jagged edges within certain frames question?
 

I just finished filming and editing a commercial for a local radio station competition and while reviewing the final product I noticed that in some of the frames it looks as if the zebra button is on the actual image (that is the only way I can explain it)

It seems as if certain objects have a fuzz to them and other sharp objects (a printer) has squiggly lines or something of the sorts.

I feel these are due to the lighting scenarios I was shooting in but am not sure so I was hoping I could get some feedback.

FYI I am writing on the XDCAM EX forum because I am using the EX1R so maybe some users have been in similar situations and could provide me a solution or alternative.

Thanks in advance.

Can you post a sample or screenshot?

That would prob. be the best thing to do let me see if I can do that.

Perrone Ford, So I went back and looked at all the clips through the camera and an HDMI cable plugged into my TV and saw none of the things I spoke of in the final video. I am using FCP 6 and am wondering now what the problem may be. I use log and transfer to import.

Any suggestions?

Well, without knowing at what steps the problem was introduced, it's hard to say. Maybe you log and transfer settings were bad. Maybe your final export settings were wrong. Maybe it was a one-time glitch. Who can say...

Well from the footage played back by the camera onto the TV I can tell it wasnt that. What settings should I have in FCP? for log and transfer I have it configured to XDCAM EX 1080p 24 frames per second.

In order to have the 1080p would I need an external blue-ray burner?

Im confused so I apologize for the all the questions.

Well, I am not a Mac user so it's hard for me to give specifics. But if doing log and transfer I'd certainly be using ProRes and not XDCam anything.

XDCAM EX is fine for your codec in FCP but you need to set it to render in Pro Res.

You have (like many people on the forum) provided far too little information for anyone to tell you what happened. If you are serious about getting help here you need to list every step of your process in great detail.

How are you viewing the finished video?

I shot everything in 1080p24 then went on to log and transfer it into fcp6 straight from the express card slot on the mac after editing it was saved to .mov and because it was quicker and we just wanted to see the finished product on tv we used Idvd to burn it onto a dvd.

In all do honesty Im not quite sure what other information to provide and yes I am serious about getting help.

It is probably the conversion from HD to SD that's causing the problem. If you start off with too much resolution and downconvert to SD you will get aliasing. You need to soften the HD before downconverting to SD. Take a look at this article: XDCAM-USER.com Getting SD from Hd and the problems of oversampling

Ok great. But if I wanted to keep it 1080p24 how would I export it or what should I do in FCP in order to keep its original resolution?

Keep in mind I have only recently started working with FCP and XDCAM EX so its a grand learning process for me. I do appreciate all the input and help I have received from DVINFO users thus far.

Alister, I respect you and your work more than you know. But I simply can't buy this. I've seen films scanned at 8k end up on DVD and look just fine. I've put RED 4K footage onto DVD and it's been just fine. I don't think it's just down to resolution.

Sure it isn't all about resolution, but it's the bulk of the problem. The frame grabs in my article illustrate it very well. It's all down to Nyquist Theory. Take a transition from light to dark that in HD uses four pixels and drop that down to just 2 pixels in SD and instead of a soft edge you have an instant hard light to dark switch, which doesn't look good or natural. Simply softening the HD image with a 2 two to 3 pixel blur can really improve the quality of the SD downconvert.

Your end result when down converting from HD to SD will depend on whether the down conversion algorithm applies any interpolation or softening. A high quality converter should do a good job, but a basic frame size conversion may just dump pixels resulting in a nasty image. It's one of the reasons DSLR's suffer from so much aliasing when used for video: Too much resolution. It's why quality video cameras have optical low pass filters that reduce the image resolution to around half that of the sensor resolution. It's counter intuitive I know but you really can have too much resolution.

Ok, I see what you're getting at. It's not that you necessarily have too much resolution, it's that poor handling of it as you move to lower resolutiuon leads to issue. Yes, we see this all the time. And it's why I do my downconversions with Lanczos or bicubic spline algorithms. They seem to produce excellent quality.

It's unfortunate these aren't available in modern NLEs, as it would solve a lot of issues.

-P