john_dennis,
I see your preview is at Preview/Half.
Scopes are only accurate at Best/Full (with both checkboxes unchecked).
If they still look out-of-gamut, maybe give them a call.
;?)

@musicvid

I adjusted the video scopes settings per your post and sampled different places in the broadcast. I included representative samples from different sources, studio, prepared video reports, a commercial, etc.

This PDF opens full screen and has multiple 1680x1050 pages for your analytical pleasure. Hit Escape to return to Acrobat. My considered opinion is, at least some sources are peaked. The commercial looks reasonable, if a little peaked in the blue.
Do you recognize the blonde lady on page 3?

@ ushere

I've seen my work look like a dog's dinner, mostly from my own ignorance. I can work on that, though. I once made a CD of a live tape and delivered a copy to one of the band members as a Christmas surprise. I was somewhat taken back that he proceeded play it in the nearest boom box. So much for making a live recording sound live.

fwiw, it appears the the "problem" area in the pdf examples is the pillarbox fill. If we crop this out, and put a 16, 16, 16 black on he bottom track, here's what I see:


...Jerry

Thanks Jerry,

It looks as if the computer generated graphics seems to be the major source of peaks. Page 3, where there is little in the way of graphics, is well within range. I sympathize with them having to blend all the various sources and still keep it legal. And they have to do it in real time.

I have a list of things that I need to understand better. Back then, I needed to get my own levels in check. I appreciate all the work you, musicvid and others have done to expose how levels can get shifted when uploaded to streaming sites and burned to optical media. So now I'm thinking a lot about the constraints of "broadcast levels".

So far, I'm up to "C", color curves. Eventually, I'll get to "S", scripting, John and Nick.

All scopes can tell you is if the image is inside legal limits, what they cannot tell you is if it's pleasing to the eye or conveys the artistic intent of how it was shot.

There's also the technical issue of out gamut colors to consider. We've had a few discussions about this over the years, one being Ushere's better half's works of art. One I've had but barely worth a mention is a particular flower that grows in a pot at the back of my house. It is impossible to capture the color or even an approximate rendition of it in full sun.

In the first screen grab the red roses look like the color is "out there" but I don't know how or if you can tell this from an RGB parade, the vectorscope might be a better tool and then as Jerry hints at above you need to use a mask to isolate that part of the frame.

As a matter of interest seeing as how this is about OTA, one of our broadcasters not so long ago was called to account by the regulators for dialing something upto 11. The intent seems to be the same as it was back in the days of AM radio when extreme compression schemes were the go. The idea then was when you spun the dial to get you to stop at the loudest station. Today there seems to be a bit of a push to get the same result when channel surfing. My probably forelorn hope is the broadcasters will realise their efforts are counter productive, at least for one potential viewer.

Bob.

"what they cannot tell you is if it's pleasing to the eye or conveys the artistic intent of how it was shot."

To some extent, learning anything at all about this tends to rein me in. I'm sure the industry wants to produce repeatable results and the regulators want to allocate scarce spectrum fairly. Not being a broadcaster or having any particular knowledge of the mechanics of transmission, I'm not certain that the constraints of the analog days applies to the digital world. It's likely that if you give broadcasters 6mHz, some will try to use 7 mHz, and if you give them 16 mHz, the same one's will try to use 20 mHz. I will acknowledge that there is a tremendous amount of equipment that makes these broadcasts happen and that the equipment evolved over 75 years of mostly analog. So, if I don't know history, I suspect I'm doomed to make the mistakes of the past.

"their efforts are counter productive, at least for one potential viewer."

Our TV has TV Guide On Screen and show selection is much more deliberate. In my case I just ask my wife which show we plan to watch.

Thanks for indulging my curiousity, I realize this is probably not of any great social or technical importance.

It could be that the erroneous levels are from some form of sharpening.

When you scale from HD to SD, you may have sharpening-like effects depending on the scaling algorithm used.
And of course something in the signal chain may add sharpening to the signal.

The broadcast colors concept is an enormous pain to me... the majority of my video goes to YouTube to promote Diving in Palm Beach, Fl. I don't want this stinking Broadcast Color reduction in color space....Vegas and many other programs practically sabotage your attempts to maintain the entire color spectrum--which is fine on the computer, and on YouTube or Vimeo.

If you really want to push a new LED TV with richer color space, try underwater videos at 1080P...
Here is one....

I don't want this stinking Broadcast Color reduction in color space....Vegas and many other programs practically sabotage your attempts to maintain the entire color spectrum
Vegas does not do that... nor does any other program for that matter?

I downloaded the 720p version of your underwater video (very creative!) and actually the levels weren't too bad -- except the text and stills, which were rendered in their native RGB space. Just throw a Studio RGB filter on these events, and you should be in good shape.

There were a few underwater shots where boosting the shadow levels to 16 and some land shots where dropping the highlight levels to 235 would give you more detail from Youtube, but again, not too bad. There is some pronounced audio clipping with the music piece at 3:25-5:25.

All in all, a good effort. Here's more food for thought:
http://www.jazzythedog.com/testing/DNxHD/HD-Guide.aspx

Musicvid, very cool material on taking video to Youtube or Vimeo.
I shoot u/w video with a canon 5 D Mark II....import the clips into HDLink, ( cineform) where they convert to cineform avis' and then if shadow areas don't have the right look, I correct in First Light ( typical in underwater video is an issue of very dark right next to very light, and where gradations of blacks in the shadows are critical to being able to see important subject matter in the video)...The correction gets saved as 4-2-2 colorspace usually, though I have played with 4-4-4 as well.
This goes into Vegas, for editing...when completed, I render a cineform avi, and then pull this into Sorrenson Squeeze 8. It allows use of the MC encoder for h264 with virtually all of the setting available to manipulate. I have much better results, when I use a Sorrenson filter to black restore, and avoid the starting at 16..this reverts it to zero....I thinl :-) The result is sharper gradations in the dark area.
I wish I had a way to send you a 1 minute or 30 second clip of a cineform avi I have, of a Goliath Grouper spawning dive at 95 feet deep--you could see the issue exactly. In this video, it is slightly dark from the time being 6:30 am, and the Goliaths are dark, over a whiteish sandy bottom, with baitfish swirlling all over the Goliaths, sometimes renedering the 400 pound fish almost invisible. My cineform avis are so much better than the h264's or BluRay, I had to use the avi's for a big trade show ( DEMA).
With the black restore, it helped alot, but still detail is lost that is essential detail.
This clip is an exageration of the issues in most underwater videos.

"I wish I had a way to send you a 1 minute or 30 second clip of a cineform avi I have"The free Dropbox is an excellent tool for doing just this. Merely, drag your clip to your local Dropbox Public folder, right-click & copy URL. Then post the URL back here.

...Jerry

btw: The link I posted above gives us both an extra free 250MB.

Music vid..the Goliath clip I was talking about is included in the tour of Palm Beach video you watched...right at about 4:44 it is running....see the darkness, but need for detail....the fish are swimming at about 3 to 4 mph, which is over twice as fast as 99% of u/w videographers are able to swim at, lot less shoot at :-). My camera is mounted to a 80 pound monster underwater scooter, which I turbo boost for more speed when needed, using 4 foot long composite freedive fins. What this means is there is no way to add more light than I am adding ( dual 50 watt HID cave lights by Apollo) without getting too much drag to follow at this speed.
Each u/w videographer has their own strengths and weaknesses...one of my strengths is that I get footage of big marine life traveling far too fast for other divers.....like the clip I have where I am inserted in to a school of wild bottlenose dolphons for close to 20 minutes, traveling with them over coral reefs, averaging close to 5 mph the whole time, even doing the rapid ascents for dolphon breathing, and the descents with them :-)

I end up with cool footage, that I will often need to alter in First Light, much like a stills photographer shooting with insufficient light, will need to go into LightRoom and add Fill Light, blacks, Brightness, contrast, and Vibrance..carefully.. :-)

Now squarely in a larger color space than I started in with the Canon, the 4-2-2 or 4-4-4 color space makes better sense, especially for the gradations in the dark areas that I need maximum detail in....

As you have been prolific in your coverage of issues very related to my issues, I am hoping you will have some interest in these underwater video parameters and how best to handle them.
REgards,
Dan

We have a fairly tight working relationship with Nick from Bubblevision (he shared a lot of input for the video tutorial), and although he personally uses entirely different workflows than you (or me), I'm sure he would love to share notes. His latest efforts on Youtube take underwater video to the realm of a separate art form, which is where I'm sure you're headed with yours.

And the notion of crushing underwater blacks on purpose for impact is not unknown as can be seen in one of his latest efforts; as I've said often, Ansel Adams is one's best friend in post, even though he never made an underwater video.

So it's nice to know how to manipulate levels and curves for creative control, and how to discard or keep the end levels according to your purposes, within the constraints of 709, which Youtube and most players expect.


http://www.bubblevision.com/underwater-video/Vegas-YouTube-Vimeo.htm

WOW!!!!!!!
Like underwater Video on Anabolic Steroids!!!
I will be using some of his shot strategies in the future :-)

I have been to Fiji.... the colors there do not seem to be even possible....it is in another Universe, really... Fortunately for Palm Beach Dive Tourism, most American divers can't afford the time to spend close to 2 days of travel in each direction :-)

Also, the majority of his optimal subjects are in shallow water, where ambient light is spectacular....video lights assist, but the lighting stage is totally unlike one of my 100 foot deep dives off Palm Beach, where ambient is not particularly helpful for colors.

I am uploading the jewfish clip I spoke of now to dropbox. It is 1.2 gigs, and will take
"a while".... For the purposes of the colorspace and light gradations, this should be a great reference clip, for anyone that wants to play with it.

I will visit your friend's site, and spend some serious time on picking up technique ideas from him. Thanks again.
Regards,
Dan

Nick does a lot of deeper dives, and shooting them comes with their own set of limitations, murkiness being but one of them. But he has a special knack for getting close without spooking the talent. I'm sure that's something one couldn't teach. Back in the day, I hand-corrected tens of thousands of underwater stills in a professional film lab (in between weddings) ;?o

But for those of us who have never had the experience of actually being there, we don't mind or even notice the problems; just the thrill of seeing it is enough.

{EDIT} I just noticed, 1.7 Million hits in just 3-1/2 months. Congratulations NICK!!!

OK, here is the same clip with the end levels optimized for web delivery, much in the way described here:
http://www.bubblevision.com/underwater-video/YouTube-Vimeo-levels-fix.htm

In the 720p version on Youtube, you will notice more shadow detail in the "Goliath" example and more highlight detail in the following aerial clip. Artistry played no role in this experiment, just the numbers. No gamma changes were made to either clip, just the endpoints. I'm sure you also know the Mainconcept encoder is prone to blocking in deep shadows, moreso than x264.

The optimized levels are on the left. This is an unlisted video on YT. I will be glad to take it down whenever you wish (it's your gig).
;?)

Looks like I will have to add another step.
Thanks.
Dan

My jewfish video -- cineform avi, is in a dropbox folder now...https://dl-web.dropbox.com/get/videos/jewfish.avi?w=1ed07007&dl=1

Dan,

Did you put this in your "Public" folder?


And did you get the URL by doing a "Copy Public Link"?


If not, there might be a way to move it from your private folders to your public folders, but I'm scared that you may have to copy it to your local "Public" folder which may require it to be uploaded again.

...Jerry

PS: You might be able to "share" this even if it is private, but it will require an email dialog.

Lets try this...the Goliath Grouper cineform avi for testing
http://dl.dropbox.com/u/57686349/jewfish.avi
( moved from video to public :-)

http://dl.dropbox.com/u/57686349/jewfish.avi

Doing this so I can download it without having to create a separate html page with the link.