Gamut errors, what are they?

[Bob Zelin]

https://broadcastengineering.com/mag/broadcasting_quality_control_digital/

A gamut error is defined as any video signal excursion outside the appropriate color space and is grounds for rejection of the material by the broadcaster.

https://www.tek.com/Measurement/App_Notes/25_15609/25W_15609_0.pdf

REPLY – simply google GAMUT error, and you will find lots of info. SO, how do you avoid Gamut errors, or illegal level error, so a waveform monitor does not “go off” and hit the Gamut Error Alarm on the scope – easy – YOU BUY A SCOPE AND USE IT TO ON LINE YOUR MATERIAL. “But I dont’ own a scope, and they are expensive – how do I correct for Gamut errors without owning or renting a scope” – YOU DONT. Having broadcast monitoring equipment is part of the game of being in the broadcast business. Modern scopes from Tektronix, Hamlet, Leader, and Videotek have Gamut indications on them, but old analog “ebay” scopes from these very manufacturers probably will not. Depends on what model scope you buy.

Bob Zelin

[Hamish Boyd]

crumbs…

I’ve just been “Zelin’ed”. 🙂

I appreciate your response.. I really do. But but but… sir….
I can explain….

Yes I agree that I should have scopes.. but I don’t. Maybe next year. Work is on the up and I should have some monitoring.

BUT I have supplied over 50 tvcs. All passed. All ok.
(And actually the tvc we are talking about here has not been rejected by the stations, so I think someone was not reading things right or something at Adstream… so I still have a perfect record. I need to clarify with them EXACTLY why they thought this one tvc was out when I get the chance.)

My asking was also to fill in a whole in my knowledge so I come with cap in hand…

Am I right when Gamut is not something that is measured with the old vectorscope etc.. but is read by something all together different?

How can one control Gamut in an edit environment when delivering via QT files? if it is not something to do with luminance or chroma?

Again apologies for the ignorance.. just need to be directed onto the right path.
Thank you.

[Bob Zelin]

YOU WRITE –

BUT I have supplied over 50 tvcs. All passed. All ok.
(And actually the tvc we are talking about here has not been rejected by the stations, so I think someone was not reading things right or something at Adstream… so I still have a perfect record. I need to clarify with them EXACTLY why they thought this one tvc was out when I get the chance.)

REPLY – Gamut can be simply put as you went too high or too low on the waveform monitor limits (under 0 or 7.5, and over 100). The only people who will “reject” your tapes are those with waveform monitors that have Gamut alarm indicators, and are looking to justify their jobs, and go “AH HA, THIS GUYS TAPES ARE OVER GAMUT – SEE I AM IMPORTANT”. The only way you are going to catch this is if you have THE SAME SCOPE (or a better scope). I was well known in NY to have “more expensive” test equipment than the usual QC guys, and would march into a QC area (where tapes were rejected), and say “LOOK YOU IDIOT, YOUR SCOPES ARE WRONG, THIS IS THE REAL EXPENSIVE TEKTRONIX STUFF, AND THERE IS NOTHING WRONG WITH THIS TAPE”. If you dont’ have the toys, you can’t play.
I am sure your levels are fine, but if someone rejects them (blanking, audio levels, gamut, vertical interval errors, etc.), and you can’t prove them wrong, you are screwed. Most stations dont care but some (like PBS and Discovery) will torture you.

Am I right when Gamut is not something that is measured with the old vectorscope etc.. but is read by something all together different?

REPLY – you see these errors on a waveform monitor. Often, little luminance peaks will set off a gamut alarm.
Some “color” waveform monitors will indicate GAMUT ERRORS by turning red in the problem area.

How can one control Gamut in an edit environment when delivering via QT files? if it is not something to do with luminance or chroma?

REPLY – how do you control it? You watch your levels on a scope, and adjust via your software, or ultimately go thru a proc amp, or legalizer, that will adjust your levels. There ain’t no magic fix. And there will be times where YOU ARE CORRECT, and the station will be WRONG, and you can’t prove them wrong unles you have A SCOPE.

Do work for Discovery Channel and watch them reject your audio because you don’t monitor on a Dolby LM100 meter.

Bob Zelin

[Paul Thurston]

Gamut errors are USUALLY, with minor exceptions, generated when the Luma of a particular video image goes over 100 IRE and those offending parts have no color representation (in other words, the info that generates color in those highlights is missing.

You may say, but wait, I know that video can actually be recorded on tape to 120IRE Paul. How come over 100 IRE Luma signals cause Gamut errors?

Let me put it this way… some equipment manufacturers are not too cool, and make equipment designed to only accept Luma signals between 0 and 100 IRE. So, while your monitor and video tape recorder may work with 120 IRE of Luma, the TV transmitter/cable headend will spew it out. YUK. Legally speaking, NTSC accepts Luma up to 100 IRE, and therefore, transmitting Luma over 100 IRE is grounds for loosing your over the air license in the USA.

When a Gamut Error is transmitted via an NTSC transmitter, the parts of Luma that are over 100 IRE will appear digital dark green. You may notice this ugliness in live transmissions where Sunlight unexpectedly creeps in the shot, the whole video interconnection is digital, and the resulting Luma level jumps to 150IRE. (Normally in outdoor events the Luma is clamped to 100IRE for these reasons.)

To fix a Gamut Error issue:
1) Artificially clamp all your video Luma levels to 95 IRE.
2) Lower your Luma levels to 95 IRE.

If you have not edited on linear based digital video systems before, these two suggestions may not mean much to you. So, think of this as reducing the white level. In 8-bit YUV video, 100IRE is the same as R=235, G=235, and B=235, so you lower these signals to say 223. In 10-bit video, 100IRE YUV is the same as R=940, G=940, and B=940, so lower it to 893. This would theoretically get you Luma levels that peak at 95 IRE.

When you only work in RGB color space, it is technically rare or impossible to go over 100 IRE. HOWEVER, if the RGB video is changed to YUV color space, it is very easy to get video with Luma levels over 100 IRE.

Notice I said that YUV reaches 100IRE when the RGB values are 235 in 8-bit video. If you raise the RGB color to 236 or higher, say RGB 255, you will imediately get Gamut Errors over those parts that are over 235. Remember, this is only in YUV color space video, not RGB color space video.

In RGB color space, 8-bit video can actually reach 255 for the RGB intensities, and you will be peaking 100 IRE. In 10-bit video (RGB color space), 1023 is the same as 100 IRE.

All digital video tape (with the exception of HDCAM SR) records in YUV color space.

To fix your problem, you may want to lower the white component of the RGB signal in those offending frames and you should be ok.

-Paul

[Patricia Ward]

Seamus,

Just came across your post while searching for info on gamut errors. Our story is strangely similar to yours so I was wondering if you’re still experiencing the pain of rejection due to gamut errors or come across a foolproof solution? We’ve been sending hundreds of TVCs out through Dubsat and Adstream for the last 5 years that we’ve done in After Effects, Final Cut, Lightwave 3D etc and about 2 months ago Adstream began rejecting them based on Gamut Errors (not dubsat, just adstream). Like you the stations never actually rejected the adds – only adstream. My theories were – 1. I upgraded to FC Studio 3 a couple of months ago and thought this may have changed something in the way colour space was rendered. 2. The move towards HD transmission was pushing the networks to tighten their guidelines. 3. Adstream installed new equipment that showed errors where no errors showed before. I’ve spoken directly to Adstream but they can’t advise a foolproof solution apart from an external vectorscope/waveform monitor but I was pretty sure the FCP RGB limiting filter would do the trick – however I used this yesterday and was rejected by adstream again. Any ideas appreciated.

[Hamish Boyd]

Hiya,

Yep, my problem was with adstream too..
In the end I did nothing, and said send it to the stations anyway, backing my judgment on it. And it was fine. No rejection.

I haven’t changed my workflow at all since then and they have not questioned any more.
I output to a QT H264 codec (as adstream recommends) from after effects. Never even take it into FCP. (15sec commercials so don’t need to)
I have always run an adjustment layer over the top of everything with a levels filter and pulled in the black and white output by about 8points. Its never given me grief, except for that once.

[John Goerner]

This was a helpful thread for me. We recently had a couple programs accepted at American Public Television for distribution on PBS. Everyone has told me PBS is possibly the toughest on these specs.

One of our programs is airing on 4×3 SD. We had no problems with this one. The other was shot SD 16×9. These anamorphic WS standard def shows are upconverted as a matter of course and thereby have to meet the HD specs for video and audio levels.

This has been a challenge for me because most of what comes out of my suite is for corporate or direct to DVD programs. I’m quite familiar with keeping an eagle eye on waveform and vector scope levels but the gamut thing was all new to me.

We are working with some good, helpful, and knowledgeable people. Hopefully we can pass it in the second try!

John

John Goerner
Non-Stop Productions

[Dave Divehall]

FYI