"I spent a bit of time hunting on Jerry's site for the embedded test videos but failed..."Nick, try these:

JW Player: http://www.jazzythedog.com/testing/nickhopelevels.html
HTML5: http://www.jazzythedog.com/testing/nickhopehtml5.html
Silverligtht: http://www.jazzythedog.com/testing/ovp-nickhopeh265.html
YouTube:
Vimeo: http://www.vimeo.com/25317240

And, just for kicks, here's a WMV render of the same Vegas Project:

Nick, you're the most thorough tester I know!!
...Jerry

PS: I'll post the codec info on my WMP tests in an hour or so.

Nick,
if that's the case then great however I have a concern over the ongoing use of the term "expanding levels". I know what you mean but there's no doubt many reading all this who don't.
The phrase sounds like it means something is happening that is wrong and needs to be fixed when it fact that's not the case. If the term "expanding levels" is replaced with "correctly displays video" then I think everyone will get the correct idea.

To put this another way, the current quest should be to find out which things display video correctly and why they do it and others don't and if they don't if possible how to fix them so they do.

I hope that makes some sense, I'm trying to see this discussion throught the eyes of someone who doesn't know what you and I know.

Bob.

"John R and Jerry, I'm interested to know what video codec your WMP is using to play this file"Okay, here's what I found. The video codec is blank. I assume that means that WMP is not using an external codec.


fwiw: My WMP tests were run with WMP 12:
ATI laptop: Version 12.0.7601.17514
nVidia desktop: Version 12.0.7600.16667

...Jerry

amendegw wrote: Nick, you're the most thorough tester I know!!

You can change that to "biggest procrastinator"!. Thanks very much for all the testing links.

musicvid wrote: I assume your results for Youtube (always expands) are consistent for Vimeo, JWPlayer, and standalone Flash Player as well?

Yes, consistent for all those players. I have updated the table I posted above to show all the results. Each player was tested in Firefox, Internet Explorer and Google Chrome. Note that I had hardware acceleration ON in the browser's Flash Player.

Fuchs wrote: Thus to me the Vegas internal preview works correct as it shows me what actually is in the signal. And the external preview works correct as I can set it to see how my video would look like playbacked on many systems later. I find it's very important to also see and proof the signal as it is.... I wouldn't mind to have a choice how to use the internal preview but I would very miss if they would kill the way it displays the video up to now.

I totally agree with this and I think it would be useful to have a prominent switch above the Vegas preview window that allows one to see it either way, without affecting the render.

Farss wrote: I have a concern over the ongoing use of the term "expanding levels"

Well I'm not totally comfortable with it either. I try to avoid referring to [0,255] or [16,235] in this context as these could often be interpreted as the values that the signal is being mapped either from or to (with opposite meanings). The problem with "correct" is that this is rather ambiguous too, especially in relation to web video. When Laurence first raised the YouTube levels issue a while back, just about everyone was interpreting the behaviour as being "incorrect" since nothing decreed that web video should behave as broadcast/DVD, and some would probably still see it that way. Anyway for the purposes of my table above, I've kept the "expands" terminology to avoid introducing further confusion.

amendegw wrote: The video codec is blank. I assume that means that WMP is not using an external codec.

I guess so. My WMP 11 won't play the video stream of the file at all unless I have ffdshow enabled. So I guess MS included a new H.264/mp4 decoder in WMP 12 and decided that is should "expand" levels no matter what one's graphic card settings are. In other words the same as what Flash Player is doing.

Just a couple comments on these results:

1) Firefox does not currently support the h.264 HTML5 <video> tag.
2) IE9 does support h.264 HTML5 video and appears to "expand" in all my tests.
3) Chrome may drop support of h.264 HTML5 in the future.

I guess it's the subject of a future study to determine the playback characteristics of .webm & .ogv playback. (I would hope all HTML5 playback would be the same, but who knows?).

Isn't the HTML5 situation a mess?

...Jerry

Nick wrote: The problem with "correct" is that this is rather ambiguous too, especially in relation to web video. When Laurence first raised the YouTube levels issue a while back, just about everyone was interpreting the behaviour as being "incorrect" since nothing decreed that web video should behave as broadcast/DVD, and some would probably still see it that way.

Well I must have missed that one. On top of that Laurence has pipped up pretty much saying the same as me.
I would most certainly want my "correct" video to show up the same on the web as it shows up everywhere else, why wouldn't anyone??

The problem is when you say "Do you want Vimeo to "expand your levels" many would say "No" because it sounds like something has going wrong. On the other hand if you say "Do you want your video to look different on YouTube?" I'll be they say "No" too.

Why would you spend hours getting it look one way to only have it show up on the web another way, sorry to be blunt but that just sounds absurd to me. I suspect the people who were saying that it shouldn't match didn't realise they'd not seen it correctly in the Vegas preview monitor in the first place.


There is a basic set of standards used in video cameras. The Y' (luma) level for black ends up in Vegas as 16,16,16 and the Y' value for white ends up in Vegas as 235,235,235. There is some overshoot to higher values due to a number of factors, ignore them for the moment, that's a more complex discussion best left for another thread. There should be no possible confusion over this. This is what we're calling "Studio RGB" Tested, checked, explained many, many times here.
On a computer monitor a different set of standards apply, Black is 0,0,0 and white is 255, 255, 255.
To get video recorded by a video camera correctly to display on a computer monitor somehow, one way or the other, the levels have to be remapped, expanded if you want. The result then is correctly displayed video. If this is not done the blacks will not be black and the whites will not be white. Sometimes this can be hard to see, cheap computer monitors are not very good at many things but that's another issue.
This is what YouTube and Vimeo are doing, they are taking the video levels and expanding them before they go to the video cards memory so they will be displayed correctly. This is a good thing, not something wrong that needs to be corrected.

Bob.

I agree that to "expand" levels is theoretically correct behaviour and I love how the web players are behaving the same as DVD/broadcast etc., so that we can grade levels equally for both.

As for the YouTube levels discussions that begun a couple of years ago, my recollection (without hunting the exact posts, which wouldn't achieve anything useful) is that some of us were viewing it as an unexpected "quirk" or "error" that YouTube was "expanding" levels. Admittedly my relationship with this whole issue is slightly skewed because for the last 3 years, until this morning, all of my offline video players have been displaying video "incorrectly", just like the Vegas preview window.

Here are the first documented discussions of this going back to 2009, as compiled by Jerry on his website.

http://www.sonycreativesoftware.com/forums/ShowMessage.asp?ForumID=4&MessageID=677560

http://www.sonycreativesoftware.com/forums/ShowMessage.asp?ForumID=4&MessageID=718695

http://www.sonycreativesoftware.com/forums/ShowMessage.asp?ForumID=4&MessageID=740894

The only real change I've seen since those revelations is that eventually Nick was able to convince us that the "expansion" occurs in the Flash player, not in the upload server processing.

Personally, I now prefer the descriptive term "expand" to the adjectives "correct" or "incorrect." The latter convey no usable information, and the readers of such statements may be merely seeking information, not wishing the imposition of a particular point of view.

The behavior of the Vegas preview is correct for my workflow, and I would be very distraught if the ability to view and scope the full range of available RGB values was taken away because others have deemed it "incorrect."

Admittedly my relationship with this whole issue is slightly skewed because for the last 3 years, until this morning, all of my offline video players have been displaying video "incorrectly", just like the Vegas preview window.What did you change this morning?

johnmeyer wrote: What did you change this morning?

I updated my NVIDIA Quadro FX1600 driver from version 175.75 (the latest Dell certified version) to the current version 275.36 from the NVIDIA website, which gave me an "advanced" tab in my NVIDIA video color adjustments control panel that hadn't previously been there. That allowed me to select a dynamic range of 0-255. If you look at the NVIDIA Control Panel screengrabs further up the thread, mine changed from the one that I posted to effectively the same as the one that amendegw posted.

> (Nick Hope): "John R and Jerry, I'm interested to know what video codec your WMP is using to play this file."

Mine shows the same as Jerry, no codec for Video or Audio so I assume WMP handles MP4 internally.

> (Nick Hope): "Also John R, please confirm your video card make and driver version."

I'm using an NVIDIA Quadro 4000 with the 270.71 drivers connected to two ASUS ProArt PA246Q displays via the DisplayPort connections. The displays are calibrated with a Spyder 3 Elite.

> (Nick Hope): "As for the YouTube levels discussions that begun a couple of years ago, my recollection (without hunting the exact posts, which wouldn't achieve anything useful) is that some of us were viewing it as an unexpected "quirk" or "error" that YouTube was "expanding" levels. "

OK, this just all made sense for me. I thought that I had to expand Studio RGB to Computer RGB for any Internet playback like YouTube, but I see now that it is already doing that expansion on playback (not re-coding) so I really don't have to do anything if my source is already Studio RGB (which coming from my Sony Z1U camera it is). That explains a lot about what I'm seeing.

> (farss): "1) The Vegas Preview monitor out of the box on an out of the box computer monitor does not display video correctly. It is absolutely vital to understand this."

Bob, you bring up a good point. I was taking a snapshot of the Vegas Preview Windows which, of course, is not calibrated for correct video playback. When I watch the test video via the Windows Secondary Display using the setup Use Studio RGB (16 to 235) the blacks and whites are indistinguishable as they should be for proper calibration. This is how I normally edit. I should have mentioned this in my original post. I was just trying to figure out where the expansion was occurring on YouTube.

Bottom line: If send YouTube Studio RGB it will correctly play back as Computer RGB with no further processing necessary.

~jr

Thanks JR.

Can you guys please confirm that if, in your NVIDIA settings, you select "with the NVIDIA settings" and then "Limited (16-235)", WMP12 still expands the levels?

If so, I'm wondering if it's only in conjunction with NVIDIA cards, that WMP12 expands levels when the graphics card settings say it shouldn't.

I'm also wondering what WMP12 does with other formats.

Can't test here as I only have WMP11.

"Can you guys please confirm that if, in your NVIDIA settings, you select "with the NVIDIA settings" and then "Limited (16-235)", WMP12 still expands the levels?"

No, it doesn't expand levels with this setting here. It does only when 0-255 is selected.

"I'm also wondering what WMP12 does with other formats."

On my system when using a WMV video WMP12 does never expand levels no matter what the grafic card setting is.

"Can you guys please confirm that if, in your NVIDIA settings, you select "with the NVIDIA settings" and then "Limited (16-235)", WMP12 still expands the levels?"Here's my test results with WMP 12 and my Geforce GT9500 GT and an h.264 render,

With the Video Player Settings: Expands
Full 0-255...................................Expands
Limited 16-235.......................... Does not expand

...Jerry

Ah OK, so it is behaving itself.

Musicvid said: Personally, I now prefer the descriptive term "expand" to the adjectives "correct" or "incorrect." The latter convey no usable information, and the readers of such statements may be merely seeking information, not wishing the imposition of a particular point of view.

Well then you have a problem. "Others" are not deeming it incorrect. It simply is incorrect by published standards. It is not a point of view at all, it is from the standards for video as defined by SMPTE, EBU etc hence the use of words such as "correct" and "incorrect" do indeed convery useful information. If your instruments are incorrectly calibrated then everything that flows on from that point will be incorrect.

From one of the previous threads on this, from Glenn Chan himself:

"So if you are outputting a MPEG-2 or MPEG-4 with Vegas' default codecs, those codes expect studio RGB levels."

Those "studio RGB levels" mean black = 16 and white = 235 in the Vegas world. For those levels to get displayed correctly the levels must be expanded into Computer RGB.

As JR points out a few posts above, setup your Vegas preview monitor correctly and then what you see in that monitor matches what he and just about everyone else sees on YouTube. If your workflow works for you, then fine, whatever floats your boat but when you're publishing something to a wider audience then you need to make certain that what you're saying is correct to the agreed international standards. Glenn Chan's work has been peer group reviewed by SMPTE, they were so impressed they gave him an award. What he has said and what I and a few others are repeating is not a matter of opinion, it is a statement of fact. If you feel that is wrong then you are entitled to that opinion but the people to go argue with are not here, go argue with the SMPTE or EBU. I wish you the best of luck getting them to change the published standards :)

Bob.

OK Bob,
Best of luck with your crusade.
I have no interest in being "correct" in your eyes, since I haven't passed judgment on either your thinking processes or knowledge.
And if I need to be lectured on the standards, I'll go to the source, which ain't you.

Having developed and published (on my own time) an almost-bulletproof method of producing 601/709 compliant video levels for encoding and upload, intended for the vast majority of Youtubers who have absolutely no knowledge of either shooting levels or video scopes, I will lay claim to more than a passing knowledge and sensitivity to accepted luminance standards.

But if my "incorrect" methodology works, how about yours? Does yours promise failsafe compliance with 16-235 luminance standards in almost every circumstance, or does it ignore the issue completely, when taken with the lack of deliberate informed user intervention (which seems to be a pretty rare commodity these days)? How can that be any more "correct?"

What you persistently fail to understand is that where you are at now with this is exactly where I was when I decided to dig into this issue deeper, and at a time when I was making exactly the same recommendations on the forum as you are making today. You too may change your mind someday, but if not, you can take comfort that yours represents one valid approach.

And while we are on the subject of compliance with written standards, how about you review these ones before engaging in any more personal insults and baiting, during the course of what should have remained dispassionate dialog. This isn't the Middle East crisis, Bob, and you really need to cool your jets.

So before you engage in more posturing, please recall that I extended a huge olive branch just a couple of days ago at the bottom of the tutorial thread, hoping to circumvent by example any passing of judgment or schoolyard brawls I might have seen coming on the horizon.
Vegas to Youtube for the masses - a Tutorial

;?)

Musicvid said: And if I need to be lectured on the standards, I'll go to the source, which ain't you.

I'd certainly hope so. If you find anything that I have said to be incorrect I'll be both gratefull for the information and more than willing to admit that I've be wrong.

But if my "incorrect" methodology works, how about yours? Does yours promise failsafe compliance with 16-235 luminance standards in every circumstance, or does it ignore the issue completely, in the absence of deliberate, informed user intervention (which seems to be a pretty rare commodity these days)? How can that be any more "correct?"

My workflow is based on previously published work and indeed an informed understanding of what I am doing. Yes, indeed it requires "informed" user intervention. The outcome will always be correct, assuming it is being viewed correctly of course, something which is a bit hit and miss, always has been, always, will be. It is more "correct" because no matter where I send the output it will be correct. And for the record, yes, sometimes I choose to ignore a lot of "stuff", I know I'm not doing it 100% right but time is money and most of my clients sadly don't care. For the ones that do obviously I get it 100% right. I have oftenly had to work from already graded masters, I need to ensure that there is no change in anything, it will have critical eyes looking at it. Yes, I've had people look at me working and complain about my monitor having "setup" because I was using the Vegas preview monitor without it being corrected.

That is of course the crunch. Your method is "foolproof" if all someone wants to do is take video from any camera and upload it to YouTube and never get an ugly mess. I specifically made that point a couple of days ago and complimented you on coming up with a great idea.

As I pointed out then I feel the first part of it regarding levels is fine for those who totally want to ignore the issue. For those who have some understanding and have taken the time to do their own research it has the risk of adding to their confusion. At your invitation I have been trying to email you for several days to offer my suggestion as to how it could be better worded. For some unknown reasons they don't seem to be getting though to you. It maybe simpler for you to drop me an email, my address is in my profile.

As for the rest of your post I am at a loss as to why you have decided to make this personal. My jets are very "cool" down here, freezing in fact :)

Bob.

Well Bob,
As I have already stated many times, my tutorial isn't for you.
It's for the other 98% who don't share any of your expertise and knowledge.
Kind of thought the phrase "for the masses" in my thread title would provide a hint.

"Your method is "foolproof" if all someone wants to do is take video from any camera and upload it to YouTube and never get an ugly mess."
Bingo.
And as a bonus, they can add effects, twiddle obliviously with colors,contrast, and levels, and it will still be "correct to the agreed international standards" and return just what they saw on the Vegas preview before adding the final filter for rendering. Just what on earth is so wrong with that?
;?)

But if language like "you have a problem," "whatever floats your boat," and "go argue with the SMPTE or EBU," represents your "cool" attitude, then I'd really hate to see you when you're upset. But I shouldn't take any of that personally, right?

In the meantime, I will check my email spam folder.
Best of luck.

"But I shouldn't take any of that personally, right?"

My apologies, no it is not meant to be taken personally
I'm an Australian, trust me when we mean for it to be taken personally, you'll know for sure. Heck mate, down here "Silly Old Bastard" is a complimentary title, one that gets bestowed on me regularly :)

Bob.

Guys,
take a look at this, nothing to do with me of course:



Switch to 720p, take a bit of a look and then switch to full screen.
I doubt they meant for the 'scope mask to end up like that.
What has me a tad puzzled is the mask looks just fine until I switch to full screen, wierd... or my eyeballs are out of calibration.

Bob.

That video looks the same at 720p whether embedded in the Youtube page or at full screen to me. What am I not seeing?