Does Bluray support 59.94 fps Progressive for 720 x 480?

[VideoFanatic]

This page says "progressive scan can go up to 1920×1080 pixel resolution at 24 frames per second, or up to 1280×720 at up to 59.94 frames per second". So can anyone comform that Bluray supports 720 x 480 at 59.94 fps?

[jagabo]

can anyone comform that Bluray supports 720 x 480 at 59.94 fps?

It does not.
http://www.hughsnews.ca/faqs/authoritative-blu-ray-disc-bd-faq/4-physical-logical-and-...ifications#4.6

[VideoFanatic]

I'm using QTGMC(Preset="Ultra Fast") with Avisynth. If I'm reading this correctly the highest FPS 720 x 480p supports is 29.97 fps. Is that correct?

So what should I add to my script to keep it at 29.97 FPS?

Also I remember you mentioned this: https://forum.videohelp.com/threads/346916-Remove-Interlacing-Artefacts-without-convert...=1#post2168471

So is 29.97 FPS going to look worse than 59.94 FPS then because it only has half the temporal resolution? Could you explain that as I don't understand. I thought both of them are progressive and 720 x 480?

[poisondeathray]

Yes it will work worse if it's 29.97p. Half the motion samples , so motion will not be as smooth. It's more noticable in sports footage

For BD, your options are either re-interlace it for SD 480i59.94 or upscale to 720p59.94

To reinterlace from a 59.94p source

AssumeTFF()
SeparateFields()
SelectEvery(4, 0, 3)
Weave()

But then why are you deinterlacing it in the first place, to only re-interlace it later? (unless something in your workflow requires it) ?

[VideoFanatic]

I don't know what you mean. I don't want it interlaced. I just want it progressive at the highest FPS that Bluray supports for 720 x 480. The interlaced video is 29.97 fps. I'm converting to progressive with the following script. What script should I add to keep the video as 29.97?

QTGMC(Preset="Ultra Fast")

[poisondeathray]

You have 2 options. They are listed above

If you don't want interlaced, you're ONLY option for highest progressive FPS (for blu-ray) is 1280x720p59.94 . If it's 4:3 content , you probably want to pillarbox it along with upscaling

720x480p29.97 isn't supported directly by blu-ray (it's encoded and/or flagged interlaced, but the content can be 29.97p)

29.97p will look a lot worse for sports videos like wrestling (1/2 the motion samples)

If it were up to me, I would leave it interlaced

[poisondeathray]

Or better yet, forget blu-ray and optical media. Just get a media player like WDTV, Asus Oplay etc...

[intracube]

So is 29.97 FPS going to look worse than 59.94 FPS then because it only has half the temporal resolution?

It totally depends on what the source footage is. If it's regular video (sports, news, general entertainment, home videos) or hard telecined film footage - yes.

If the source is native 29.97p content (possibly flagged/encoded as a 59.94i stream) - no.

You say you don't want it interlaced. I'd argue the source footage should decide that.

[manono]

What script should I add to keep the video as 29.97?

QTGMC(Preset="Ultra Fast")

SelectEven()#or SelectOdd(), sometimes one looks better than the other

[Cornucopia]

can anyone comform that Bluray supports 720 x 480 at 59.94 fps?

It does not.
http://www.hughsnews.ca/faqs/authoritative-blu-ray-disc-bd-faq/4-physical-logical-and-...ifications#4.6]

One inconsistency I notice in that link is that it says you can have 2nd Video (PIP) using 24p (23.976p) 480 video. But the spec REQUIRES that any 2nd Video stream MUST MATCH the primary video stream. It doesn't list 24p as supported for Primary, but I have a feeling that it actually IS supported.

Of course, the OP asked about 59.94p 480 video. I think we can rule that one out with confidence!

Scott

[jagabo]

One inconsistency I notice in that link is that it says you can have 2nd Video (PIP) using 24p (23.976p) 480 video. But the spec REQUIRES that any 2nd Video stream MUST MATCH the primary video stream.

Wouldn't 720x480p24 match 1920x1080p24?

[jagabo]

So is 29.97 FPS going to look worse than 59.94 FPS then because it only has half the temporal resolution? Could you explain that as I don't understand. I thought both of them are progressive and 720 x 480?

Play the video in this post on a 60 Hz monitor -- full screen for best effect. The middle row is 30p. The bottom row is 60p.

https://forum.videohelp.com/threads/307004-Best-framerate-conversion-%28eg-23-97-to-30-...=1#post1888926

Notice how the 30p row is jerky and flickery? The 60p row is silky smooth. The top row is 24p. It has the inherent jerkiness of 24p plus the judder of 3:2 frame repeats.

[VideoFanatic]

Yes I see what you mean, 30fps looks like crap. Just wondering though, why would Bluray support 1280 x 720p 60 fps but not 720 x 480 in 60 fps?

Surely 60 fps is the same no matter what resolution so wouldn't 720 x 480 play in 60 fps?

[intracube]

Yes I see what you mean, 30fps looks like crap. Just wondering though, why would Bluray support 1280 x 720p 60 fps but not 720 x 480 in 60 fps?

Because historically 720x480 video has been interlaced. If 720x480@59.94p doesn't exist out in the wild, why make provision for it on Bluray?

Another comparison would be 59.94i vs 59.94p. There will be much less difference here, but the latter can be slightly shaper with less artefacts (like line jitter).

You still haven't said what your original footage is. I'm assuming it isn't HD as you wouldn't want to downscale it to 720x480. So if the source is 720x480 it's going to be 59.94i or possibly 29.97p. Either way you should be able to put that on a bluray disc as-is. If your TV is a modern flatscreen it can convert 59.94i to 59.94 progressive.

I don't see what the problem is.

[VideoFanatic]

The source footage is 720 x 480 interlaced. I can see interlacing artefacts in it on my TV. If I convert to 60 fps progressive those artefacts are removed. However Bluray doesn't support 60 fps and 30 fps looks like crap.

[jagabo]

The source footage is 720 x 480 interlaced. I can see interlacing artefacts in it on my TV.

Then you did something wrong. If you had done it right there would be no comb artifacts. Every Blu-ray player and TV can deal with 480i.

[VideoFanatic]

The source footage was like that. I never did anything to it. I can't imagine anyone converting VHS-to-DVD would cause interlacing artefacts. I even took a sample into a professional TV shop and they said that interlacing artefacts on NTSC broadcast TV is normal (from that era of around 1996).

The interlacing artefacts aren't over the whole picture all the time they only occur in certain times and sections of the video.

[intracube]

I even took a sample into a professional TV shop and they said that interlacing artefacts on NTSC broadcast TV is normal (from that era of around 1996).

They're probably referring to things like line jitter (which is subtle from a normal viewing distance). It sounds like your TV is displaying both fields at the same time - which it shouldn't do.

Maybe if your video is incorrectly flagged as progressive, you might get the artefacts you describe.

The interlacing artefacts aren't over the whole picture all the time they only occur in certain times and sections of the video.

It'd be worth posting the details of the video with mediainfo.

Another possibility is the source has been through non-interlace aware scaling at some point which could have damaged the field structure.

[VideoFanatic]

Here's the clip so you can see for yourself what it looks like on your TV:
https://dl.dropbox.com/u/57941257/Original%20few%20secs.mpg

Here's the MediaInfo:

Format : MPEG-PS
File size : 22.4 MiB
Duration : 39s 973ms
Overall bit rate : 4 696 Kbps

Video
ID : 224 (0xE0)
Format : MPEG Video
Format version : Version 2
Format profile : Main@Main
Format settings, BVOP : Yes
Format settings, Matrix : Custom
Format settings, GOP : M=3, N=15
Duration : 39s 973ms
Bit rate mode : Variable
Bit rate : 4 411 Kbps
Maximum bit rate : 6 500 Kbps
Width : 720 pixels
Height : 480 pixels
Display aspect ratio : 4:3
Frame rate : 29.970 fps
Standard : NTSC
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Scan order : Top Field First
Compression mode : Lossy
Bits/(Pixel*Frame) : 0.426
Stream size : 21.0 MiB (94%)

Audio
ID : 189 (0xBD)-128 (0x80)
Format : AC-3
Format/Info : Audio Coding 3
Mode extension : CM (complete main)
Muxing mode : DVD-Video
Duration : 39s 968ms
Bit rate mode : Constant
Bit rate : 192 Kbps
Channel(s) : 2 channels
Channel positions : Front: L R
Sampling rate : 48.0 KHz
Bit depth : 16 bits
Compression mode : Lossy
Stream size : 937 KiB (4%)

[intracube]

It's late in the evening and I might be wrong about this, but it looks like the interlaced video might have been lossy encoded as a progressive stream at some point (although it's flagged as interlaced now according to mediainfo)

When encoding an interlaced content as a progressive stream, as well as requiring a higher bitrate, any compression artefacts/macroblocking could have partially damaged the interlacing.

Processing/converting the stream to progressive with QTGMC might have a 'restorative effect' on the picture quality.

So what you're seeing when you put this interlaced video on a disc might not be an inherent limitation with interlaced content, but instead be because of the way this footage has been processed/encoded at some earlier point in the chain.

[VideoFanatic]

This is weird as well. I tried QTFMC on Ultra Fast with Yadif and Vinverse and found that it removed more interlacing artefacts than Slow does!

Mpeg2Source("L:\Original few secs.d2v", CPU=6)
TTempSmoothF(maxr=3, lthresh=8, cthresh=5, strength=4, interlaced=true)
Load_Stdcall_plugin("C:\Program Files\AviSynth 2.5\plugins\yadif.dll")
QTGMC(Preset="Ultra Fast") 
Vinverse()
SeparateFields()
SelectEvery(4,0,3)
Weave()

[jagabo]

That MPG clip is normal TFF interlaced NTSC video. Of course, it's from a poor VHS source and it's very poorly encoded with too little bitrate.

You should add AssumeTFF() to your script. Or Info=3 to your Mpeg2Source() commands.

Vinverse() just blurs any remaining comb artifacts. I think QTTMC() delivers better results than QTGMC(preset="ultra fast").VInverse(). It retains more detail and there's less edge buzzing.

[VideoFanatic]

I added AssumeTFF() but it didn't see any difference to the picture. I also can't see any difference between Ultra Fast and Slow but what's obvious to me is that Ultra Fast removes more interlacing artefacts. This must be because of Yadif. Is there no way to use Yadif in speeds slower than Ultra Fast?

Also what does Info=3 do?

[jagabo]

I see no comb artifacts when I play your MPG file on my TV with a WDTV Live. I didn't bother making a DVD but I'm confident it would have looked about the same from a DVD or Blu-ray player.

I don't see QTGMC() leaving a lot of artifacts either. QTGMC(Preset="Ultra Fast").Vinverse(), and QTGMC() side by side:

[VideoFanatic]

I also play the file on a Media Player and you can clearly see interlacing artefacts when the guy applies the hold - look at the top right in the crowd. When the view changes look at the top-left in the crowd and the hold-appliers face.

If you can't see it then try this script which deinterlaces and re-interlaces the video then you can see the difference compared to the original:

Mpeg2Source("L:\Original few secs.d2v", CPU=6)
TTempSmoothF(maxr=3, lthresh=8, cthresh=5, strength=4, interlaced=true)

AssumeTFF() # or AssumeBFF() depending on the source

# deinterlace

Load_Stdcall_plugin("C:\Program Files\AviSynth 2.5\plugins\yadif.dll")
QTGMC(Preset="Ultra Fast") 
Vinverse()

SeparateFields()
SelectEvery(4,0,3)
Weave()

[jagabo]

I also play the file on a Media Player and you can clearly see interlacing artefacts when the guy applies the hold - look at the top right in the crowd. When the view changes look at the top-left in the crowd and the hold-appliers face.

I don't see what you're talking about. Give me some frame numbers. And/or some sample images.

[VideoFanatic]

Between the following times you can see interlacing artefacts: 10 secs - 31 secs. You must have a really good deinterlacer on your media player or TV then if you can't see the interlacing artefacts. Or have you converted to progressive? The sample you gave me was progressive so of course you won't see any interlacing artefacts. My script kept the video interlaced. Perhaps you didn't set the video as interlaced in the encoder you used such as HC Encoder?

Here's my script again:

Mpeg2Source("L:\Original few secs.d2v", CPU=6) 
TTempSmoothF(maxr=3, lthresh=8, cthresh=5, strength=4, interlaced=true)  
AssumeTFF() # or AssumeBFF() depending on the source  

# deinterlace  

Load_Stdcall_plugin("C:\Program Files\AviSynth 2.5\plugins\yadif.dll") 
QTGMC(Preset="Ultra Fast")  
Vinverse()  

SeparateFields() 
SelectEvery(4,0,3) 
Weave()

[jagabo]

I've watched your original MPG video on my WDTV Live and HDTV. There were no significant comb artifacts. Just a little flicker on the horizontal ropes. I see no significant interlacing artifacts on the QTGMC() MKV I uploaded -- In post 21 you said QTGMC() left more artifacts than QTGMC(UF).VInverse().

[VideoFanatic]

So the video in this post before you used a script on it, you're saying you can't see any interlacing artefacts? Intracube says that he can see the interlacing.

[jagabo]

So the video in this post before you used a script on it, you're saying you can't see any interlacing artefacts?

That's the one.