Quite correct, and good catch!
I always level at the event or track level, and I can see where the advice might conflict with the basic information in the tutorial. My post above has been appended with your advice.

After adding a blocking track to a different video (an even larger one) HB opened it. I'm not going back to re-render the other one, but it does seem to work.

One question about blocking: I have one track which has several fades and is made up of generated media (.com overlay for the length of the video, ending copyright info.) As I was watching the fades, I noticed that during them the luminance RGB histogram shows the black levels bumping up to 32. I tried several methods to keep the levels at 16 16 16, including adding a levels FX with Computer RGB to Studio RGB, and scrapping that idea and changing the generated media so all blacks in it (background, text color) is 16 16 16. Still have the same results though. If I mute the video channel, the fade stays at 16 16 16. The video channel does have a Computer RGB to Studio RGB level applied to it. Not sure what to do here to keep the black levels from bumping up.

As robwood pointed out above, if the levels filter is on the output, set the blocking track (and the others) to 0-0-0 since the filter already takes care of it.

That's not what I'm doing, I've only applied the filter to the specific video track. I think for the next one I'll apply the filter to the output instead so everything else can be 0-0-0.

Edit: Just set it for the output. That was way easier than what I was doing. Thank you!

Adam, look at your video on the Histogram with all the filters removed and Preview at Best/Full. If your native levels are 16-255 or thereabouts, you should use the modified levels settings shown at 9:05 in the tutorial instead.

Just checked that, and the levels are definitely smacking 0 through the whole deal. Once this render is complete, I'll render a small slice of the video for you to check out. The original footage's color was off, so I did my best (which is probably substandard for you guys being pros) but it is by far the best footage I've produced despite my white balance being off when I shot.

Short sample of the results using the methods in the tutorial (and other advice given in this thread):

Best viewed in 720p.

Adam,
The sharpness and motion in that clip are terrific, and it's amazing how far you've come in a short time.

Viewed on a calibrated monitor, your levels are still a bit hot, at least what was downloaded from Youtube.



The Knee of the luminance curve should just kiss 16, letting the noise below that go. The luminance should be capped at 235, to keep all the highlight detail and whites within playback range.



Color looks just a bit on the cool side, but that is an entirely personal choice. Maybe it works better for action footage. Will look forward to seeing the finished product on Youtube.

235 cap, got it! Thank you, musicvid. Without this tutorial and other advice (I lurk a lot) I'd still have very substandard footage. :)

One of the problems with Vegas is (correct me if I am wrong) that Vegas doesn't tell you if your footage is REC 601 or REC 709.

If your camera shoots REC 709 (such as the EX1) then the Levels preset "Computer RGB to Studio RGB" should not be used.

John

Almost all HD video cameras and portables (except MJPEG models) use REC 709 codec primaries and 601/709 coefficients. That doesn't mean they shoot the correct luminance levels; in fact most of them do not.

The EX1, properly set up and exposed, will shoot at the correct levels. But it can also record other levels too, as it often does in the hands of people whose wallets are larger than their brains.

I recently viewed some EX1 footage that was a wide shot of a stage production, and the whole center of the scene, along with the actors, was one radioactive fuzzy white blob. As a general rule, getting the correct levels is a result of purposeful interaction with the equipment..

So the critical piece is the scopes. If you have them, they should always be used to check the output. If you don't, a combination of your eyes and some vanilla corrections will produce failsafe, if not necessarily optimal results.

In the context of this tutorial, any source being rendered to a YUV intermediate and destination should be done so at failsafe levels, to preempt clipping at one or both ends on playback.

i can't speak to the EX1, but there should be no difference in the black/white point with REC 601/709 for playback/delivery.

DV, DVD & Bluray want 016-235.

I've seen camcorder footage go above/below these specs, but the playback devices used to show these formats expect 016-235. I get that specific colors within each of those spaces may be unique and not have a direct counterpart in the other space, but the low and high of both should be the same.

then again, i just started sipping my 1st coffee so i may have missed the point, or missed the boat completely... i'm sure someone will step in to clarify if i'm still on the pier.

One of the problems with Vegas is (correct me if I am wrong) that Vegas doesn't tell you if your footage is REC 601 or REC 709.

You're looking for hard and simple rules THAT DON'T EXIST. What you should do depends on:
1- What the decoding codec is decoding to... either computer RGB or studio RGB.
2- What the encoding codec is expecting.
3- Whether your Vegas project is 32-bit floating point/full range, 32/video, or 8-bit. This will change the behaviour of some but not all of Vegas' codecs (affecting 1 and 2).
4- Whether or not your camera records illegal values above white level.

Rec. 601 versus Rec. 709 has NOTHING to do with what you should do in Vegas.

DV, DVD & Bluray want 016-235.
This is kind of confusing because you are not specifying the units. These formats want the image to between 16-235 Y'. There is no way to specify Y' in Vegas. You can only see RGB values in Vegas. So knowing that these formats want 16-235Y' is not entirely helpful; it is more useful to know whether the codec (that Vegas will use) expects studio or computer RGB levels.

If you are encoding to MPEG-2 (for DVD), sometimes the main concept codec is expecting studio RGB levels (if in an 8-bit project) and sometimes it expects computer RGB levels (in a 32-bit floating point project and with the full range setting on). So you can't even say that these formats (or the codec used to deliver for that format) expect studio or computer RGB, because it isn't always true.

Unfortunately, there are no simple rules for dealing with levels in Vegas.

This is the correct way of doing things:
http://www.glennchan.info/articles/vegas/v8color/vegas-9-levels.htm

And if you don't understand any of this stuff... I would recommend that you file a feature request. Almost every other NLE on the market handles this stuff for you.

HOW IT WOULD/SHOULD WORK:
Vegas should automatically convert everything to studio or computer RGB levels for you. (It could give the user the option to turn this off.)
And when you encode, it will apply the proper levels conversion.

In rare situations, you will get banding because Vegas may be applying a levels conversion twice. If you have FX on top of that, you are doing 3 or more operations on the image. Vegas should let you switch to a 32-bit floating point project (or 32-bit floating point rendering... same idea) to avoid these banding issues, without changing your image.

While we're at it... why not make the video preview window correct all the time.
And why not fix the scopes. On the waveform monitor (and histogram), any levels that are illegal should be colored red. And of course the waveform monitor should understand that Vegas is converting everything to studio or computer RGB (right now you have to manually put in the right setting).
And the background track should change to 16 16 16 RGB if everything is converted to studio RGB. (Heck, I don't even remember to do this myself.)

There should be an option to force a clip to studio/computer RGB for the rare situations where Vegas isn't familiar with the codec in question.

"...sometimes MainConcept mpeg-2 codec expects computer RGB levels (in a 32-bit floating point project and with the full range setting on)." - Glenn

good to know that one. and +1 to your suggestion list, they'd all be great to have.

All comments are absolutely welcomed and appreciated here, and it might be helpful to mention at this point that the video tutorial itself serves the narrow purpose summarized in the title and uppermost post. It was never intended to be an all-encompassing dissertation covering every imaginable editing and encoding situation, nor as a critique of Vegas itself.

Among the more interesting topics that are not (and will not be) addressed in the video tutorial "for the masses", are video scoping, RGB intermediates to x264, colorspace differences, luminance vs. luma, 32-bit float params, MPEG-2 or AVI output, Vegas preview characteristics and differences, or even correct shooting levels, simply because most consumers don't have the means to recognize or muster them. Touching on these and other advanced subjects here from time to time when they relate to a specific question or related method is nonetheless welcome. BTW, the Youtube version hits 2,500 views today ;?)

Glenn Chan's tutorials are certainly an excellent place to start for those wanting to dig deeper into the academics of these subjects, and I refer to them often.

added Sony Levels filter to the video output (master ) and vegas wont render = error 0x800 10 105 (message missing). UPDATE: reboot fixed it. could have been tied to having QT player and/or Handbrake open at same time as vegas? not sure...anyhow...happy rendering

settings in newest version of handbrake are not the same as in tutorial

should it be constant frame rate or VFR?

"should it be constant frame rate or VFR?Constant, 29.97 fps assuming you're in NTSC-land.



...Jerry

thanks, in reading both threads I keep seeing mention of "could use 2 pass 2Mbps if render time doesn't matter'

in my case render time doesn't matter so is CQ19 better then using 2 Pass?

2Mpbps is defined as 2000kbps and not other setting change?

does one or the other stream on the web with less buffering?

I'm not sure I understand those questions, but if you're destination is YouTube or Vimeo then use the CQ19 setting.

If your destination is a custom player on your own webpage (such as JW Player), then you need to do some trial and error to get the lowest bitrate with the acceptable quality (I like to start at CQ30).

...Jerry

yes i'll be using JW Player and JW HD Plug-in so there will be a low and high rez version. does anyone here share handbrake presets?

i was referring to if using Constant VS Average 2 Pass effects web playback / buffering?

regarding data rate, I used to do only progressive dl, I will begin using Amazon S3 / CF which has RMTP (if using the streaming option). so I am concerned about data rate more then i might have been before. of course its just a guess since streaming is new to me but I assume the data rate is the key to performance in a streaming delivery method.

i just tested and my Constant (RF 20) was 3.24 , the Average 2 Pass (2000) was 2.21 according to QT.

another thing I wondered about was in Advanced tab setting are very different for some of the presets - in your tutorial you only circle a couple sections of Advanced...is there a general consensus as to what is good to use for Advanced tab in different scenarios?