Capture Guide Suggestions Part 3

[Wilbert]

Hi,

A new version of the Capture Guide is ready and put online. The guide can be found here: http://www.doom9.org/capture/start.html

Please give us feedback and tell us things you don't agree or didn't get or what should be added for version 4.0.

I also want to thank the many helping hands, especially steVe (killingspree), zisoft and ppera2. Please refer to the "Appendix" in the guide, if you want to know more.

Several issues are improved/added:


The PAL/NTSC resizing part is improved, making a clear difference between 'ITU compliant capturing' and 'capturing with a horizontal scaler'.

The VirtualVCR capturing (based on WDM) part is added by zisoft and translated by Steve. The BT8x8 Tweaker part is added by ppera2.

Regarding AviSynth: the logo removal part is extended,

removing/correcting of rainbows and chromashifts is explained.


We tried to make this guide a bit more technical than the previous versions, so that people can a better understanding of the workings of a capture card. Although this will please some people, it will also be a bit harder to follow for people who just began with capturing. We hope that we have found a good mix. Some input on this is very much appreciated!

[Koepi]

In the sound-encoding section, BaronVlad claims that ogg vorbis is "invented" by Tobias Waldvogel.

Ogg and Ogg vorbis are developed and "invented" by Monty or better the XipH organisation. Tobias was the one who developed direct show filters for muxing/demuxing video and audio into an ogg stream (more popular known as OGM in that combination ;) ).

Else I'm very impressed by the guide, it's really helpful and much to chew on for a beginner :)

Regards
Koepi

[Kika]

The PAL/NTSC resizing part is improved, making a clear difference between 'ITU compliant capturing' and 'capturing with a horizontal scaler'.

Yeah, i like this. :)
But a small suggestion. Maybe it's a good idea to tell the Readers, what different Resolutios like Half-D1, 2/3-D1 or 3/4-D1 are based of.

As an Example:
2/3-D1 (480x576) is based on 720x576.
Half-D1 is based on 704x576.

That may be helpfull to find the right way to capture an do a proper resizing.

[DoctorRad]

The VirtualVCR capturing (based on WDM) part is added by zisoft and translated by Steve. The BT8x8 Tweaker part is added by ppera2.


The link to this appears to be broken just now.

Matt...

[Wilbert]

@Koepi,

I will change that the next time :)

@Kika,

Good idea, I will look at that ...

@DoctorRad,

Thanks (will ask Doom9 to change that). For now, you have to use the link in section 2.

[High Speed Dubb]

Hi Wilbert,

I checked the color crosstalk section of the guide. It’s both clear and correct, but there are a few points which might be able to use some further clarification.

- The guide says
Whenever you have an edge (= high frequencies) in luma, you have rainbows, even faint. Whenever you have a great change in chroma (typically logos), you have dot crawls.

That’s true, but only if your input card has a poor comb filter. It’s often possible for a good card to correctly separate luma and chroma, even when there are sharp edges.

- Later on...

PAL and NTSC should be treaten differently when reducing rainbows, because NTSC has a 4 field period, and PAL an 8 field period. But I don't understand this yet, so I can't give an explanantion
Yeah, that’s going to be difficult to describe with a short explanation. For the most part, users can just take this one on faith — There probably isn’t a need for a technical discussion of carrier signal standards. If you really want a quick explanation, you could say something like...

NTSC and PAL both use an amplitude modulated signal to carry color information. The line by line offset for the NTSC carrier signal is 1/2 wavelength; For PAL it is 1/4 wavelength.

With both PAL and NTSC, every frame (i.e., every other field) has an extra line of picture information. Because of the carrier offset, this results in a phase difference of 1/2 per frame for NTSC and 1/4 per frame for PAL. So in other words, it takes two frames (four fields) to get back to the same phase with NTSC, while it takes four frames (and eight fields) to return to the same phase with PAL.

- You should also mention that SECAM uses a completely different method — frequency modulation — for its color information. So all these artifacts don’t exist on SECAM. (I’ve read that SECAM has its own color artifacts, though — And much SECAM video is processed as PAL at some point during transmission, and therefore might potentially end up showing the same artifacts, regardless.)

It’s too bad the pictures in the “TV Comb Filters” link aren’t loading at the moment — It looks like a really nice description of comb filters from a user perspective

[Boulder]

Maybe it would be better to say in pros and cons of the different resolutions that 'deinterlacing might be necessary' as it is not necessary in all of the destination formats. That one had me wondering for a while - although this is nothing new:o