BBC HD Testcard...

22 replies [Last post]
StevenBagley
Offline
Joined: Aug 14 2000

BBC HD now has an HD version of the Testcard (complete with 5.1 beeps) as part of it's preview reel. Also, it seems that the new detective series 'Wallander' was acquired using the RED one setup.

Steven

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Correct. Wallander is the BBC's first RED ONE production.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

StevenBagley
Offline
Joined: Aug 14 2000
Alan Roberts wrote:
Correct. Wallander is the BBC's first RED ONE production.

So I gather, although the BBC4 Barbara Cartland piece snook out on air first :)

Steven

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Correct. Wallander was quite a long time in post.

I just saw ep 1 in HD. Nice. Contrast handled well, but the depth of field was a long way too short for me in several shots, rather like in "Criminal Justice" (Arri D21). My test doc on it might just get slightly modified before publication, depending on a few questions being answered soon.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003
Alan Roberts wrote:
I just saw ep 1 in HD. Nice. Contrast handled well, but the depth of field was a long way too short for me in several shots, rather like in "Criminal Justice" (Arri D21).

Regrettably I saw it in SD (still waiting for Humax PVR box :)), but by and large nothing really stood out at me technically, which is probably a good sign. :)

One of the close ups of a face did seem soft, I don't know if this was related to the dof point?

The only other technical point I did notice was what seemed to be a good demonstration of the rolling shutter effect, a scene when a train goes through a level crossing. I noticed it - but I doubt if anyone outside the industry would have, even though it was quite a severe example, since there was so much other movement happening in the shot at the same time. My feeling after that is reassurance - the effect exists, I've now seen it on a BBC1 drama, but it didn't bother me in the way some other camera issues do.

I'm afraid I didn't watch beyond the first 15 minutes, but that was down to script etc - not the Red, and certainly not rolling shutters. ;)

Gyr
Offline
Joined: Jan 17 2005
infocus wrote:
I'm afraid I didn't watch beyond the first 15 minutes, but that was down to script etc - not the Red, and certainly not rolling shutters. ;)

You weren't the only one. I thought this series sounded promising, but our household lasted 30 minutes and then gave up. Technical issues had nothing to do with it - Wallander just didn't engage our attention - we had no interest in or sympathy for his character.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Certainly, the story line didn't fascinate, we'd worked it out quite early on. But the pictures were generally lovely. Particularly the scenery shots. There were a few very noisy shots, presumably at 2000ASA, shot at night, but otherwise no problem. I didn't notice the leaning train.

The ony issue I have with RED ONE is that it's too easy to record in an unsatisfactory format, 2k. To get a good HD or 2k performance you need to record in 3k or 4k and then down-convert using Lanczos or better.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

infocus
Offline
Joined: Jul 18 2003
Alan Roberts wrote:
I didn't notice the leaning train.

It was a very brief shot! But even more so, I just didn't find it unpleasant or jarring in the way some camera charecteristics can be. You can see it for the next few days on I-player http://www.bbc.co.uk/iplayer/episode/b00fy2zw/Wallander_Sidetracked/ and the effect is around 6:40 in (just at the end of a cafe scene).

That Alan didn't notice it also makes me feel that rolling shutter issues have been somewhat blown out of proportion. ;)

I can't post attachments, but I'll try and send you a frame grab, Alan.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

I suspect I didn't notice it because I was more interested in the story, which is what 99.99% of the viewing public would normally be doing. The effect on then train is clearly visible (thanks for the shot), but not disastrous by any means.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

delphiplasma
Offline
Joined: Dec 5 2008

Great to see the BBC HD test card. Is it only shown on the early morning 1am-2am loop? As this is the only time I've seen it.

Also, regarding the scanning parameters. If I adjust the monitor display for 0% overscan, using the HD widescreen test card, the other SD digital channels display with black vertical boarders on either side of approx 0.5cm. Is this due to the fact those channels are SD? I had assumed that the digital SD TV channels all broadcasted in the 1.78:1 aspect ratio?

In fact, the overscan tolerances seem different for the majority of digital channels.

I see that there is a channel that’s broadcasting colour bars (Astra 2A 28.2E 11260V 27500).

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

It isn't oversccan that's changing, it's the pictures themselves.

It's all because of international standardisation on digital imaging. When that happened, we unified on 13.5MHz sampling for SD, but the existing analogue line length was 52 microseconds, which works out at 702 samples. There's some slop in the specs for blanking, so it can be anywhere bwteen 702 and 704. To keep life tidy, the puictures are fitted into a 720-2ide array, but the outer 18~16 pixels in the 720 are not officially part of the picture. It made lots of sense to do this at the time.

Now that we can set overscan in the display to prescise 0, we can see all of the transmitted pictures, i.e. the full 720. So there should be black either side if it came from an analogue source, it's only when yoyu get SD from a digital source that it goes out to the full 720, and that doesn't often happen anyway. You'll also notice that the top line can sometimes be missing, and replaced with chain dots, when the pictures have come from an old vtr, and the bottom two or three lines may also get mangled. All this is perfectly normal, and has been going on since tv started in 1936, it's only now that we can fix the display precisely that we can see it all happen.

Test Card: is this going out on BBC HD? I haven't seen it yet, but Andy Quested is right at this very minute compiling a set of instructions for how to use it (he just phoned me for some advice :) )

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

delphiplasma
Offline
Joined: Dec 5 2008

Thanks Alan. I’ve kind of made a compromise and overscanned the image by approx 2% to reduce the effect (Not something I like to do. However, the display is a plasma and suffers image retention) mainly to get rid of those vertical black bars. This is more annoying when the image shift varies from broadcaster to broadcaster!

On my old CRT monitor, I use to have a ‘data blanking’ switch to blank out those data lines across the top of the image.

I’ve seen the test card only once, spent too much time on image scan size, then got a quick glimpse to see that I wasn’t clipping the ‘black level’ and off it went, after 90seconds!! And that was at 01:30!!! Bleary eyed to say the least. At the end of the day I think all you can really test, using the test card, is the image scan size, Black/white level clipping and maybe, at a push, after a few calculations, the Gamma response. Probably be able to check the grey scale tracking on LCD displays using a colorimeter.

I’m still unsure about Blu-Ray, not just by the lack of media, but more because I can still obtain a level of quality from a standard SD DVD outputting a progressive 562 (Or something) lines. Why does SD broadcast images look so poor (A scaler only slightly improving the image)? Is it due to the fact the image is ‘interlaced’ rather than ‘Progressive’? I notice a big improvement when I switch from ‘Interlace’ to ‘Progressive’ on the DVD output.

Thanks

StevenBagley
Offline
Joined: Aug 14 2000
Alan Roberts wrote:
Test Card: is this going out on BBC HD? I haven't seen it yet, but Andy Quested is right at this very minute compiling a set of instructions for how to use it (he just phoned me for some advice :) )

Yup, I have it captured on my SkyHD box, it's part of the preview loop but only appears for 90s out of the 2hrs or so of the loop. It's just after the electric proms and just before the 'Indian Food made easy' sections. Unfortunately, that's harder said than done -- the preview loop is flagged as unrecordable on SkyHD so I ended up having to manual program the box to record 15min chunks until I got it... At only 90s, its very inconvenient to have to grab it like that.

Unfortunately, I've a sneaking suspicion that my new Panasonic 42pz81 is cropping superblack material which makes the PLUGE a bit harder to use. Certainly, I could see the two spots on my old LG LCD, but not on the panasonic.

Steven

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

The sub-black is always a problem, because most of the coders around automatically clip at level 16. Andy Q knows about this and is busily writing stuff on what you can do with it. I've explained the Congo Blue trick to him, and how to use it on a normal test card (it's a darn sight easier on SMPTE/ARIB bars, but probably too late to change things now, he tells me).

I checked my PZ81 with SMPTE and ARIB bars a while ago, and the sub-black was definitely there. I'll check again to be sure, this was via analogue components and HDMI. It could be that the MPEG4 transmission coder's clipping it off. Dunno.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

delphiplasma
Offline
Joined: Dec 5 2008

Preview loop 2hours!!! I thought it was 1 hour. Pity I haven't got a 'PVR'!

StevenBagley
Offline
Joined: Aug 14 2000
Alan Roberts wrote:
The sub-black is always a problem, because most of the coders around automatically clip at level 16. Andy Q knows about this and is busily writing stuff on what you can do with it. I've explained the Congo Blue trick to him, and how to use it on a normal test card (it's a darn sight easier on SMPTE/ARIB bars, but probably too late to change things now, he tells me).

Can we have a free congo blue filter in the Radio Times? :)

Quote:
I checked my PZ81 with SMPTE and ARIB bars a while ago, and the sub-black was definitely there. I'll check again to be sure, this was via analogue components and HDMI. It could be that the MPEG4 transmission coder's clipping it off. Dunno.

I'll check again -- it could just be I have my TV set up properly, its a lovely image (Except it turns out I'm susceptible to Plasma rainbows :)). It also doesn't help that the SkyHD box has settings which alter the levels that output over HDMI.

Steven

delphiplasma
Offline
Joined: Dec 5 2008

Blacker than black will probably show on the component output, as Alan says something to do with the decoder output on the HDMI output. There may be some settings in your Sky box that will enable BTB output.

Blue only mode, if you have it on your plasma, is better than the Blue filter. However I still see discrepancies when I check with the colorimeter.

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

Yes, the Congo Blue trick isn't ideal, but it;s the best you can do if you can't turn of R an G drives. And if you want a bit of filter, get it the same way I did, from a dealer who stocks Lee Lighting Filters :) Doesn't cost much and you get a full sheet.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

delphiplasma
Offline
Joined: Dec 5 2008

Hi All,

Got to see the test card, for the second time, last night, at around 01:15. This time I put the image on ‘freeze’. Trust me, when I say I didn’t especially stay up all night waiting to see the test card!!

I have to say, I wasn’t too impressed. The resolution gratings didn’t resolve too well in the lower frequencies and the higher frequency gratings weren’t very distinct (I don’t expect to see the highest frequency grating to be resolved, as my display only has a native resolution of around 864 lines). The HD images, however, are excellent. As you can probably understand, I’m a bit perplexed as to why the gratings weren’t clearly resolved. I’ve got DVD SD test patterns which resolve the gratings clearly. Might be my Sat box? Anyone else seen any problems?

Also, the black level clip test. I myself could only see one dot, and my sat box is outputted via component. However, I feel pretty confident setting up the black level using just the one dot.

I know Alan is due to get further info from Andy, with regards the test card. However, I was just wondering what frequencies each of the 6 grating segments represent (Do they start at 1 MHz or 3 MHz?)? What sorts of level are those black level dots suppose to be? 2% or 4% above black video level? To me it looks more like 10%! What are the ‘ire’ levels of the greyscale boxes? Are they 20,40,60,80 and 100?

Thanks

backyard
Offline
Joined: Sep 13 2003

When you "freeze" on your sat box it may only be displaying a single field - SKyHD works that way and there is a loss of resolution.

For anyone interested there's a thread on digital spy about the testcard with some input from Andy Quested. see here
http://www.digitalspy.co.uk/forums/showthread.php?t=942385

regards

John

Alan Roberts
Alan Roberts's picture
Offline
Joined: May 3 1999

That's the right place to go for info.

Get my test cards document, and cards for 625, 525, 720 and 1080. Thanks to Gavin Gration for hosting them.
Camera settings documents are held by Daniel Browning and at the EBU
My book, 'Circles of Confusion' is available here.
Also EBU Tech.3335 tells how to test cameras, and R.118 tells how to use the results.

delphiplasma
Offline
Joined: Dec 5 2008

Thanks, Backyard. That link sure answers a lot of questions. Looks like the BTB spot has been fixed. However, I think I'll give it a miss, far too tired...Yawn...!

StevenBagley
Offline
Joined: Aug 14 2000

Andy Quested's guide is now online.

I'm wondering if pointing our DV cameras at the TV and recording the audio sync test might not be a bad way of measuring it. We can then load it onto the time line and measure the error there, although that at worst we could be 20ms out. Unless we use camera with a rolling shutter :)

Steven