I've a broadcast monitor with a 75% or 100% switch for component input, if I am giving it a feed from a Z7 should it be set to 75% or 100%? Would this differ from a feed from Final Cut through a Blackmagic Intensity Pro component cable?
Have you got access to a waveform monitor, either in hardware or software? If the RGB values go from 0% to 100%, it's 100% bars; if they go from 0% to 75% they're EBU 75%; if they go from 25% to 100% they're BBC 75% also known as 95%. If I had to guess, I'd plump for 100%.
Cheers Alan, sorry for my laymens description, but:
The left most long box is 75% grey and right most small box is 75% grey, captured in DVcam or HDV from Z7. The large white box on bottom row second from left is 100% white.
The PAL standard colour bars within Final Cut seem to have 100% white where the 75% boxes are in example above.
What should I feed my broadcast monitor for calibration?
Cheers
Jon
I don't know, you need to tell me the RGB values for the colour bars themselves, not the "white".
With the waveform:
Red is: 22%
Green is: 43%
Blue is: 8%
ie. hitting all the preset targets.
Is the phase dial supposed to work on PAL broadcast monitors?
Ah, I see. What you're measuring is the luma value for each colour bar, on a composite or YUV coded signal. What I meant was the RGB values, the values of the red green and blue signals themselves, when decoded from YUV.
If you're looking at SD PAL, then phase is not relevant, that's the whole point of PAL, phase errors don't matter (much). If you're looking at HD, neither phase nor PAL has any meaning.
Alan your a mine of info, thanks. I was just about to complain about a monitor I bourght on ebay, for the phase not working with PAL! :o
Not sure I can work out the RGB values I only have a waveform in Final Cut, thanks for your advice anyway.
Jon
OK, here's how to do it the easy way.
1 - Record you colour bars in the camera.
2 - Ingest the bars as a clip into you favourite NLE software (e.g. FCP).
3 - Find one frame of bars and export it as a TIF or BMP file (not JPEG).
4 - Import that frame into your favourite graphics software (e.g. Photo Shop).
5 - Run your cursor over the frame, and write down the RGB values you get for each of the colour bars.
With those numbers we can work it all out. I'll be happy to do that here if you get me the numbers, and you'll see how it all works, far easier than explaining.
From still in tiff format:
Grey: R 179, G 179, B 179
Yellow: R 179, G 179, B 0
Cyan: R 0 , G179, B 181
Green: R 0, G 178, B 0
Magenta: R 179, G 0, B 182
Red: R 179, G 0, B 1
Blue: R 0, G 0, B 179
I can imagine there shouldnt have been blue in the red channel
OK, allowing for a bit of noise, what we've got are signals having values of 0 and 180. But the most important thing to notice is that those are the only two values there. And that means that the bars are normal 75%, can't be anything else. But there's another thing to notice as well (and it's why FCP is no use to me as an engineer):
Standard video coding (whether 8- or 10-bit) always accords to either ITU Rec 601 or ITU Rec 709. In fact, they are identical as far as this point is concerned. Concentrating on the 8-bit version, the specs say that black level is at code value 16, white level is at code value 235 (values 0 and 255 are disallowed, as they're reserved for sync pulses). So the coding difference between white and black is 235-16=219 levels. Your example is showing the colour bars with black at 0 and white at 180, a coding range of 180 levels. FCP always does this code translation when importing or exporting stills, and it means that you can't actually get to the real video signals, there's no way. They do it this way so that you don't have to understand about 601/709 coding levels, and it makes some sense I suppose, but it would be much more helpful if you could turn that aspect off and see the real video signal levels. What FCP has done is to subtract 16 from each data value, and then multiply by 255/219, stretching the signals out so that white gets to 255. The values you've got don't quite fit the bill because 180/255=70.588%, not 75%, but it's close enough for jazz.
Anyway, rant over, you've got 75% bars. Easy, or what? 'Nuff said :)
I had a look at bars recorded on my Z7 via compact flash and when I looked on final cut pro most of the lines were at 75% with two of them at 100% The vectorscope round thingy had the RGB at 75%
Now being a sound bod anything that aint PPM meters is magic science to me!
Thanks for that explaination Alan, 75%, I'll flick that switch on the back of my monitor and calibrate to that. Cheers
The white bar is always 100%, irrespective of the type of bars. It's the RGB values for the colours that changes. There's a ITU spec for bars anyway, ITU-R. BT 471 (hands up all those who didn't think I knew that one :) ).
The exception to this rule is SMPTE bars, where the white bar is 75% as well, but there's a peak white (100%) block in the bottom part, along with the PLUGE + and -2% blocks.
Sorry Alan I thourght I understood it.!! but I don't. There are two types of colour bars in final cut
Top type (75%) is what was captured from Sony Z7. should I feed this to my monitor from final cut for a PAL dvd delivery? even though the 100% bars are called PAL in the program?
Called HD in Final Cut, with left bar at 75%
Called Pal in Final Cut, with left bar at 100%
The top set is proper SMPTE (the de facto HD standard for bars), all the RGB values for the bars across the top are 0 and 75%, the square on the bottom row is 100%. The lower set is non-standard.
Fresh, thanks for that Alan