Problems with image shooting f-3 s-log to Pix 240

JASYN

New member
We shot a music video using the F-3 at 720 50p recorded to a Pix 240 external recorder.

Now in post we are watching the footage back and are noticing that the close ups have (where objects move fast passed the lens) strange "gamma banding" around the edges. It almost looks like a loss of information. Even when conformed to 25p it looks the same.

What would cause the footage to degrade like this? The strange thing is that we recorded to SXS cards as a back up and that footage looks totally fine.

I have attached an image to show what I mean. Please help! Screen Shot 2012-08-27 at 12.03.40 PM.jpg
 
Is this supposed to be slowmo?
Did you raised the shutter speed to at least 1/100?
Did it record 720 or 1080? (With the KiPro Mini you have to record out of the SDI Out to get 720. If you record out of the Dual Link A you get 1080 even if the camera is set to 720).
 
Yes, the image I posted is a screen grab from a conformed clip from 50p-25p. We had the shutter set to off. Our main concern is that there is something strange happening, especially on the close-ups, whenever something is moving quickly past frame. It isn't motion blur that concerns us it is a loss of image quality (very difficult to describe).
 
Yes, the image I posted is a screen grab from a conformed clip from 50p-25p. We had the shutter set to off. Our main concern is that there is something strange happening, especially on the close-ups, whenever something is moving quickly past frame. It isn't motion blur that concerns us it is a loss of image quality (very difficult to describe).
 
Yes, the image I posted is a screen grab from a conformed clip from 50p-25p. We had the shutter set to off. Our main concern is that there is something strange happening, especially on the close-ups, whenever something is moving quickly past frame. It isn't motion blur that concerns us it is a loss of image quality (very difficult to describe).
I think that the shutter off is part of your problem. For slow motion you need at least to double the shutter speed. (50p=1/100) I even speed it up even more to 60p = 1/250.
 
I agree it's conventional to use a fast shutter on footage like this, but if that was the problem, why would the footage be looking "totally fine" off the SxS cards?
 
Same frame has to look the same. When I first started testing the camera and I had the same problem until I realized that even tho I had set my shutter to 180 degrees my shutter switch was not set correctly.
 
You could be seeing the difference between XDCAM and ProRes at handling this type of video. In my opinion, XDCAM is almost always the superior codec and at fraction of the file size of ProRes.
 
You care to tell us how you came up to that conclusion?
Then 8 bit 4:2:0 XDCAM is better than 10 bit 4:2:2 ProRes HQ.
If you are right anyone that has an external recorder got ripped off by Sony first and all the other experts and recorder manufacturers second.
That is not my experience on a theater, specially when pushing the limits on the edit suite.
 
Did I say anything about 8 bit 4:2:0 XDCAM vs. 10 bit 4:2:2 ProRes HQ? No I did not.
Obviously those ProRes specifications are superior to XDCAM on paper, but I'm talking about the algorithms used to calculate the compression. It all comes down to math and processing power, and I've seen XDCAM look better than ProRes on many occassions when motion was involved. I have no way of knowing if that is the case here or not, but it is worth mentioning since I have seen it with my own eyes more than once. Not all image compression schemes are equal, and the differences between different codecs goes beyond the technical specifications.

BTW, Sony does a damn fine job with image processing and that's why you'll almost never see a visible difference between 8 bit 4:2:0 XDCAM and 10 bit 4:2:2 ProRes HQ unless you get into heavy grading.
 
So, in answer to the original problem, we're either saying that Jasyn was mistaken when he said the SxS card backup was noticeably better, (because it has to be the same) or, the XDCAM EX codec at 35Mb/s is a lot better at handling movement than ProRes HQ at 220Mb/s ? I must be honest and say I'm having trouble with either of these solutions without seeing the footage.

Can we confirm Jasyn, that the PIX was connected to the SDI output on the F3 and not the Dual connectors A or B? I presume you set the video format of the F3 to - PAL Area - 1280 x 720 Progressive? Was File Resolution/Rate on the PIX set to "Same As Video Input" or did you dial in the 720p50 option manually? I'm assuming that the F3 will only output a PsF signal but note that the PIX manual indicates it records a true progressive rate at this resolution. Could there possibly be a mismatch or hiccup with the conversion from PsF to P ???

Because the PIX is so flexible and up/down converts both rates and resolutions very well, it's quite easy to have the camera running at a different setting to the PIX and record both pictures successfully, but slightly differently, if you see what I mean...

Just trying to think of alternatives to Jasyn being mistaken in what he's telling us!
 
Last edited:
well here is a long shot , but one possible cause of image degradation on the PIX:

If the SDI was taken from the Monitor out instead of the A output in the back ,
- and the menus were setup so that the viewfinder info was supposed to come out over the monitor SDI out ,
- and the operator got rid of the menus by pressing the button on the outside of the camera that gets rid of the menus

Then you'll have a weird pattern that looks like a fine screen on the picture that may have other consequences as well.
Just a thought.
 
Hello all,

i have experienced the same issue. Didn't have time to check for solution yet, but that happened for me when i was recording 720 50p.

This is an issues between formats in PIX for sure. SXS material looks great and does not exhibit any similar issues...
 
Jensen is speaking the Sony-babble. Sony's 4-2-0 vs. ProRes HQ is not even in the same league. To state that you didn't specify 4-2-0 is silly since IT'S WHAT THE F3 RECORDS!! To say that XDCam looks better than higher data-rate ProRes is misleading at best. At a similar data-rate, it may be true in some circumstances, but the advantages/quality of ProRes are known and adopted by almost the entire industry. You are comparing apples and oranges.

Since the original poster attached an image that appears to be SLOG, I'm sure he intends to grade it, which is likely why they chose a 10-bit file. I grade material ALL the time with both, and I can assure you the difference is readily apparent, not just "when doing heavy grading".
 
Back
Top