I have dozens of AVI files i would like to burn onto a DVD.
I am very new with the whole process, and i'm unfamiliar with even the basics.
My avi's are 24 minutes long and 190 megs in size, 640x480 in dimensions.
When creating a VCD with a standard CD-R, i'm able to fit 3 episodes onto 1 cd.
How many episodes can i fit onto a DVD with near/better quality then that of a VCD?

All i know is that i have to encode the AVI into DVD format? I'm using TMPGenc to do this; however my files end up turning out to over a gig. When i decrease the bit rate, the lowest it will go is 500 mb.

Other then what i've listed above... i don't know what other steps are involved in the task that i'm trying to accomplish.

Any suggestions? help would be really appreciated thanks!

By using Half-D1 (352x576/352x480), you can store up to 14 Episodes on one DVD, i suggest to store only 12 Episodes.

By using Full-D1 (704x576/704x480) you have enough Space to store up to 10 Episodes on a DVD.

I'm running Win2k Pro sp4 with 768MB physical RAM, so you'd think that memory problems would be the least of my worries...

Updated to 2.521 and found that about 20 minutes into encoding (mpeg-1 to smaller bitrate mpeg-1) I'd suddenly get endless disk thrashing and the system would slow to a crawl. Did some checking and found that available physical memory starts out at over 500MB but slowly shrinks down to <5MB at which time the disk thrashing starts. Processing speed goes way down and estimated time just keeps going up. Also, total memory used goes from about 200MB to over 1000MB as soon as I start encoding.

After I figured this out, I suddenly realized why cut editing the mpeg-1 files was taking so long. Lots of disk thrashing during that, too.

Went back to 2.56 and found that available physical memory stays pretty much constant as I encode and total memory stays at about 250MB. Processing time was about what I remember from before and cut editing is snappy.

According to the revision history, just after 2.56 there was a change in how mpeg-1 files were read, now using decoder by Microsoft through directshow. I'm assuming that's the cause of the problem.

My questions are: Has anybody else noticed this problem? Does anybody know of a VFAPI or other method that can let me use some other mpeg-1 decoder with 2.512? Or should I just stick with 2.56?

I like to use the CRI Sofdec that is built into TMPGEnc to decode MPEG-1. To make sure TMPGEnc uses it, go into "Environmental Settings" from the options menu and raise it to the top prioriy. This will prevent any directshow mpeg-1 decoders from being used.

I could not find anything like CRI Sofdec listed in my Environment Settings. I found another thread which says that the CRI selection does not come with the free version which is what I am using. I probably should have mentioned before that I was using the free version but the link I clicked to get here said "To the bulletin board of the free version of TMPGEnc" so I thought that would be assumed.

I have set the Microsoft MPEG-1 decoders to the lowest priority and am processing a test file now. I assume either/both the Ligos or Cyberlink MPEG-2 decoders (the only other MPEG decoders listed) are decoding the MPEG-1 file for me.

I had assumed that it was included in the free version. (I have not used the free version in such long time and as there is no forum for the plus version this is where most people post.)

If you don't want to use the Microsoft decoder then just install ffdshow and configure it to decode mpeg1 if you don't want to use ligos or cyberlink.

Seems strange that you are having this problem with memory hogging even when cutting MPEGs.
When cutting the MPEG decoder is not used (or so I believe). The MPEG is cut at binary level and therefore doesn't need a decoder.
The decoder is only used when placing the cut points.

You could try just raising the priority of the MPEG1 decoder to highest in the list.

Ashy: by "cut editing" I mean when I'm cutting commercials out of MPEG-1 files (640x480 24fps 4000, captured from cable TV with Asus Digital VCR software) using the Source Range filter under Settings, so that I don't see commercials or credits when I recode at a lower bitrate. I assume that the MPEG-1 decoder is used to show me the still pictures I use to determine where my cut points are, and that's where the slowdown is. What I mean to say is that with 2.56, I could drag the slide bar around and pick pretty much any point in a 3GB file and have it display in less than a second. With 2.521, such a move takes several seconds. With 2.56 I can hop forward 10 frames almost instantly and backwards 10 fairly quickly. With 2.521 both operations drag.

Recoding appears to be about the same speed until I run out of physical memory and the disk thrashing starts. Changing priorities and trying to use different codecs doesn't seem to have any effect on this problem. Nor did downloading an evaluation copy of 2.521plus and trying the CRI Sofdec that David suggested. (Thanks anyway, David!) My test runs from over the weekend turned out to be useless as they apparently the files weren't big enough to trigger the problem. (Size of the output file appears to be what sets it off, though at this point I'm not sure of much of anything.)

The fact that I still have the same problem even with the plus version seems to indicate I'm doing something really stupid with configuration or there's something really screwy with the codecs on my system.

I've gone back to using 2.56 for now. The only annoyance with 2.56 is that when I cut out commercials, the actual cut appears to be a few frames later than what I specify. So I just tell it to cut a few frames earlier than I would otherwise.

Just in case anyone is interested, I eventually discovered the source of the runaway memory problem. All my source mpeg-1 files were created by Asus Digital VCR v2.2 or so. I have since gotten a new video capture card with new software and am now creating mpeg-1 files with Intervideo's WinDVR. TMPGEnc 2.512 seems to have no trouble at all working with these files. I can only assume that there is something wrong with the mpeg-1 files that Asus Digital VCR creates, even though they seem to play or convert well enough with most other programs, including earlier versions of TMPGEnc.

Sorry,
i don't undertand (with TMPGEnc)how to convert in video source:two(or more) files AVI (file1.avi,file2.avi..);into output file name: one file MPEG filex.mpg.If is possibe what can i do? I try to search in BBS but I couldn't find anything about.

Thanks for your help.

Just name your files something like File001.avi, file002.avi...and so on then just load the first AVI in the list and the rest will automatically load in sequence.

You will have to enable the 'Open sequence files as a movie' option in the General settings first.

How to create MPEG Files which can be accepted by PHILIPS VideoCD 2.0 Tool Kit. Because when I tried to make a MPEG file using TMPGEnc and use it in the Authoring program it says It is not a valid MPEG File. the Same MPEG is accepted by Nero as well as WinOnCD.

What are you aiming to do just mux it?
If so you need to load seperate elementary Video and audio streams.

Had some bad clusters on my harddisk so TMPGEnc would not start. Moved files with scandisk /r and now when reinstalling I only get a message "can not remove old version, 1610 contact your support."
Removed all entries TMPGEnc in registry and made a search for files. Removed all files manually, but still, same message when I try to reinstall.
Any hints??

Disally

Just download the non install version.
It isn't really required to install TMPG. You can simply run it from the unzipped archive.

I've a problem with TMPGEnc! I want to convert a *.avi-file to a *.mpeg-file, but my problem is that when I preview my project the picture is up-side-down, why?!
This phenomenon only seems to happen with a *.avi-file, not with a *.mpg-file!!!

Why does this happend and what can you do about it???

//niclas

That's a problem with your Codec/DirectShow-Installation. Use ffdshow as Decoder for your AVIs. It has an Option to correct this.

I have cut a clip out of a computer-readable MPEG file, but TMPGEnc seems to only encode in those SVCD, VCD and DVD formats. How can I get it to play on a PC like a normal MPEG?

You say you have cut a clip out of the MPEG. How did you acheive this?
Did you actually use the MPEGtools to simply cut the clip out or did you indeed re-encode the part of the MPEG you wanted using the wizard?

If it was the latter then you needn't have carried out that step.
You can cut clips from MPEGs using the Merge&Cut feature in the MPEGtools.

However if you just want to create a normal MPEG1 MPEG then you will need to unlock the settings and not use the wizard.
Simply load the unlock.mcf template by clicking the 'Load' button in the main window.
You will find it in the 'Extra' folder.

I have cut a clip out of a computer-readable MPEG file, but TMPGEnc seems to only encode in those SVCD, VCD and DVD formats. How can I get it to play on a PC like a normal MPEG?

I have a ~150GB avi I am encoding for DVD which is captured from a 2-hour VHS tape. I was hoping to use the bitrate calculator in the project wizard to figure out a good bitrate to fit two such mpg-encoded avi's onto one 4.7GB DVD.

My own lame calculations get me about 2000Kb/s, which doesn't provide very good quality.

When I try and use the project wizard, I get stuck at the "Judging Field Order" process... the program is not frozen (task manager shows it still reading in bytes), but it sits there at 7% forever. My fear is that it will try and read in 150GB to figure out what I already know, that the field order is Top A field first.

So, two questions:
1) What's up with the "judging field order" routine? The avi's are captured at 640x480, YUV encoding, PCM audio. Does it need to read through the whole file to make the field order determination? If so, I wish I could that part.
2) Can anyone suggest a good average bitrate or good bitrate calculator? I want to go VBR, with mpeg audio. The beginning of the project wizard seems to indicate I can get "235 minutes" of such video onto a 4.7GB DVD, which would be ideal, but I wonder at what bitrate.
2b) Can I use a bitrate other than 384Kb/s for the audio and still use it on a DVD?
-Andrew

First of all have you never heard of compression?!
I mean 150GB?
There are codecs out there such as the HUFFYUV codec which are lossless if you are looking to maintain maximum quality.
However many lossless codecs such as MJPEG and MPEG2 can maintain very high quality with no noticible loss in image quality at a fraction of the filesize which I think is your problem in TMPG.
The file size is so large it will take quite a long time for it to scan the file.

Secondly in regard to maintaining quality, if your system is powerful enough, I would have captured to the resolution for your target output. Resizing causes a loss in image quality. You will have to resize the image twice. Once initially from the standard VHS resolution and again to the standard DVD resolution.
You likely have chosen 640x480 as this is what you have been advised by some on the net. While this resolution would be OK for downsizing later I wouldn't recommend it for upsizing. You would notice a difference in the quality if you had captured to the standard DVD resolution.

Also I don't put too much faith in TMPG's resizing abilities. Far better quality is obtained resizing with AVIsynth's bicubic resizer.

Having said all this I have 2 suggestions.
One is to install the FFDSHOW codec and enable it to decode RAW video. This codec handles uncompressed AVI far better than the native windows renderer can and uses much less CPU resources.
The second, if the problem persists, is to frameserve this file from either Virtualdub or AVIsynth to TMPG.
However if you only rerquire to find out the field order then simply extracting a small clip form the AVI using Virtualdub and then loading that into the wizard should give you the info you need.

As for a bitrate calculator there are many, but the easiest and most accurate I have used can be found here: http://www.dvdrhelp.com/calc.htm

Thanks for the information.

Of course, I've heard of compression. My reason for not capturing the video this way is two-fold: 1) I was having problems with frames being dropped when capturing real-time video (i.e. the compression time was too much for the real-time capture) 2)None of the compression codecs I tested worked well to capture, then compress using MPEG. I didn't like the idea of compressing twice.
You mention some options, I'll look into those. :)

I captured at 640x480 because I thought that was my target resolution, considering the 4:3 aspect ratio of the source. I didn't think creating the resulting mpeg-2 video at 720x480 was really resizing, but just kind of padding the sides... maybe I'm wrong on that. Unless you're suggesting I was resizing to something much smaller, such as 352x240, which I'm not. The quality at the reduced resolution was quite bad, even though it seems like it should be enough to reprsent VHS information.

I'm kind of new to this, and don't read many instructions online (just muddle through :) so I appreciate your help, I'll check into each suggestion.

The reason that 352x240 is not a high enough resolution for VHS and the quality is lower is because even though the resolution is the same as standard VHS resolution, VHS draws each frame twice because the frames are interlaced and split into two fields. First the odd scan lines are drawn then the even scan lines and thus creating 480 lines for one image.
If you encode to 352x240 you are throwing away half of the scan lines and therefore affecting the image quality.
With a resolution of 704x480 or even 352x480 you are retaining all the scan lines in the image.

As for your target resolution being 640x480. You could use this if you wanted and then add padding in the form of black borders at the sides, but who would want to watch a movie with 32 pixels of black area at each side?

DVD resolution for a 4:3 image should be 720x480 using a pixel aspect ratio of 1.25:1
The resolution you are using 640x480 is using a 1.33:1 pixel aspect ratio and a display aspect ratio of 4:3 however DVD resolutions do not use a 1.33:1 pixel aspect ratio they use a 1.25:1 pixel aspect ratio.
This means the pixels are not as wide and therefore more pixels are used to create the 4:3 image and hence create a higher quality image.

About the freezing of the "Judging Field Order" process: go to "options"
-> "enviromental settings" -> "VFAPI plug in ". Then move the "directshow" to priority level 1 or 2. That should do it.

Thanks Taurus, that definitely took care of the Judging Field Order issues. I assume bumping the priority up means it will use DirectShow over some other decoder that was not working as well for large files?

and ashy, I had a question or two for you regarding some of your suggestions. Maybe I'm just showing my ignorance to all of this, but I thought the video resolution determined the aspect ratio. For example, I was capturing at 640x480 because I wanted to maintain the 4:3 ratio of the source. When I capture at, say, 352x480 and play back the captured file, it is stretched vertically. I understand what you were explaining about the interlacing and the alternating fields and wanting to capture all the interlaced data, but is there some way I need to specify to TMPGEnc that this is what I've done to the source video? Will setting the source as 4:3, Interlaced be enough?
I'll play around with it more, but if you have any remarks that might help I'd appreciate it.

Also, using MJPEG as a lossless capture format didn't seem to make an appreciable difference in file size over the YUV2. I'll see what HUFFYUV can do.

There are 2 types of aspect ratio, Pixel Aspect Ratio (PAR) and Display Aspect Ratio (DAR)
PAR defines the actual shape of the pixels while DAR is the actual shape of the image. This is why both 640x480 and 720x480 can both have a 4:3 DAR aspect ratio however they have different pixel aspect ratio's.

As for 352x480 or 'Half D1 resolution' as it is called being stretched, you could simply resize this with TMPG to 704x480 this would retain the same Display aspect ratio. Even so the MPEG2 DVD standard has support for this resolution natively as long as the output is 4:3, however many DVD players will also handle a 16:9 output with this resolution no problem.

If you wish to play a 352x480 MPEG2 file on the PC then you will need a DVD compliant software player such as WINDVD. Media player does not respect the display aspect ratio flags in the MPEG and therefore the image will be distorted.

Thanks ashy. For any one else who's trying to get the basics down, this primer at doom9 was helpful:
http://www.doom9.org/aspectratios.htm

Hello, is there any way to be able to read PNG files as a bitmap input?

Thanks.

You will need the PNG plugin available here:
http://news.povray.org/povray.binaries.utilities/15796/

So, basically, did a search on converting outputting MOVs from TMPGEnc... Nothin.

Is there a plugin to do this for TE? I want to go straight from the streams into MOV format, are there any plugins for this?

Joe

Seriously? What the heck? My post keeps REposting, its not intentional...

Joe

Don't refresh, use the 'View article' link at the bottom of the page.

What are the relative merits of encoding a widescreen video in 4:3 aspect with black bands top and bottom versus anamorphic for a DVD.

There seem to be various pros and cons and I can't decide which is the better option.

Cheers,

Bob.

If the source is 16:9 and the output is for DVD then you should always encode as 16:9.
The will give a much better quality picture as the resolution is not wasted on black bars. It will also give the user a choice of which format he prefers when viewing.

Thanks very much, that's what I'll do then. I may have to come back here to ask what settings I should use in TMPGenc if I can't work it out for myself.

Cheers,

Bob.

I may have to do this the long way round but if there's a quick way:

The video file I have is 4:3 with black bands top and bottom. What I need to do is remove the black bands and resize the 16:9 picture in the middle to 720x576.

I can do this if I crop the picture and then choose the right output format, but is there a way of getting TMPGenc to automatically produce an output video which removes the black bands and resizes the 16:9 to be an anamorphic 4:3 video?

Thanks,

Bob.

No, you have to use TMPG to crop the bars out first. You will need to set the input aspect ratio to 4:3 and the output ratio to 16:9 and obviously the output resolution will have to be the standard DVD resolution.

Ok, thanks. That's what I thought I'd have to do.

Bob.

One final question I can't find the answer to. Given that a PAL DVD is 720 x 576 pixels for a normal 4:3 aspect ratio what is the vertical height in pixels for 16:9?

Thanks,

Bob.

It depends on what you mean.
A 16:9 DVD resolution is still 720x576, but because of the 16:9 flag that is encoded with the MPEG data the image will be resized to the correct aspect ratio by the DVD palyer otherwise the image would be stretched as you could imagine.
The actual height of the image when the correct aspect ratio is applied would infact be 432 with the rest being padded with black bars by the DVD player.

Thanks, that's what I needed to know in order to produce titles and animations to go in front of a 16:9 video.

Cheers,

Bob.

And finally, many thanks indeed for your advice. I ran two encodes overnight on a 16:9 file with black bands top and bottom. The encode was to change the 8MB/S bitrate to 3.5MB/S so the whole video would fit on a DVD.

Version 1 was a straight forward encode, version 2 cropped the picture to 16:9 to remove the black bands and then resized to anamorphic 4:3. I then burned a test DVD with clips from each version.

I was amazed at how very much better the quality was of the anamorphic version. I suppose it should have been obvious because having the black bands in the MPEG and VOB meant that nearly half the space was taken up with black bands as opposed to picture information.

I just wanted to pass this on to anyone else doing something similar: If your picture is 16:9 for DVD it's definitely worth artificically making it anamorphic.

What a pity you can't do this for VCDs.

Cheers,

Bob.