I have somewhere in my pile of organized paper work ABC's recommended settings for the Harris/VideoTek DL-860 family of legalizers. I believe from memory they wanted 721mV to -21mV on Gamut limits. I don't exactly know why Tek revised the Tek default limits from 721mV -21mV to the EBU values. I recall at a seminar it was suggested that EBU was the only official body with a real tangible standard spec and that was why the opted to go the EBU route.
Over the years I've never received what I would call an error log readout QC report so I don't know if in practice people are using things like the Tek scope alarm error log. Given how finicky I find it I can see where they wouldn't want to hassle people over transient errors that really effect nothing in the distribution food chain.
The only concrete concern I've heard about spike illegal levels is that it will whack mpeg encoders making the rest of the picture soft. I personally haven't experienced that.
--- In Avid-L2@yahoogroups.com, Bogdan Grigorescu <bogdan_grigorescu@...> wrote:
>
> let's not mix things up:
>
> -one thing is the HDCAM codec, which, as Pat mentioned, is subsampled and SONY wants to show you pretty pictures, so they crank up kerning and sharpening on the decode/PB stage. In the process they introduce overshoots that were not there originally, potentially generating illegal levels. So even if you record a signal with perfectly legal levels(as output from Avid or even a legalizer), the HDCAM tape, when played back, could have illegal levels. This mainly depends on how close to the legal limits your source signal is and how much detail there is in it. No easy way around it, except to revisit the offending spots and dial down the video levels up to the point the codec no longer produces the spikes. I hope this explains what you are seeing on your end and the need to turn area to 1% in order to shut up the alarms on tape PB
>
> -now a different thing, but potentially connected, is what you're seeing on the tape between normal playback and pause. Stopping the tape seems to excacerbate the illegal levels. I suggested this could be caused by the deck being set to display one of the fields(now that we confirmed we're looking at an interlaced tape), which means repeating the same content on the second field, so errors in one field get doubled on pause, potentially triggering alarms.Â
>
> -the other thing is Tek and how they deal with alarms and low pass filtering for transients
> first of all, EBU103, which is definitely available as a preset in Tek, is too lax. So beside the 1% area, it calls for -35 to +735 for RGB limits and -7 to +721 for luminance. We had to dial down the trigger levels to -10 to 710 in order to consistently pass QC(while keeping the area at 1%). Even with this, we've got some odd programme rejection, every now and then, especially on HDCAM. It's less and less the delivery format, luckily.
> The biggest challenge, IMHO, is the fact the QC houses won't publish their alarm limits and criteria - it's purely their own interpretation of the standard, and I tell you it ain't EBU103...
>
> BG
> http://www.finale.tv
>
>
>
> ________________________________
> From: johnrobmoore <bigfish@...>
> To: Avid-L2@yahoogroups.com
> Sent: Saturday, February 8, 2014 11:01 PM
> Subject: [Avid-L2] Re: Video Levels jump in pause on Sony F-500 HDCam
>
>
>
> Â
> Project is 1920x1080 1080i/59.94. Looking at the main output of the deck no cross convert.
>
> I recall you have mentioned before that nothing lower than 1% area is available. The only reason I go to 1% area for the diamond alarm is that at 0% everything triggers the error alarms when set to Tek Default values which use to be High=721mV and Low=-21mV and 0% area IIRC. I noticed on my recently purchased, about a year ago, the Tek Defaults are High=735mV and Low=-35mV and 1% area.
>
> At the last Tek seminar I recall Steven mentioning the Tek defaults mirrow EBU standard as SMPTE doesn't have a comparable published standard. I will have to review the material to get the exact values if they were actually mentioned.
>
> So I will once again state that as an operator out of the Avid I can pass a sequence with Tek default alarms and 0% area but same video back from the deck I have to go to 1% area. Like it or not that's the only recipe I've found that seems to make sense from an operating stand point for tape QC. I'd welcome other people's approaches to the conundrum. In my experience all the alarms on these modern scopes seem to be overly sensitive to be practical with out backing them off in some way shape or form. Be it the alarm levels or the sensitivity however that is measured on a particular unit.
>
> --- In Avid-L2@yahoogroups.com, Bogdan Grigorescu <bogdan_grigorescu@> wrote:
> >
> > You did not mention the video format yet, nor did you mention which output of the deck is monitored(mainHD, crossconvertHD, downconvertSD, etc.)... if, for instance, the hot levels are on one field, when pausing tape you effectively double their area by repeating the field.
> >
> > Also, keep in mind that Tektronix refuses (or is technically unable) to offer an in between 0% and 1% step for their scopes and 1% is huge at 192x108 pixels. So 0% is way too low(essentially any sample above the alarm level will register) whereas 1% is too coarse. I discussed this numerous times with both Steve Holmes and Michael Waidson, starting back in 2007 - no solution so far.
> >
> > BG
> > www.finale.tv
> >
> >
> > ________________________________
> > From: johnrobmoore <bigfish@>
> > To: Avid-L2@yahoogroups.com
> > Sent: Saturday, February 8, 2014 6:47 PM
> > Subject: [Avid-L2] Re: Video Levels jump in pause on Sony F-500 HDCam
> >
> >
> >
> > ÂÂ
> > I'll take a look when I'm back in next week on the tape fields. There is a very pronounced difference between play and pause for these over shoots. The Tek scope alarm triggers in pause and not in play. That would indicate that it's not there in play but in pause as far as I know. My understanding is the 1% area relates to what percentage of the frame that is in error and not how many frames in a row have and error. I have seen Harris scopes with a Consecutive Error Sampled parameter but again I think that is across a line of video pixels and not across frames to my knowledge. Perhaps I have this wrong? The area percentage setting is 1% and that doesn't care if the frame is still or moving although when parked I didn't take note of which field the deck was showing the character generator output. The difference isn't very subtle so I doubt I'm just missing in play but of course more will be revealed and I do miss the better raster clarity
> with the
> > older CRT scopes. Given my safe color limit and the external legalizer I don't think this is anything in the original SDI output from Avid. I've delivered 4 seasons of this show through this system and never once have they flagged anything for this kind of specular transition overshoots. Also these shots were in last weeks show and didn't get flagged, so the polar axis didn't shift and the network didn't go off the air and most importantly the ratings didn't dip with these spikes.
> >
> > --- In Avid-L2@yahoogroups.com, Bogdan Grigorescu <bogdan_grigorescu@> wrote:
> > >
> > > The spikes are probably there even in playback, just less visible.
> > > Try to increase the trace intensity and use line select to position yourself in the offending areas(hot spots, car headlights, fire, flames, etc.)
> > > Then see if you can spot them at normal speed playback.
> > >
> > > Another factor is interlacing. If the format recorded on tape is interlaced and the deck is set to display one field, it could make a difference when you're paused.
> > >
> > > As a last trick, you can set trace persistence to infinity, let the tape run at normal speed, then come back and see the video level swing (both under and over shoot), then decide if you need to intervene.
> > >
> > > BG
> > > http://www.finale.tv
> > >
> > >
> > >
> > > ________________________________
> > > From: John Moore <bigfish@>
> > > To: Avid L2 <Avid-L2@yahoogroups.com>
> > > Sent: Saturday, February 8, 2014 12:03 PM
> > > Subject: [Avid-L2] Video Levels jump in pause on Sony F-500 HDCam
> > >
> > >
> > >
> > > ÂÂÂ
> > > Got a QC bounce for levels exceeding 720 millivolts. I have safe color limit set RGB 0 to 700 etc... plus an ensemble external legalizer also set 0 to 700mV. I am aware how both HDCam and to a lesser extent HDCamSr will exhibit specular/transient spikes on level due to their form of compression. What will run error free through my Tek WFM 7020 set to 0% gamut diamond error and Tek default alarm setting will light up like an Xmas tree playing the same sequence back from tape. In consulting with Tektronix setting the alarm area to 1% eliminates the unavoidable alarm triggers on tape playback.ÂÂÂ
> > >
> > >
> > > Well today's QC of transient levels of 720mV is not on my Avid output only on the HDCam tape playback and it really only happens in pause. Playing through the flagged section is fine but once you hit pause the transient levels appear. Now pre digital I understood more about why in pause levels might shift, still frame chroma adjustment on a VPR-2 anyone ;-), but I'm curious why in a digital format the HDCam jacks the levels a bit in Pause/still. Given the levels are fine in play I told the QC Monkey to kindly pass the show, leaving out the part where I wanted to tell him/her to learn to use the scope properly and stop wasting my weekend time.
> > >
> > > So can anyone enlighten me as to why in our HDCam F-500 transient levels appear in pause? The are little specular like excursion that pop up. It's not like the whole video signal level is jumping up overall. I figure it's the nature of the beast but why?
> > >
> > > ÂÂÂ
> > > John Moore
> > > Barking Trout Productions
> > > Studio City, CA
> > > bigfish@
> > >
> >
>
| Reply via web post | Reply to sender | Reply to group | Start a New Topic | Messages in this topic (13) |
No comments:
Post a Comment