This forum is for users to exchange information and discuss with other users about a TMPGEnc product.
In case you need official support, please contact TMPG Inc.
Pegasys Products BBS [ Sorted by thread creation date ]
I rip DVD with Smart Ripper then get 2dv file with DVD2AVI, then get MPEG file with TMPGenc but video-audio come out of sync, I know tempgenc has audio gap utility but I dont know how to calculate the audio-video gap, is there any way to fix it besides trial and error?
Just rip the audio again from the DVD to a wav with a program called VOB2AUDIO. DVD2AVI has a habit of doing this and usually solves the problem.
By the way Smartripper already creates a d2v for you wherever you have ripped the VOBs to.
when i try to use it tells me when i go to source range that it could not find audio source.......even when i use .wav files.............what can i do???????
:O(
The guy that needs help all the time
2003-04-06 22:06:42 ( ID:aa1vkte.qxc )
[ Delete / Reply with quotation ]
Hello, i was wondering how i would make a wide screen avi into a wide screen mpg. I encoded it with TMPGEnc but when i pop it into my dvd player it comes out as a full screen (with the image stretched from top to bottom). I know this might be a stupid question with a really easy answer, sorry to bother you guys.
Under the 'Advanced' tab you need to set your 'source aspect ratio' to 16:9 NTSC or PAL whichever is appropiate, the 'Video arrange method' to 'Full screen keep aspect ratio' and then under the 'Video tab' set the output aspect ratio to 4:3.
I use tmpgenc to make vcd's and svcd's. Great software, very easy to use and fast.
One thing I notice on the svcd's is that the quantize matrix blocks appear to be visible...
I capture my original video with an MJPG capture card at 640x480, and then encode it at 2520kbps with tmpgenc onto a single svcd. The source material is only ~42 minutes long, so the quality is very good, except for these matrix blocks showing up. It's very subtle, but once you know what to look for, it's VERY annoying to watch either on TV or on the computer. It shows up in both places.
I'll to to explain better what I mean by them showing up. Once you know where the matrix lines are (the 8x8 blocks) then anytime there is motion that traverses a matrix block, the edge of the block distorts the edge of the motion object a little bit. If the object traverses a horizontal edge of the block, it tends to create a bump in the edge of the object inwards (toward the center of the object). If the object traverses a vertical edge of the block, it tends to create a bump in the edge inwards again (toward the center of the object). It seems to be the same behavior, but whatever algorithm is deciding how to deal with the motion-edge-block-distortion is consistently choosing to create a distortion that makes an inward bump on the object. This is very obvious on curved objects like cars, human faces, spacecraft, furniture, that sort of thing.
My question is, what will get rid of this?
2 pass VBR? (I've just used manual VBR so far)
Slower motion search (higher quality?)
CG/Animation Quantize matrix? (I've only used default)
MPEG Standard Quantize Matrix? (I've only used default)
>I'll to to explain better what I mean by them showing up. Once you know where the matrix lines are (the 8x8 blocks) then anytime there is motion that traverses a matrix block, the edge of the block distorts the edge of the motion object a little bit. If the object traverses a horizontal edge of the block, it tends to create a bump in the edge of the object inwards (toward the center of the object). If the object traverses a vertical edge of the block, it tends to create a bump in the edge inwards again (toward the center of the object). It seems to be the same behavior, but whatever algorithm is deciding how to deal with the motion-edge-block-distortion is consistently choosing to create a distortion that makes an inward bump on the object. This is very obvious on curved objects like cars, human faces, spacecraft, furniture, that sort of thing.
It seems like you have probably copied and pasted this info from somewhere as it is somewhat technical.
Anybody who would have such an indepth understanding of Matrix's would know that the blocks you are experiencing, which are called Macro blocks, are simply due to a shortage in the bitrate.
All you need do is raise the bitrate high enough until the blocks disappear.
Standard SVCD bitrate is not usually high enough to produce block free images especially in high motion scenes. In high motion and high detail scenes there isn't enough bits to produce the image and the only way to cure it is to raise the bitrate or lower the resolution.
You would be better using the (CQ) VBR method of encoding and setting your Min bitrate to 1800kb/s and your Max bitrate to 4000kb/s and quality at 65%. Also check the 'soften block noise' option under the 'quantize matrix' tab.
This should allow you to fit your movie on to an 80min disk.
Interesting. I wrote that paragraph from what I observed, I did not cut and paste it from anywhere. I'm flattered it indicates knowledge I don't have, but I still have the same problem. :)
And, as I mentioned, the bit rate is set to 2520 already as a maximum, although I did not up the minimum from 300. Would increasing the minimum improve the situation?
As I didn't get an *immediate* response, I started doing some encoding testing. It appears that 2 pass VBR improves the 'hiding' of these blocks as does choosing 'Highest Quality' under the 'Motion search precision' drop down.
I will continue with my encoding testing, and try raising the minimum bit rate, although this will likely require that I drop the maximum bit rate, if raising the minimum forces the creation of a larger file size. Using the wizard with tmpgenc and a value of 300 min / 2540 max creates a file that is exactly 80min/700MB, unless I'm doing something wrong.. That's why I didn't think this was a strictly bit rate related problem.
>I will continue with my encoding testing, and try raising the minimum bit rate, although this will likely require that I drop the maximum bit rate, if raising the minimum forces the creation of a larger file size. Using the wizard with tmpgenc and a value of 300 min / 2540 max creates a file that is exactly 80min/700MB, unless I'm doing something wrong.. That's why I didn't think this was a strictly bit rate related problem.
Do not use the wizard.
Just load the SVCD template then load the 'unlock.mcf' template located in the 'Extra' folder and set the values I gave in the previous post. Don't bother with the 2pass or 'highest quality' this will just increase the encode time for no benefit.
Use the CQ method as I have described and the 'High quality' setting for the motion search precision, 'Highest' will not give any visible improvment over 'High' but will dramatically increase the encoding time.
Yes the minimum bitrate DOES make a difference to the output. This is the minimum bitrate you can use without blocks becoming visible in regular scenes.
These settings should easily fit your 42mins movie on to an 80min cd.
>Using the wizard with tmpgenc and a value of 300 min / 2540 max creates a file that is exactly 80min/700MB, unless I'm doing something wrong.
Do not use the wizard. These values are for the 2pass VBR method with padding added.
Best quality vs file size is gained using the Constant Quality(CQ) method. Ensure the 'Enable padding' option is NOT checked.
Trust me, I and many others in this BBS have been using these settings for donkeys years.
>Du NOT use Soften Block Noise, or, if you need it, set it to 0/0 or 10/5.
Look, at low bitrates like this the soften block noise option makes good sense to use and the settings you have posted are pointless, the default setting is quite a good setting to use at low bitrates.
The soften block noise may soften the picture ever so slightly, but is hardly even noticible on TV any way and is far less obvious than ugly block noise popping up all over the goddam image.
I'm sure most people would rather watch a slightly, hardly noticible softer image than an image filled with ugly macro blocks.
I would agree at high bitrates there is no need for this option, but don't go posting contradictions to my posts unless you explain the reason why first so the user can make his own mind up.
Soften Block Noise is acting a little like a low pass filter. If you use it with too high parameters, you will GET Blocks and not prevent them. Especially in Szenes with dark colors or low contrast.
Using this Feature has nearly the same effect as a Matrix with higher Quantisation in the High Frequency Part.
You are right, if you say, that the picture will be softened. But it will be softened Block by Block and not overall. This may help, if you using very low Bitrates or absolutly noisy Video, but for High Quality Video it is nonsense.
If you don't belive me, try it by your self. To see is to belive:
Use a Video with a slight blue Sky and use different settings for Soften Block Noise. The difference WILL be visible.
The B-Frame Settings are not new. TMPGEnc gives them not enough Bitrate, setting the Spoilage to a lower state, will solve this Problem (which is known since more than a Year!)
So if you attack my postings, maybe you should learn more about MPEG-Enocding in generally?
I already have tried myself a long time ago. I have tested and adjusted almost every setting I can think in TMPG in the many years I have been using it.
I always check how dark colors look when making any sort of adjustment to bitrates and associated settings as blocks can and do become visible at low bitrates in dark scenes.
I also check with high detail scenes such as backgrounds with lots of trees or other foliage and also with movement as this also gives a good indication of the amount of blockiness thats occuring at different settings and as far as I'm concerned when using low bitrates checking the blocknoise reduction is beneficial. Anything below 20 doesn't even seem to have an effect, so to me 35 is a good setting.
Maybe it is just personal preference and is probably dependent on the actual encoding methods and bitrates we are using, so you stick with your settings and I'll stick with mine, but at least now the users can make their own minds up.
As for your picture spoilage settings, if they work for you then fine, but I wasn't refering to these settings.
ALL my posts are based on my own experience and NOT something I have read on the net and don't necessarily expect my setting preferences to work for everyone. Each situation is different and may require different tweaks, but at least I have enough knowledge and wide enough experience in encoding to guide people in the right direction.
I don't like it when someone blatantly contradicts my guide without reason and this was my point. No offence intended.
i'm converting an avi. The source was made at 23.7 fps. If i encode it in tmpgenc at 25fps, the picture isn't very smooth, i.e its jerky. If i encode it at 23.7 it plays fine on the computer, but when i try to burn it to vcd using nero, nero says it must encode it for it to work.
am i doing something wrong. If you don't know, could you recommend other vcd burning software that might work for me?
I don't where you got that AVI from, but whoever encoded it at that frame rate did not know what they were doing.
Nero will accept this frame rate if you select the option 'Turn off standard compliance and continue' when asked, but it unlikely it will play in any player correctly.
Yoy will need to adjust the frame rate to 23.976 fps or 25 fps using AVIfrate and then use a program such as cool edit to shrink the audio to sync it with the video.
Alternatively you can use AVisynth to alter the frame rate. With this method you won't need to alter the audio.
Is it possible to add, in a future version, the ability to specify a time offset for the audio when muxing separate audio/video files into a program stream. This would greatly help sync a/v files that have become unsynced after de-muxing. A real time preview would also help greatly.
I'm having a problem with tmpgenc (current version) that I can't seem to figure out. I've got an AVI that gSpot reports as being MPEG-1 Layer 3 audio and video codec of S-MPEG 4 Version 4 (FOURCC: MP43). The AVI plays fine in Windows Media Player (both audio and video work), but when I run it through tmpgenc, the resulting .mpg has no audio. I've tried everything I can think of. These are the things I've tried with no success:
1) I used virtualdub to strip the audio out to a .WAV and then used that for my audio input in tmpgenc--didn't work.
2) I used virtualdub's full interpretation mode for the audio and converted the audio of the .AVI to a huge raw PCM file and used that for my audio input--didn't work.
3) I set virtualdub up as a frame server and and told tmpgenc to pull the data from the frameserver--same result, only video, no audio.
When I look at the resulting .mpg with gSpot, it simply tells me that the I have a video/mpg file, but nothing is filled in for the FOURCC or the audio settings. When I play the mpg, I just get a silent version of the movie.
What could this possibly be? Why does it work as an AVI in media player, but I can't seem to get the audio stripped out to work with tmpgenc? Is my copy of tmpgenc screwed up somehow?
>I'm having a problem with tmpgenc (current version) that I can't seem to figure out. I've got an AVI that gSpot reports as being MPEG-1 Layer 3 audio and video codec of S-MPEG 4 Version 4 (FOURCC: MP43). The AVI plays fine in Windows Media Player (both audio and video work), but when I run it through tmpgenc, the resulting .mpg has no audio. I've tried everything I can think of. These are the things I've tried with no success:
>
>1) I used virtualdub to strip the audio out to a .WAV and then used that for my audio input in tmpgenc--didn't work.
>2) I used virtualdub's full interpretation mode for the audio and converted the audio of the .AVI to a huge raw PCM file and used that for my audio input--didn't work.
>3) I set virtualdub up as a frame server and and told tmpgenc to pull the data from the frameserver--same result, only video, no audio.
>
>When I look at the resulting .mpg with gSpot, it simply tells me that the I have a video/mpg file, but nothing is filled in for the FOURCC or the audio settings. When I play the mpg, I just get a silent version of the movie.
>
>What could this possibly be? Why does it work as an AVI in media player, but I can't seem to get the audio stripped out to work with tmpgenc? Is my copy of tmpgenc screwed up somehow?
>
>Thanks for any help anyone can provide.
Well, I've helped myself a little I think. I downloaded Graphedit and opened up the AVI that renders ok. I then removed the DirectShow audio renderer from the resulting graph and replaced it with a wave file renderer and finally a file writer. Then I played the whole works and ended up with a garden variety wav file, although I'm not sure what format it is. Then I ran tmpgenc with the new wav file set as its audio input. Voila! I now have an MPG with audio. The strange thing is that when I look at it with gSpot, it still reports it the same way as before. When I look at it with Graphedit, it shows that it's using the MPEG-1 audio decoder. That's not right for VCD is it? I thought tmpgenc would convert it to MPEG-2 with its built in MPEG-2 engine. I thought I read that the correct audio format for VCD is MPEG-2... no? I hate to go to all the trouble to create an mpg and burn it to VCD only to find it won't play in my DVD player. I swear, I've never had so much trouble with an AVI before...(I've probably done this a dozen times before without a hitch).
>I'm having a problem with tmpgenc (current version) that I can't seem to figure out. I've got an AVI that gSpot reports as being MPEG-1 Layer 3 audio and video codec of S-MPEG 4 Version 4 (FOURCC: MP43). The AVI plays fine in Windows Media Player (both audio and video work), but when I run it through tmpgenc, the resulting .mpg has no audio. I've tried everything I can think of. These are the things I've tried with no success:
>
>1) I used virtualdub to strip the audio out to a .WAV and then used that for my audio input in tmpgenc--didn't work.
>2) I used virtualdub's full interpretation mode for the audio and converted the audio of the .AVI to a huge raw PCM file and used that for my audio input--didn't work.
>3) I set virtualdub up as a frame server and and told tmpgenc to pull the data from the frameserver--same result, only video, no audio.
>
>When I look at the resulting .mpg with gSpot, it simply tells me that the I have a video/mpg file, but nothing is filled in for the FOURCC or the audio settings. When I play the mpg, I just get a silent version of the movie.
>
>What could this possibly be? Why does it work as an AVI in media player, but I can't seem to get the audio stripped out to work with tmpgenc? Is my copy of tmpgenc screwed up somehow?
>
>Thanks for any help anyone can provide.
Well, I've helped myself a little I think. I downloaded Graphedit and opened up the AVI that renders ok. I then removed the DirectShow audio renderer from the resulting graph and replaced it with a wave file renderer and finally a file writer. Then I played the whole works and ended up with a garden variety wav file, although I'm not sure what format it is. Then I ran tmpgenc with the new wav file set as its audio input. Voila! I now have an MPG with audio. The strange thing is that when I look at it with gSpot, it still reports it the same way as before. When I look at it with Graphedit, it shows that it's using the MPEG-1 audio decoder. That's not right for VCD is it? I thought tmpgenc would convert it to MPEG-2 with its built in MPEG-2 engine. I thought I read that the correct audio format for VCD is MPEG-2... no? I hate to go to all the trouble to create an mpg and burn it to VCD only to find it won't play in my DVD player. I swear, I've never had so much trouble with an AVI before...(I've probably done this a dozen times before without a hitch).
I merged several of my mpeg files together in order to form one long one, but when i output it the movie's audio is way out of sync... How do i fix this?
Could you be a little more specific? Where do i get those & how do i use them? And i'm not multiplexing(nor do i know how), i'm just going to "Merge & Cut", adding my videos to the list then hitting "output".
Multiplexing is a process of combining the Video and audio components of an MPEG in to one file. De-Multiplexing is a process of seperating the Video and audio components of an MPEG.
First you will need to de-multiplex the MPEG with TMPEG then use BBMPEG or MPEG2VCR to Re-multiplex the two components back together. Both of these programs will allow you to use an offset for the audio to re-sync it with the video when you multiplex. Find them on Google, BBMPEG is free, MPEG2VCR is not.
If Minion sees this hopefully he will post you a link to MPEG2VCR.
It is Because the audio format in the AVI file is Not Supported By Tmpgenc or you don"t have the Proper Codecs or Filters to decode the audio, Generally this Happens with DivX and XviD files that have AC3 audio, what you have to do is useing something like "AVI-Mux-Gui" to Extract the audio from your AVI file to WAV format then use the WAV file to encode in Tmpgenc....
I just upgraded my Windows Media Player 7.1 to 9.0 However, I can no longer convert any of my WMV files in my hard drive to MPG. The error message indicates the file is unsupported or can't be open. I had no problems converting my WMV files when I was running WMP 7.1 Does someone have any ideas?
Thanks!
You should NOT have installed "Media Player 9" it causes Big Problems with useing Tmpgenc and Many other Codec dependant Programs and the only thing you can do to get Tmpgenc to work Properly again with your files is to delete Media Player 9 by doing a System Restore.....
You can Not Properly Convert NTSC to PAL With Tmpgenc, if you try the Video will Be Jumpy and the audio might not be in Sync..You Can use AVISynth to do the Frame rate conversion while Frame serveing but you have to have a Good Grasp on AVISynth Script Writing..Besides that there are Programs that can do it But they don"t come cheap, doing Proper NTSC/Pal conversions takes fairly sofisticated software,some Programs that can do it are "ProCoder" "Vegas Video 4" and "Advanced NTSC/Pal Converter" and there are ways to do it with mostly Free software accept for a good audio editor but these methods take a Good Knowlege of the Software involved..Try going to "http://www.dvdrhelp.com/" and look for a Tutorial on converting between NTSC and Pal....
It can.
Change the frame rate to 25 fps in TMPG and make sure you check the option 'Do not frame rate conversion' under the advanced tab. Do not encode the audio.
You will then need an audio editor like Cooledit to change the length of the audio to match the Video then encode the audio and multiplex it with your video.
Be aware if this AVI is 29.97 fps and you convert it to 25 fps that the movie may seem like it's playing slow.
To avoid this the best way to format convert is with AVIsynth.
>IThe way you descripted is only usable for PAL-Speedup from Film-Source (23.976 fps).
Yes I do know this. I'm getting pretty fed up with your attack on every post I make!
If you read my post again, you will note that I already warned the poster that if the frame rate is 29.97 rather than 23.976 that the movie will seem like it's playing slow which is why I advised using AVIsynth.
Here it is again. This time have a good read of it.
>Be aware if this AVI is 29.97 fps and you convert it to 25 fps that the movie may seem like it's playing slow.
>To avoid this the best way to format convert is with AVIsynth.
Which was exactly the point in my post.
I was trying to give him the simplest way I could think of for converting the movie.
I did warn him that with this method there would be playback artifacts if it was 29.97. I didn't need to mention about the movie being 23.976 because if it was and he used the above method then the resulting movie would have been fine anyway.
If it likely was 29.97 and he used the above method but didn't like the results then he could use AVIsynth as I said, but I didn't want to baffle, as you say a newbie, with technical details about a program such as AVIsynth which isn't really newbie friendly if he was happy after the process above.
If not then he could post back asking for more details about how to use AVIsynth, to which I would have duly replied with the info.
My aim is to try and consider the posters knowledge and give whatever info may be relevant based on that without trying to confuse him/her.