What is the Maximum Capture Bitrate i should use ??

[VEBouto]

I use TmpgEnc DVD Author to produce my DVD's.
I usually capture to Mpeg-2 720x576 at a pretty high bitrate using my Leadtek WinFast Capture card. Anywhere from 8000 Kbps to 9800 Kbps. Anyway when i load the finished Mpeg's into TmpgEnc it always comes up with this message that the combined bitrate is too high and this will produce an out of standard DVD that might not be compatible with some DVD Players.
It says that the combined Video Bitrate should be no more than 9848 Kbps.
So my question is !!
If i was to subtract the 1536 Kbps that i am using for my 48,000 Hz LPCM Audio track, from the 9848 Kbps Maximum allowed, that will leave me with a maximum 8312 Kbps Bitrate for Capturing the Video.
So in reality, this should keep my finished Mpeg within the DVD Video standard.
Does this calculation make sense?
Or does that 9848 Maximum Bitrate refer to the Video only without the Audio?
And because i am Capturing at such a high Bitrate, should i keep doing so to get better results, even if it does produce out of standard DVD's, which still play in my DVD Player?
Hope it doesn't sound too confusing.

[guns1inger]

Do you need uncompressed LPCM audio ? Simplest solution is to demux the audio and encode it as AC3 @ 256 kbps, preserving the highest bitrate you can for the video. You can capture the way you have been, and re-encode the audio before authoring. FFmpeggui is a simple, free AC3 encoding front-end. Or you can buy the AC3 plugin for TDA and do it at authoring time.

Personally, that would be the way I would do it.

[c627627]

But he is asking what the maximum allowable video bitrate is for 1536 Kbps 48,000 Hz LPCM Audio.

https://www.videohelp.com/calc.htm
Bitrate Calculator lists 8227 as maximum video bitrate.

Web sites list different max bitrates. Some sites even say you should go as low as 7 Mb/s for video to be fully compatible with all DVD players.

I would go with 8227 kbit/s that the Bitrate Calculator suggests but it sure would be interesting to know what the maximum constant bitrate is for *full* compatibility with *all* DVD players if you use 1536 Kbps 48,000 Hz LPCM Audio



EDIT: The DVD specification says that the maximum multiplexed bitrate of a video/audio stream must be no greater than 10.08 Mbps (10080 kbps), so:

10080 kbps - 1536 kbps if LPCM wav audio is used = 8544 kbps for video bitrate.

About 500 kbps should account for most video bitrate fluctuations so about 8000 kbps should be selected as Constant Video Bitrate for 60 minutes or less. On older capture cards, more than 500 kbps may be needed.


Use Dolby Digital AC3 audio to allow for higher Video Bitrates to be used.
This is because Dolby Digital audio is compressed without significant decrease in sound quality and is compatible with DVD players whereas MPEG audio is compatible with some, but not all DVD players.


10080 kbps - 448 kbps (if 448 kbps Dolby Digital AC3 audio is used) = 9632 kbps for video bitrate.

About 500 kbps should account for most video bitrate fluctuations so about 9100 kbps should be selected as Constant Video Bitrate. This can be lowered to be on the safe side, because once the video bitrate approaches 9000 kbps, it is more difficult to notice the differences in video quality.


Use bitrate calculators to reduce video and/or AC3 audio bitrate to fit videos longer than 60 minutes on one single layer DVD.


Note that some software such as Sony DVD Architect 3 may incorrectly report video bitrate of a captured file being too high.

• ATi TV Setup: ATi TV Setup

[guns1inger]

Given there are so many other variables, bitrate is only a small factor in compatibility across players. Media is the single biggest factor, cover the quality of the actual media itself, how well it was burned, and the compatibility of the media with any given player.

Also, 9800 kbps is the maximum video bitrate. The actual maximum bitrate for combined video and audio stream is 10080 kbps.

[c627627]

Yeah? So with LPCM wav audio at 1536 kbit/s and assuming you are burning 20 or 30 minutes only onto an excellent quality DVD disc without menus, using an excellent burner and with the media being most widely compatible with all DVD players (and assuming all your other factors are covered positively)

if you want that DVD disc to be compatible with *all* DVD players, what is the maximum video capture bit rate that can be used with LPCM wav (not AC3 or any other kind of audio).

We are debating if even 8000 kbps video capture bitrate is safe in this thread:
https://forum.videohelp.com/viewtopic.php?t=321598

7700 kbps is said to be the safest bet.

I am leaning towards using 8000 constant video capture bitrate with LPCM wav because the
https://www.videohelp.com/calc.htm
Bitrate Calculator lists 8227 as maximum video bitrate.

So lowering it to 8000 should cover most situations but I guess the absolute safest might be 7700.

[guns1inger]

If you want it to be compatible with *all* DVD players, you get it professionally mastered and pressed. Otherwise you just take your chances with the rest of us.

I regularly encode to a combined bitrate in the high 9000's without any so-called bitrate related playback issues. Personally I think the 7700 or 8400 or whatever limit holds about as much water as the burn at 2x or 4x limit for maximum compatibility. There is no absolute. You can err on the cautious side if you want - that's what personal choice is about. But I think too many people still think the restrictions of early, poorly made media, burners and players mean these arbitrary guesses (and that's all they were - there was no science behind it) are some sort of law. And put simply, they ain't.

[c627627]

None of your three posts contain your opinion addressing the actual question

You are correct when posting about related topics of media quality, using different kind of audio, and the rate you use personally without experiencing issues personally.

But that was and still is not the question


The DVD specification says that the maximum multiplexed bitrate of a video/audio stream must be no greater than 10.08 Mbps (10080 kbps), right?

So: 10080 kbps - 1536 kbps for LPCM wav audio = 8544 kbps

(Is this equation correct and appropriate?)

so now we have to account for the bitrate going up and down a bit from the target even at constant video capture bitrates, right?

So I figure with a 544 kbps buffer, 8000 kbps should be more than safe but going above 8544 kbps for the video capture
[10080 kbps - 1536 kbps for LPCM wav audio = 8544 kbps]
would be asking for trouble with some older DVD players, correct?

[guns1inger]

Actually, my first post indicates that I would leave the video alone, and re-encode the audio down to AC3 at a lower bitrate. This removes the need to re-encode the video, or alter the OP's current practices. My last posts indicates that I use higher combined bitrates where the running time allows.

Unless the OP has audiophile grade playback kit and pristine source, LPCM stereo is a gratuitous waste of bitrate, IMO.

In black and white - keep the video in the high 8000's - low 9000's and encode the audio to 256 kbps 2 channel AC3. If the running time permits such high rates, or the OP is willing to go DL or split the discs.

[c627627]

However much you are correct that AC3 may be the way to go, we are discussing, for informational purposes, LPCM wav audio.

So, I'll accommodate you by saying: should for whatever strange reason someone decide to go with LPCM audio,

Can you post the reason why going over 8544 kbps for constant video capture bitrate is not going to be a problem with older DVD players, since:


The DVD specification says that the maximum multiplexed bitrate of a video/audio stream must be no greater than 10.08 Mbps (10080 kbps), so:

10080 kbps - 1536 kbps for LPCM wav audio = 8544 kbps

[jman98]

c627627 - When you use any capture card, it will ignore whatever value you give it as the maximum bit rate to use. Just try it. It will ALWAYS go above it. Maybe it only goes 100 Kbps above. However, with an older ATI card I have, I have seen it spike almost 1000 Kbps (!!!) above what I told it to use for the max bit rate on a few captures I did.

Why are you pushing the envelope? Do a test between 8000 CBR and 8400 CBR. Can you honestly tell a difference? Because I don't think you will be able to. You've been advised to stick to 8227 and now you are wanting to do 8544. Seriously, once you get to 8000 Kbps for DVD, you can't really tell the difference if you go higher. Try it if you don't believe me. I'd advise sticking to 8200 or so just to prevent problems. This way your max bit rate can spike a little to 8400 or so and you should still be OK. However, if you want to capture at 8544, feel free to do so, but when your captures actually spike above this value and/or the DVD authoring program just says your bit rate is too high, even though you think it isn't, well, we warned you.

[c627627]

Thank you for posting. No actually, if you read my posts, "all" I want is 8000 kbps (not higher) and am asking why go lower than 8000?

And you gave the answer because some capture cards may drasticly overshoot. Thanks for that.

But yes, 8000 is also what I would use with LPCM.


This is just for infirmative purposes, I'm of course going to go with AC3 for reasons guns1inger posted and am thinking how much lower to go with AC3 than the advised 9x00?

[c627627]

Check this out, I captured a file at 8020 kbps, and look at how different apps are telling me what the actual video capture bit rate was:

[c627627]

Having discovered jut how wrong the software can be in detecting the actual video capture bitrate and seeing that it can be off by a 1000 kbps or even 1500 kbps (!), these are my conclusions:


The DVD specification says that the maximum multiplexed bitrate of a video/audio stream must be no greater than 10.08 Mbps (10080 kbps), so:

10080 kbps - 1536 kbps if LPCM wav audio is used = 8544 kbps for video bitrate.

About 500 kbps should account for most video bitrate fluctuations so about 8000 kbps should be selected as Constant Video Bitrate for 60 minutes or less.


Use Dolby Digital AC3 audio to allow for higher Video Bitrates to be used.
This is because Dolby Digital audio is compressed without significant decrease in sound quality.


10080 kbps - 448 kbps (if 448 kbps Dolby Digital AC3 audio is used) = 9632 kbps for video bitrate.

About 500 kbps should account for most video bitrate fluctuations so about 9100 kbps should be selected as Constant Video Bitrate. This can be lowered to be on the safe side, because once the video bitrate approaches 9000 kbps, it is more difficult to notice the differences in video quality.


Use bitrate calculators to reduce video and/or AC3 audio bitrate to fit videos longer than 60 minutes on one single layer DVD.

[guns1inger]

I trust very few apps to correctly disclose the actual encoded bitrate. Many are lazy and just reader the header, and many lazy encoders just put the maximum allowable video bitrate, or some other arbitrary number. Bitrate viewer is one of the better apps for checking. DLP's built-in bitrate analysis tool is OK as well. Anything that just pops out a number but can't give you a graph to back it up should be considered suspect.

[c627627]

Right, so according to Bitrate Viewer pictured in the post above yours, 7372 kbps was the peak bitrate even though I captured at constant 8020 kbps, is that the way to read it?

Even though Bitrate Viewer also says that Nom. bitrate: 8020000 Bit/Sec


Does that mean that since 8020-7372=648, the actual video captured bitrate was off by 648 kbps, and that if I chose to recapture the same content, I could add that much to previously set 8020 kbps?

[guns1inger]

What bitrate viewer is saying is that your video has a peak bitrate of 7372, and average bitrate of 6156 kbps, but the encoder wrote your chosen CBR bitrate of 8020 kbps into the header.

I have a programme stream on my disk at the moment that is directly taken from an OTA digital broadcast. No extra encoding etc. The stream header (and bitrate viewer) both say the nominal bitrate is 10000kbps. Which is patently untrue as the 1:28:14 running time fits comfortably on a DVD5 when authored.

Using a bitrate calculator with a custom size option, I get an average bitrate of around 6278 kbps to reach the actual file size I have.

Bitrate Viewer reports a peak of 8426, and an average of 6265.

G-spot 2.60Beta1 reports and average bitrate of 6309.

A quick look at DLP's bitrate viewer showed similar numbers, but also included the audio bitrate. I didn't let the scan finish, so I won't publish the actuals.

Bottom line - none agree, all ballpark. The nominal bitrate as stored in the mpg file proved to be meaningless.

So what does this mean for your file ?

You asked for 8020 kbps, so that is what your encoder put in the header. In reality, it could not maintain that as a CBR encoding, and appears to have been unable to encode at a constant rate, period.

How are you capturing ? Is it software or hardware encoding ?

[c627627]

.mpg file was captured using ATI Multimedia Center, software set Constant 8.02 M Bit/sec video bitrate.

I'm not sure I understand your post. What is the actual video captured bitrate of the file?

What would you use looking at the Bitrate Viewer screenshot as the video bitrate if you had to recapture the same source?

In other words, if 7372 is reported as peak and I set 8020 constant, does that not mean that the video bitrate could be actually increased because the capture program is undershooting the set constant bitrate?

Can you believe how wrong Sony DVD Architect 2 and 3 are? They are reporting off the charts high bitrates for the same file and are warning it will not be compatible with DVD players if encoded as such?!


Here's another screen shot of it:

[guns1inger]

G-Spot is almost exactly the bitrate you asked for - does this mean that BitRate Viewer is way off as well ?

This is why, when people ask which program to use to determine the bitrate of an mpeg file, they are told that none of them can be trusted.

If I feed the file size and length into a bitrate calculator, then select 1536 kbps audio, I get an average bitrate of 7969 kbps, which is closer than bitrate viewer got, but still less than G-Spot.

I guess my appraoch is different, as I don't capture to mpeg2 unless I'm taking directly from the digital stream, which means I don't have to worry about bitrate.

If I am encoding to DVD then I use a bitrate calculator, work out the average bitrate needs for the space I have available, and use a good encoder that will fill that space appropriately. As I hit the target, within a couple of MB, every time, I know that the bitrate I requested is being delivered.

Personally, I think if you want to capture directly to mpeg-2, you might as well use a DVDR instead.

[manono]

If you ever want to know the exact average bitrate for any sort of MPEG file, open it in DGIndex, go Options->Playback Speed->Maximum, and then File->Preview. In the new Information Screen that opens up, it'll give you the current and average bitrate as it plays your file. It'll take maybe 15 minutes for a complete movie, but it's accurate.

And any authoring app that warns of a too high combined bitrate when you first open the files should be completely ignored. There's no way for it to know that it's going to give buffer underflows until it does the actual authoring. TMPGEnc DVD Author is notorious for this. Go ahead and author. You may or may not run over the limit, but the opening warning isn't any indication of that.

Also, I take my combined bitrates to the limit, often using a max video bitrate of 9800 when encoding for DVD (if the audio+video+subs+overhead will come in under 10080). Unlike jman98, I do believe there's a difference if you use a lower max bitrate than possible. During complex scenes the average quant might skyrocket when using a lower max bitrate, and produce mosquito noise, color smearing, and maybe even macroblocks, artifacts that might not appear were the max bitrate to be set as high as possible. Just my opinion.

All DVD players are required to conform to the specs. Mine does. What do I care if some older players don't, and start to stutter at lower max bitrates? Any player that can't play in-spec DVDs is a piece of junk and should be consigned to the scrap heap. Yes, I know the bit about burned DVDRs giving more trouble than pressed ones. The only time I've had any trouble with momentary freezes is after my authoring app has reported buffer underruns, I was expecting problems, and burned and tested to confirm. Otherwise, when they've authored OK, even with maxed out max bitrates, I've had no problems at all.

To be able to get away with this, though, you should know the abilities of your capture card, your encoder, and your player. CCE is good, so I've had no problems. Other encoders or capture cards don't keep with the max bitrates set, and might not author correctly.

[Nelson37]

The problem with what you are trying to do is that ATI cards do not adhere closely to the specified settings. Furthermore, the amount and direction of the variance will change significantly with the particular type of video being captured. Basically, the darker it is, the lower the bitrate will be.

On all types of files, I have found that the actual bitrate, calculated simply from file size, is lower than that specified. This has held true from MMC 6.xx thru 9.12, 8 to 10 different versions and five or six different cards. Many hundreds of captures. I have NEVER seen an ATI card go HIGHER than the specified bitrate. This is in total average, I do not often view a detailed graph for spikes.

Your "cushion factor" of approx 500 is not large enough to account for the possible variance, I would double it. There is also the possibility of occassional "spikes" in bitrate which could cause problems with some players. These rarely exceed maximum specified bitrate, though again I rarely view them on a continous graph.

With MMC 9.12, VBR, a max rate of 10.0 with an avg of 9.5, audio set to 48k 384 bit, Attack of the Clones at approx 2 hrs 26 minutes yielded a file of 3.5 Gb.

I just check the schedule and cap it again if I go too high or too low. Trying to calculate what MMC will do with a particular file and settings is just too much of a crapshoot and frankly a total waste of time. The best you can do is an educated guess.

[Lula07]

The problem with what you are trying to do is that ATI cards do not adhere closely to the specified settings. Furthermore, the amount and direction of the variance will change significantly with the particular type of video being captured. Basically, the darker it is, the lower the bitrate will be.

On all types of files, I have found that the actual bitrate, calculated simply from file size, is lower than that specified. This has held true from MMC 6.xx thru 9.12, 8 to 10 different versions and five or six different cards. Many hundreds of captures. I have NEVER seen an ATI card go HIGHER than the specified bitrate. This is in total average, I do not often view a detailed graph for spikes.

Your "cushion factor" of approx 500 is not large enough to account for the possible variance, I would double it. There is also the possibility of occassional "spikes" in bitrate which could cause problems with some players. These rarely exceed maximum specified bitrate, though again I rarely view them on a continous graph.

With MMC 9.12, VBR, a max rate of 10.0 with an avg of 9.5, audio set to 48k 384 bit, Attack of the Clones at approx 2 hrs 26 minutes yielded a file of 3.5 Gb.

I just check the schedule and cap it again if I go too high or too low. Trying to calculate what MMC will do with a particular file and settings is just too much of a crapshoot and frankly a total waste of time. The best you can do is an educated guess.

Hi Nelson,

I saw some of your posts in "How to rip closed captions from a DVD?" and I got particularly interested in the application, you created, called ATI2SRT.

I tried to get in touch with you through PM and post in the topic but I got no reply.

I know this this not the place to post for this, but I thought this was the only way of reaching you.

Could you send me the ATI2SRT to my e-mail? I would really appreciate it?

My e-mail is: ricardovas@yahoo.com

Thanks and sorry for any inconvenience.

[c627627]

Thanks for the advice. I just want to say that so far, I have only used constant bitrates because I usually capture shows that are under an hour.

I didn't know the exact length of the show but I knew it was under an hour, allowing for high bitrates even with wav audio.

10080 dvd spec - 1536 kbps wav audio = 8544 kbps
So I knocked off about 500 kbps for this and captured at 8.02 Mbps.

I retrospect, I should have used 448 kbps Dolby Digital AC3 audio:
10080 dvd spec - 448 kbps AC3 = 9632 kbps.
I would have then used, maybe 9100 kbps for video.


So in the end, my conclusions are

1. Ignore any app that spits out a bitrate without running the entire clip through.

2. Even when apps analyze the entire clip, variations between them cast doubt on accuracy, so I believe it then comes down to this statement:

Your "cushion factor" of approx 500 is not large enough to account for the possible variance, I would double it.
(...)
The best you can do is an educated guess.

I'm leaning towards an opinion that the size of the protecting buffer ("cushion factor") would depend on the capture card itself. Older cards may fluctuate widely going above the 500 kbps cushion factor.



P.S. Following manono's advice, I found DGMPGDec Version 1.4.8 (which is the DGIndex app) and it ran the entire 52 minute mpg file in 15 minutes to show the Bitrate (Avg) being 8.020 Mbps, spot on. (By the way, this was a clip I captured from a High Definition channel, that's why the frame type was set to 'Progressive' instead of 'Upper Field First' ATi cards usually require for captures.)

But anyway, Bitrate Viewer is popular and widely praised so it is strange that it is showing considerably lower actual average video bitrate for the same file.


That is why I think that "an educated guess" means having some experience, knowing your capture card and lowering the set video capture bitrate by the amount specific to you.

[Nelson37]

For only one-hour files, assuming you intend only one per disk, Bitrate is less important, only issue is not to exceed maximums. If PC playback is intended, then bitrate considerations go away altogether.

Try the same settings on a variety of video types and you should see some variation, though it should be reduced by using CBR. This is just too inefficient for longer encodes, but works well for short ones.

You might also want to try the 3:2 pulldown removal to arrive at true progressive, 24 fps video. It will have the pulldown flag already embedded, possibly in some not-entirely-standard way. This gives some programs major issues with editing and authoring. The IVTC effectiveness varies by source, and also with different MPG encoder settings, but usually works quite well. This may not be usable for some HDTV broadcasts which are not movies.

[c627627]

Hm. I capture to burn to DVD to watch on both standard & HD TVs.

You mean CHECK: Inverse 3:2 Pulldown for Progressive?

I usually have it
• 'Encode Interlaced' for Standard TV source or 'Progressive Source' for HD signal source
• UNCHECK: Inverse 3:2 Pulldown

but for a High Definition signal, I should CHECK: Inverse 3:2 Pulldown?




EDIT: The next four pages of this thread are a struggle to finally get this:

The Pros & Cons of using the Progressive Source setting and CHECKING Inverse 3:2 Pulldown

Inverse 3:2 Pulldown should only be CHECKED in combination with Progressive Source setting. This should only be used if the source is standard TELECINE with continuous cadence from either a standard or a high definition signal.

* Pros: *

• Smaller files (with more bits available per frame) created quickly on the fly, allowing multiple movies to fit on one DVD at higher quality.

• Captures may look better on computer displays.


* Cons: *

• The source may have a break in cadence, in other words there may not be a continuity of pattern required for this setting to be used, resulting in frame disturbance.

• These errors may be due to a temporary field reversal (bad edit) or due to special effects that weren't shot on film. Therefore, if the source is not pure, noticeable pauses or stutters in the video can result.

• This may not be as useful for some Progressive DVD players and HDTVs which are designed to switch between telecine and adaptive deinterlacing quickly.

• If this is not done on the fly and is instead done by reencoding later, original capture can be preserved to fix any errors.

• Resulting errors may look far worse than Telecine viewed at 1x speed on computer displays this is designed to help with.


TELECINE can be recognized by using a program that can show individual frames. Look at five individual frames in a raw to see if interlacing artifacts show on the 4th and 5th frame. There should be 3 clean, artifact-free frames followed by two with interlacing artifacts. If interlacing artifacts are visible in all five frames (or not visible at all), the source is not telecine.

Telecine with continuous cadence is found on video that has been transferred from film. Telecine sequence (cadence) is interrupted by commercial breaks, special effects, cut edits, title overlay, conversion and speed up done to allow for insertion of more commercials or to fit the program into a given time slot.


Remember that Interlace artifacts clearly visible on a computer display do not show up when a DVD player is used with standard TV or an HD TV.

Deinterlacing software player can help with displaying minimal artifacts on computer displays.


Read this:
ATi TV Setup

[Nelson37]

Yes, absolutely for an HD source.

Examining the AVI captures of these sources thru the s-video feed reveals at least two different types of display, some simply repeat the fourth frame and some appear to do a standard telecine. Most are NOT interlaced in the conventional sense, meaning the interlacing artifacts are not visible except for the fourth and fifth frames. Unfortunately, many one-hour shows do seem to be conventionally interlaced, though not all of them.

The "remove pulldown" works well for most movies, tho it does not seem to adapt well to changes or irregularities in the telecine pattern. Previous versions of MMC would give a framerate of 24.1x, while since 9.03 or 9.06 the pulldown removal has resulted in exactly 23.976 framerate.

The newer version files also seem to be more editable, previous ones gave a wierd variety of problems mostly dealing with audio synch. Be prepared to try a few different programs for muxing, demuxing, and cutting.

The pulldown removal, as with many MMC settings, seems to react differently with different capture settings. Higher bitrates help, other settings have varied in their effect thru different versions of MMC. Have not thoroughly tested 9.12 with each of the options, generally MPG2-DVD with bitrates above 6000 with aspect ratio set same as source is a starting point.

9.06 would correctly crop and set 16:9 when making a VCR file, so far 9.12 is not cropping correctly and I have not yet tested for AR setting. This is easily changed but it was nice to have it already done. That plus the correct cropping, I may go back to 9.06. Just gotta love that wonderful MMC.

[c627627]

Well here's a chance to prove things one way or another:

Progressive Source, 8000 kbps Constant Video Bitrate


• UNCHECKED: Inverse 3:2 Pulldown: withoutpulldown.mpg


• CHECKED: Inverse 3:2 Pulldown: withpulldown.mpg


Please test them and post what you see as differences.


I am standing by my claim that MMC 9.06 & 9.06.1 will crash every system attempting to capture with a custom AVI setting as confirmed by ATi, my tests and people posting on the internet about it.

I am using the latest ATi Encoder with MMC 9.03, I believe that there is not difference in video capture quality between MMC 9.03 and later versions of MMC. Additional features (bugs) is mostly what you get with later versions plus I understand VideoSoap feature was actually taken out from latest, most recent versions of Multimedia Center, so that's one feature loss, not gain.

[Nelson37]

Both clips appear to have fourth frame duplicated, this is fairly common. Can definitely confirm that I have removed this on some videos with the pulldown setting. Can also confirm that certain versions and/or certain settings either did not work at all or worked on a random basis. Both cropping settings, type of MPG, 720 and 704, letterboxing setting and bitrate, also the VCR filetype can affect this.

As far as the 9.06 AVI bug, this may well be so. As I do not use this setting currently, other than for testing, and no longer use that version, it does not matter to me. Same with VideoSoap, I very rairly used it. With the hi-quality sources I prefer, it has no real value to me and I do not miss it.

While I saw no real difference in 9.03 to 9.06, the 9.12 is most definitely using a different encoder or at least is handling settings differently. Same video recorded with same bitrate settings has significantly different size. Appears to possibly be handling dark areas better, where older versions tended to display blocks in these areas. The 9.12 version has a different set of files listed on ATI's website, first one with a seperate Encoder file. May or may not be some significant difference from earlier versions. They have added the ability to record closed captions with an MPG file. There is also supposed to be a new audio capture mode in some versions that holds the possibility of real-time AC-3 5.1 capture thru SPDIF.

Have not yet done real frame-by-frame testing of the two versions as the 9.06 is no longer installed.

Would recommend further testing of the IVTC with several source types and settings. It is definitely possible to capture real, progressive, full-res, 23.976 material with no interlacing or duplication artifacts.

[c627627]

1. I didn't get if you said to YES, CHECK: Inverse 3:2 Pulldown for high definition signals or if you said leave it UNCHECKED like we should for standard signals.

2. On to a very important question you raised: Opinions may vary but same system test captures are a different thing:
https://forum.videohelp.com/viewtopic.php?t=321598

I am standing by to test any version of MMC to see if there is more than just additional functions and bugs in later versions.

As you know MMC 9.12 is not available anywhere but 9.13, 9.14, 9.15 and 9.16 are. They all show same quality captures without any differences. I can retest of course, but it is important to definitively prove if capture quality is affected by MMC versions, so what should I use to retest because side by side screen caps of same frames should not show consistent differences in quality.

3. The different set of files you're referring to (including encoder files), can be used by any version of MMC. As a matter of fact, I do use the latest version of encoder files with my MMC 9.03.

4. Regarding, MMC 9.06, it is important everyone knows about the bug without having doubts because it is a major bug that can be replicated on every system. If you create a custom AVI recording preset on MMC 9.06 & 9.06.1 and try to use it, it will crash your system per official ATi list of bug fixes in the follow up version 9.08 and per my tests.

[manono]

Forgive me if I'm misunderstanding some things. I've never captured anything in my life. Am I to understand that this came originally from a Hi-Def source? 1080i maybe? And if so, wouldn't it have been interlaced to begin with?

Therefore, hasn't this already been deinterlaced, as well as having been resized?

1080i is 29.97fps interlaced, which makes me think that's what your source was. 720p is 59.94fps progressive. This isn't 59.94fps.

Therefore, the source has already been filtered (resized, made progressive, and probably deinterlaced). You certainly don't want to keep all those duplicate frames in there, as they'll make the video play jerky. When reencoding (you are reencoding, aren't you?), you want to remove the dupe frames, encode for 16:9, and use VBR encoding, if this is a movie (which it is, King Kong), and is being recompressed to fit on a DVD5.

Inverse 3:2 Pulldown is for interlaced sources. I don't know for sure, but I doubt if turning on that setting will remove the dupe frames. There's no difference that I can tell, between your 2 samples.

To remove the dupes, AviSynth is far and away the best method, using such filters as Decimate (part of the Decomb package) or TDecimate (part of the TIVTC package). You'll also want to remove that nasty looking black bar along the right side, and resize. Then you reencode for 23.976fps progressive, and apply pulldown during or after the encoding stage, before then sending it off to your authoring app.

Please correct me if I'm misstated anything.

[Nelson37]

Hi-def source, usually 1080i if I can trust the indicator light. What exactly comes thru the S-Video cable to the capture card varies. It may be true interlaced, usually only for one-hour series or documentaries.

However virtually all movies show some form of Telecining, which shows up as either interlacing ONLY on the fourth and fifth frames, or an exact duplicate fifth frame. In both cases it is desirable to remove these artifacts and arrive at true progessive, 24fps material for both quality and bitrate usage.

This removal, "remove 3:2 pulldown", Inverse Telecine, or IVTC, can be accomplished by checking the box in MMC. You also select Progressive source as this is what you have AFTER the IVTC is performed. This procedure should NOT be performed on a true interlaced source.

The functionality will be specific to the particular video you are capturing. HD or not is not relevant. The S-Video feed is not really HD, anyway, though it is most definitely better than regular SD feed.

This is a real-time function, used with real-time encoding. It will adapt, but will not correct for poor editing. Most recent, unedited movies show smooth playback throughout, higher bitrates are possible because of the corrected framerate. Having used Avisynth with at least three different IVTC filters, and a few other programs, I would say the MMC is at least as good as most. I gave up on the whole AVI-filter-IVTC-re-encode bit several years ago, when real-time mpg got almost as good and worthy of more tweaking. With HD source, real-time is better, cause you can't tell the difference and it's already done. YMMV.

As for the difference in versions, 9.12 is definitely making smaller files than 9.03. 15-20% smaller. With VBR settings higher in 9.12. It may simply be a different scale, the seperate encoder package may just be the way the files are grouped. Then again, 9.03 did NOT include these files, and if you have installed them, and they are in use, which encoder, and effectively which version of MMC, are you using and testing? 9.03 or 9.12 or above?