and off as it will be ideal for focus.

I will post photos in a new thread when it is complete.
Finnegan's Wake Films
Director/ Cameraman
Offline
User avatar

Tom

  • Posts: 1626
  • Joined: Wed Aug 22, 2012 5:08 am
  • Location: Manchester, UK

Re: Zebra issue on my BMCC

PostTue Sep 10, 2013 6:52 pm

Patrick Finnegan wrote:Thanks Tom for the results of the test. I will use 200 ASA when shooting raw for framing purposes as it will make it much easier, now confident that the sensor data will be exposed at 800 ASA.

I will go back and play with grading the (apparently) clipped images and see if I can full recover the highlights. I appreciate all of your feedback. This is an example of this forum at its best, when used to share knowledge and discuss possibilities rather than bemoaning disappointments!

As for shooting in Pro Res, I guess I will just have to get used to looking at a very bleached out screen if I want to maximize dynamic range.

As a side note, I own a Sony F3 and at a certain point Sony provided a firmware update which allowed users of the camera to dial an exposure index (EI) on the viewing image which did not affect the recorded image.





Well in ProRes mode, you just do the pulling down in camera before the data is compressed, so in the same lighting situation, you would also film in Iso 200.
Tom Majerski
http://tetragrade.com/
http://www.imdb.com/name/nm5157752/
Offline

Patrick Finnegan

  • Posts: 210
  • Joined: Wed Aug 29, 2012 11:43 pm
  • Location: Los Angeles, & Santa Fe, New Mexico

Re: Zebra issue on my BMCC

PostTue Sep 10, 2013 10:08 pm

Tom wrote:
Patrick Finnegan wrote:Thanks Tom for the results of the test. I will use 200 ASA when shooting raw for framing purposes as it will make it much easier, now confident that the sensor data will be exposed at 800 ASA.

I will go back and play with grading the (apparently) clipped images and see if I can full recover the highlights. I appreciate all of your feedback. This is an example of this forum at its best, when used to share knowledge and discuss possibilities rather than bemoaning disappointments!

As for shooting in Pro Res, I guess I will just have to get used to looking at a very bleached out screen if I want to maximize dynamic range.

As a side note, I own a Sony F3 and at a certain point Sony provided a firmware update which allowed users of the camera to dial an exposure index (EI) on the viewing image which did not affect the recorded image.





Well in ProRes mode, you just do the pulling down in camera before the data is compressed, so in the same lighting situation, you would also film in Iso 200.


I assume you mean in a situation where I have control of lighting and not a wide dynamic range I need to cover ( for example, a high contrast day exterior). In the high contrast situation, I could temporarily dial 200 iso to get a less washed out image for framing, but once I am ready to pull the trigger, my understanding is I would have to shoot at the base iso of 800 to maximize dynamic range...?
Finnegan's Wake Films
Director/ Cameraman
Offline

CaptainHook

Blackmagic Design

  • Posts: 1603
  • Joined: Wed Aug 22, 2012 4:50 am
  • Location: Melbourne, Australia

Re: Zebra issue on my BMCC

PostWed Sep 11, 2013 12:46 am

Tom wrote:As you can see in both examples, they result in an identical waveform and image.

They do appear similar at first glance. I can see differences in the midrange on the scopes (and if you feel like it, i wonder if applying a LUT/further processing would exaggerate those differences?) and to my eye a slight colour shift between them. Very subtle (this IS a log image) but would love to see if a grade applied (same to all 3) minimized the differences or exaggerated them.

To test if i were seeing things, i pulled your image into photoshop, duplicated the the top image and changed the blend mode to difference, then copied it over the bottom two. I added a curve to make the differences more obvious:

Image

Click to view large.

The trees etc outside i would put down to movement.. but i see what looks like colour, noise and detail differences. The scopes show a similar story if you really inspect the midrange areas (which again are compressed in this log image). I guess to rule out sampling variation we would need two frames at the same ASA and difference them to see if this would produce similar results. But it appears from my above image that ASA200 is more different to 800 than 400 is, which stands to reason. Not ideal to do this on JPG screenshots though, so exporting some TIFs from Resolve might show less/more difference.

I'm still not sure if this makes a difference practically either, but i think it warrants further investigation (even just to satisfy curiosity). By the way i also tried this the opposite direction with similar results, eg taking the 200ASA frame and putting over the other two in difference mode.

I would still also like the option to choose the ASA curves from the camera in Resolve (which i've requested), as they may be useful at times to pick from as a starting point.

Clipping 1600ASA seems to be the only big issue (very obvious colour shifts when i tested).
http://www.captainhook.co.nz/blackmagic-cinema-camera-lut/

**Any post by me prior to Aug 2014 was before i started working for Blackmagic**
Offline
User avatar

Tom

  • Posts: 1626
  • Joined: Wed Aug 22, 2012 5:08 am
  • Location: Manchester, UK

Re: Zebra issue on my BMCC

PostWed Sep 11, 2013 8:43 am

CaptainHook wrote:
Tom wrote:As you can see in both examples, they result in an identical waveform and image.

They do appear similar at first glance. I can see differences in the midrange on the scopes (and if you feel like it, i wonder if applying a LUT/further processing would exaggerate those differences?) and to my eye a slight colour shift between them. Very subtle (this IS a log image) but would love to see if a grade applied (same to all 3) minimized the differences or exaggerated them.

To test if i were seeing things, i pulled your image into photoshop, duplicated the the top image and changed the blend mode to difference, then copied it over the bottom two. I added a curve to make the differences more obvious:

Image

Click to view large.

The trees etc outside i would put down to movement.. but i see what looks like colour, noise and detail differences. The scopes show a similar story if you really inspect the midrange areas (which again are compressed in this log image). I guess to rule out sampling variation we would need two frames at the same ASA and difference them to see if this would produce similar results. But it appears from my above image that ASA200 is more different to 800 than 400 is, which stands to reason. Not ideal to do this on JPG screenshots though, so exporting some TIFs from Resolve might show less/more difference.

I'm still not sure if this makes a difference practically either, but i think it warrants further investigation (even just to satisfy curiosity). By the way i also tried this the opposite direction with similar results, eg taking the 200ASA frame and putting over the other two in difference mode.

I would still also like the option to choose the ASA curves from the camera in Resolve (which i've requested), as they may be useful at times to pick from as a starting point.

Clipping 1600ASA seems to be the only big issue (very obvious colour shifts when i tested).



Worth noting, the camera was not tripod mounted for these tests, but was being held down onto the arm of a chair. Slight movement could account for these differences - but I agree that It is worth further investigating to be sure.
Tom Majerski
http://tetragrade.com/
http://www.imdb.com/name/nm5157752/
Offline
User avatar

Tom

  • Posts: 1626
  • Joined: Wed Aug 22, 2012 5:08 am
  • Location: Manchester, UK

Re: Zebra issue on my BMCC

PostWed Sep 11, 2013 8:45 am

Patrick Finnegan wrote:I assume you mean in a situation where I have control of lighting and not a wide dynamic range I need to cover ( for example, a high contrast day exterior). In the high contrast situation, I could temporarily dial 200 iso to get a less washed out image for framing, but once I am ready to pull the trigger, my understanding is I would have to shoot at the base iso of 800 to maximize dynamic range...?


I mean in a situation where you are not clipping anything essential.

If you are lighting a scene and you have complete control over the lighting - lighting a scene for base 800 iso will let you maximise you dr. But if you do not have control, but nothing is clipping at 800, but the whole thing looks too bright on your screen - shoot at a lower iso.
Tom Majerski
http://tetragrade.com/
http://www.imdb.com/name/nm5157752/
Offline

SarahChana

  • Posts: 45
  • Joined: Sat Aug 24, 2013 4:19 pm

Re: Zebra issue on my BMCC

PostWed Sep 11, 2013 12:36 pm

Are you sure it's not the lens. Reminded me of what things looked like when I had a lense with a wide open aperture.
Offline

Dmitry Kitsov

  • Posts: 339
  • Joined: Fri Jan 25, 2013 8:47 pm
  • Location: Walnut Creek, Ca

Re: Zebra issue on my BMCC

PostWed Sep 11, 2013 2:28 pm

SarahChana wrote:Are you sure it's not the lens. Reminded me of what things looked like when I had a lense with a wide open aperture.

Sarah, it truly doesn't matter if it is the lens or the shutter speed or anything else. The ultimate problem is that the OP doesn't know if the footage was indeed overexposed or merely it's representation as interpreted by default in resolve seems to be overexposed.
Dmitry Kitsov
Offline

Patrick Finnegan

  • Posts: 210
  • Joined: Wed Aug 29, 2012 11:43 pm
  • Location: Los Angeles, & Santa Fe, New Mexico

Re: Zebra issue on my BMCC

PostWed Sep 11, 2013 7:01 pm

Tom wrote:
Patrick Finnegan wrote:
Tom wrote:Please share a dng showing the problem.

I dont follow exactly what the problem is.


i am sorry. I don't mean to be confusing. Everyone's input for shooting in raw and dialing back the iso more or less solves the issue for raw...

but the problem remains for pro res

It is simple, when using zebras at 100%:-

1) the screen is bleached so white I can only visually detect the zebras by using a loupe in a darkened room,

2) I can barely see anything of the subject I am shooting because the screen is bleached so white

3) using the zebras at anything above 90% as exposure guides results in clipping


Tom, I can supply a dng but the raw is really not the issue as I can use 200 asa to see my scene.
The screen shot I showed in my last post shows a pro res test image I shot at 95% Zebras (which should be safe exposure) clipping in the red and blue channels.

So, in essence, the only exposure guide I have with the camera (zebras) is giving me false readings and in order to even use them I can barely make out my subject in the bleached image on the screen.

i am trying to find out if this is normal as I don't have another BMCC to make a comparison with

i of course can compensate by using only 90% zebras but find it odd that everyone else seems to be able to use 100% zebras on their BMCC's just fine while mine clips.

thanks for all of your valuable input concerning shooting raw


The waveform might show clipping, but try recovering the data - I am fairly sure you will find that there is detail still there.


After Tom's feedback about this issue I went back into Resolve to reexamine my scopes. You are right, Tom, I was able to pull down the highlights and I believe recover most, if not all, of the information there. I remember now that for my Sony F3 when shooting Log that the camera actually maps up to 109 ire and I was looking at scopes mapped for Rec 709 which maxes out at 100 ire. As I am new to Resolve I didn't realize that with the three way color corrector there is a little bar in the upper right hand corner where you can select Log which I think allows you to set High and Low range values to map your white and blacks against as well as an offset wheel. I need to research more into this to fully understand all of the functionality this adds for grading log (or Black Magic Film).

Once I took my raw files and reexamined them against the camera raw function in Resolve mapped to Black Magic Film the "clipped" highlights immediately dropped into normal levels and left me plenty of room in the grade.

Of course, I will still have the issue of looking into a bleached out monitor when shooting pro res, but at least I can feel more confident that my highlights are being protected. After numerous tests I have resolved to shoot at 95% zebras with the zebras just appearing in the hottest parts of the scene.

I am assuming that if for any reason I am shooting in video mode on the BMCC , for instance if I am shooting to supply directly to edit for a client who has no budget for color correction, I will have to be much more careful about making sure my highlights don't exceed 100 ire.

Perhaps in that case I will drop the zebras setting to 85-90%.
Finnegan's Wake Films
Director/ Cameraman

Return to Cinematography

Who is online

Users browsing this forum: No registered users and 238 guests