Problems with image shooting f-3 s-log to Pix 240
- Thread starter JASYN
- Start date
Doug Jensen
Carbonite Member
You don't have to agree with me, but I stand by my post. It is nearly impossible to detect any visual difference between ungraded XDCAM and ProRes. I have done countless tests in a half-dozen different kinds of shooting situations and I have shown split screen comparisons to literally hundreds of people and nobody has ever been able to tell me which is which or point out any differences. In fact, unless I put border on a moving vertical wipe you cannot even tell where the split is happening. On paper you'd think the differences would be huge, but in the real world, it's almost nothing. If you have evidence that shows something else I would certainly love to see the raw clips. Until then, I stand by my opinion. It doesn't matter to me what the "entire industry" thinks because I have actually taken the time to do my own testing. How many people have done that?
Lenilenapi
Carbonite Member
I went back and looked at prior posts. Yes they did record out of SDI out instead of dual link A. I'll bet they had the show-menus setting on in the internal menu and switched them off on the outside camera button. That would screw up picture quality badly and is an easy mistake to make if you don't know the camera well.
RTe What Doug is saying: He is referring to ungraded footage, not how well it holds up under grading and I think he's largely right, XDCam looks fine any differences are minor . Maybe on extreme blowup you could see it but i never have.
RTe What Doug is saying: He is referring to ungraded footage, not how well it holds up under grading and I think he's largely right, XDCam looks fine any differences are minor . Maybe on extreme blowup you could see it but i never have.
What kind of light were you using? Looks like hmi flicker.
Could you persuade the BBC to accept your findings? (or ITV, Sky, Channel 4 etc)... because at the moment they won't accept XDCam 35Mb/s for mainstream HD production. It's a pain but unfortunately, the "entire industry" in the UK sets the standards we have to meet in order to work. Personally, I can't spot the difference either, but that's irrelevant to my clients!
Doug Jensen
Carbonite Member
Paul, I completely understand the need to comply with client specifications even if we don't agree with them. Some clients can be educated, and I do that all the time, but obviously there's no chance to even have that conversation with a big broadcaster. If they say jump, you've gotta jump.
LeavingTheCandy
Well-known member
Maybe your perception or deliverable is different than mine, but in "real world" testing of my own--shooting both formats all the time (I have hundreds of hours of footage shooting to XDCam and ProRes simultaneously) I can think of many instances where XDCam was inferior, particularly when shooting SLOG. Depends upon your workflow and deliverable, but for mine, ProRes gives me a better result and a lot more latitude in grading.
Are you saying 4-2-2 or 4-4-4- are not necessary and it's overkill for everyone?
Viddovation
Veteran
I don't think Doug is saying that at all. His primary point has been that, in his testing, XDCAM-EX handles motion better than ProRes does. He goes on to claim that unless you do heavy grading, the two codecs are usually indistinguishable in terms of color.
We can argue about how much grading is heavy, but I personally would agree with Doug that ungraded XDCAM appears visually identical to ProRes, in my experience.
Obviously if you want to grade heavily and leverage the full dynamic range of the camera, XDCAM is not the best choice. But I know one broadcast News station that has PMW-500 cameras that can shoot both to XDCAM-EX (35Mbps) and XDCAM 422HD (50Mbps), and they've chosen to archive in 35Mbps, because they can't see the difference, and it it makes archiving cheaper. News organizations tend to do very minimal (if any) color correction, so this doesn't surprise me.
Doug Jensen
Carbonite Member
Did I say anything about S-LOG? Now you're mixing apples and oranges. Heavy grading is absolutely required for S-LOG and I would not recommend anyone shoot S-LOG with anything less than a 10-bit 4:2:2 external recorder. Obviously XDCAM would be inferior in that type workflow because it's only 8-bit. But that's not what I'm talking about.
Is 4:2:2 or 4:4:4 necessary? Only if you are going to grade. Otherwise you will see no visible advantage of ProRes over XDCAM, and you might very well see some motion artifacts with ProRes as illustrated by the OP of this thread. I have seen it myself more than once. Codecs are about more than just the raw numbers.
When I shoot S-LOG I use my Samurai or rent an Cinedeck EX. I don't even bother to record on the SxS cards even as a backup. But if the footage I'm shooting is not going to undergo grading, then the SxS cards are just fine.
Last edited: