What bitrate to choose ?

[Mickey79]

Ave,

I've encoded and authored more then 150 DVDs, so i can't say i'm a newbie, although my question sounds like it. I've had a 90% sucess rate with my work. I'm trying to work on that 10%.

Earlier i used the BitrateCalc application to calculate the bitrate. As from what i understand, the length, the audio bitrate, and the resolution are the 3 prime factors which help you determine the bitrate.

Now i use Tmpgenc 3.0 XPress, which calculates the bitrate for you, based on total length & audio bitrate you specify for the video you are encoding.

Like i said, while 90% of my work has been fine, that is, excellent quality, both audio & video.. there has been 10% which has not been that great, that is, some distortion or less then excellent video quality.

What i've had a difficult time with is figuring out why. Because i've encoded videos at 2000 to 2500 bitrate and they've turned out Excellent in Quality, and i've encoded videos at 5500 to 6000 and they've turned out crummy.

My first question is, if there is a definite answer to this, what is the bitrate below which you should never go? If there is such a bitrate? And secondly, is there a max amount of bitrate Over which you should never go?

Usually i keep the lenth of material no more then 3 hours on a DVD, and it keeps it excellent quality. Sometimes the entire content is no more then 1 hr 30 mins. Recently i wanted to encode something which would be about 4 hrs in length... but i'm afraid to do it as i'm presuming, with a 448 kbps audio bitrate (for 5.1 AC3), my video bitrate will need to be encoded at something like 1500 or 1800.

I would just love to understand this whole thing a little better so as to minimize those 10% failures that i have with my encoding & authoring.

Thanks,
Mickey

[TB Player]

And secondly, is there a max amount of bitrate Over which you should never go?

I believe that anything over 8000 can cause problems on some players.

[lenti_75]

I wouldn't go below 4000 which is ~2hrs usually I keep it around 6000 ~1.5hr, and not higher than 9000, 9500 it's the max, I think for dvd players.
and try to convert VBR with 2 passes, not CBR
you can always go lower, but never higher.... keep this in mind.

[Cornucopia]

If your player can't play up to 9800kbps for video, return it: it's defective.

It is mandatory in the DVD-Video spec that ALL players be capable of correctly playing up to that level.

re: what to choose?
Use a bitrate calc and always give your (total combined on disc) programs the max (within DVD legal limits) that it can take. The only reason for not doing this is if you know ahead of time that a portion/clip will later be used in a longer compilation/composition or you're doing DVD-VR recording on settop using multiple sessions.

By MAX, I mean highest CBR level or highest AVG VBR level. (Peak VBR-->9800kbps, Min VBR-depends)

Scott

>>>>>>>
edit: and I concur about the 2-Pass VBR...

[TBoneit]

I suspect the difference in output is due to the source. You did not indicate if they are all the same specs before converting. IE AVI's of 640 by 480 by xxx bitrate. Even then if they are coming from differnt peoples encodes they can be different quality. I've seen videos from specific websites that they provide for legal download that are of differing quality(in case anybody is wondering), that they have created for download. You can find some links here via a search. You could try looking at the source before encoding and try some filtering to improve it.

Good Luck

[monzie]

quote="Mickey79"]Ave,

I've encoded and authored more then 150 DVDs, so i can't say i'm a newbie, although my question sounds like it. I've had a 90% sucess rate with my work. I'm trying to work on that 10%.

Earlier i used the BitrateCalc application to calculate the bitrate. As from what i understand, the length, the audio bitrate, and the resolution are the 3 prime factors which help you determine the bitrate.

Now i use Tmpgenc 3.0 XPress, which calculates the bitrate for you, based on total length & audio bitrate you specify for the video you are encoding.

Like i said, while 90% of my work has been fine, that is, excellent quality, both audio & video.. there has been 10% which has not been that great, that is, some distortion or less then excellent video quality.

What i've had a difficult time with is figuring out why. Because i've encoded videos at 2000 to 2500 bitrate and they've turned out Excellent in Quality, and i've encoded videos at 5500 to 6000 and they've turned out crummy.

My first question is, if there is a definite answer to this, what is the bitrate below which you should never go? If there is such a bitrate? And secondly, is there a max amount of bitrate Over which you should never go?

Usually i keep the lenth of material no more then 3 hours on a DVD, and it keeps it excellent quality. Sometimes the entire content is no more then 1 hr 30 mins. Recently i wanted to encode something which would be about 4 hrs in length... but i'm afraid to do it as i'm presuming, with a 448 kbps audio bitrate (for 5.1 AC3), my video bitrate will need to be encoded at something like 1500 or 1800.

I would just love to understand this whole thing a little better so as to minimize those 10% failures that i have with my encoding & authoring.

Thanks,
Mickey[/quote]


If your source is DIVX/XVID avi's then start here:

https://www.videohelp.com/forum/viewtopic.php?t=255103

What you need to understand is 'BITRATE' is NOT an answer and was NEVER INTENDED TO BE A ANSWER (theres so much MIS-INFORMATION in these and other forums that 99% of of people BELIEVE that bitrate is THE answer to all mpeg2/dvd encoding problems)...well I can tell you that ITS NOT... its purely the amount 'AVAILABLE' for a blank DVDr over any given time REGARDLESS of it source.

My own experiences of BITRATE indicate that 2200 is a good VBR/OPV MINIMUM for a typical avi (2.3:1) but that figure INCREASES as the picture ratio gets closer to 1.33:1 (fullscreen) where 2700 is a more realistic minimum......BUT....we need to take into account the 'source'.....I cant help you on this NOR CAN ANYONE ELSE, as only YOU know the source that you are encoding and wether or not it is bright and full of colour..or dark with lots of 'black and grey', is the camera moving a lot?..the outcome DEPENDS on you...ultimately.

[monzie]

Double post

[Cornucopia]

monzie, I don't know what you're smokin', but BITRATE is always a part of the answer, whether you like it or not.

If keeping the BITRATE down was unimportant, you wouldn't need video to be compressed at all. The pipeline constraints determines that max bitrate, and the Length of program and bucket size constraints determine the overall average bitrate.

And the need for QUALITY is what gets us to keep the BITRATE up as high as possible.

Quality is determined by a combination of: Resolution (W x H), Scene Complexity, and BITRATE (overall, per GOP, and per Frame).

Scene complexity is determined during original production, and optionally with a few post-production filtering techniques. That's all you can do there.

Resolution is Either Full D1, 1/2D1, or SIF(1/4D1). Couple paths to take, but otherwise, that's all you can do there.

BITRATE allocation is the only thing that you can have a good deal of control over at such a late stage in the game.

Rule of thumb:
Assuming all of your sources are very high quality, and medium to low action/editing/complexity...
Minimum Bitrate for D1 ~ 2500-4000 (Max 9800) kbps
Minimum Bitrate for 1/2 D1 ~ 1200-1800 (Max 9800)
Minimum Bitrate for 1/4 D1 ~ 650-1100 (Max 1860)

Scott

[monzie]

monzie, I don't know what you're smokin', but BITRATE is always a part of the answer, whether you like it or not.

If keeping the BITRATE down was unimportant, you wouldn't need video to be compressed at all. The pipeline constraints determines that max bitrate, and the Length of program and bucket size constraints determine the overall average bitrate.

And the need for QUALITY is what gets us to keep the BITRATE up as high as possible.

Quality is determined by a combination of: Resolution (W x H), Scene Complexity, and BITRATE (overall, per GOP, and per Frame).

Scene complexity is determined during original production, and optionally with a few post-production filtering techniques. That's all you can do there.

Resolution is Either Full D1, 1/2D1, or SIF(1/4D1). Couple paths to take, but otherwise, that's all you can do there.

BITRATE allocation is the only thing that you can have a good deal of control over at such a late stage in the game.

Rule of thumb:
Assuming all of your sources are very high quality, and medium to low action/editing/complexity...
Minimum Bitrate for D1 ~ 2500-4000 (Max 9800) kbps
Minimum Bitrate for 1/2 D1 ~ 1200-1800 (Max 9800)
Minimum Bitrate for 1/4 D1 ~ 650-1100 (Max 1860)

Scott

Well Scott/Cornucopia if you READ my post you would understand/know that I was SPECIFICALY (spelling?) talking about divx/xvid/mpeg4 avi's converted to DVD/MPEG2......you seem to fail to realise that AN AWFUL LOT (no, I cant put an EXACT figure on it but lets say 35-45% for arguements sake) of all video information is lost in 99% of all conversions from (say) mpeg2 or DV to divx/xvid REGARDLESS of the bitrate used..much more when xvid/divx bitrate IS taken into account (and RESIZING)....which allows RE-ENCODING of these files to be MUCH lower as there is LESS info to RE-ENCODE to start (you cant 'create' video info from info that IS NOT THERE TO BEGIN WITH)..the old 'for this bitrate use this DVD res' arguement is BULLSHIT in mpeg4 to DVD conversions....as for 'scene complexity'..what scene complexity?...the original conversion to divx/xvid has ALREADY dealt with this isuue...are you suggesting that MPEG2 encoding from a frameserved xvid/divx source RE-CREATES some sort of bitrate bottleneck?..Does it re-create the original source (pixels/colours/complexity et al?)?.......me finks..GO BACK TO SCHOOL...those resolutions/bitrates MIGHT have been true in 1996 but arent EVEN CLOSE today..stick to the DV forum where your 'enlightenment' MIGHT (just) hold true...and take 'the blue gnome' and the aussie with you. GOODBYE (or wake up to the 21st century).

[junkmalle]

What bitrate to choose ?

Use a bounded constant quality mode and stop worrying about bitrate.

[jimmalenko]

@Mickey79:
I think your issue might be related to you letting TMPGEnc Xpress do a bit of the hard yards for you. I say this mainly because of this statement you made:

Because i've encoded videos at 2000 to 2500 bitrate and they've turned out Excellent in Quality, and i've encoded videos at 5500 to 6000 and they've turned out crummy.

What I THINK is happening is that when your bitrate is 2000 to 2500, TMPGEnc Xpress is automatically changing your frame size down to Half D1 (352 X 480/576) for you (you using the wizard, yes ?), and this is contributing to your video loking 'good' at that bitrate. IMO you really need to take a bit more control of the situation, as the frame size, rate control mode and the bitrate all need to be "compatible" (for want of a better word) with each other in order for you to get the best results.


@monzie:

are you suggesting that MPEG2 encoding from a frameserved xvid/divx source RE-CREATES some sort of bitrate bottleneck?..Does it re-create the original source (pixels/colours/complexity et al?)?

Yes. All encoders decode your source back to RAW before encoding, so therefore you need to assign enough bitrate to retain the original quality of the source. Once we start talking quality, we start talking frame size as well (I stand by my statement of never upsizing), and there obviously comes a time where there's no need using more than a certain bitrate for a certain frame size. I know you're of the mentality of using the exact optimum bitrates - I'm of the "use the whole disc" ilk.

[vhelp]

My first question is, if there is a definite answer to this, what is the bitrate below which you should never go? If there is such a bitrate? And secondly, is there a max amount of bitrate Over which you should never go?

This is very hard to answer because it is dependant on the source file
being encoded.

Most of this is all related to Source being processed previously
by another user (aka, File D/L) etc:
.
* How much noise in video
* 60fps vs. 30fps vs. 24fps
* Interlace vs. Progressive vs. Film(telecine)
* Resolutions
* Source aspect ratio
* Audio
* Source origin
* Source compression
* Source poorly encoded (ie, bad IVTC; Wrong fps; NTSC vs. PAL; ect)
* Source placement (ie, was source encoded, and then re-encode, and.. etc
.. by another user)
* ..and a few other phenomina's etc
.
So, it varies from source to source. Sometimes, you strike luck, and
others, you don't. So, you can't apply the sample (shall I say, template)
to every source. Over time, you learn a given number of "source styles"
and "feel" your way around various "encoding techniques" for each style.
This skill takes time to master
.
To note, those who *have* such receipes (those that demo or proclaim
receipes) are those users who have "learnt" this skill, but have not
come to realize that you can't just slap a piece of paper (guide) in
front of a user (newbie) and expect them to understand the paper (what
you've learnt (skill) over the long haul - years)
.
The new user (those who are in receipt of this paper guide, also have
to learn the skill. They can follow the guide, and for some Sources,
the techniques might work, but in most others, they won't. But, then
they come back to you with questions. And, only *you* (the guide man)
knows the answers, because this guide man has the skills (or vision)
to see what the Source requires.
.
People miss this point. I guess its the equivalent of "the blind spot".

I'm with the belief that you don't have to fill the disk in every
project. I have many disks that are half filled (for reasons I do
not have to explain to you all) and others, I have filled.
.
But IMO, the closets approx one might want to consider in dvd projects,
is something along these lines:

If the source ( > 1 hour; and pure interlace) then:
* break project into more than one disk, and devise your bitrate equally

If the source ( =< 1 hour; and pure interlace) then:
* use the maximum bitrate your - player will handle; and fill the disk

If the source ( =< 1 hour; and Film[telecined] perfectly) then:
* use the maximum bitrate your - player will handle; and fill the disk

If the source ( > 1 hour; and Film[telecined] perfectly) then:
* break project into more than one disk, and devise your bitrate equally

break project into more than one disk, and devise your bitrate equally

What I mean by this, is that you distribute your bitrate high enough to
maintain the level of quality your govern, and apply it to the multiple
of disks you use.. (ie, if 2 disks are needed for a 1 1/2 hr source, then
use a high bitrate like 9000 for both disk set, not 9000 for disk 1,
and 4000 for disk 2 - the technique is to maintain the same level of quality
so that the source does not look like you skimp on one of the disks. Try
and maintain equality or fluidness)

Also, (though most won't agree) when it comes to DVD's, I prefer to use
the maximum (or close to it) and use high bitrate w/ CBR in mind. But,
that is up to you. 1 hr sources would probably be better off anyways
with a 9000 CBR encode, since you're gonna fill the disk anyways.
Why apply a 2-pass to fill 1 hr source onto a dvd disk (even when the
right bitrate calculation) and as a result, skimp your bitrate ??
Your 1 hr source will look better at 9000 CBR vs. a 2-pass VBR (bitrate
calculated) for that one disk
.
Go figure

-vhelp 3296

[Mickey79]

Ave,

I have been playing around with a little bit, from all the knowledge that i have gathered from all these replies.. and i must thank everyone for contributing. I certainly do still have questions and concerns... but let me bring in one more important question that is seeming to stay on top right now...

This dates back to an old post i had made when i was first learning about encoding.. and if i'm not mistaken (I could be, don't quote me on this), i think it was Baldrick who had educated me a little on this matter. From what i learnt at that time from that post was that :

If it's PAL .. Choose "Progressive"
If it's NTSC .. Choose "3:2 Pulldown When Playback"

... and this is what i have been doing ever since.
Now i'm not sure wether this is how it should be, and would love someone has any other input on this matter. I have never chosen "Interlace" for an encoding, even if TmpgEnc Xpress would bring "interlace" in the option box, i would change it to either of the two depending on PAL or NTSC, as i mentioned above.

I have one more important question. I have always, and i really mean Always, encoded every single (NTSC) video at resolution 720 X 480, regardless of it's original frame size, and i have never encoded anything at 352 X 480 ... for the simple fear that "i believed" that 352 X 480 won't be right for a normal TV playing a DVD. Now i'm forced to re-think this issue, and thus i ask, would encoding at 352 X 480 come display on the TV just as good as any other resolution? HOw can the width be smaller then the height, for a TV display? I don't get it.

Thanks again,
Mickey

[junkmalle]

would encoding at 352 X 480 come display on the TV just as good as any other resolution? HOw can the width be smaller then the height, for a TV display? I don't get it.

The DVD player simply stretches the image as it plays it.

You should use 352x480 or 352x240 if your source frame is very small. For example there's no point in stretching a 320x240 image up to 720x480 for DVD -- all your doing is making more pixels for the MPEG encoder to encode.

[Mickey79]

The DVD player simply stretches the image as it plays it

Ave,

Would it display properly/clearly (assuming other encoding attributes were optimum) ? Wouldn't there be distortion by the "Stretching" ?

Also, most video sizes i encode are originally 640 X 480 ... is it better to upsize to res 720 X 408 or downsize to 352 X 480 ?

Thanks,
Mickey

[junkmalle]

A 352x480 image stretched to the full width of a 4:3 TV screen will not be as clear as a 720x480 image on the same screen (assuming the image therein was clear enough to require 720x480 to start with).

I usually resize 640x480 up to 720x480 -- unless the picture is very blurry to start with. For example, a 640x480 AVI made from an old VHS tape is blurry enough that reducing the image to 352x480 won't hurt image quality much.

The other time to consider 352x480 is when you're trying to fit a lot of video on one DVD. A 3 hour movie may show severe macroblocks at 720x480. It might be less clear at 352x480 but there will be much less macroblocking.

[Cornucopia]

Well Scott/Cornucopia if you READ my post you would understand/know that I was SPECIFICALY (spelling?) talking about divx/xvid/mpeg4 avi's converted to DVD/MPEG2......you seem to fail to realise that AN AWFUL LOT (no, I cant put an EXACT figure on it but lets say 35-45% for arguements sake) of all video information is lost in 99% of all conversions from (say) mpeg2 or DV to divx/xvid REGARDLESS of the bitrate used..much more when xvid/divx bitrate IS taken into account (and RESIZING)....which allows RE-ENCODING of these files to be MUCH lower as there is LESS info to RE-ENCODE to start (you cant 'create' video info from info that IS NOT THERE TO BEGIN WITH)..the old 'for this bitrate use this DVD res' arguement is BULLSHIT in mpeg4 to DVD conversions....as for 'scene complexity'..what scene complexity?...the original conversion to divx/xvid has ALREADY dealt with this isuue...are you suggesting that MPEG2 encoding from a frameserved xvid/divx source RE-CREATES some sort of bitrate bottleneck?..Does it re-create the original source (pixels/colours/complexity et al?)?.......me finks..GO BACK TO SCHOOL...those resolutions/bitrates MIGHT have been true in 1996 but arent EVEN CLOSE today..stick to the DV forum where your 'enlightenment' MIGHT (just) hold true...and take 'the blue gnome' and the aussie with you. GOODBYE (or wake up to the 21st century).

monzie, I read your post. I also read the original request. It doesn't specifically state ANYTHING about re-encoding or cross-converting from mpeg4 files. He just wants to know a good way to calc good quality and good rules of thumb.
You do have a few good things to say...
You're right in that when you compress once that alot of the more difficult (to an encoder) material is thrown out. And this does make it somewhat easier for the requirements of the encoder if re-encoding. But that only changes a few constants in the same time-tested equations about encoding. It doesn't EVER throw them out the window, as you seem to think.
Even if the rules are quite different with MPEG4 (which they aren't that much), this guy is talking about standard DVD material--that's MPEG2. Whether his sources are live, DV, other AVI's, MPEG4, etc., it still will need to be encoded to MPEG2 to get onto the DVD and all the USUAL rules still apply. As jimmalenko said, MPEG4's must still be decoded to the full, raw (YUV?) picture before being worked upon. The "recompression makes it easier for the encoder" idea may help somewhat, then you always have to watch out for ARTIFACT BUILDUP, and this will make you want to keep the bitrate up again.
Going back to school sounds like fun if I had the $$$ , but I think that my advanced degree from one of the top 5 Radio-TV-Film schools in the country (US), and my record of helping people here in a number of forums speaks for itself.
BTW, which top 5 RTF school did you get your degree from?

back on topic...

Mickey79, you have a number of good things to consider that different posters have referred to...

1. Interlacing or Not (including IVTC if warranted)
2. Quality constraints--Framesize/Resolution balanced with sufficient bitrate allocation
3. Filesize/Bitrate constraints of the medium--whether to fill or not, Maximums and Minimums

monzie is right in saying that it's never too late to learn a few new tricks, and you sometimes have to go back to the drawing board (and this board is a great place to start).

Scott[/i]

[Mickey79]

Mickey79, you have a number of good things to consider that different posters have referred to...

1. Interlacing or Not (including IVTC if warranted)
2. Quality constraints--Framesize/Resolution balanced with sufficient bitrate allocation
3. Filesize/Bitrate constraints of the medium--whether to fill or not, Maximums and Minimums

Ave,

I'm definately understanding the few important things to consider a lot better now then i used to, and that probably will help me to get that 10% failure rate down to none. I'm still however not very certain about the "Interlace" and "3:2 Pulldown When Playback" issue that i mentioned up there in my post. Still looking to understand that a little better, i'm sure it has an important part in the output as well.

Thanks again,
Mickey

[Cornucopia]

re: interlacing or not...

Some may argue with this statement, but I recommend that you do as little progressive/interlace conversion as possible. That means, if you start with an interlaced source, and it's going to be shown on an interlaced display, KEEP it interlaced. Same with progressive. That part seems easy. The problem usually is when you've got a [P] source and you're displaying [I] or vice-versa.

What do you primarily watch this stuff on? Computer Monitor? (Progressive) StdDef TV? (Interlaced?), HiDefTV? (could be Either)

What's your source? Home video? (Interlaced) TV shows? (Either, but usually stored as Interlaced), Film? (Progressive, but usually stored as interlaced), Animations? (Either)

IVTC should be used ONLY to undo an already telecined Film-type source (Progressive-->Interlaced, 24fps-->29.97fps), otherwise you're mucking things up.

If/When you do convert, it should be done ONCE, just before/during MPEG encoding.

Check the guides and how2's on this and other sites for further info.

HTH,
Scott

[junkmalle]

The first thing you need to know: on standard definition NTSC TV you never see a full frame of video. You see 59.94 fields a second. By the time a field is being drawn the previous field has faded away.

Each field consists of half the picture -- every other scanline. First one field is painted, then 1/59.94 seconds later the other field is painted to fill in the other half of the scan lines.

When viewing a live video source (camera, sporting event, etc) each field is from a different point in time. You see 59.94 unique half pictures per second.

When a computer captures this interlaced video it captures pairs of fields together and weaves them into frames. If the two pictures came from different points in time, and anything moved during that time, the pictures won't line up and you will see comb lines.

Movies are usually made on film. Film is shot at 24 frames per second. To display movies on TV the film must be shown at 59.94 fields per second. This process is called telecining. It is done by first slowing the film down to 23.976 frames per second then showing each frame for either 2 fields or 3 fields, alternating between the two values. So, on average each film frame is displayed for 2.5 fields. 23.976 frames per second times 2.5 = 59.94 fields per second.

Most film based movies on DVD are left as progressive 23.976 frames per second and marked for 3:2 pulldown. This instructs the DVD player to perform the 3:2 pulldown to create 59.94 fields per second for the TV.

Some DVDs have the video with 3:2 pulldown already applied to the MPEG (VOB) file. In this case the DVD player simply displays each field of the frames once. 29.99 frames per second (on the DVD) * 2 fields per frame = 59.94 fields per second.

Any time you capture off-air or cable broadcasts you are recording 59.94 fields per second. The computer (or other digital recording device) weaves the fields together into 29.97 frames per second. If you are going to view this later on TV you can leave it as 29.97 frames per second and it will automatically be sent to the TV as 59.94 fields per second.

If you plan to watch only on a computer monitor (a progressive display) you might consider deinterlacing. Even better, leave the video interlaced and use a player that performs a bob deinterlace on the fly. All forms of deinterlacing are destructive -- except IVTC:

Lastly, the telecine process that is used to create 59.94 field per second video out of 23.976 frame per second film can be reversed back to the original progressive film frames in a process called inverse telecining (IVTC). But for this to be done perfectly requires that there be no breaks in the telecine pattern. Commercials usually cause a break in the pattern. Even if you edit them out you will probably be left with a break. Dropped frames while capturing can cause breaks in the pattern. Sometimes movies are edited on tape (or maybe on a computer) after being telecined leading to breaks. Most software that performs IVTC has an automatic mode that's supposed to compensate for the breaks but I haven't seen any that works perfectly. Often some interlaced frames get through, or entire frames are dropped or repeated leading to jerks in the video.

[Mickey79]

Ave,

Many new and interesting finds in this thread for me... very enlightening. I do have some information/question regarding the Progressive/Interlace issue to try and understand better what options i should choose.

if you start with an interlaced source, and it's going to be shown on an interlaced display, KEEP it interlaced. Same with progressive.

My first question... how do i find out what type of source i have, wether it's Progressive or Interlace? I usually use GSpot to find out Audio/Video Codecs & other details information about a source... is there something like that which i can use to find if a video source is Interlace or Progressive?

What do you primarily watch this stuff on? Computer Monitor? (Progressive) StdDef TV? (Interlaced?), HiDefTV? (could be Either)

I have a Sansui 21" TV. Standard Def (It's not Hi-Def). I have no cluse wether it's Progressive or Interlace. It's a normal standard TV i went and bought at Best Buy. I don't know if i even have the manual to it. But i watch all my stuff on it. Not really on the Computer.

I have a Panasonic SC-HT670 Home Theater System which i use to play my DVDs.

What's your source? Home video? (Interlaced) TV shows? (Either, but usually stored as Interlaced), Film? (Progressive, but usually stored as interlaced), Animations? (Either)

Primarily, Film. Mostly NTSC (23.976 or 29.970), sometimes PAL (25.00). Mostly Xvid's and DivX's. Again, i don't know how to find out if the source AVI is Interlace or Progressive.

I usually Choose "3:2 Pulldown When Playback" for NTSC and "Progressive" for PAL. That's what i have been doing thus far. But i'm not sure if that needs to be the rule of the thumb. I choose the option in the Tmpgenc Xpress encoding settings after i have chosen the source video.

Any suggestions really appreciated.

Thanks,
Mickey

[junkmalle]

My first question... how do i find out what type of source i have, wether it's Progressive or Interlace?

Just look at it! Open the file in VirtualDubMod. Find a section where there is some motion and step through the frames. If you see interlace comb lines -- it's interlaced. If you have a 29.97 fps video and see a pattern of 3 clear frames followed by two interlaced frames -- it's telecined film. 23.976 fps is progressive (unless it was made by someone who had no idea what they were doing). 25 fps could be either -- generally, film based movies will be progressive, live events (like sporting events) will be interlaced.

I have a Sansui 21" TV. Standard Def (It's not Hi-Def). I have no cluse wether it's Progressive or Interlace.

All SD TV is interlaced.