#1  
09-15-2005, 01:34 AM
bosteis bosteis is offline
Premium Member
 
Join Date: Sep 2005
Posts: 16
Thanked 1 Time in 1 Post
I see at http://www.digitalfaq.com/dvdguides/...timpeg/atimpeg. htm that you suggest to capture at 352x480. isn't DVD MPEG-2 quality at 720x480? is there benefit to capturing at the best possible quality (as a general rule/best practice) so that you're always editing with a higher quality file (b/c you can always compress the video to a lower quality later)? or, is it just better to capture using the settings you've suggested b/c it saves steps and time later and saves hard drive space in general?

also, is it best to capture as avi? I know you've said that avi is best for projects where you know you're doing heavy effects and editing. could you give examples of these kinds of projects?
Reply With Quote
Someday, 12:01 PM
admin's Avatar
Ads / Sponsors
 
Join Date: ∞
Posts: 42
Thanks: ∞
Thanked 42 Times in 42 Posts
  #2  
09-15-2005, 06:08 AM
admin's Avatar
admin admin is offline
Site Staff | Web Development
 
Join Date: Jul 2003
Posts: 4,334
Thanked 644 Times in 454 Posts
There really is no such thing as "DVD quality". That's a term often misused. A DVD, specifically the DVD-Video format, is merely a digital video standard with a wide range of options. I've seen a great many DVD's that contained essentially butchered video due to negligence or lack of care/knowledge by the author. We're talking about something so messed up that a grandma (one who cannot even set the clock away from blinking 12:00) with a decades-old VCR could have created a higher quality video on a VHS tape.

The DVD-Video specifications (often called the "DVD spec") is a range of options. Resolutions, bitrates, encoding styles... a lot of leeway and options.

As shown on this page http://www.digitalfaq.com/dvdguides/...tandsource.htm, most sources are lower than most people think. And then the DVD spec is laid out quickly on this page http://www.digitalfaq.com/dvdguides/...tandsource.htm

The NTSC DVD spec allows for 352x240 (SIF), 352x480 (Half D1), 704x480 (Broadcast D1), and 720x480 (Full D1). USA is NTSC video, so I'm not listing PAL resolutions (unless you need them). The spec also speficies bitrates, audio formats, etc... lot of variables are controlled. Even MPEG-1 is supported on the lowest res, not just MPEG-2.

Now, in general, if you're using an MPEG capture method, with the intention of doing minor editing, go for the desired resolution right away.

"Minor" editing is basically digital versions of scissors and tape, cutting off stuff you don't want, and taping the rest together into a new file. Removing commercials from tv recordings is an example.

You can also add very minor fade-in's and fade-out's, maybe a transition or two. Simple stuff. With the correct software, the MPEG is not re-encoded to cut out unwanted footage, and only the small segment that is transitioned or faded is re-encoded with a minimal loss. Although AVI work may still yield better quality on anything more complex than fade-in/fade-out.

"Major" editing is what I do when I create advanced motion menus for DVDs. Joining many clips, crossfades between clips, add matte images over portions of the video. Complex work that involves lots of filters, tools, and steps.

Major editing is where AVI is needed, and where high res captures are suggested. You can "encode down" later on if desired. (The war here is encoding down may lose the sharpness of newly rendered content (the images, text, etc), but encoding down to the resolution of the source will give it a uniform look rather than having sharp/detailed items appearing on top of a rather muddled lower res VHS source.)

Restoring video is also a situation where you want to capture high resolution and encode down later on (maybe). There are pieces of hardware out there that can increase the resolution and detail a bit, so it's being restored and enhanced greater than the source, so higher res is then desireable.

Converting interlace to progressive video is another area where higher resolution is often desireable, to minimize artifacts from the IVTC or adaptive methods. This is not something that need be done often, in general leave interlaced video as interlaced.

.
.
.

Now that's the "simple" answer.

The truth of the matter about resolutions captures, for those that are very picky (usually those who own a 60" tv screen in HD and stand 2 feet away from it), is that resolution differences in quality can also depend largely on the capture card, the editing/encoding software, and even the DVD player used. While 352x480 is technically at/above VHS/tv resolution, and will absolutely maintain all of it's clarity, some of the capture chips (BT8x8, for example) are unable to process the lower res at good quality. DVD player MPEG decoder chips can also perform better at the higher res, mostly the cheapo units you get at Walmart/etc for $30-40, as 720x480 is the "standard" of Hollywood (because studios have high grade film source and use the max resolution), so little care is given to the rest of the DVD spec.

Most people need not worry about this more complex answer, as sensible folks tend to sit 10+ feet away from a tv of normal size and tend to buy nicer DVD players. The hard part with them is always getting them to buy a good capture card or DVD recorder, not the cheapo or "popular" junk that performs poorly (such as $50 "on sale" capture cards, or the Panasonic DVD recorders).

.
.
.

That was a lot of info. If it caused any confusion, just ask for clarifications.

- Did this site help you? Then upgrade to Premium Member and show your support!
- Also: Like Us on Facebook for special DVD/Blu-ray news and deals!
Reply With Quote
  #3  
09-16-2005, 08:59 AM
bosteis bosteis is offline
Premium Member
 
Join Date: Sep 2005
Posts: 16
Thanked 1 Time in 1 Post
someone I work with just scared me by saying, "oh no, definately capture at 720x480. you will notice the difference, especially as TVs go to high definition over time. basically, TV is a bunch of rows of light with black rows inbetween. when you cut it in half, you're relying on the TV to make the DVD look good."

personally, I don't trust this statement. is there truth to any of that comment?

this prompts other questions...

[1] what resolution are store bought DVDs at?

[2] if I want to capture at 720x480 to archive it at high quality, is that a bad thing? what settings in MMC do you recommend for that setting?

[3] I captured something on TV using the default ATI DVD MPEG-2 setting. what do I need to do to make 110 minutes of video captured at that setting fit on 1 DVD?

[4] how much video will fit on 1 DVD with 720x480 and 352x480 video?
Reply With Quote
  #4  
09-19-2005, 11:33 AM
admin's Avatar
admin admin is offline
Site Staff | Web Development
 
Join Date: Jul 2003
Posts: 4,334
Thanked 644 Times in 454 Posts
"definitely capture at 720x480"

You're wise to not trust that statement. Because it's simply not true. That sort of statement is born from using bad methods in most cases (and/or inferior hardware).

The truth to resolution capturing is that you want to, at minimum, merely retain what already exists. If you look at the resolution charts on this site, it shows the resolutions for many sources, including that of broadcast, cable, satellite, VHS, and a number of others. Most traditional home sources are medium resolution, and will likely stay this way for a long time.

"Retain what already exists" is usually the hard part for most people. You have to be sure you have hardware and software that is adequate to perform this function. More on this in a minute...


HDTV:

The truth about HD technology (used in HDTVs) is that most channels are still SD quality over HD bandwidth. Only some select newscasts, sports events or other major events are going to be HD. Some movie channels are using ED

HD = high definition (sort of a "super resolution" right now)
ED = enhanced definition (720x480 Full D1 would be here, modern "high resolution")
SD = standard definition (Half D1 and others, modern "medium resolution")

It need be stressed again that you can only view as great as the least common denominator. For example, if you record a HDTV broadcast on VHS, it'll never look better than VHS resolution, even if viewed on a HDTV again. Even transferring from VHS to DVD, it would still look like VHS (unless you add sharpness with a hardware detailer or res booster, at the time of capture, but even then, not anywhere near HD). Once detail is lost, it cannot be regained. Same is true for detail that never existed to begin with.

In other words, don't worry about HD. Leave it to the future. What you have now will always be the same resolution.


How to retain resolution

Getting back to retaining resolution, you need good hardware. Capture cards have chips that grab video at a hardware-specified size and then sample over to the requested resolution. Most of them contain smart hardware tech that do a great job. As long as you request a resolution that is equal to or greater than the source, the video capture should be relatively transparent to the source, in terms of detail. Nothing should be lost.

Some people insist on maximum captures and sampling down later in software or leaving it at max resolution.
- High res on low res source accomplishes nothing but wasted disc space, meaning you have to put only 60-90 minutes on a disc before you start to compress. A Half D1 resolution can do about 120-180 minutes before you start to really compress.
- Software downsizes are often no very smart, and you will lose detail and sharpness. Bicubic and bilinear resize methods are the ones used most, neither of them a smart methods (like lanzcos and others).
- Upsizing video from medium res to high res can also increase the visibility of flaws in the source. Therefore you lose more quality, rather than gain it.

If that got a bit to techie, I apologize. In layman's terms, buy a good card. Good cards can capture any medium or high resolution with no visible flaws. So the need to "capture as high as you can" is a voided and false theory.

An ATI All In Wonder card is an excellent capture card. It does a beautiful job at 704x480 and 352x480 captures. I prefer 704x480 over 720x480, for this card. It's seemingly closer to the native chip size.




[1] what resolution are store bought DVDs at?

It depends. Many of your "big name" movies are 720x480, always have been. Other movies can go both ways, at least on the "old days" of DVD, not so much now. Discs are cheaper to make now, so they go the extra mile. TV shows are really split in half. Some are Full D1, some are not. Many of your imported non-USA discs (which are still awesome quality) are Half D1. The only reason Full D1 is good on movies is (A) plenty of space on a disc, and (B) the studio film/digital sources are high resolution these days.

Be we at home do not have that kind of source. It's all about the source.


[2] if I want to capture at 720x480 to archive it at high quality, is that a bad thing? what settings in MMC do you recommend for that setting?

It can be. As discussed earlier. Upsizing can be as bad as downsizing. And it requires higher bitrates to look good. I suggest 704x480 (or 720x480 if you want) and a bitrate of 6000-8000k VBR. No less, higher is better. No more than 8000, however. Be sure to read the MPEG bitrate allocation charts on this site.


[3] I captured something on TV using the default ATI DVD MPEG-2 setting. what do I need to do to make 110 minutes of video captured at that setting fit on 1 DVD?

Well, by using the default, you lost interlace. So motion may be jerky, and you will likely have some jaggy edges (aliasing) on straight lines that skew away from 90 or 180 degrees perfectly straight (perpendicular lines flat to the horizon). If these jaggies are bothersome, we can re-encode the video with filters. That software filter will soften the video slightly, while sampling the jaggies together, an anti-aliasing effect.

If it looks alright as is, just use it. Author to a disc, you're done, easy as that. And don't do that again. A mistake I've made once or twice, kick myself for it.

The bitrate needed for 110 minutes is basically 5500k VBR with a max of about 7000k. That should fit somewhere just under 120 minutes. You've got a little bit of play, always guess down, never get too close, as you may have to try again because the encode is too big. Wasted time.


[4] how much video will fit on 1 DVD with 720x480 and 352x480 video?

- In high quality, and a good capture card/encoder, you can put about 60-90 minutes on a DVD in 720x480.
- If you apply some more compression, generally 2-2 hours is alright.

- In high quality, and a good capture card/encoder, you can put about 2-3 hours on a DVD in 352x480.
- If you apply some more compression, generally 3-4 hours is alright.

You never want more than 4 hours on a single layer disc. The bitrates needed for that res generally drop it down to non-interlaced 352x240, and that's no good.

- Did this site help you? Then upgrade to Premium Member and show your support!
- Also: Like Us on Facebook for special DVD/Blu-ray news and deals!
Reply With Quote
  #5  
09-24-2005, 05:40 AM
bosteis bosteis is offline
Premium Member
 
Join Date: Sep 2005
Posts: 16
Thanked 1 Time in 1 Post
no, you didn't lose me in tech talk. same rule applies for devices on a SCSI chain... you're only as fast as your slowest device.

concerning the stuff I captured at the default ATI DVD setting, I am not fussy right now about how the final product will come out. I am most concerned about making 110 minutes of video at this setting fit onto 1 DVD so I can let my bandmates see the show. I can live with whatever outcome. it doesn't need to be perfect.

2 questions:

[1] even though it was set at 720x480, it looks like the card captured 704x480 of video as specified by the chipset into a 720x480 file and therefore created a gutter on the right side. does this look true as per this screenshot?




[2] how do I convert this video properly to half D1 or 352x480 so it will fit? I'm guessing I would do this in TMPGEnc, right? should I edit out commercials first and then convert after? here is a GSpot screenshot of the file:

Reply With Quote
  #6  
09-28-2005, 10:29 AM
bosteis bosteis is offline
Premium Member
 
Join Date: Sep 2005
Posts: 16
Thanked 1 Time in 1 Post
so I tried a new set of captures for this week. are these settings correct?

* Format = MPEG-2 (I didn't use MPEG-2 DVD b/c you used it for a cartoon capture example, so I thought it was less robust. should I be using MPEG-2 DVD instead with "Record Cropped Video" also enabled?)
* Standard = NTSC (525)
* Resolution = 352x480
* De-interlacing Options = Encode Interlaced
* Inverse 3:2 Pulldown = disabled
* Record Cropped Video = disabled
* Audio Format = 48.000 KHz, 16 Bit, Stereo
* GOP = 4-2
* Closed GOP = enabled
* Video Encoding Parameters:
* VBR
* Max Bit Rate = 4.00 Mb/S
* Target Bit Rate = 3.90 Mb/S
* Motion Estimation Quality = 99
* Audio Encoding Parameters = 256 Kb/S
* VideoSoap = None
* Max File Size = 4 GB (I use FAT32 partitions)

what's irking me is that I'm still seeing a gutter on the right side of the video. is that supposed to be there? I'm assuming that it's supposed to look squashed in Windows Media Player b/c it's at 352x480, not 720x480. for some reason, I watched a DVD using Windows Media Player recently and now there's always a black gutter underneath any video I watch. I can live with that. some glitch in Windows Media Player I guess.

I hoping that I didn't screw up the settings or that something is going on with my drivers or hardware. if you tell me that the gutter on the right side of the video is ok, I will be relieved, but I'm suspicious.


here is a screenshot of the recent video:




and here is a screenshot of GSpot on that same video:




thanks!
Reply With Quote
  #7  
09-28-2005, 02:44 PM
admin's Avatar
admin admin is offline
Site Staff | Web Development
 
Join Date: Jul 2003
Posts: 4,334
Thanked 644 Times in 454 Posts
You've made two more posts since my last reply, so I'll address each one separately. This is the reply for the first one....

The very first capture I ever made on my ATI AIW card was on the morning of Sept 11, 2001. I didn't pay attention to settings, I turned on the computer, picked a setting that looked good, and hit capture. Four hours later my hard drive was full of MSNBC coverage and the capture stopped. I started to capture a few minutes after the first tower collapsed. The computer was a few days old and I had not even used the ATI card yet. I had maybe turned it on twice before that, mostly to load some office and photo software. The video was DVD compliant, as I selected one of the DVD templates, and it looks fine. There are a handful of dropped frames at the beginning of the captures (it was segmented into 4GB chunks on a FAT32 drive), and some obvious deinterlacing, but it's fine. The thing about cartoons and graphics is deinterlacing is very noticeable, while live action is not so bad (but still not optimal). Basically, I just told you all this to assure you what you have will be fine, it's not perfect, but it's far from being ruined.


Q: [1] even though it was set at 720x480, it looks like the card captured 704x480 of video as specified by the chipset into a 720x480 file and therefore created a gutter on the right side. does this look true as per this screenshot?

A: Broadcast is sized for tv viewing. Remember the concept of overscan (which is explained in the UNDERSTANDING YOUR SOURCE guide on this site). Broadcast is sized to fit the screen, not the full frame which has a 7% loss of image shown on screen. This is really outside a concept of resolution, it's merely occupying a small portion of the frame.

Q: [2] how do I convert this video properly to half D1 or 352x480 so it will fit? I'm guessing I would do this in TMPGEnc, right? should I edit out commercials first and then convert after? here is a GSpot screenshot of the file:

A: Don't. What you're showing me right there is perfect. It's less than 4.38GB, so it fits on a single disc. And then it has DVD-Video compliant specs. There is nothing at all wrong with it. Personally, if this were my project, I'd drop the file into Tsunami DVD Author, make a menu-less disc, and add chapters (either randomly or at specified timeframes, like the beginning of songs or natural breaks in scenes). You're set to go, re-encoding is neither necessary nor suggested.



- Did this site help you? Then upgrade to Premium Member and show your support!
- Also: Like Us on Facebook for special DVD/Blu-ray news and deals!
Reply With Quote
  #8  
09-28-2005, 03:29 PM
admin's Avatar
admin admin is offline
Site Staff | Web Development
 
Join Date: Jul 2003
Posts: 4,334
Thanked 644 Times in 454 Posts
And now for the second post reply...

On the format, when dealing with off-air broadcasts, you have to make a personal choice.
- There is nothing really different about cartoons or live action. It's still video. The only observations I would make here is that deinterlace artifacts are more obvious on the clean linear style of a cartoon. And then because it has large areas is identical color, you can often get away with using a slightly lesser bitrate, maybe 500k or so (half megabit) to achieve the same visible cleanliness of a live capture. As far as using the MPEG-2 or MPEG-2 DVD setting, still just video, no consideration of content.
- The MPEG-2 DVD setting is a forced DVD-compliant resolution. You're not allowed to use a non-DVD res, and the canvas will be painted black if you crop the video (which effectively crops out the overscan). The only real reason to crop out the overscan is for viewing on a device with no overscan (watching DVDs on a computer), or for eliminating noisy portions of the image that suck up your bitrate. Blacking it out will allow more bitrate to be used on the actual important image area of the video frame. With a tv broadcast, you don't have much noise in the overscan (although CC noise at the top of the screen can still steal bitrate bandwidth). ON a VHS tape, your overscan is completely saturated with noise most of the time, so you'd certainly want to crop it.

So that setting is mainly intended to be used with the crop settings activated.

This is where the choice comes into play. For me, recording the overscan on off-air broadcasts is more of a morbid curiosity than anything else. I like to see if the cameraman was on the ball or not, see what may have been cropped out of the frame and pushed into the overscan. Aside from something like this, there's really no reason to record the overscan, it won't be shown on a tv anyway.

Everything else about your file is perfect. I'd be surprised if you had anything other than a clean encode that looked great.

You've done well. Author and burn.






- Did this site help you? Then upgrade to Premium Member and show your support!
- Also: Like Us on Facebook for special DVD/Blu-ray news and deals!
Reply With Quote
  #9  
09-29-2005, 04:34 PM
bosteis bosteis is offline
Premium Member
 
Join Date: Sep 2005
Posts: 16
Thanked 1 Time in 1 Post
Quote:
Originally Posted by admin
Q: [2] how do I convert this video properly to half D1 or 352x480 so it will fit? I'm guessing I would do this in TMPGEnc, right? should I edit out commercials first and then convert after? here is a GSpot screenshot of the file:

A: Don't. What you're showing me right there is perfect. It's less than 4.38GB, so it fits on a single disc. And then it has DVD-Video compliant specs. There is nothing at all wrong with it. Personally, if this were my project, I'd drop the file into Tsunami DVD Author, make a menu-less disc, and add chapters (either randomly or at specified timeframes, like the beginning of songs or natural breaks in scenes). You're set to go, re-encoding is neither necessary nor suggested.

I know the file is less than 4.38GB, but I'd like to fit 4 more shows at this size on 1 DVD. that's why I'd like to re-encode them to 352x480 so they'll fit. how should I go about doing that?
Reply With Quote
  #10  
09-29-2005, 06:47 PM
admin's Avatar
admin admin is offline
Site Staff | Web Development
 
Join Date: Jul 2003
Posts: 4,334
Thanked 644 Times in 454 Posts
Ah, now that's different.
Which encoder are you planning to use?


- Did this site help you? Then upgrade to Premium Member and show your support!
- Also: Like Us on Facebook for special DVD/Blu-ray news and deals!
Reply With Quote
  #11  
09-30-2005, 01:47 AM
bosteis bosteis is offline
Premium Member
 
Join Date: Sep 2005
Posts: 16
Thanked 1 Time in 1 Post
I'll use whatever encoder you recommend.

the latest batch of shows I've recorded are at the 352x480 setting as per your instructions for how to set it up through ATI MMC. I guess those settings would be optimal as well.

[1] in general, is it best to re-encode first (if you need to) and then edit out commercials, or the other way around?

[2] it looks like you recommend Womble instead of TMPGEnc for editing b/c of instability, right? this forum thread from mitch kinda worries me. should I be worried? was it b/c he was using "MPEG Video Wizard" when he should've only used "Womble MPEG2VCR"? I think if you buy "MPEG Video Wizard", they give you "Womble MPEG2VCR" for free.
- bosteis
Reply With Quote
  #12  
10-04-2005, 05:38 AM
admin's Avatar
admin admin is offline
Site Staff | Web Development
 
Join Date: Jul 2003
Posts: 4,334
Thanked 644 Times in 454 Posts
Sorry for the slow response on this, lot going on past few days.

The encoder I suggest as being the very best is probably too expensive for most people. I highly recommend CANOPUS PROCODER, but it runs about $500 or so.

Failing that, TMPGENC PLUS is a good choice for quality, options, and price (only $37). So we'll look at using that.

Yes, those ATI MMC settings are still relevant. Good DVD video is reliant on a good MPEG encoder, proper settings, and good bitrate allocation.

For 352x480 live action, a bitrate of 3500-4500k would be optimal. So a 4000k average and a 4500k max should work fine. Since we'll look at using TMPGENC for the encoding, probably be safe to set a minimum of 2000l also.

This guide here should suffice for most of the setting information:
http://www.digitalfaq.com/dvdguides/...mpgenc/tmpgenc. htm
If anything is confusing, just ask me to clarify for you. I would also refrain from guessing at settings, never be afraid to ask for verification on something.

In TMPGENC, to get the bitrates I suggested, just use math. It is about a 90% CQ setting of 4500k. This means the average bitrate is about 90% of 4500k, or in plain numbers, about 4000k. TMPG is CQ optimized, so it will often look better than CBR or VBR. The only catch is it will not be able to calculate a precise file size, but it'll be close.

[1] in general, is it best to re-encode first (if you need to) and then edit out commercials, or the other way around?

Very good question. It depends on your source conditions. In this case here, your source is a clean MPEG file, not corrupted. Edit first. This will remove unneeded junk and then make your encode time less when you go to make the new file.

[2] it looks like you recommend Womble instead of TMPGEnc for editing b/c of instability, right? this forum thread from mitch kinda worries me. should I be worried? was it b/c he was using "MPEG Video Wizard" when he should've only used "Womble MPEG2VCR"? I think if you buy "MPEG Video Wizard", they give you "Womble MPEG2VCR" for free.

Not as much stability as precision. TMPG is not a very good cut tool. Womble is accurate to the frame.

I really don't see what could have happened to mitch there. I feel as if I'm missing a key piece to the puzzle. It sounds like the PTS (timestamps) was corrupted either in or before the editor. It amy very well have been MVW doing it. I have only used MVW 3-4 times total, as I don't like it near as much as Womble MPEG VCR. I prefer VideoRedo over MVW also. MPEG-VCR is my favorite, hands down. VideoRedo has trouble with video files sometimes when it really should not, requiring the fixstream step. I can edit in MPEG-VCR and then author with no issues whatsoever.

I've never heard about the buy one get other one free deal from Womble, but I've not paid attention to their pricing or promos lately either.


I think this covers all you asked....



- Did this site help you? Then upgrade to Premium Member and show your support!
- Also: Like Us on Facebook for special DVD/Blu-ray news and deals!
Reply With Quote
  #13  
10-06-2005, 07:15 PM
bosteis bosteis is offline
Premium Member
 
Join Date: Sep 2005
Posts: 16
Thanked 1 Time in 1 Post
some more questions (thanks for answering them)...

[1] it looks like capturing with the MPEG-DVD 2 setting and cropping is a best practice for me b/c I don't care about overscan. BUT, if I captured with overscan left in and author it to DVD, will it be visible if I view the DVD on a computer (I understand that it won't be visible on a TV)? will it be visible if rip the same DVD later? seems like another reason to capture cropped, then I'll never have to worry about it. ALSO, how do I safely crop out overscan if I've already encoded a file with it left in?

[2] you've said there is no benefit to capturing higher than your source, but what if you want to encode a 352x480 video for viewing on a computer and wanted to blow it up to a higher resolution than 352x240?

[3] is the reason you recommend capturing at 352x480 instead of 352x240 b/c it preserves interlacing better by doubling the height of the video?

[4] you said in Introduction to Video Capturing that you gain no benefits from capturing at AVI only to encode to MPEG later, unless you're doing "advanced editing". how advanced is advanced? can TDA do fade outs and simple text overlays? should I only be using Premiere instead? does it matter that much if I'm doing simple editing like that with MPEG-2?

[5] my brother does his work in Final Cut Pro on the Mac and gives me QuickTime files. if I want to convert these files to other formats, do I need to use QuickTime or can TMPGEnc do that for me if I have QuickTime already installed on my computer?

Reply With Quote
  #14  
10-06-2005, 08:22 PM
admin's Avatar
admin admin is offline
Site Staff | Web Development
 
Join Date: Jul 2003
Posts: 4,334
Thanked 644 Times in 454 Posts
>> [1] it looks like capturing with the MPEG-DVD 2 setting and cropping is a best practice for me b/c I don't care about overscan. BUT, if I captured with overscan left in and author it to DVD, will it be visible if I view the DVD on a computer (I understand that it won't be visible on a TV)? will it be visible if rip the same DVD later? seems like another reason to capture cropped, then I'll never have to worry about it.

If you capture overscan, you will see it on the monitor. If you mask (which is when you cover it up with black) the video, you will simply see black edges around the video. This is far preferred to noise.


>> ALSO, how do I safely crop out overscan if I've already encoded a file with it left in?

You have to be smart about this one. Realize right now most people use the wrong terms all too often. So settings in software may do different than the word means.

Part of the UNDERSTANDING YOUR SOURCE guide discusses the issue, you may want to refer to it.

Anyway, "cropping" is when you remove outer edges of something. In video, it most often means you remove the outer edges and then resize what is left to fill the canvas. This is bad most of the time. Only do this for restoration or effects, not general video work.

"Masking" is covering over the bad parts with something, usually plain black color. It can also be referred to as "clipping". This is what you want.

Simply mask the video. TMPGENC can do this with it's crop filter, and then enable the mask. Add about 11 pixels on top/bottom and about 5 on the left/right. Never do more than 15-16 on all sides.


>> [2] you've said there is no benefit to capturing higher than your source, but what if you want to encode a 352x480 video for viewing on a computer and wanted to blow it up to a higher resolution than 352x240?

Whether you use more pixels to display the data, the fact will always remain that the image cannot look any better than the source. In fact, if you use inadequate bitrate, or resize with an imperfect filter (bilinear, bicubic), it will look worse.

Plus, for viewing on a computer, you want a 1:1 aspect, not a 4:3 flagged one. The DVD format is 4:3 at all times. Computers do not have to respect this, and many softwares do not. You need to change size to something like 320x240, 640x480, 400x300, 512x384 for viewing on a computer, and encode with 1:1 pixels. This, of course, means that your DVD and computer encodes will be different. The computer also needed a smart deinterlace. I assume this is for web/download content intended for PC viewing only, so advice given as such.


>> [3] is the reason you recommend capturing at 352x480 instead of 352x240 b/c it preserves interlacing better by doubling the height of the video?

The reason to use 352x480 is because it BOTH maintains interlace of the source, as well as retaining all the video resolution it is fed on that axis. The x240 is both deinterlaced and lower resolution than any analog source. There is no "doubling" of any kind.


>> [4] you said in Introduction to Video Capturing that you gain no benefits from capturing at AVI only to encode to MPEG later, unless you're doing "advanced editing". how advanced is advanced? can TDA do fade outs and simple text overlays? should I only be using Premiere instead? does it matter that much if I'm doing simple editing like that with MPEG-2?

Fade outs can be done in any number of MPEG editors, like Womble MVW or Womble MPEG-VCR.

But text overlays, not really. That would probably be more towards the realm of advanced editing. You may be best served capturing directly to AVI as NLEs work best with uncompressed or less compressed AVI, as well as remove the loss of re-encoding video.

If you wish to salvage any current MPEG videos, and convert them to AVI for NLE work, simply open them in VirtualDub MPEG2 edition and save them as an uncompressed AVI file (or a HuffYUV or MJPEG or something to that effect).


>> [5] my brother does his work in Final Cut Pro on the Mac and gives me QuickTime files. if I want to convert these files to other formats, do I need to use QuickTime or can TMPGEnc do that for me if I have QuickTime already installed on my computer?

FCP is a nice piece of software. Quicktime is a container like AVI is (holding DV, for example), although Quicktime can also be it's own format of video too. So be sure you know what sort of video it is, whether it's acting as a container for something else, or if it's a pure QT format movie (Sorenson).

TMPGEnc has a environmental setting codec that can be installed. It's free. Simply go to the TMPGEnc environmental settings (as shown on the guide) and enable it, as well as raise it near the top (right click). Disable when not using QT files, as it tries to overpower MPEG and MS codecs, greedy little thing.


Some links to software mentioned:

VirtualDub MPEG-2 edition:
http://fcchandler.home.comcast.net/stable/index.html

TMPGEnc Quicktime Plugin (QTReader.vfp):
http://www.geocities.com/SiliconValley/4942/svcd.html


As always, if you need anything re-explained or clarified, just give the word.


- Did this site help you? Then upgrade to Premium Member and show your support!
- Also: Like Us on Facebook for special DVD/Blu-ray news and deals!
Reply With Quote
Reply




Similar Threads
Thread Thread Starter Forum Replies Last Post
352x480 mpeg Input / Output Virtual Dub deter Encode, Convert for discs 6 12-03-2010 02:49 PM
Image sizes for menus? 720x480, etc admin Author, Make Menus, Slideshows, Burn 30 11-04-2009 11:19 AM
Capturing fortunec00kie Capture, Record, Transfer 2 08-03-2005 03:12 AM
Can't capture AVI at 720x480 with ATI 9600XT? Shadow Capture, Record, Transfer 3 01-10-2005 09:54 PM
Capturing 16:9 Hambcastle Capture, Record, Transfer 1 08-14-2004 02:35 AM

Thread Tools



 
All times are GMT -5. The time now is 01:22 AM