Setting the fade color to 16/16/16 and add a fade-to-color envelope to one track.

Marco

"What would folks suggest as the quickest method of creating a black track"

Or... Media Generator > Solid Color > Black. Drop it on a track, set it for 16-16-16 and stretch it across the entire project.

[i]Media Generator > Solid Color > Black. Drop it on a track, set it for 16-16-16 and stretch it across the entire project.[/i[

And then save it as a preset for future use!!

Mike

Wouldn't make any difference.
The video is encoded the way it is in the camera. Which codec you use only comes into play when you encode video in Vegas and the Vegas encoder is superior to the MS one.

Bob.

The codec handles both decoding and encoding...

When decoding, Vegas' codec will decode to 16-235 levels while MS' will decode to 0-255. With 16-235 levels, the values outside that range carry headroom and footage... virtually all cameras will record stuff in the headroom area. With 0-255 levels, all the headroom and footroom is clipped off.

On encoding, different codecs will want to see different levels. So you would apply a color corrector preset to convert between studio (16-235) and computer (0-255) range if need be.

2- I would stick with the Vegas default DV codec to be on the same page as most other people. It's also advantageous because it doesn't clip off headroom values (which are useful in case you overexpose).

Feed MS DV with a DV signal range from 16-235 and after decoding MS DV presents it as 0-255.

Feed MS DV with a DV signal range from 0-255 and after decoding MS DV will clip blacks and whites.

Feed MS DV with a grafic signal range from 0-255 and after encoding with MS DV it will be 16-235 if it is decoded with a DV codec like Sony DV (the Vegas one) or Canopus DV or any typicall hardware dv codec used in cameras or players.

Feed Sony DV with a DV signal range from 16-235 and after decoding Sony DV presents it as 16-235 - just as the original.

Feed Sony DV with a DV signal range from 0-255 and after decoding Sony DV presents it with 0-255 without any clipping, just like the original.

Feed Sony DV with a grafic signal range from 0-255 and after encoding with Sony DV it will be still untouched 0-255 (without clipping) if it is decoded with a DV codec like Sony DV (the Vegas one) or Canopus DV or any typicall hardware dv codec used in cameras or players.

Note that any dv camera probably will have levels up to 255 white recorded. I don't know any DV camera which doesn't (if you didn't precisely avoid when shooting). If you use Sony DV for decoding such a recording nothing happens to that signal range. If you use Microsoft DV for decoding you will have clipped whites from the beginning of your editing.

I'll leave it to you which codec you would use in which case.

Marco

Thanks all for the tip regarding the media generator, with a pre-set that's a really speedy solution.

If this isn't going throw another spanner in the works... how does this discussion relate to black levels in footage other than DV? For example, HDV, which is MPEG-2 in its native format and is likely to be saved as something other than DV (CineForm, QuickTime or Windows Media)?

Feed MS DV with a DV signal range from 16-235 and after decoding MS DV presents it as 0-255.
Something weird is happening there? The MS DV codec from what I've seen will encode and decode for 0-255 levels. Otherwise you will quickly get major generation loss/errors.

2- HDV seems to decode to 16-235 as the default. Ditto for MPEG2. Not sure about Cineform.

Quicktime tends to be 0-255.

VfW codecs tend to decode to 0-255, though not all of them do this.

[i]"Something weird is happening there? The MS DV codec from what I've seen will encode and decode for 0-255 levels. Otherwise you will quickly get major generation loss/errors."[i]

Indeed there is a bigger generation loss using MS DV. Though the biggest loss happens in the first decoding process when feeded whith DV footage with whites going up to 255 (which is almost always the case).
After that first decoding the losses are much minor and of another kind but still existent and visible if you go through a high cycle count of ten or twenty generations.

Worst case of all is if you always use one kind of codec for one part of the render process and another kind of codec for the second part of the process.
If you use Sony DV for decoding (nothing done to the signal) and MS DV for encoding (reducing the signals range) and back again decoding with Sony DV, encoding with MS DV you will have a grey-in-grey signal after a few cycles. Cause is the range reduction done by MS DV when encoding.

Other way round - you use MS DV for decoding a signal ranging up to 255 and it will result in clipping. You decode with Sony DV and you'll keep the widest ranges from 0-255 if present. Decode this one with MS DV and it will clip again black and whites. Do this in a few cycles and you'll have almost anything which wasn't a middle grey clipped.

You won't notice it if you always use MS DV - both for decoding and encoding. In this case only the first decoding will cause some clipping when extreme whites were present. When you encode this signal with MS DV it will press down the whites to 235. So if you decode again in another cycle it will decode to 255 but without clipping.
If your first input is a grafic with whites going to 255 MS DV is not used for decoding but when encoding it presses it to 235. Decoding then to 255 without clipping and so on. When looking at this resulting DV signal it will always look just like you do everything with Sony DV.
Similar when first feed is DV with whites up to only 235 (which is unlikely to happen). MS DV will decode it to 255 without clipping, encode it then to 235 again. And so on. Decoded with MS DV this will always look fine and same or very similar to if you'd used Sony DV for any process. The only difference here is MS DV will lose a few percent of luminance after somewhat of 20 render generations which Sony DV won't and blocking etc. is bit more appearant when using MS DV but nothing being rather visible

So if you ALWAYS use MS DV you probably won't notice at all what's going on there except of clipping in the very first cycle (if footage was DV). But talking of generation losses I'd always prefere Sony DV over MS DV.

Feed MS DV with a DV signal range from 16-235 and after decoding MS DV presents it as 0-255.

Feed MS DV with a DV signal range from 0-255 and after decoding MS DV will clip blacks and whites.

We tested this very often with different methods. The most simply one is creating a DV file which shows pure black (RGB 0/0/0) in the left half and pure white (RGB 255/255/255) in the right half. Text with RGB 16/16/16 over the pure black left half and text with RGB 235/235/235 in the pure white right half. You must be sure the resulting DV file keeps the range 1:1 so you can't use MS DV to create this file.

Now when you take this DV file and preview it you can easily see which way the used DV codec decodes the signal:
Use Sony DV or Canopus DV or a similar one for decoding and you can read both the text in the black as well in the white part (of course your monitor must be adjusted and able to difference the range between RGB 0 and RGB 16 and the range between RGB 235 and RGB 255).
Use Microsoft DV, Panasonic DV or a similar one for decoding and the text is gone. There is only pure black and pure white left without any text in it.

Marco

Marco, I think that your methodology might be wrong. You need to specify your units (Y' or RGB) and be aware of what decoder is in use.

RGB values get converted to Y'CbCr values in a DV file. Those Y'CbCr values can get decoded in different ways depending on what DV decoder you are using.

Of course it's not to be recommended to mix two codecs which work in a different way. I only mentioned it because this makes very clear what happens when using what codec. It's a worst case scenario which I hope will never happen to a user.

But at least there is one case where a MS DV user can't avoid mixing. It is because the DV encoder used in cameras does usually work they way Sony DV does, not the way MS DV does. At least I never saw DV footage coming from a DV camera that behaved like MS DV. You can't avoid clipping the headroom highlights when using MS DV here. Use Sony DV and the levels (and the headroom) stay untouched.

So when working with DV I can't see any reason to use MS DV over Sony DV. Use Sony DV to be sure you'll keep your luminance range. And use a filter like Broadcast Colors in the last step if you really need to get head- and footroom ranges into 16-235.

Maybe some of you remember times when users of other NLEs complained about they have luminance jumps at beginning and end of transitions when working on DV. Again this was caused by the way MS DV worked and still works. In the past many NLEs used MS DV internally and also did smart-rendering. So users sometimes had parts of clips which were smart-rendered (so NOT really rendered) and other parts - the transitions - where the MS DV codec worked and pressed down the luminance level. In the end playbacked with a device which had a regular hardware dv codec which works same way as Sony DV does and suddenly they saw slightly darker transitions then rest of the stuff was. This was not always visible dependend on what the footage looked like and what kind of transition and what speed the used. But I read a hundreds of such complaints and it's easy to reproduce when using MS DV.

Marco

Interesting to note that a capture from VHS tape (PAL) via a Canopus external converter, results in a file that has black levels down to 0, and when composited on the timeline with a bottom track that is 16-16-16, on playback (rendered to MPEG) the difference is quite clear between the two levels.

Not entirely sure where that leaves us! ;-)

If you used Sony DV for decoding it just states levels are left untouched. Canopus DV works the same way Sony DV does - speaking about the luminance levels. So even in your final output the difference between to deepest 0-level range of the input and the 16-level range of the internal media keeps being visible - asumed your monitor allow showing differences between RGB 0 and RGB 16.

In this certain case if you'd use MS DV for decoding you still would see a difference in your final output but you would have clipped black details (if there were any). MS DV would not work on the internal media which is RGB 16 in your case. But if the RGB 16 media would have come from outside being DV (replacing your RGB 16 bottom track) there would have been no difference no more.

Marco

Thanks Marco. So that begs the question, if a file (assuming it is being decoded by the Sony DV CODEC) has levels that are below 16, is there any need for that lower 16-16-16 generated media?

I think you need to be aware of how signals are encoded for your local standard and region. For instance NTSC black level in the United States is supposed to be 7.5 IRE. On the other hand NTSC black level in Japan can fall to 0 IRE. Evidently PAL also uses 0 to 100.

Let's also get straight that when you talk about 16-235 you're really talking about 7.5 to 100 IRE. Values below 16 should equate to IRE below 7.5 when using the Vegas DV codec.

I have to assume that Vegas' background black is designed to be compatible with PAL and Japan's version of NTSC. Editors here in the US have to add generated media to the bottom track. And, Yes, it'd be nice if you could configure the background level.

Imagine the hell that would break out amongst the PAL users here if Vegas' background defaulted to 16-16-16!

Rob Mack

A fairly interesting page that gives a lot of useful information:

http://www.execulink.com/~impact/scopes.htm

Rob Mack

The 7.5 IRE setup will be removed by the A->D converter if it's told to do so and added when going D->A. There is NO setup in digital NTSC.

If the A-> D converter is sending you black at 0 then match your gen media to that. For output then legalise the whole thing.

Bob.

Let's also get straight that when you talk about 16-235 you're really talking about 7.5 to 100 IRE. Values below 16 should equate to IRE below 7.5 when using the Vegas DV codec.
No...! 16 16 16 RGB in Vegas doesn't always get translated into 7.5 IRE. It depends what conversions are going on. There are two things happening:

1- The R'G'B' values get converted into Y'CbCr values / gets converted into DV. This behaviour depends on what DV codec you are using. Either 0 0 0 RGB will get mapped to Y'CbCr black level (16 Y'), or 16 16 16 RGB will get mapped to Y'CbCr black level (16 Y'). Y'CbCr black level is always 16 Y'; this is the standard.

2- The Y'CbCr values then get converted to an analog format... such as composite NTSC or PAL. Composite levels are measured in IRE. For North America, Y'CbCr black level should get mapped to 7.5 IRE. But most equipment will map Y'CbCr black level to 0 IRE.

For Japanese NTSC and PAL, Y'CbCr black level gets mapped to 0 IRE.

3- Everybody, PLEASE specify your units... whether it's RGB, Y' (or Y'CbCr), or IRE.

If things are still unclear, please see http://glennchan.info/articles/technical/setup/75IREsetup.html



For PAL that can be appropriate if you're using the default Sony Vegas codecs (for DV, HDV, MPEG2).

So that begs the question, if a file (assuming it is being decoded by the Sony DV CODEC) has levels that are below 16, is there any need for that lower 16-16-16 generated media?
...yes. The first has little to do with the second though.

If you have levels below 16 RGB (assuming Sony DV codec), they aren't in legal range and will get clipped. You may or may not want to bring them into legal range.

Stuart is giving PAL examples with blacks falling down to zero on the vegas waveform. You're saying that he still needs to add a 16-16-16 black track under it all?

Rob Mack

yes. (If there are fades to black, gaps in the timeline and such.)

Stepping through this a little more then...

If Stuart burns a PAL DVD and I burn an NTSC DVD, and we both add a black layer, then both DVDs will have blacks falling to 16-16-16 on fadeouts, etc.

But because his DVD player is a PAL player, it'll output that 16-16-16 black at 0 IRE while my NTSC player will output it at 7.5 IRE? True?

And what happens when levels fall below 16-16-16? In my case the levels just drop closer to zero. What happens in Stuart's case?

Rob
(leaving work now, I'll check back in 3-4 hours)

For those still confused by the whole black level issue, this tutorial from JVC may help.
And trust me when I say that things were MUCH simpler in the days of analog video :-)

Mike