Need advanced broadcast safe levels help

[Nate Weaver]

I've been doing onlines at home with rented decks (DVCPRO HD, Digibeta) for a while now, and have had little trouble until a 720p24 master I delivered came back with QC problems (video levels). I have my first HDCAM layoff coming up, and I have questions. Here's what I know:

1-Of course I know IRE should not pass 100. The place doing the QCing takes a hardcore stance on this...not over, not for even a frame (if they catch it).

2-Chroma can be up to 130.

I already know the scopes in FCP are dubious at best, because they don't show you an end result...they show you things BEFORE it's moved to a deck, and that can be a problem.

One of my big questions has to do with the saturation overlay on the FCP scopes. If "saturation" is off, my luma levels are limited well via BS filter to 93-95ish. If I add the "saturation" scope option, my levels go way higher than 100.

The BS filter seems to squash luma great. If I add the saturation option, it even squashes the sat ALMOST to 100 with the scop in sat mode. But the saturation still has little bitty peaks higher than 100, and only in places. Does this count as over 100? Or do these saturation peaks count in the chroma dept, where things are safe to 130 (or 120 or whatever)?

{p.s. I used the BS filter in "Extremely Conservative" mode, and still got a levels reject. That's what started all this}

[William Hohauser]

What I am curious about is the link between the computer and the recording deck. Perhaps there is something happening there that is pushing the signal over the QC limits. The FireWire port shouldn't be the problem but if you have an SDI card or something similar, it may be out of alignment.

Rent an stand-alone waveform/vectorscope combo for a weekend, it'll be worth it. Make sure it has the same ports as your video card (the AJA or Blackmagic). Connect the card directly to the waveform and compare FCP's assessment of the signal to what the card is putting out. To check the FireWire output you'll have to connect the waveform to a deck, not the best but it'll do.

Of course you might be running into a situation where your tape is being rejected for unreasonable reasons, that's a harder problem to solve. I was once the studio manager/engineer for a facility with a direct feed to the local cable company. One day the engineering staff at the cable company decided that all the facilities with direct feeds had to supply a continuous steam of 100% video or our feeds would be disconnected! I asked if they really meant a calibrated 1 volt sync signal wih 7.5 black since 100% video is pure white. No, they demanded 100% video all the time. I was prepared to argue the ludicrousness of the request when the owner of the facility said, "Just agree with them". I did, that made the cable company happy and we went on to feed the same signal as before. Who knows what sparked the request.

Unfortunately if your card is defective, you might be forced to go to a dub house to get an adjusted dub of your master.

[Nate Weaver]

First, thanks for responding. This isn't an issue of defective hardware, it just my defective knowledge.

I guess here's my question:

My video has lots of highly saturated stage lights going all the way to camera clip (unavoidable, but it happens in a very eye-pleasing way).

If I add and tweak the broadcast safe filter, I can get luma down to whatever I want. But if I view the "saturation" option in FCP's scopes, everything goes way over 100. I think what it's showing me is what the composite signal would look like. So anything that is over, say, 88 IRE that happens to be very saturated then pegs the waveform much higher. Here' some screenshots to maybe clarify my question:

[William Hohauser]

Did the facility specify the frame numbers that upset their delicate constitutions? What is it that they can't deal with themselves? I am amazed that a facility that is complaining about this doesn't have a method to cope with it.

The screen caps show me nothing really unusual. The vectorscope might show more. Is there any stylistic filtering going on here that might push the chroma over the edge?

[Glenn Chan]

1- It may simply be a Final Cut Pro bug.

http://forums.creativecow.net/cgi-bi...&postid=909453

When you render, the BS filter doesn't necessarily get applied.

The thread suggests one workaround (using the 3-way CC manually on each clip).
Another potential route would be to use a hardware video legalizer.
Or figure out how to get the broadcast safe filter to render.

2-
The resulting analog NTSC signal is a composite signal, which consists of the luma signal and the chroma signal combined. Anyways, this would be the terminology I would use.

In Final Cut Pro, the "saturation" setting would show you the resulting composite signal instead of the resulting luma signal. However, I don't think it's very useful since:
A- It assumes 0 IRE setup, which is wrong for NTSC (except Japan).
B- You can see much above "110". The unit is presumably IRE, so you can't see above 110 IRE.

When the broadcaster talks about chroma, they probably really mean the composite signal instead.

3- The scopes in Final Cut Pro aren't always accurate. It changed in one of the latest releases of FCP. I haven't followed this closely, but hopefully it's accurate now.

I haven't checked to see if it has the superblack clipping problem. [ To check for THAT, put the color bars generator in the timeline. Add the 3-way CC. Adjust the black luma level slider slowly. The waveform shouldn't clip the reading at "0" in FCP units.]

[Nate Weaver]

Originally Posted by Glenn Chan

1- It may simply be a Final Cut Pro bug.

http://forums.creativecow.net/cgi-bi...&postid=909453

When you render, the BS filter doesn't necessarily get applied.

The thread suggests one workaround (using the 3-way CC manually on each clip).
Another potential route would be to use a hardware video legalizer.
Or figure out how to get the broadcast safe filter to render.

Yeah, that one is going around. I'm definitely not having that problem, I can see the results of the BS filter in my renders.

Originally Posted by Glenn Chan

2-
The resulting analog NTSC signal is a composite signal, which consists of the luma signal and the chroma signal combined. Anyways, this would be the terminology I would use.

In Final Cut Pro, the "saturation" setting would show you the resulting composite signal instead of the resulting luma signal. However, I don't think it's very useful since:
A- It assumes 0 IRE setup, which is wrong for NTSC (except Japan).
B- You can see much above "110". The unit is presumably IRE, so you can't see above 110 IRE.

When the broadcaster talks about chroma, they probably really mean the composite signal instead.

Finally, an answer to my question. "Saturation" in this case means luma and chroma overlaid.

So my very very important question that comes after that, is is this composite signal the one that can't go over 100? Or is it just the luma alone that can't go over 100? My uncorrected signal has high saturation values in the high IREs, but luma alone is 80-90ish.

Originally Posted by Glenn Chan

3- The scopes in Final Cut Pro aren't always accurate. It changed in one of the latest releases of FCP. I haven't followed this closely, but hopefully it's accurate now.

If you're referring to the scopes only sampling 16 lines, that was JUST fixed in the latest update.