There's no such thing as a black level control for tracks, but you could apply an effect like levels to an entire track to tweak your levels. Empty tracks are just that, empty. They should render as 0/0/0 black, which is too black for NTSC.

As bob says, there are settings for the internal waveform tool and if it's misconfigured then it could be telling you your blacks are too high.

If it's not that then the next thing to ask is how you're acquiring the media. The black levels could get cooked in if you're capturing from an analog source. PD150s, for example, are capable of adding setup to their analog output, something you really don't want to do.

What project template? What type of media? how are you acquiring it?

Rob Mack

When editing for NTSC I always add a solid color on the bottom track set to 16,16,16. This will give you NTSC black for your background. Also make sure your scope is set to Sudio RGB for it to read properly.

Marc

As Marc pointed out, sometimes you may need to add a solid color generator on the bottom track to get that aspect right.

----
See these articles about getting setup and your levels right:
http://www.glennchan.info/articles/technical/setup/75IREsetup.html
http://www.glennchan.info/articles/vegas/v8color/v8color.htm

Key points:
- Vegas' scopes may not be accurate!! If you are making broadcast masters, get hardware scopes. Get ones that are engineered correctly and can read SDI output off your decks.
- Setup doesn't really exist in the digital domain. There is one standard black level (Y' = 16). Some cameras add fake digital setup, which is confusing and can get you into trouble (it's a case of trying to make two wrongs equal a right; useful in very obscure applications, but dangerous in most applications).

Glenn,

Are you saying that Vegas scopes are less accurate than other software programs like Premiere, Color Finesse, Avid etc.?

Thanks, Marc

1- Ideally, your scopes should measure off the end of your signal chain. If you didn't, the scopes may not be accurate since other things further down the signal chain might be wrong.
So in this sense, most software scopes should not be trusted! (Unless you are outputting to data/computer formats and not a VTR.)

7.5 IRE setup is one good example... some 'professional' equipment won't correctly convert from digital->analog. Software scopes can't figure it out because they aren't measuring the analog signal.

2- A lot of the scopes built into editing programs are/can be inaccurate.

Some editors have bugs where the broadcast safe filter won't actually be run when you perform a render (FCP), not every line is read for the scopes output, overscan is ignored (FCP), weird superblack clipping occurs (FCP), etc.
As far as Vegas goes, you have to get the levels conversions right, there is RGB gamut clipping on what the scopes show (incorrect for Y'CbCr output), lack of 7.5 marker is confusing/dangerous, the vectorscope trace for Rec. 709 is not adjusted accordingly, etc.

Haven't looked closely into Premiere, Color Finesse, Avid's scopes.

Given that most people here are working with DV or HDV I don't see how they can monitor at the end of their signal chain. Maybe there is a scope that'll accept firewire, closest I've seen though is a firewire option board for some of the Sony monitors. That said I can't see much chance for anything to happen between when the bits leave Vegas and hit the heads of the VCR.

In reality I can't see too many Vegas users having the funds for hardware scopes. With a bit of education the scopes in Vegas are adequate for the purpose and certainly better than nothing.

I'd also mention that a hardware scope can be just as misleading. Forget to teminate a composite feed and you can really get yourself in a knot. I've seen people swear by scopes that haven't been calibrated in decades and the same goes for A>D converters. At least for most here who stay in the digital domain that's a few issues that they don't have to worry about.

I'm not saying that Vegas scopes are perfect and indeed there's a few traps. But learning to avoid them is no harder than learning how to read a hardware scope.

Bob.

Hi Rob,
Thanks for the input. I'm shooting on a brand new XDCam EX1, and while I don't suspect a camera problem, my engineer at corporate HQ thinks a camera setting may be amiss.

We're shooting in 1080/60i, and it looks great in the camera.

If you're working with DV and HDV then hardware scopes probably don't make sense. But you should at the very least take a look at your final DVD and monitor that (being careful about whether or not your DVD player puts black level at 0 or 7.5 IRE, and what your monitor is calibrated to/for). Actually pop the DVD into a DVD player and check it, not just simply assume that what you see on firewire out is what you get (which can be a big mistake with 32bit).

While doing color correction you can't do that, but you should use the DV device as the external preview device and monitor the analog outputs off that. It's the closest thing to what your DVD will look like. (And then actually watch your DVD when it's done as a final pass of QC.) You'll be able to spot things that you wouldn't see on a computer monitor (e.g. field order). If your QC standards are not high then I suppose this is being overly picky.
*This may not work for 32-bit projects.

- In a broadcast situation, you can get DV decks with SDI out (and most post facilities doing broadcast work I've seen have the $15k+ decks with SDI out).

That said I can't see much chance for anything to happen between when the bits leave Vegas and hit the heads of the VCR.
If the deck has bad/old/worn heads, then the master will have occasional dropouts on it. You'd see this on the SDI output off the deck if confidence heads are turned on (I can't remember if the expensive DV decks have it). You can also see any sync errors as the picture would be totally munged. Heh, stuff can happen to those bits!
Granted, you don't need scopes to spot that but if you are making broadcast masters then you need scopes anyways. Or in the case of FCP... its scopes show that the broadcast safe filter is being rendered but you wouldn't know that FCP has/had a bug where it wasn't being rendered when going to tape.

That's good info. Glenn could speak to this much better than I can but Setup as far as I know is a phenomena of NTSC SD. A digital playback device can be configured to add setup to it's analog outputs to bring blacks a 0DB up to 7.5DB. Normally you don't want to do this. Just leave it off.

You're in the HD realm so I'd suspect something a little different, like remapping between different color spaces or something.

The only analogous situation I experience is with DV we shoot here at work. If I capture DV straight off the tapes through my DSR11 over firewire then the bars at the head of the tape fall exactly where they should in Vegas. If, however, I take the same footage that was captured from the house DV deck, over SDI, and into Matrox Axio hardware and finally into Premiere, and then open the clip in Vegas, the blacks are darker and land down at zero DB. They're fine in Premiere, but wrong in Vegas. (Using the house chain gets us footage in a matrox codec, not DV-avi)

So, it could be that the opposite is happening going from Vegas to *what-ever*. Maybe. Or it could be some funky colorspace conversion happening in the chain...something like that. I'd maybe start by just recording bars and walking it through the chain to see where things are going wrong.

Rob

For years I've been saying Vegas is not handling levels properly. But, to date, it's fallen on deaf ears.

By all means, step in here and solve Dave's problem. :-)

Rob

The scopes are wrong, the preview window is unreliable. The codec you chose to render to may or may not change the levels on its own without warning you, you won't know until you render it out. If you use other third party software, AE, TMPGENc, etc, they will all handle luma in different ways depending on whether its RGB or YPbPr. So what's the right answer? I think it's called trial and error with the levels FX. Make a guess, render a test clip, make sure you repeat the exact workflow when you finally get a test clip to come out right.

You didn't really think it would be easy, did you?

There's a few issues that bring things undone.
All the D>A converters will convert luma values from 0 to 255 to superblack to superwhite. Some analog VCRs do use superblack, we now have a couple of SP decks that had a pretty standard mod done to them for this purpose.
On the other hand DVD players seem to have a mind of their own, depending on whose brain they're using and how they talk to the monitor.

I have an el cheapo one connected composite to a Sony monitor. I also have a Sony DVD player connected component RGB via SCART to a 16:9 TV. The same DVD looks quite different on both setups.The el cheapo seems to legalise the output on the composite output. The Sony sends the full set of values to the TV and the over legal highlights end up looking pretty horrid but the legal blacks end up grey which makes 4:3 DVDs look pretty bad as the blacks don't match the black bars on the side of the screen.

Then there's the whole field order issue. I thought I had this under control but something truly funky is going on with HD. Somehow I'm seeing interlace combing from progressive content. It looks like somehow, somewhere in the pipeline frames are being created from fields pulled from alternating frames. Anything with motion ends up looking uber horrid as the scalers make a real mess of it or is it the de-interlacers in some HDTVs bobbing.

Bob.

I've never seen my EX1 record blacks below legal however the camera can very easily create an image with more solid blacks than you'd see from most other cameras and there is also quite a few controls that affect what happens to the blacks. Just to confuse you more Sony gives them different names to what you might be used to.

Bob.

In other words, run a test clip with supposedly known values through your chain to figure out where things go wrong. I assume that camera can make bars so that's the place to start.

Rob

Heh, some cameras can generate incorrect bars. (It gets back to the whole two wrongs make a right thing... the first wrong is the D->A not converting per standards, which you can "fix" by doing the second wrong: raising digital black level. On some cameras this affects the bars too. Confusing and dangerous.)

Dave, I doubt this has been helpful so far. If you're still interested, maybe you could outline how the media gets into vegas, how it exits vegas, and at what point they're looking at it and telling you the black levels are too high?

Rob

If you're working with DV and HDV then hardware scopes probably don't make sense.

Sorry but I could not disagree more. Without a hardware scope downstream of the video from your computer e.g looped through your broadcast monitor or out from your DV deck, you have no idea what levels etc. you are putting out, and also for accurate color correction. Software scopes (apart from histograms) in most NLE packages are only convenient, free, low resolution, processor power sucking and not worth looking at. Buy a used Tektronix wfm/vectorscope for a reasonable price and then you can put out video and audio hand on heart correct.

Well, it really depends on what sort of hardware suite you've got. Many people here are only viewing things on computer screens or flat panel TVs via HDMI. Those people don't have a way to feed a scope and have no intention of spending the money to do it.

I agree that if you're monitoring professionally and have SDI output then putting a scope in the chain makes good sense. Assuming the output from your Xena or Decklink card matches what's in your project.

Most software scopes take a small sample of your preview, maybe just an upper or lower field, maybe at lower resolution. That's surely good enough to tell if your black levels are off, but it's not the best scope you could hope for.

Rob Mack

Dave,

What exactly are you observing that makes you say, "Black video levels appear high"?

Are you looking at your computer LCD? A Calibrated VIDEO monitor? TV fed by a DVD player playing your rendered footage?

We need to know to give you competent advice.

Some editors have bugs where the broadcast safe filter won't actually be run when you perform a render (FCP)

That is not a bug. I would be really pissed off if it did otherwise. As you know, the broadcast safe filter is just a clamp, and it can really make footage look substandard, compared to using say curves to get things right (on a broadcast monitor, not a computer LCD of dubious calibration and black level capability).

Using the BC filter "to be sure levels are not exceeded" is not a good idea for anyone who cares about the quality (still, an amazing number of pros do it).

Those old Tektronixs scopes certainly add a bit of bling so I guess they serve some function. They also use a lot of power and generate a fair bit of heat. As for being accurate, they couldn't be less so. Analogue video is dead or at best on it's last gasps. No one routes composite video anymore and all the gear that was once used for that is now scrap metal.
But more to the point by design they must be less accurate than measuring using software. The software is measuring the binary values that will be written to tape or encoded using some other codec. Those old hardware scopes can at best only measure voltages ouput from digital to analogue converters. Adding more processing into a signal chain always decreases accuracy at best, at worst you can get gross errors as the D>A converters will not know what standard the binary values represent. This problem is very real in HD.
Where this becomes a silly discussion is every modern scope that I've seen is software based anyway. It might be in a box with BNC connectors but have a look inside the box and you'll find it's another computer, running software. That means if you're still monitoring legacy analogue yet another source of error, the scopes A>D converters.

I'm not knocking the worth of good metering but to say a hardware scope is inherently better is plain wrong. If you don't know how to use it, how to read it, keep it calibrated and exactly what it's measuring it's just a bit of bling taking up space.

Bob.

Bob, you're right that the interest in analog video has dropped dramatically.

But Glenn and others here are correct that the NLE software scopes are less accurate (with digital signals!), because they take shortcuts by subsampling. They don't look at every pixel, but average across x*y pixels per value.

This works OK most of the time but not all of the time.

On the other hand I don't much feel like spending $12K on Tektronix' latest little digital field WFM/vectorscope/etc. with HD-SDI inputs.