Particular vid looks worse Blu-Ray to HDTV than on computer monitor

[_Al_]

mediainfo or ffprobe read flags in video, they cannot analyze how is video structured correctly,
the best for these DVD's , even if you think it is simple interlaced footage, is to just after creating avisynth script is to add assumetff().separatefields() , then to load it into VD or MPC-HC and step forward to look what video does, if weird, or jumping back, then change it to assumebff().separatefields() and watch again step by step. If neither video is fluid and there are repeated frames or it is screwed up in some way it needs additional fixing , it could be de-delecine and other things ..., dgindex should fix telecine DVD, not sure I do not work with commercial DVD's

[_Al_]

When I looked at the DV file MediaInfo was saying it was BFF even though the VOB is TFF.

DV avi is BFF only.

[LMotlow]

Btw, I'm glad you mentioned this - as it happens as part of the process I've been first converting the VOB into a MainConcept DV .avi file since the deinterlace, resize & reinterlace script was choking if the starting file is HuffYUV. When I looked at the DV file MediaInfo was saying it was BFF even though the VOB is TFF.

Sure enough, when I verified the field order with the internal Virtualdub deinterlacer looking for the back and forth mambo when the field order is set wrong, it verified it as TFF. I hadn't been aware that MediaInfo could be in error like that - good to be aware of. Thanks!

Is there an info utility that's more solid?

I give up, man, sorry I wasted my time here. All you do is find more ways to Phk up video. Don't quit your day job.

[brassplyer]

When I looked at the DV file MediaInfo was saying it was BFF even though the VOB is TFF.

DV avi is BFF only.

Are you saying that's what the spec calls for or that's the only way it can exist? Here's 60 frames of the file - you tell me if it behaves like BFF. When I test it with the onboard VirtualDub deinterlacer and specify TFF I see correctly advancing progressive frames. The video I made that involves deinterlacing and reinterlacing where I specified TFF looks fine - no wrong frame-order waviness.

There's an option in the codec settings for reverse frame order which isn't selected. What you're seeing is how the file got created when I went from VOB to DV.


https://drive.google.com/file/d/0B0zvAZXgfLgiNF9mNzBVc1dRZ0E/view?usp=sharing

[jagabo]

When I looked at the DV file MediaInfo was saying it was BFF even though the VOB is TFF.

DV avi is BFF only.

But there's nothing to stop you from encoding TFF or progressive frames with it.

[_Al_]

I see it now, even Vegas allows to change template to upper field first, op most likely did it overthere

[awgie]

Maybe I'm nuts, and maybe I'm just tired, but...

As I go back and re-read this thread, a couple thoughts strike me...

What exactly is your ultimate goal? Why are you converting a DVD up to a BD? You're starting with a DVD that is of uncertain quality to begin with. It came from VHS, so it's already gone through at least one generation of recoding just to get it onto the DVD. Then you're running it through who knows how many different programs trying to improve it, although every time it's processed, it will tend to lose quality, not gain it. Yes, you can play with settings to try and improve the picture along the way, but it's still kind of like using a Xerox machine and making a copy of a copy of a copy. And if you're upscaling to HD size, you're not going to actually have HD video, especially since it originated from VHS. You're only going to make the poor quality easier to see.

If I were trying to work with VOB files, the first thing I would do is turn it into a compliant MPG file using something like VOB2MPG. If you're trying to process the VOB files directly, it's a crap shoot whether you'll ever get the results you're after, since VOB files are not necessarily what you think they should be. The IFO files on the DVD instruct the DVD player what to play and how to play it. Granted, homemade DVDs are not as convoluted as commercial DVDs (which jumble their VOB files specifically to deter copying), but without using the IFO file, you're still just guessing.

[brassplyer]

Maybe I'm nuts, and maybe I'm just tired, but...

I suppose both are in the realm of plausibility.

What exactly is your ultimate goal? Why are you converting a DVD up to a BD?

More than one goal - the project I'm working on is a Doc Severinsen sampler video for someone, spreading fandom. To that end this one particular segment was presenting some issues. And of course always looking to expand my awareness both out of necessity and curiosity. I've got a bunch of video I want to clean up.

I find that video upscaled to HD size looks better than video burned to DVD. As of my last awareness Youtube gives preferential treatment to video that's been upscaled to HD dimensions.

If I were trying to work with VOB files, the first thing I would do is turn it into a compliant MPG file using something like VOB2MPG. If you're trying to process the VOB files directly, it's a crap shoot whether you'll ever get the results you're after, since VOB files are not necessarily what you think they should be.

What I normally do is rip the whole DVD using Pinnacle Studio which automatically creates an Mpeg. I just happened to be working with a VOB file here because I knew that particular file had the segment I wanted.

[brassplyer]

I give up, man, sorry I wasted my time here. All you do is find more ways to Phk up video. Don't quit your day job.

Off to spread sunshine elsewhere eh?

[awgie]

Maybe I'm nuts, and maybe I'm just tired, but...

I suppose both are in the realm of plausibility.

Touché 😏

I find that video upscaled to HD size looks better than video burned to DVD. As of my last awareness Youtube gives preferential treatment to video that's been upscaled to HD dimensions.

In my own experience - and mind you, I've been editing video for over 30 years - it's an uphill battle to begin with if you're trying to improve any video when the original is already poor. If you're upscaling it as well, it's even more difficult. And each time you re-process it, you push it closer and closer to impossibility.

And YouTube - as an entity - may or may not give preferential treatment of some sort if you upload a video of higher frame size. However, YouTube viewers - as a whole - tend to be rather abusive of uploaders who try to pass off a video that is clearly from VHS as being "HD".

I just happened to be working with a VOB file here because I knew that particular file had the segment I wanted.

Again, if you're trying to process a VOB file directly, you're only guessing that it's a compliant MPEG2 video.

But hey... do whatever floats your boat. And have a happy New Year.

[brassplyer]

In my own experience - and mind you, I've been editing video for over 30 years - it's an uphill battle to begin with if you're trying to improve any video when the original is already poor.

In this case I wouldn't call it poor, I'd call it typical. The guy I got it from told me he recorded it from the broadcast with a consumer VCR and then captured to DVD with no processing other than whatever his software did to convert it to VOB. I'm guessing it was off cable rather than OTA. Obviously seriously downgraded from the original studio tape but in the realm it was done in it's as close to first generation as you could get unless there was a commercial video available and I'm sure there wasn't at the time. There's something there to work with - 8500 bitrate, fair amount of detail all things considered.

[poisondeathray]

And YouTube - as an entity - may or may not give preferential treatment of some sort if you upload a video of higher frame size. However, YouTube viewers - as a whole - tend to be rather abusive of uploaders who try to pass off a video that is clearly from VHS as being "HD".

Youtube does for sure . Comparisons have been posted before. It's one of the very few cases in the video world where upscaling is definitely beneficial. The more tangible benefit is 60p. But you need 720p for 60p treatment . It currently doesn't give higher framerates for SD.

[LMotlow]

I give up, man, sorry I wasted my time here. All you do is find more ways to Phk up video. Don't quit your day job.

Off to spread sunshine elsewhere eh?

No, my good man. Off to look for some. This thread is just repetition. How many times are you going to repeat the same mistakes and create new ones (that's a rhetorical question, I know)?
Off to better information, brassplayer. And better video. A lot of good experience and advice is being wasted here. I do wish you'd pay attention, but you'll be back with a new thread about the same old stuff.

[brassplyer]

Off to spread sunshine elsewhere eh?

No, my good man.

Lem old boy - you're back!

Off to look for some.

By putting in your .02 on a question I asked in a different forum.

So you came back to sign the portrait you painted of yourself as a miserable tool.

[LMotlow]

Your video work is ugly, bud. Always was. Plain and simple. You're a lost cause. YouTube material, for sure, that garbage bin that'ss today's quality standard. You're depressing.

[awgie]

In this case I wouldn't call it poor, I'd call it typical. The guy I got it from told me he recorded it from the broadcast with a consumer VCR and then captured to DVD with no processing other than whatever his software did to convert it to VOB. I'm guessing it was off cable rather than OTA. Obviously seriously downgraded from the original studio tape but in the realm it was done in it's as close to first generation as you could get unless there was a commercial video available and I'm sure there wasn't at the time. There's something there to work with - 8500 bitrate, fair amount of detail all things considered.

Several points:

• Unless you're using extremely good quality commercial (i.e. TV studio grade) equipment, then Broadcast -> VHS -> DVD is going to be poor quality.
• And judging from the thumbnail you posted, it's very poor quality. I'd be truly amazed if it came from cable TV. If that's your idea of "typical" quality, I'd really hate to see what you would consider "poor".
• 8500kbps is totally irrelevant. It's a typical bitrate for a standard DVD, but it doesn't equate to more detail, it merely means more bits are used to show you what little detail is there. I've seen 400kbps videos with a clearer picture than that. Even when you jack it up to 20mbps like you did, you still have to deal with the fact that you started with Broadcast -> VHS, so you're still stuck with a poor quality original that you're trying desperately to enhance artificially.

Every time you re-process a video segment, you will lose detail. Yes, you may enhance the colours or contrast, or deinterlace it (and speaking of which, why would you want to re-interlace the video once you've deinterlaced it? Why not just leave it as progressive?), but you will still lose something along the way. It's the nature of the beast. If you're trying to create a perfect video from that source, you're only setting yourself up for a lot of frustration and disappointment. Do what you can with what you have, but you have to accept that there is only so much you can do.

[brassplyer]

If you're trying to create a perfect video from that source, you're only setting yourself up for a lot of frustration and disappointment.

There's no such thing as "perfect" video. There are varying degrees of not as good as seeing it in real life. The most expensive equipment and software in existence creates video that's a compromise....with hyped color and contrast. 4K looks like there's Vaseline on the lens compared to looking around.

As to why I want to reinterlace it, in part as an exercise in doing it. Also, there's no such thing as 1080 60p on Blu-Ray. FYI SD isn't the only video I've worked with.

Do what you can with what you have, but you have to accept that there is only so much you can do.

Never said or suggested otherwise.

But there are definitely improvements that can be made and I don't presume to be aware of all of them at this point.

8500kbps is totally irrelevant.

You think it doesn't mean there's a lot more information? Such as the quality of the noise for noise removal software to grab onto?

If that's your idea of "typical" quality, I'd really hate to see what you would consider "poor".

Typical in the realm of consumer transfers - I assumed it was understood that's what we're talking about. Presumably you're aware that video can be a lot more degraded than that video. Much of the video people have sent me is multi-generation consumer VCR to consumer VCR using tired machines that have never been cleaned. Tracking glitches, dropouts, smeary detail, audio problems. Getting video that's Broadcast > VHS > high bitrate DVD is actually a lot better than typical.

I've seen 400kbps videos with a clearer picture than that.

Point me to an example of 400kbps SD video that looks clearer at HD dimensions than what the video example I've been talking about can be cleaned up to.

[awgie]

Your video work is ugly, bud. Always was. Plain and simple. You're a lost cause. YouTube material, for sure, that garbage bin that'ss today's quality standard. You're depressing.

I hate to admit it, but I owe you an apology. You're quite right. And it only took one day for me to figure it out. It's like talking to a brick. Or maybe more accurately, like throwing nickels down a gopher hole as if it were a wishing well. The gopher gets all that money, but still doesn't have a clue what to do with it.

Typical in the realm of consumer transfers - I assumed it was understood that's what we're talking about. Presumably you're aware that video can be a lot more degraded than that video. Much of the video people have sent me is multi-generation consumer VCR to consumer VCR using tired machines that have never been cleaned. Tracking glitches, dropouts, smeary detail, audio problems. Getting video that's Broadcast > VHS > high bitrate DVD is actually a lot better than typical.

Best of luck in your fruitless endeavors, brassplyer, but I'm not throwing you any more nickels. You clearly think you're smarter than me, and I'm not going to argue with you any longer.

[brassplyer]

I'm not throwing you any more nickels.

Can't say you've offered any solid information here, not even a nickel's worth. A few have offered genuinely helpful input - what you've done is a bunch of naysaying. Your so-called contributions have amounted to "why are you bothering to try and clean up video?" Consumer SD transfers with consumer gear won't look as good as pro video watched on pro gear - you don't say?

You haven't been any more helpful than Lemmy the lout.

Now you're all butthurt because I questioned some things you've said that I don't believe are correct. By all means show me 400kbps SD video that looks clearer at HD dimensions than what the example video can clean up to. I don't think you can. If you can back that claim up I'll be the first to say I was wrong.

I'm not going to argue with you any longer.

So what we've learned from awgie today is that bitrate makes no difference, is "totally irrelevant". Okee doke.

[davejavu]

No, what we have learned it that have already made your mind up and are wasting peoples time asking for help and then throwing it back in their faces when it doesn't tell you what you want to hear. If you start with garbage, sure you can probably improve it a bit, but you will still end up with something that is just less smelly garbage. Something about pigs ears and silk purses comes to mind here.

Just as an experiment I took one of my over the air DVB recordings, equivalent to a DVD with Mpeg-2 video at 720x576 anamorphic widescreen starting at about 3000kbps, rencoded it to x264 using 2-pass at 400kbps SD and then took that new video but rescaled it up to HD without any bitrate constraints. The original SD broadcast wasn't the greatest quality to start with and the result is fuzzy/blurry, but it still looks better than the POS you're working with.

[brassplyer]

No, what we have learned it that have already made your mind up and are wasting peoples time asking for help and then throwing it back in their faces when it doesn't tell you what you want to hear.

I don't have my ego tied up in the fact that the "POS I'm working with" is what I have to work with. People insulting the file just sound stupid.

I asked a specific question - what are reasons this file doesn't look the same on a TV as it does on a monitor. At no time did I ever say anything even remotely resembling "Hey how do I make a noisy SD video file look like HD video shot with a pro HD camera?" Saying I'm burning it to Blu-Ray isn't the same as saying I expect it to look like HD Blu-Ray. What I said is I want to retain as much of what quality is there as possible and I like the idea of me doing the upscaling instead of the hardware doing the upscaling. For some reason this apparently bothers awgie's sensibilities.

What I "want to hear" are any specifics as to how a TV screen might process an image differently than a computer monitor. A couple of people have offered suggestions and other helpful information and I appreciate it.

If you think Motlow wading in here with his informationless haranguing BS because he has a burr up his wazoo and has nothing better to do is fine, you have a very different perspective on the universe than I do. Awgie saying things like "bitrate is irrelevant" is just wrong. A higher bitrate copy of a noisy VHS is better for cleanup purposes because the reproduction of the noise is higher quality.

As for your test files - are you watching it on a cell phone? The heck it looks better. Viewing it on both a 22" monitor and a 42" HDTV I see glaringly obvious blocking, aliasing, no edge detail, fuzzy halos around moving elements. Might that be related to, oh I dunno, bitrate?

Screenshots of both at max zoom as viewed on Media Player Classic. This is your higher bitrate 1080p vs my 720p version. Your 400kbps version certainly isn't an improvement over the higher bitrate version. View both at max zoom and tell me it looks better. Even at thumbnail size you can see the issues on yours.

[davejavu]

What I "want to hear" are any specifics as to how a TV screen might process an image differently than a computer monitor. A couple of people have offered suggestions and other helpful information and I appreciate it.

OK. There are dozens of possible reasons for a difference but here a just a few of the ones I would have looked at.

----------

For the TV:

Is the TV 'FullHD' or 'HD Ready' and is the playback device sending a signal at the native resolution of the TV (if it isn't then the TV is rescaling even if there isn't anything in the settings saying that)

Is the TV rescaling or reprocessing the video after it is being fed from the playback device in addition to scaling to the native resolution.

Is the playback device applying any processing/scaling before sending it to the TV.

Are the cables connecting the playback device to the TV working properly/damaged/faulty.

Is the playback device feeding a colour space signal into a port on the TV that only accepts a different colour space (RGB in S-Video or any multitude of combinations of output/input thereof)

Is the playback device working properly.

Is the TV calibrated properly or is it at factory defaults, etc.

For the computer:

What is the actualy physical resolution of the monitor and what resolution is the OS/graphics card set to

Is the monitor calibrated properly

Is it applying any sort of 'enhancement' via a pre-defined operating mode built into the monitor.

Is the playback software/graphics card drivers applying any sort of processing.

What colour space is the computer/playback software/graphics card set to, and to they match.

Are both sets of hardware set to read, process and output the video in exactly the same way. If not then you will obviously get variations in appearance on each. Even if they have exactly the same settings the 2 different screens will probably look different depending on the quality of the device/manufacture, and there isn't much you can do about that but adjust both until they look the same.

The computer monitor and TV screen may differ in physical size but have the same number of real pixels making up the image, which give the same effect as zooming in or blowing up the image whcih, as you noted in your post above, just helps to exagerate any faults in the picture.

----------

As for the quality of my samples, you are right that they are rubbish which I stated quite plainly when I posted them. My 'HD' sample was created by taking the fuzzy 400kbps SD version and rescaling it up to an HD resolution without any image processing or cleaning/enhancement filters, so if course it still looks bad. I wouldn't want either in my permanent video collection, but still don't think mine is any worse to watch than yours. It is just bad for a different reason. Your video might have fewer artifacts/blocks but has been so heavily processed it looks indistict and smeared, almost like a posterising filter in photoshop.

The point that doesn't seem to have gotten accross is that you should aim to process/reencode video as little as possible as every time you do this you lose information and risk degrading quality, and that you sometimes have no choice than to accept a compromised end result from a poor quality source. No amoiunt of bitrate or filtering is going to turn a poor quality video source into anything other that a slightly less poor quality video.

[usually_quiet]

I don't know if anybody else experienced this, but the uploaded clip from the original VTS_01_2.VOB file shows a run time of about 56.9 seconds but only about 22 seconds of it played. What happened to the rest of it? According to MediaInfo, the uploaded clip from the original VTS_01_2.VOB file is 24.5 MiB and about 56.9 seconds long. The overall bitrate is 3611 Kbps. Based on the reported bitrate and a 56.9 sec run time, the reported file size looks right.

After playing the first 8 seconds showing the band, Potplayer's progress bar jumps to 42 seconds, where Doc begins playing and continues to play normally from that point on. MPC-HC plays 22 seconds of video, about 8 seconds showing the band, immediately followed by 14 seconds showing Doc, then skips to 56 seconds. PowerDVD refused to play the clip. VLC only played the first few seconds.

The editors I tried also show discontinuities. When scrolling through the video, MPEG-VCR jumps from frame 254 to frame 1284. VideoReDo's and Mpg2Cut2's time display show a frame at about 8 seconds then skips to about 42 seconds when I advanced to the next frame. I re-exported the clip from VideoReDo as an mpg, and it shrank to 22.7 MiB and the overall bitrate increased to 8713 Kbps. VideoReDo removed 5 bad video frames, and re-encoded only the GOPs adjacent to the removed frames.

[brassplyer]

There are dozens of possible reasons for a difference but here a just a few of the ones I would have looked at.

----------

For the TV:

Is the TV 'FullHD' or 'HD Ready' and is the playback device sending a signal at the native resolution of the TV (if it isn't then the TV is rescaling even if there isn't anything in the settings saying that)

Is the TV rescaling or reprocessing the video after it is being fed from the playback device in addition to scaling to the native resolution.

Etc.

Good list of things to check - thanks!

[brassplyer]

I don't know if anybody else experienced this, but the uploaded clip from the original VTS_01_2.VOB file shows a run time of about 56.9 seconds but only about 22 seconds of it played. What happened to the rest of it? According to MediaInfo, the uploaded clip from the original VTS_01_2.VOB file is 24.5 MiB and about 56.9 seconds long. The overall bitrate is 3611 Kbps. Based on the reported bitrate and a 56.9 sec run time, the reported file size looks right.

They're sections cut out using Mpg2Cut2 which I first heard of maybe 10 minutes before I uploaded the file. I assume the video retained original frame numbers, maybe original header info?

[awgie]

I don't know if anybody else experienced this, but the uploaded clip from the original VTS_01_2.VOB file shows a run time of about 56.9 seconds but only about 22 seconds of it played. What happened to the rest of it? According to MediaInfo, the uploaded clip from the original VTS_01_2.VOB file is 24.5 MiB and about 56.9 seconds long. The overall bitrate is 3611 Kbps. Based on the reported bitrate and a 56.9 sec run time, the reported file size looks right.

They're sections cut out using Mpg2Cut2 which I first heard of maybe 10 minutes before I uploaded the file. I assume the video retained original frame numbers, maybe original header info?

Incidentally, this is an example of one reason you don't work directly with VOB files.

[jagabo]

As was pointed out, VOB files often aren't parsed correctly without the IFO files.

[brassplyer]

Incidentally, this is an example of one reason you don't work directly with VOB files.

Not my normal process, I did it here probably for the first time ever to save time rather rip the whole DVD, and also to have an example of the raw VOB.

In your PM you cautioned against using Pinnacle. I'm not in love with it and don't have an objection to using something else - other than having paid good money for it and like getting *some* use out of it. But for technical clarification my assumption was that it just strings the VOB files together and labels it as a continuous mpeg without doing anything to the actual audio/video data. You feel this isn't a safe assumption?

You also mention using Vegas to do everything. The version of NeatVideo I have is specifically for Virtualdub so that's what I use. I find that it's a lot more reliable to do only NeatVideo during that pass. I also find it good to have the cleaned file as a base to then test other things out on. Also, I use QTGMC for deinterlacing - all the Avisynth stuff I've done has been through Virtualdub. If it's possible to run Avisynth through Vegas it's not something I've done or explored. Another assumption is that QTGMC is about as good as it gets - I know it beats the heck out of anything that's onboard VirtualDub. In fact that's how I got into fooling around with Avisynth in the first place, I wasn't satisfied with the deinterlacing in Virtualdub. I don't recall if I've ever even tried the onboard Vegas deinterlacing.

You and others have mentioned degradation of the video going through multiple passes. I mostly use HuffYUV. I've also tried Lagarith - I thought they're lossless codecs that shouldn't be degraded by more than one processing stage - is that not the way it is?

[jagabo]

In your PM you cautioned against using Pinnacle. I'm not in love with it and don't have an objection to using something else - other than having paid good money for it and like getting *some* use out of it. But for technical clarification my assumption was that it just strings the VOB files together and labels it as a continuous mpeg without doing anything to the actual audio/video data. You feel this isn't a safe assumption?

I don't know if Pinnacle does that or not. But it doesn't matter. VOB files are not necessarily accessed sequentially. The IFO file controls the order data is accessed in the VOB files. It's possible for a single title to jump around in the VOB set. This is why you sometimes have to use a program like VOB2MPG to extract titles from the VOB/IFO set.

For example, I have one commercial DVD that has two titles interleaved withing the VOB set. A few seconds of the first title, a few seconds of the second title, a few seconds of the first title, etc. The only way to de-interleave the two titles is to use VOB2MPG (or some other IFO cognizant extractor).

[JVRaines]

Also, I use QTGMC for deinterlacing - all the Avisynth stuff I've done has been through Virtualdub. If it's possible to run Avisynth through Vegas it's not something I've done or explored.

The Avisynth Virtual File System (AVFS) is a very cool package that I use to connect Vegas to Avisynth. You just double-click an AVS script and it launches as a virtual media file in the C:\Volumes path. Then you can put it on the Vegas timeline. A bonus is that you don't have to worry about mixing 32- and 64-bit apps. It just works.