Hi all,

when creating a SVCD from a DVD, I create MP2 files from the relevant AC3 tracks. These MP2 files are already in SVCD format (224KBps, 44,1KHz).
Is it ok to just mux the track with the video mpeg-stream?
Would save a lot of time and HDD-space rather then creating a temporal wave file which has to be reincoded.

Thanks.

SP

In a Word...YES, This is what you should be doing as re-encodeing the File will Just degrade the Quality so you Should just encode the Video then Mux the audio to it after...

Hi All,

I transfer the video from my Victor DVHS recorder to my pc. The original source
is 1920x1024 20Mbps 29.9fps interlance. The size too big and can't play smooth at my pc. So I try to use Tmpgenc to encode it to 1024x576 at 15Mbps CBR.

I have some questions about the encoding.

1)can i encode it to non-interlance? I have try to set the encoding method in non-interlance. But it have some "wave effect" in fast motion.

2)can i encode the original souce to 59.9fps to make the video more smooth.

Please advice.

Thanks,

Chisato

First all can I ask where you intend to view this MPEG and what for?

>1)can i encode it to non-interlance? I have try to set the encoding method in non-interlance. But it have some "wave effect" in fast motion.

The artifacts you are seeing are interlace artifacts.
You can encode it to Non-interlace(progressive), but you will have to de-interlace it first. To be honest there is no point as you will lose quality unless you intend to watch this MPEG on a PC or other progressive display.
Depending on how your DVHS records the material you may simply be able to IVTC the material to return it to a 23.976 fps progressive source. This will allow you to encode as progressive and will prevent any interlacing artifacts and won't affect the quality.

Secondly the bitrate and resolution you are encoding to are quite high and will take quite a powerful machine to render. So unless this MPEG is only for archiving purposes and not for general viewing for others, then I wouln't use such a high setting.

>2)can i encode the original souce to 59.9fps to make the video more smooth.

In a word NO. Firstly it WON'T make the video play more smoothly, no matter what fps you set it at, as a matter a fact if you change the frame rate from the original it will play less smooth because of the conversion.
Besides 29.07 fps is high enough to produce a smooth playback any way and your playback problem is not due to the frame rate, but your high resolution and bitrate settings.

Hers's an answer from someone who does this ALL THE TIME. Me.

As stated, 1920x1088x60i (interlaced) is a very cumbersome format and is overkill when the goal is to archive for DVD and/or view on computer Mpeg players.

I convert 16:9 video source to 960x544x60p (progressive). A bitrate of 9Mbs seems adequate and saves file space. The result is stunningly smooth AND crisp when played with WinDVD. Furthermore, this format can be easily converted to 720x480x60i for authoring on DVDs.

The key is AVISynth. I use a new version of DVD2AVI (DVD2AVIT3.zip) and the mpeg2dec3.dll plugin to source transport streams.

A very subtle point is that video (non-film) source material MUST be "Bob()" deinterlaced in AVISynth first. This fixes the 1/2 pixel jitter which would occur if you just did a "SeparateFields()" to go from 60i to 60p.

Here is a script outline (assuming mpeg2dec3.dll is used):

Mpeg2Source("video_file.ts").YV12ToYUY2().AssumeTFF()
Bob() # convert to 59.94p
BicubicResize(960,544)

For film, the script needs to do an inverse telecine:

Mpeg2Source("film_file.ts").YV12ToYUY2().AssumeFrameBased()
SeparateFields().SelectEven() # convert to 29.97p
BicubicResize(960,544)
Decimate() # convert to 23.976p

(This is not my usual coding style so the above syntax may be off a bit.)

Legacy 4:3 broadcast material which is "windowboxed" can be cropped and sized down to 720x480 instead of 960x544 without any loss of resolution.

Now to the original question. Absolutely you should encode at 59.94 to get video which is smooth. For this you need to select the MP@HL profile in TMPGEnc. Extend the number of frames in a GOP to about 36 and select a bitrate of about 9Mbs for normal motion video. TMPGEnc will automatically select progressive when loading source at that frame rate.

Film material which has been reduced to 23.976fps can push to even lower bitrates.

The real fly-in-the-ointment is audio sync. You will probably be muxing the original AC3 back into the video. Frames can be trimmed or black frames added to the start of the video by AVISynth in order to compensate.

Or use AC3Delay.exe to adjust the audio file itself.

Converting to 720x480x60i for DVD also requires some subtle manipulations to put back the 1/2 line offset for each field. My script is not handy but goes something like this:

vid = Mpeg2Source("file_960_544_60p.mpg").BicubicResize(720, 479)
vid_ff = vid.SelectEven().AddBorders(0, 1, 0, 0).VerticalReduceBy2()
vid_sf = vid.SelectOdd().AddBorders(0, 0, 0, 1).VerticalReduceBy2()
vid_new = Interleave(vid_ff, vid_sf).AssumeFieldBased().AssumeTFF()
return vid_new.Weave()

(I may have the AddBorders() parameters swapped. Experiment with this. Note that adding 1 line to 479 lines gets you to 480 and reducing that to 240 gets you the needed 1/2 line offset for each field.)

My AVISynth scripts shown above failed to include a ConvertToRGB24() at the end which is needed by TMPGEnc.

Even with that, these scripts are sometimes rejected by TMPG and I have to go through VirtualDub first.

> First all can I ask where you intend to view this MPEG and what for?

I will play my video at projector and pc. :D

Thanks ashy and wcpaul. I will try in coming hoilday.

>Now to the original question. Absolutely you should encode at 59.94 to get video which is smooth. For this you need to select the MP@HL profile in TMPGEnc. Extend the number of frames in a GOP to about 36 and select a bitrate of about 9Mbs for normal motion video. TMPGEnc will automatically select progressive when loading source at that frame rate.

Hmm...
Are you referring to originally encoding the source to 59.94 or actually converting the original 29.97 source to 59.94?
If this is the case this is bad advice.
Surely you realise any kind of frame rate conversion is undesirable and will cause playback artifacts and won't make the playback any smoother. Even if you did convert the frame rate to 59.94. It would still actually be 29.97 material.

"Hmm...
Are you referring to originally encoding the source to 59.94 or actually > converting the original 29.97 source to 59.94?
If this is the case this is bad advice."

Ashy, have you ever worked with 1920x1088 HDTV source material?

Each FIELD is 1920x544. And they arrive at a rate of 60 FIELDS PER SECOND!!!! That is FAR HIGHER resolution than DVD at twice the frame rate.

Resizing these fields and running them progressively results in 960x544 video at 60fps. No computer screen is 1920 pixels across so the horizontal resizing is not significantly lossy.

Viewing my 960x544x60p material at full screen with WinDVD looks as good as any Hi Definition TV display I have ever seen because, when viewed at proper distances, the apparent resolution of HDTV is in this ballpark.

When you compare a Leno Show music performance at 960x544x60p (31 MPixels/sec) to any DVD film material at 720x480x24p (8 MPixels/s) the result is totally amazing. 31Mpixels/sec IN NO WAY goes beyond what the eye and brain can process so you get a picture with 4x the information (along with kick-ass AC3 audio).

Right, first of all, this wasn't the god damn question!
I will quote the question:
>2)can i encode the original souce to 59.9fps to make the video more smooth.

Ok? My answer was to this question. Not to the fact that using Avisynth scripts and BOB de-interlacing the source would double the frame rate thus resulting in 59.9 fps

I was referring to frame rate conversion and not simply turning each field into frame and doubling the frame rate using the BOB filter.

This is your answer:
>Now to the original question. Absolutely you should encode at 59.94 to get video which is smooth. For this you need to select the MP@HL profile in TMPGEnc. Extend the number of frames in a GOP to about 36 and select a bitrate of about 9Mbs for normal motion video. TMPGEnc will automatically select progressive when loading source at that frame rate.

You did not state in the above answer that the frame rate should be set at 59.94 in TMPG only when using your script.

I was assuming that the source was going to be directly loaded into TMPG and you were suggesting to simply set the frame rate at 59.94.
As any body knows frame rate CONVERSION is a bad thing with TMPG or any other process for that matter.

It's fine to use 59.94 if loading your script into TMPG, but you did not make this clear and this is why I posted the reply I did.

>Each FIELD is 1920x544. And they arrive at a rate of 60 FIELDS PER SECOND!!!! That is FAR HIGHER resolution than DVD at twice the frame rate.

Don't you mean 60 frames per second. 60 fields per second would be the same as DVD seeing as NTSC video material is 29.97/30 fps which equals 60 fields per second.

Maybe it's just a misunderstanding, but maybe you should clarify yourself in future before having a pop.

Well, I think it is fairly obvious in my post that I was advocating conversion to 59.94fps progressive outside of TMPGEnc then using TMPGEnc to encode without any tranformation. Hence the script examples and instruction to use MP@HL.

I never use any of TMPGEnc's crude filters or conversions and did not think to make a distiction up front.

None of us have the time to write dozen page technical white papers on these techniques. We can just hope to point people in a direction and let them sort out the low level details.

I am having problems with mpeg tools. Last night it worked fine but today it just hangs. What i am doing is trying to merge to files together and after i load them and hit edit it just freezes/goes white. Anyone have any suggestions ?

Thanks in advance

TMPGEnc Plus v2.520 (2.520.54.163) Release Notes:

Added the MPEG decoder CRI Sofdec. Now, input of MPEG-2 files is a standard feature. Plus, the picture quality (brightness, color) shown on your pc display will be close to the one on your TV. So the editing will be easier.

*when using CRI sofdec:
remove the check on [ Output YUV data as Basic YCbCr not CCIR601 ]
[ Setting ] => [ Quantize matrix ] => [Output YUV data as...]


Does this mean that TMPGEnc Plus v2.520 now supports YUY2 and/or YV12?

If this is true, then I guess I can stop using ConvertToRGB24() in my AVISynth scripts as long as I make sure CRI Sofdec is set to the highest priority (correct me if I'm wrong).

Yes you are wrong.

The setting is not referring to the format of the input files, but is related to the output colour space. What it is referring to is the the colour and luma range values of the output.
For instance TV luma ranges are 8-235(CCIR601) and PC is 0-255(basic YCbCr)
If you have a source such as DV which has a range of 0-255 and wish to view this on a display that has the same range then you would use this option if you want the values to remain the same in the output.
However for most sources and output to TV 8-235(CCIR601) would be more appropiate as these values are identical to your TV's luma range. Use the wrong setting and it can affect the brightness and contrast of the output.

>Does this mean that TMPGEnc Plus v2.520 now supports YUY2 and/or YV12?
TMPG has always supported YUY2 and YV12 as input but it will convert it to it's native RGB format first.
However TMPG outputs to YV12 which is the native colour space for MPEG and is what the settings above relate to.

In conclusion you should still use the RGB command in your scripts.

Thanks for your quick and detailed response. So, does this just mean that the new version does a better conversion to RGB as an input, which should result in a higher quality output file (better output color)?

I think what it means is that the codec uses the CCIR601 colour space as imput, which of course most MPEG2 material does therefore to maintain the same luma levels you need to set the output to CCIR601 also hence removing the check.

Hope someone can help me here.....

I'm new to this kind of thing, but I've successfully copied from DVD to VCD (I need to be able to play the discs on my stand alone DVD player), using SmartRipper on the DVD, then DVD2AVI to create separate AVI and WAV files, then TMPEnc, and finally burning the VCD using Nero. No problem with that. (although there must be an easier way!)

Now I want to copy some stuff to the same kind of disc format (VCD), but from a VCR. My ATI All-in-Wonder (9000 Pro) card captures it OK from the VCR, in some kind of MPEG format. But whatever I seem to try, using the same media (Sony CD-RW) that worked OK for the DVD to VCD copy, I can't get this into something that my DVD player will play.

I figure that, as the last parts of the chain, if I can get it encoded by TMPEnc and then burned with Nero, it will work. But I can't get from the ATI capture file into something that TMPEnc will work with. I can convert to AVI OK, but then TMPEnc won't recognise that file for the audio. Nero claims to be able to create a VCD from the AVI file, but that didn't work (and it takes forever!)

I can't use DVD2AVI, cos whatever MPEG format its captured in by the ATI card doesn't equate to the MPEG-2 that DVD2AVI wants (vob, mp2, m2c, mpv)...its just mpg, which DVD2AVI won't seem to recognise.

I figure there must be something in the settings of one or more of these programs, or possibly in the ATI capture itself, that I need to tweak, but I can't work it out

Help!!!

Thx
Brian

Yes you are doing everthing wrong.
Firstly when using DVD2AVI you don't need to create an AVI. This is just a waste of time and quality.
You should have selected to create a project file and a wav and then imported these into TMPG.

When using your ATI card to capture to MPEG you have to capture to the right format.
You are probably capturing to MPEG2 which is not VCD compliant and are using all the wrong settings.
VCD has strict standards which you must follow to produce a fully compatible VCD.
To open MPEGs in DVD2AVI you simply have to make sure that the format is MPEG2 then just select 'All files' from the drop down box.

Also NEVER use Nero to encode you MPEGs the result will be awful.

My advice to you is to set up your ATI capture card to capture at full resolution using as high bitrate to MPEG2 then use TMPG to re-encode to MPEG1 VCD.
This will give you the best quality, but if you would prefer to use your card to capture directly to VCD format then follow this guide here:

http://www.dvdrhelp.com/ATI_AIW_Guide.htm

<Firstly when using DVD2AVI you don't need to create an AVI. This is just a waste of time and quality. You should have selected to create a project file and a wav and then imported these into TMPG>

I guess I expressed myself poorly...this is actually what I did do.

<When using your ATI card to capture to MPEG you have to capture to the right format. You are probably capturing to MPEG2 which is not VCD compliant and are using all the wrong settings.>

I just checked the capture settings in the ATI setup, and it was using the inbuilt "Video CD" settings, which are MPEG-1, 352x288, PAL 625, 1.13MBits/sec, with the audio at 44.1. That's correct for VCD, isn't it?

<<My advice to you is to set up your ATI capture card to capture at full resolution using as high bitrate to MPEG2 then use TMPG to re-encode to MPEG1 VCD>>

OK, I've created a custom capture profile, using MPEG-2, 720x576, PAL 625, 1.18 Mbits/sec....is that likely to be OK? I want to be able to get the resulting file (about 50 mins) onto one disc.

One other thing that might be affecting the results. I copied some DVD to VCD last week, and it worked OK, except for one of the discs (there were 10), which won't play in the DVD player. The rest are OK, and all 10 play OK in the drive on the PC. All 10 were copied using exactly the same process, using CD-RW media from the same brand, supplier, and even from the same box. I don't know why one of them wouldn't work, unless the disc itself was somehow at fault. So I re-copied one of the discs that HAD worked onto a new disc (same brand and supplier as before, but from a different batch - and actually the SAME batch that I'm currently having a problem with) and guess what....it won't play. I'm gonna try and re-copy the one that hadn't worked onto a new disk and see if it does - if so, I'm going to be blaming the media rather than the capture format, although that may also not be working OK.

Thanks

Brian

>OK, I've created a custom capture profile, using MPEG-2, 720x576, PAL 625, 1.18 Mbits/sec....is that likely to be OK? I want to be able to get the resulting file (about 50 mins) onto one disc.

No this NOT ok. The bitrate is far to low. The result will be awful. You will end up with a blocky MPEG.
The bitrate should be quite high. If you have enough space on your drive (at least 8 - 10gb) then set the bitrate at something like 8000kb/s or as high as you can get it without dropping frames in the capture.

I use an ATI AIW 8500DV and the Way I do it and get awesome Results From VHS Tapes is to Capture to Mpeg2 useing WinDVR 3 But you can use ATI"s MMC, and set the Capture Bitrate to the Maximum which is 15mb per sec or 15,000kbs and set the Quality slider all the way up to 100% that is if you PC is Fast enough, In WINDVR3 the Bitrate goes all the Way up to 20,000kbs and the Quality is the Best I have seen...Then take the Mpeg file you capture and Load it into DVD2AVI (Choose "All Files" from the Dropdown Menu or Just drag and Drop the Mpeg2 file into the DVD2AVI window as DVD2AVI will Accept mpg files)...Now Just make a D2V project file with DVD2AVI and it will Demux your Audio ,But Just leave the audio for now, Now load the D2V file into Tmpgenc and Make sure the "Stream" setting is set to "Video Only" ,now just encode it with the VCD settings and encode it ,you will be left with a "m1v" Mpeg1 video file, Now just go to "File" to "Mpeg Tools" to "Simple Multiplex" and Load the M1V file you just encoeded in for the Video and load the Audio that DVD2AVI extracted from your Captured File and Load it in for the Audio, then Choose "Video-CD" from the Dropdown Menu then click "Run" and it will join the Audio and Video together into a VCD Compliant Mpeg1 file that you can Burn as a VCD with Nero or some other such Program....good Luck...

Thanks for that...I'm OK up to the streaming options in TmpgEnc, but when I get there, all the options are greyed out, with the button for "audio and video" checked. How do I get to where I can chanage that option to "video only"?

Brian

Thanks for that...I'm OK up to the streaming options in TmpgEnc, but when I get there, all the options are greyed out, with the button for "audio and video" checked. How do I get to where I can chanage that option to "video only"?

Brian

Click the 'Load' button and locate your TMPG template folder. Inside the Template folder is another folder labelled 'Extra' inside is a file called 'Unlock.mcf'
Double click this and all the options will be unlocked.

hi, i was just wondering, once a 2 hour vcd has been encoded with vbr and the bitrate is right and all that, how do you record it to a vcd disc, is it still not the time that matters?

You can only fit 80omb on an 80min disk. If your vcd is latger than 800mb then you will have to split the MPEG.

thats bulls**t! as long as the file is not larger than the disc you have no problem burning it as a vcd. and it even plays in most dvd standalone

So what the hell's your point?

What did I just say or can you not read!

I already said that you can burn up to an 80min disks maximum capacity.
800mb is the maximum capacity on an 80min disk.
If the file is any larger then you will need to split it or do you still not understand the size limitions and characteristics of burning files to disk. Maybe you are some sort of magician and have special way that we would all like to know about of burning larger files than 800mb to an 80min disk, so please share it.

Is it possible to make auto start and endless repeat with the software. It will be usefull for fairs and exibitions.
Mike

Yes it is possible.

I have tried to use TMPGEnc several times to convert AVI to MPEG2 - the process takes 5 or 6 hours but the movie always cuts off after 20 mins or so - and the sound is missing - am I missing an obvious point?

If the Video Cuts off after a While that can usually be fixed By going to "Options" to "Enviromental Settings" to "Vfapi Plugins" and raise the "Direct show File reader" to "2", But if the audio isn"t there then it is Because the Audio format in the AVI file is of a Format that is Not supported by Tmpgenc, You will Need to Evxtact the audio from the AVI file to Wav format and Use that as the audio source in Tmpgenc, you can use Virtual Dub or Even Better use "AVI_MUX" to demux the audio to Wav format...

Hi all,

when I convert avi to svcd it seems that the colour red is missing, people are blue in their faces - that also appears in the previe with tmpgenc. What could the reason for this be? Thanks in advance for helping.

This sounds a Little Like a Well Known Problem with a Codec called "Angel Potion" check you device manager under Video codecs for a Codec called "Angel potion" or it might be under "apmpeg4v3.dll" if you have it Delete it and re-encode the File...But if this is not the Problem you can try Raiseing the priority of the "Direct show" in the Vfapi Plugins, or you Can use the "Advanced Settings" under "Color Correction"(Simple Or Custom) to change the Balance of Colors...

Thanks for your reply, but
- I have no angelpotion on my machine
- already raised directshow to 2

The thing is that it looks fine in tmpg itself, but I am using dvd2svcd with fitcd and autofitcd and with that the preview already looks awful. Any other idea? Anything I am missing? Any other codecs causing trouble, any software missing that is mandatory? Weird. The avi is dvx4 coded, but that shouldn't really matter, right? Thanks again for helping.

regards, Ed

update: just installed CCE as alternative to Tmpg, faces are still blue. Any dvd2svcd specialist around here? Any other hint?

THANKS!

I wanted to burn a VCD last time and I learned to use the overburn method. But I also fooled around with other things and I don't know how to replace it. It's the Status bar under Nero's preferences menu, in "General". What is the default min. that the yellow and red bars be set? I want the default so I'll know when to use the overburn and when not to. Thanks!

Well it a Genius to Figure out when you use Overburning and You don"t need Nero to tell you, When your Mpeg/VCD file is Over say 795-800mb then you will need to use Overburning..But I think the red Line starts at 700mb as that is How Much(In Nero) a CD-R holds...

You can put 800 + 13% on a 800 Mb VCD disc. Error correction is omitted on a VCD because MPEG already takes care of that.
I've burned 1 Gb on a 900 Mb VCD. No problem with nero if your equipment can handle it.

Sory Yeppie But there is No way that you can get 904mb(800+13%) on a single 80Min 700mb CD-R even with OverBurning, I don"t Know about 90min CD-R"s But I know that a 700mb 80min CD-R will Fit Maybe a Max of 830mb but usually closer to 820mb..I can see 700mb +13% but not 800+13%...

@minion: the question was about the status bar in nero. My point is that the bar behaves differently with a VCD or a data cd. The amount of Mb's on the CD cover 74 min/800 Mb is related to data/audio cd content. This content has error checking burned with it. When you choose in nero for a (S)VCD and drop a 800 Mb mpeg file in it the status bar doesn't go all the way to 800 mb but 800-13% =696 Mb. So on a 800 Mb cdrom you can burn 800+13% = 904 Mb mpeg files without overburning. Give it a go. You need nero v5.5.9.14 or higher. Checkout: http://cd-rw.org/software/version_history/nero.cfm

;O)

@Sakuya: defaults are 74:00:0 and 80:00:0

Are you Talking about 700mb 80minute CD-R"s??, If so then you are Totally wrong..I agree you can Fit 700mb+100mb on a CDR in (S)VCD Format, But if you encode a File and Windows says the File is 900mb you are NOT going to be able to Fit it on a 700mb 80min CD-R..and 74 Minute CD-R"s are only 650mb + about 90-100mb for error correction(depending on the Brand/Quality of the CD-R) But if you are Talking about those 90 or 99minute CD-Rs then I"m sure it is Possible, But on a Standard 80min CD-R you aren"t going to get 900mb on it no matter what....

Yes you are right. Slip of the mind. ;O)

>When you choose in nero for a (S)VCD and drop a 800 Mb mpeg file in it the status bar doesn't go all the way to 800 mb but 800-13% =696 Mb. So on a 800 Mb cdrom you can burn 800+13% = 904 Mb mpeg files without overburning.

How the heck did you arrive at this conlusion?

Look it impossible fit anymore than 823mb on most 80 min disks and none will let you burn anymore than 830mb.
The only reason the bar drops down to 696mb from an 800mb MPEG is to take into account the 100mb of error correction parity data which is not needed for MPEGs.
Basically a 700mb disk will allow 700mb + 100mb = 800mb of raw data to be strored on the disk.
Nero takes this into account and adjusts the status bar accordingly.
I think what you mean to say is that Nero will allow you to burn 700mb+ + 13% = 791mb. The 13% being the extra space not required for parity data.

There is absolutely no way you can fit 904mb on any 80 min disk. Not unless you are using one of the new special CDR writers which can actually fit up to 1.5gb on an 80min disks using a higher pit compression technique.

Um...thanks a lot! :D

I have an MPEG2 movie file and at a certain point there's a small glinch and sound gets shifted. I no longer have the original... Is there any way to fix the file or should I just download it again?

Well you Can Try to fix it But it might not work Depending on the Glitch...What you can do is Cut the File exactly were the Glitch happens then take the Out of sync Part of the File and De-Mux it, Then use a Program called "Mpeg2VCR" and use it to Mux the audio and Video Back together But use the "Sync" feature to adjust when the audio Starts and try to sync it up..It will take Some Trial End error to Get right But after syncing it back up you can Join the 2 files back together..The Problem with this is Mpeg2VCR isn"t free ..

yikes, forget it, it's not worth it.. thanx anyway

There is an easy way to find the correct audio skew setting.
Do as minion says and cut the MPEG just after were it goes out of sync then load the part you need to re-sync into VirtualdubMod.
Click streams then select stream list from the menu. Right click on the audio stream and select Interleaving.
Where it says audio skew correction you can add a delay to the audio or remove a delay by putting a minus before the value you set. The delay is in milliseconds, usually around + or - 250 ms is enough to correct sync, but you can check it each time just by clicking the first play button. Once you are happy with the sync just use that value in BBMPEG which is free or MPEG2VCR when you multiplex the audio and video back together after you have demultiplexed it.

Ok, I've got to the point where I have the interleave value... Now how exactly do I "use that value in BBMPEG"? I don't see anywhere to put the value in

Never mind, I downloaded the wrong program...

When I encode some avi files of say 80 minutes, via AviSynth2:

SetMemoryMax(384)
AviFile("h:/video.avi",true)
ConvertToRGB32

using Wizard PAL (TMPGenc Plus latest version) VBR (Highest quality (very slow)) fit to DVD 4.3 Gb (about average 6 mbs), the analysis goes on for 10 hours (remaining time) slowly up to almost 30 hours (remaining time) to stabilize after one hour. Then the first pass ends slowly after 15 hours which is correctly half the 30 hours predicted.

The good new is, the second pass goes flying by at the speed of the video, ie. a little longer than 80 minutes (instead of the predicted 15 hours).

This gives me a total of ca. 17 hours, and curiously, this is the time I needed with the trial version of CCE SP (1950 USD) using the nine pass setting.

Mind you, it is better this way, than the other way round, were the program tells you the encoding will be finished in two hours and the job's not finished after another seven hours.

That is Strange Cuz for me the Pass is Faster than the Encodeing, Not By Much But still a Bit Faster...Just One Little Hint on your Script, You don"t Need to Use RGB32 as RGB32 is Exactly the same as RGB24 But RGB32 has 8 Totally Blank Bits, the Extra Bits do Absolutly Nothing accept Probably Take Longer to Convert from YV12(Or YUY2 if you are Useing a version of AVisynth older than 2.5)..So with RGB24 you will get the same Quality and Cut down on the encodeing Time....Cheers

I'll try RGB24 next time (I still have quite a lot of them to backup).
I made the same experience with a NTSC source of the same length.
I do not activate the preview as it takes a lot longer during encoding.
The sources are LaserDiscs (95% opera therefore not too much mouvements in the picture), captured with S-video with Director's Cut (firewire) and Pinnacle Studio 8.5 producing DV-avi. At the end, the DVD looks almost as good as the LaserDisc (sound has got down a little bit, I can live with that).
When encoding some VHS (or SVHS)I am usually satisfied with the High Quality (slow) motion search precision and have not observed the previously reported behaviour yet, I mean using the newest version of TMGEnc (2.520.54.163) which I managed to download last Saturday Morning (19.7.2003) through the pegasus site, (the tmpgenc site still has the old version (21.07.2003)).

Forgot to mention: the avi-file is on drive H: (big firewire disc), TMPGenc workspace and additional swapspace sit on drive F: (HD of about 13 Gb, with 10 Gb free space), drive C: for the programs, swapspace and the encoded video (m2v + wav/mp2). The encoding (pass two) is really going on fast.
In the Environmental setting, under Cache Setting, the Saves analyzing result of multi pass VBR to cache is selected with a Create cache up to 8192 MB. I understand this cache of 8 Gb is going to the TMPGenc workspace on my drive F: but when I check with the performance manager, I cannot notice much. May be I am missing something there.
This cache buffer is supposed to save the analyzing result of the first pass (a bit like the feature in CCE, and the encoding process should be boosted by using the result at the 2nd (or later !) pass(es). Will there be more passes in the future ?

Jaysus theres nothing like overkill is there!

I really don't think it's necessary to have so many optimizations enabled.
For a start off 9 pass?!
Did you never hear that it's commonly realised in most encoding forums that you don't get any benefit at all with half decent sources after 4 passes. Even at 4 it's very much doubtful you'll see any difference from a 3 pass.

RGB32? I'm not sure about this one because as I understand it TMPG and it's filters work in RGB32, so using the ConvertToRGB32 command is ok.


Next the Highest quality setting in TMPG?
It is well known by most TMPG experienced encoders that there is no benefit over the High quality setting in TMPG except for doubling the encode time.

Anothert point is that 2 pass isn't always best. It usually Constant Quality(CQ) which gives the better results with good sources such as laser disk and seeing as you will be using a high bitrate anyway because you are encoding to DVD then using all the above is simply a waste of time.

When you have a good source and are encoding to a high enough bitrate all these optimizations aren't necessary and actually can sometimes make the result worse.
Don't ask me why because I don't want to go into a big explanation as to the reasons of why using something like the motion search precision on highest is not the best for some sources.

Do yourself a favour and cut some of these out and save yourself some time. Your results may look good, but you could probably obtain the same results with less overkill.
Remember sometimes less is more.

Your "DV-avi" is probably a 4:1:1 color format and should be avoided.

Resampling 4:1:1 color to 4:2:0 during Mpeg encoding, results in a 50% loss of color samples. (Both have 4:1 luma-to-color sample ratios but the conversion ends up being 8:1.)

Capture using Huffyuv Lossless Codec which is 4:2:2. Buy an analog capture card if necessary.

A lot of magazines recommend running SVHS and Hi8 quality analog material through a DV conversion step but that just means they value ease-of-use over quality.

Thanks, I will try all this.