Jump to content


Photo

I'm trying to figure out the cause of this CCD imager issue.


  • Please log in to reply
41 replies to this topic

#21 Mike Krumlauf

Mike Krumlauf
  • Basic Members
  • PipPip
  • 63 posts
  • Cinematographer
  • OG from Chicago, Currently In Denver

Posted 16 September 2018 - 12:06 AM

Here is what the waveform looks like on the color bars on my camera. If anything looks strange, please let me know.

 

 

Attached Images

  • Screen Shot 2018-09-15 at 11.02.50 PM.png

  • 0

#22 Mike Krumlauf

Mike Krumlauf
  • Basic Members
  • PipPip
  • 63 posts
  • Cinematographer
  • OG from Chicago, Currently In Denver

Posted 16 September 2018 - 12:09 AM

Here is the vectorscope on the color bars as well.

Attached Images

  • Screen Shot 2018-09-15 at 11.03.10 PM.png

  • 0

#23 Bruce Greene

Bruce Greene
  • Basic Members
  • PipPipPipPip
  • 638 posts
  • Cinematographer
  • Los Angeles

Posted 16 September 2018 - 03:12 AM

There seem to be a lot of dots for a test color bars signal.  Can you post an image of the bars themselves?

 

It's also possible that it's a cable issue.  Or it's your Ki-Pro.  Try another recorder, or record directly to a capture card in your computer.


  • 0

#24 Frank Wylie

Frank Wylie
  • Sustaining Members
  • 117 posts
  • Other
  • Culpeper, VA

Posted 16 September 2018 - 02:15 PM

Yeah, that doesn't look kosher to me.  Unless there are custom markers the camera maker placed in the test signal ( and to be fair, there could be), I would say those dots indicate a noise problem.


  • 0

#25 Mike Krumlauf

Mike Krumlauf
  • Basic Members
  • PipPip
  • 63 posts
  • Cinematographer
  • OG from Chicago, Currently In Denver

Posted 16 September 2018 - 09:54 PM

What steps can I take to calibrate the signal noise without damaging the ccd or a board in the camera?
  • 0

#26 Bruce Greene

Bruce Greene
  • Basic Members
  • PipPipPipPip
  • 638 posts
  • Cinematographer
  • Los Angeles

Posted 17 September 2018 - 04:02 AM

What steps can I take to calibrate the signal noise without damaging the ccd or a board in the camera?

I would start by changing SDI cables connecting the camera to your recorder.


  • 1

#27 Frank Wylie

Frank Wylie
  • Sustaining Members
  • 117 posts
  • Other
  • Culpeper, VA

Posted 17 September 2018 - 06:16 AM

I would start by changing SDI cables connecting the camera to your recorder.

 

I agree.  Start simple, move to complex and only change one thing at a time and evaluate.


  • 0

#28 Mike Krumlauf

Mike Krumlauf
  • Basic Members
  • PipPip
  • 63 posts
  • Cinematographer
  • OG from Chicago, Currently In Denver

Posted 17 September 2018 - 02:15 PM

Well ive tried 8 different BNC cables and still same results so Im leaning on that its an internal problem. Maybe the signal to noise ratio is adjusted (is that possible?). What internal steps can I carefully take to reduce noise in the video signal?


  • 0

#29 Frank Wylie

Frank Wylie
  • Sustaining Members
  • 117 posts
  • Other
  • Culpeper, VA

Posted 17 September 2018 - 03:45 PM

Eliminate the converter as a potential problem before you go cranking around on the camera! 


  • 0

#30 Mike Krumlauf

Mike Krumlauf
  • Basic Members
  • PipPip
  • 63 posts
  • Cinematographer
  • OG from Chicago, Currently In Denver

Posted 17 September 2018 - 11:56 PM

Well i've found that the converter is not the problem. Even with the miranda box taken off the camera and monitoring using the component bncs on the side of the body, the noise still looks a bit high. Is that port that the miranda box gets its information from just a rewrapping of what the component outs on the side are getting or is it a totally different channel? I tried black balencing the camera a few times as well to see if that could get rid some of the noise to no avail.


  • 0

#31 Bruce Greene

Bruce Greene
  • Basic Members
  • PipPipPipPip
  • 638 posts
  • Cinematographer
  • Los Angeles

Posted 18 September 2018 - 12:00 PM

Well i've found that the converter is not the problem. Even with the miranda box taken off the camera and monitoring using the component bncs on the side of the body, the noise still looks a bit high. Is that port that the miranda box gets its information from just a rewrapping of what the component outs on the side are getting or is it a totally different channel? I tried black balencing the camera a few times as well to see if that could get rid some of the noise to no avail.

Well next step is to eliminate the Ki-pro mini...


  • 0

#32 Mike Krumlauf

Mike Krumlauf
  • Basic Members
  • PipPip
  • 63 posts
  • Cinematographer
  • OG from Chicago, Currently In Denver

Posted 18 September 2018 - 01:05 PM

I've eliminated everything but the camera! The camera is just plugged via component on the side into a 55in monitor and nothing else but a lens and the noise still seems a tad high in lower values and against lights or edges of heavy contrast, those horizontal or vertical lines appear from T2.3 (which is wide open on my lens) and disappear at T4. Now what can I look at in the camera? If its just the gains being too high, how do i bring them down. If its CCD Clock noise, how do you calibrate the clocks/timers correctly?


  • 0

#33 Frank Wylie

Frank Wylie
  • Sustaining Members
  • 117 posts
  • Other
  • Culpeper, VA

Posted 21 September 2018 - 03:15 PM

That I couldn't tell you.  I still suggest you try to have a broadcast engineer look at the signal and give you advice.  HD is the norm in broadcasting now...


  • 0

#34 Mike Krumlauf

Mike Krumlauf
  • Basic Members
  • PipPip
  • 63 posts
  • Cinematographer
  • OG from Chicago, Currently In Denver

Posted 25 September 2018 - 08:02 PM

I wanted to add that I recently sat down and watched both Youth without Youth and Collateral. Both of these films were shot on F900/3's (collateral being the HD900F) and both of the films had that line issue i am having! I could not believe it! Multi million dollar films and the lines are there in scenes clear as day. Both of these movies went out HD SDI to an HDCAM-SR Deck. Yet, films like once upon a time in mexico, star wars ep2, and spy kids 2 have the issue and those films shot on HDCAM tape inside the camera's deck. I am really beginning to believe that the HD-SDI signal feed either from a miranda box like i have or sony's own HDCA 901 is not meant to be used as a recording feed, and more a monitor out only.. seeing these image artifacts in major productions made me realize maybe this is just the way the camera was made and the F900R which has HD SDI fixed this issue? Doesn't help that 3 different F900s had the issue as well.. Sticking to HDCAM tape and digitizing that to prores sounds like the fail safe way to go. :)

 

And just to clear something up.. when the problem is showing on films shot with Zeiss digiprimes and Panavision digital primos, i have to rule out that its a lens issue. I know my Optex lens is nowhere near the quality of glass as either of those lenses but its still "HD Spec". The reason I mention this is because i thought maybe it was a strange artifact from the lens not resolving lines of information or something from the CCD block.


Edited by Mike Krumlauf, 25 September 2018 - 08:06 PM.

  • 0

#35 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 20317 posts
  • Cinematographer
  • Los Angeles

Posted 25 September 2018 - 08:33 PM

Most of "Collateral" was shot on a Thomson Viper camera but in Rec.709, not log.


  • 0

#36 Mike Krumlauf

Mike Krumlauf
  • Basic Members
  • PipPip
  • 63 posts
  • Cinematographer
  • OG from Chicago, Currently In Denver

Posted 25 September 2018 - 08:40 PM

Most of "Collateral" was shot on a Thomson Viper camera but in Rec.709, not log.

I know that it was a mix of both but judging off the american cinematographer article, it seemed like they used the panavised F900s a bit more than the vipers. Anyhow, two scenes which I know used the F900s (the scene when the body hits the taxi cab and they put it in the trunk and when tom cruise is chasing jada's character in the law building) both showed to some extent the same issue i was having with the lines by bright lights.

 

I should add i just looked at some Varicam F & H videos i shot out of the HD SDI to the kipro and to my surprise, that line artifact issue was present in night shots around the street lights. It has to be just how cameras of that time were made. How can it be constantly popping up in every single 3CCD ENG camera i've used over the past 4 years and now im finding it in high budget films.

 

David, i know you have quite extensive use of the Panavision HD900F that I have (trying to find a DVD copy of the quiet btw ;) ), have you ever seen anything like this/used the old varicam or f900 via HD SDI and not tape because like i said, these problems are non existent when you use the camera's tape decks and digitize later.. its only when recording the direct live feed out of the HD SDI.


  • 0

#37 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 20317 posts
  • Cinematographer
  • Los Angeles

Posted 25 September 2018 - 09:11 PM

I'm not as much of a pixel peeper as you are, I'm sure there is an underlying structure to the image if looked at closely.  All my F900 features were recorded to HDCAM tape for budget reasons, so I never tested recording to external systems.  As you know, the problem with HDCAM is that it is a compressed 8-bit 3:1:1 recording, 1440 x 1080.  

 

At the time, the F900 was pretty much at the top of what was available for most people so there wasn't much point in looking for artifacts at the pixel level since there wasn't something better that I could afford to use instead.  The only thing I noticed is that the Rec.709 color matrix did add some edge artifacts that affected chroma key work, so many of my F900 features were shot with the color matrix turned off, which was Panavision's recommended way of using the camera.  The matrix also caused some odd color-clipping artifacts in saturated lighting.

 

My F900 movies sort of pre-date HD home video and digital streaming, so most were transferred to film for release. And most of them are only now available on DVD (there was one horror movie called "Infested" on blu-ray but the HD post was very cheap on that and whatever master they used was terrible, I have no idea why it looks so bad.) "Jackpot", "When Do We Eat?", "New Suit" all look fine on DVD, but that's standard def. Maybe some are now streaming in HD.  My first digital TV series, "United States of Tara", was shot on the Genesis but only released on DVD as well (though broadcast in HD by Showtime).

 

"Akeelah and the Bee" was shot in 35mm anamorphic and was telecine transferred to HD from a 35mm IP, which included some shots done on an F900 and cropped to 2.40, but that footage went through a 2K D.I. and went out to 35mm anamorphic IN, cut into the o-neg, and an IP struck for HD transfer, so it is somewhat removed from the original HD recording.


  • 0

#38 Mike Krumlauf

Mike Krumlauf
  • Basic Members
  • PipPip
  • 63 posts
  • Cinematographer
  • OG from Chicago, Currently In Denver

Posted 25 September 2018 - 11:11 PM

So the quiet dvd is not from a digital master? its from a 35mm print scanned back to digital? I remember all of InDigEnt's DV films ending up on dvd back in the day like that and i always wondered "why"?


  • 0

#39 David Mullen ASC

David Mullen ASC
  • Sustaining Members
  • 20317 posts
  • Cinematographer
  • Los Angeles

Posted 26 September 2018 - 02:31 AM

“The Quiet” is probably from an HD tape master, not a film-out transferred back.
  • 0

#40 Daniel D. Teoli Jr.

Daniel D. Teoli Jr.
  • Basic Members
  • PipPipPip
  • 204 posts
  • Other
  • Born in L.A....NYC is 2nd home...Rustbelt is home base.

Posted 26 September 2018 - 09:55 AM

Do the movies look good or not? If you blow up anyhting big enuf and pixel peep it looks bad.


  • 0


Tai Audio

Ritter Battery

Glidecam

Willys Widgets

CineLab

rebotnix Technologies

New Pro Video - New and Used Equipment

Gamma Ray Digital Inc

Technodolly

FJS International, LLC

Wooden Camera

Visual Products

Media Blackout - Custom Cables and AKS

Rig Wheels Passport

Broadcast Solutions Inc

Metropolis Post

Aerial Filmworks

The Slider

Abel Cine

CineTape

Paralinx LLC

Media Blackout - Custom Cables and AKS

rebotnix Technologies

The Slider

Paralinx LLC

FJS International, LLC

Abel Cine

Aerial Filmworks

Rig Wheels Passport

Ritter Battery

New Pro Video - New and Used Equipment

Wooden Camera

Technodolly

Metropolis Post

CineTape

CineLab

Visual Products

Broadcast Solutions Inc

Gamma Ray Digital Inc

Willys Widgets

Glidecam

Tai Audio