Hi8/VHS to DVD: which bitrate do you recommend?

[Luis Neves]

bbeiler, take note that bitrates of different media can not be compared directly. If you encode to mpg2 DVD the image size would be 720x576 (PAL) against 480x576 (SVCD) and 352x288 (VCD). This means that for the same bitrate, lets say 1100, a VCD would be ok and a DVD will look lousy. Anyway, there is something wrong with your system if you can not obtain a good quality with standard settings for SVCD! Which encoder are you using?

We can not compare experience because our systems (hardware) are probably different. However, let me say that with my inexpensive PCTV card I found the following results when capturing my home Hi8 videos (encoding with TMPGEnc):

1) standard VCD did not satisfyed me completely, the result is just reasonable: it does not matter if capture is carried at 480x288 or 720x576, final quality is the same.

2) SVCD provided a very good result. After many tests I noticed a slight improvement when capturing at 480x288 instead of 352x288; no difference between 720x288 or 720x576 relatively to 480x288! The problem of SVCD is that with a standard CQ only carries 35-40 minutes in one disk, and specially, as I realized later, is not DVD compatible, whcih means that when migrating to DVD we have to capture and encode everything again!

3) When capturing and converting directly to DVD format, I have already tested and tested and again my capture resolution stayed at 480x288. I can´t see no improvement with higher capture resolutions and the final result is perfect. As I stated before, I have made tests with lots of different bitrate settings; for home videos I found that 2700 with CQ or 2000 with 2 pass VBR provide 3 and 4 hours of film in one disk, respectively, and quality is not different of a 4000 CQ or higher setting (note: a have a flat 82 cm TV and in some of my tests I have even used the zoom feature in one part of the image to compare!)

Happy capturing!

[yg1968]

Luis Neves said:

I have made tests with lots of different bitrate settings; for home videos I found that 2700 with CQ

Given that CQ does not have an average bitrate, what do you set your maximum and minimum bitrate for CQ and what percentage?

[bbeiler]

Luis Neves, I really don't think there is anything wrong with my hardware. I have the all-in-wonder radeon with the S-Video in. Like you mentioned, even the cheap cards work well. Most of the quality comes from encoding.

My config. as previously mentioned are:

I use Virtual Dub, with Huffy compression at 352X480 for end results of the same 352x480. I also use 480x480 in virtual dub w/ huffy if SVCD is my end result. And so on. My encoder is TMPG. I also, understand that if the end result is a higher resolution than higher bitrates are in order to avoid pixalation. So it sounds like we are on the same page. However, I think my lower quality results are not based on the average bitrates (VBR situations-average, CBR, or CQ at 100). As I mentioned I have used similar average bitrates to what you are using, however, my maximums can never be over 3000 (video + audio) on a CDR. I think that this is where my quality is suffering. At least, I'm hoping That way I can get 3 hrs. on a disk if I simply change the max.
I have tried config. with 2500 the average with min=300 and max=2800. with VBR. It sounds like you have even used lower averages than 2500 however, your max is up to 4000. I hoping that is the difference.

One more thing I noticed that you mentioned in you last post is that you capture at 480x288. What do you encode at to stay DVD compliant?

Yet another thing, I too, have zoomed my TV to check for clarity. Aren't we pathetic

[DJRumpy]

With cam sources, VCD will not satisfy your needs no matter what you do. The bitrate needs for cam sources is too high. SVCD is better, although still bit-starved. CVD is better still.

Very Important! Assuming your suffering from bitnoise, make sure you clean your source first. Use a good 2-D filter, or a temporal smoother in VirtualDub or AVISynth to process your AVI before encoding. This will help clean up any noise in the video, which eats up bitrate. I can't stress this one enough.

Know your correct FIELD setting. Get it wrong, and your video will be jerky.

Use the maximum (slowest) setting for Motion Search. A regular encode will see little benefit from the slowest setting. A CAM source however needs all the help it can get.

Use 2-Pass VBR. Extra Passes means better motion detection, and bitrate allocation. Min: 0, Avg: 2400, Max: 2776
This assumes your player specs are indeed maxed out at 3000. You may also need to fudge with the MIN setting. Some players can't handle MPEG's encoded with TMPGenc set to a MIN of 0. CCE doesn't seem to exhibit this problem. If a min of 0 doesn't work, try 300.

Good luck!

[Luis Neves]

Yg1968: sory, my mistake, I was reffering to CBR and not CQ.

bbeiler, the search for perfection sometimes is a little pathetic indeed!

When using 2 Pass VBR I usually set the maximum bitrate to 4000, maybe this explains some difference. However I found a good result with CBR 2700, which is a setting that your player should be able to handle. The alternative of DJRumpy seems also an alternative to consider.

When encoding the captured file (AVI, Huffyuv, 480x288) with TMPGEnc, this encoder automatically resizes the resulting Mpg-2 to DVD standard (720x576 for PAL). In order to use such a low minimum bitrate (like 300), however, it is necessary to unlock the DVD template first.

The problem of 2 pass VBR is the time it takes to encode (7 hours on a Pentium IV for each hour of film). I understand that CCE is a good and faster alternative, but how do we process an AVI file with this program? And the result is not an mpg-2 file, I guess? DJRumpy, some help on this?

Happy new year (and encoding)!

[DJRumpy]

CCE is an excellent alternative to TMPGenc. Searches on google.com should yeild results on the things you'll need for it, and that's all I'll say about that...

CCE yields much faster times than TMPGenc is capable of. On my P4 2.4Ghz, 720x480 video compresses at about 1.4 times the playback speed of the movie or ( 60 Minutes / 1.4 ) = 42 minutes to encode each hour of video. At SVCD, and CVD resolutions, CCE will encode an hour of video in about 20 minutes.

As an added benefit, CCE will also allow you to use more than 2 passes for your encoded video, maximizing bitrate allocation and scene/motion detection. I have not seen any benefit beyond 3 passes though besides a minor reduction in your MPEG size. Your results may vary. Try a small clip with multi-passes to determine if it's worth the extra time.

CCE will accept AVI input, as well as frame served input from any source. It performs best using avi input, and very well using frame served source material from AVISynth, and slightly slower from VFAPI. It will encode to both MPEG-1, and MPEG-2 formats in a fraction of the time that TMPGenc takes. It does require more work on your part. AVISynth is almost a must, as you do not have any 'click' filters in CCE. All of your pre-processing should be done in VirtualDub, or using AVISynth (preferred). You will need PULLDOWN.EXE if your source material is 23.976 frames per second. Get it in the tools section, or you can also get it with DVD2SVCD as part of the program. You'll need this to add the pulldown flags into your final video, if your working with FILM framerates. These will tell your DVD player to telecine your material back to 29.97 frame per second. You should also get BeSweet GUI, SSRC, and TooLame for your audio needs if your a bit of a perfectionist. You can find all of these in the TOOLS section to your left.

[sammie]

I'll spout a bit of heresy here. I was firmly convinced by the "you can't create information" from a VHS source" argument too and have been capturing and encoding everything at CVD resolution. But over the last couple of weeks I have been converting several tapes. I accidentally captured one at 720x480 and was amazed at the difference. I went ahead and converted at that size as well. It did make a significant positive difference, I posed the question a few weeks ago but no one had any answers. Since then I’ve done a few experiments. 5 of 5 viewers looking at 3 separate clips all picked the higher resolution clips 100% of the time. I’ll admit it wasn’t even close when looking at it. All clips were captured with Virtual Dub and encoded with TMPEGEnc 2 pass, highest quality VBR with approximately 5000 average bit rate then burned onto a DVD-R. This should have meant the CVD clip had bit rate out the yin yang. Just to be fair I did one CVD at around 2500 too, just in case somehow “too high” a bit rate had a negative effect (ok, I was grasping at straws).

I’m not arguing the logic behind no need to increase your capture size, but in practice didn’t seem to hold up. That means either there is indeed some additional information that is captured, or I (and I presume others) are doing something else wrong. I have at least one theory I’ll throw out. Further digging on the web and even on this site indicated that there is indeed likely information above 352. There is no exact number for composite or sVideo horizontal lines. That the estimate here is somewhere around 384 and I’ve read elsewhere that good cables, source material, players, signal amps, sVideo!!, etc. can boost that number higher. It’s web information so take it perhaps with a grain of salt, but capturing above 352 does seem likely to provide additional information. Maybe not much, but some and it with my set up it seems noticeable.

I haven’t tried capturing at a higher resolution but then encoding back down to CVD resolution. Maybe that is a good solution, worth a try at any rate. Perhaps the other way to look at CVD is simply for what it is, a good compromise. It’s not much different, and it allows capture with fewer frame drops on many slower computers, which is a worse problem. It plays on a lot of sVCD compatible standalone boxes. And you can stick a lot of video onto one DVDr that is nearly the same quality as full DVD.

[energy80s]

It all depends on whether you want a copy of your VHS source or a virtual clone of it! As VHS is pretty poor to begin with, any copy will look a damn sight worse than the original. The only way to combat this is to copy it to a higher resolution format. This gives you a virtual clone of your tape rather than a 2nd generation copy. The picture on a VHS tape has a high amount of chroma noise on it, just turn down your colour control to see the image sharpen up a bit! All these filters and reduced frame sizes try to eliminate this noise. Unfortunately the noise is seen by the human eye as part of the picture sharpness, so once you reduce it you also reduce the sharpness of your image. MPEG encoding doesn't like noise (as it is too random) and so needs a lot of bitrate to code each frame. This is not going to happen at standard SVCD bitrates, so to get a clone of your VHS tapes you are looking at using DVD bitrates and frame sizes. CVD can be a good comprimise as standard TV sets don't give very sharp images anyway, but if you are planning on watching it on a PC monitor or a video projector, you will notice the difference in horizontal resolution.

The funny thing is that the better your source material the less bitrate is needed to code it! DVD's need less bitrate than Satellite. Sat requires less bitrate than cable/laserdisc. VHS (with the worst pciture quality of all) requires the most bitrate!

[sammie]

All true energy80s but it means there is additional info there to capture. I'll try some higher res captures then encode back to CVD resolution to see how that looks. Might be an even better compromise. At least for those of us lucky enough to have a computer that can capture at that size without dropping frames.

[energy80s]

If all else fails, use a DV camcorder to passthrough the analogue signal and cap as a DV AVI file. That's what I do now. It won't drop frames like an analogue capture and will give a damn sight better quality to boot. Have recorded nearly four hours of TV today alone using this method. Then convert to VCD/SVCD/CVD or DVD with TMPGenc or CCE at your leisure.

[Luis Neves]

DJRumpy, big thanks for the advice. I have now completed sucessfuly the steps of capturing with Virtualdub, then resizing and frameserving also with Virtualdub to CCE and producing the final DVD files to burn with Maestro. But...as usual a problem arises!

As you said CCE is expected to be much faster than TMPGEnc; some months ago I have tested CCE and at the time I also noticed that is it faster. But now I found that with a 2 pass VBR setting 1 minute of film takes around 7 minutes with CCE to encode and only 6 with TMPGEnc (same settings!).

After some tests I found out the reason for this. I capture at 480x288, and because CCE does not resize, I have to do it previously with Virtualdub. So the AVI file to encode already goes to CCE with the standard PAL resolution of 720x576; on the other side, TMPGEnc probably encodes the file at the original resolution and only after that internally resizes to 720x576. This way, capturing at 720x576 will not solve the problem for CCE and will further slow TMPGEnc.

Any advice on this problem?

[cooel]

I am quite new to computer video, but plan to learn fast, hence reading this debate with much interest. I have an ATI All In Wonder 128 pro card and have been experimenting with capture rates, codecs etc. Being a Video technician the point I think that has been missed is when importing VHS video into a computer, the quality of the VHS playback machine is one of the most important factors. Some machines will have a lot of noise and have poor resolution compaired to others, and as I have access to many VHS players, I have compaired many to see what works best. To my great surprise, a "Thompson" VHS recorder I tried was better than a "JVC" S-VHS recorder, and an "Amstrad" recorder was as expected a joke. Many other machines gave mixed results. So if you own more than one VCR, I suggest you try them to see which is best.

[bbeiler]

Luis Neves said:

After some tests I found out the reason for this. I capture at 480x288, and because CCE does not resize,

I was not aware the CCE does not resize. That would explain why I was getting such weird outputs. If it does not resize, then what is the point of some of the options like half horizontal and half vertical? Also, dumb question, is frameserve and resize the same thing?

Sammie: Interesting information!!! about the 720x480 encoding on VHS. I would be really interested in hearing how well it turns out capturing at 720x480 and then encoding at CVD. I was planning on doing the same thing only capturing at 704x480 since this is exactly twice the resolution as CVD then encode to CVD.

Luis Neves: After the ongoing search for perfection and through about 10 different 2 minute samples on a DVD+RW, I have found that VBR_CQ at 100 with a video max. of 3800 has served the best for me. This allows for about 2.5 hours on one DVDR. I don't know what it is about VBR and my Hi8, but I am having much better success with the CQ at 100. Any lower and I am seeing a reduction in quality from the original. Any higher is not necessary. (These are CVD's captured with virtual dub at CVD resolution). I am curious if capturing at 704x480 will help things the encode to CVD with a slightly lower bitrate.

Before anyone mentions about my VBR_CQ at 100 vs. CBR. I agree that mathematically they are probably the same, but my eyes don't lie to me, and VBR_CQ at 100 is consistantly better than CBR at the same bitrate.

One last thing, at least for now , for hi8 captures and encoded to DVDR (like above) I cannot get any sound out of my optical digital link. I have to hook up analog "RCA" type cables and turn off digital input on my receiver to hear anything. I never had this problem with SVCD's, or any other format for that matter, on CDRs. I just wondered if anyone else has had this problem.

[thxkid]

well here is my experince on using hi-8mm , since I have been using the 8mm format since 1986, SVHS tapes since 1991, so I know the difference in quality here, tried 352x240, 480x480 and 704x480 and on my 36 in. JVC TV the 704x480 on a DVD+R is way much more pleasing to the eye than the other resolutions, the 352x240 was always on a 2520 bitrate VCD the 480x480 was on a 3500 bitrate SVCD, the 704x480 is using 8000 bitrate all captured in the huffy nocompress AVI, edit with Uleads Media Studio Pro with hollywood fX effects and TMPGE encoded. But since each persons preception of quality is different need to try out what you like best, trial and error will always be there at the begining. Hey I have not used VHS( use SVHS very seldom) for over 6 years, its too much crap. While other people can't see any difference from a VHS to SVHS recording. Crap you still have folks out there using EP on a VHS tape, hurts my eyes even to watch that junk, but to them its fine.

[sammie]

See the summary under the capture forum.

Bottom line the higher the resolution capture and higher the bitrate the better the picture. Guess that comes under the heading of doh once your read it.

If you want to resize using CCE to encode just use AVISYNTH and frameserve using one of the good quality resize filters.

[bobmitch]

I am still a bit confused about this post. Here's my question. I have an AIW 9700 Pro with ATI MMC 8.0. I have an analog Hi8 hooked up through the S video and the audio straight into my Sound Card. I encode on the fly to Mp2 720 x 480 with 44 Khz Stereo sound. The capture looks great for the first 45 minutes...then I get "lightening' artifacting on the top of the screen that goes on and off, but does go to the final product. I normally dub mp2 and then use either Pinnacle Studio 8.4 or MyDVD 4 to burn to DVD (my Sony DRU-500A). What am I doing wrong that would cause the lightening effect at the top of the screen only. I have done other dubs off my VHS at the same resolution that don't have the lightening. Can anyone help me???

Please!!!!!

Thanks

Bob M

[DJRumpy]

First off. bobmitch. Post a new thread. Your posting your problem on top of someone elses, confusing this post beyond the mess that it's already in. Start a new thread for a new topic.

bbeiler, frameserving and resize are not the same thing. Resizing simply resizes your video to a new resolution (i.e 480x480 to 720x480). Frameserving takes one format, and hands it, or 'serves' it to your encoder/editor, frame by frame. You do this to save time, hard drive space, or to convert one format directly to another.

Luis Neves, CCE does resize, as was noted. The 'half' options will reduce your output either vertically, or horizontally by half, giving you almost every valid resolution via point and click, depending on your source. I never use these, as I resize everything in AVISynth. If you plan to spend any effort with this as a hobby, learn AVISynth. Your wasting time editing, or re-saving a resized avi with VirtualDub.

Use BicubicResize, or BilinearResize in AVISynth to enlarge, or reduce respectivly. Also, make sure your working in YUY2, and not RGB for a colorspace. RGB is too slow for encoding. Assuming your file is an AVI, use AVISynth, and CCE. Here's a simple sample script for AVISynth, which can be dropped, or opened in CCE (or TMPGenc) for encoding. Just save the file as a .AVS file (i.e C:\temp\video.avs )

# Begin Script
# Note: anything after a pound symbol is ignored or 'remarked'
# Change the paths below to match your source AVI file
AVISource("c:\directory\youravifile.AVI")
# ConvertToYUY2() will be ignored if your file is already YUY2.
# there is no space between the parens ()
ConvertToYUY2()
# You can remark this out witha pound sign if you don't need to
# resize or change it as needed to resize to any resolution.
BicubicResize(720,480)
# End Script

Alternately, some prefer a script a bit cleaner and easier to edit..here it is in a different way, without the comments, and with audio:

Video="C:\directory\myfile.avi"
Audio="C:\directory\wavfile.wav"
newwidth=720
newheight=480
samplerate=48000
vid=AVISource(video)
aud=WAVSource(Audio)
AudioDub(vid,aud)
ResampleAudio(samplerate)
ConvertToYUY2()
BicubicResize(newwidth,newheight)

Last but not least..same script, but with letterboxing for a 2.35:1 aspect ratio movie

Video="C:\directory\myfile.avi"
Audio="C:\directory\wavfile.wav"
newwidth=720
newheight=360
samplerate=48000
box=60
vid=AVISource(video)
aud=WAVSource(Audio)
AudioDub(vid,aud)
ResampleAudio(samplerate)
ConvertToYUY2()
BicubicResize(newwidth,newheight)
AddBorders(0,box,0,box)

The above scripts can be dragged and dropped directly onto CCE, assuming you have installed AVISynth. The new installer will configure everything for you. I would also suggest you get MPEG2DEC (see the tools section). That will allow you to open DVD2AVI project files, which gives AVISynth access to any MPEG formats, like VOB, MPG, M2V, M1V, etc. It has a nice TemporalSmoother to clean up noisy video, as well as a few de-interlacing addons for AVISynth.

In regards to your performance problem. It sounds like your editing in VirtualDub first, and saving the AVI, and then dropping this into CCE? Can you clarify what type of file your encoding in CCE? Definately something wrong there. Encodinga minute of video in CCE should take only seconds.

[Luis Neves]

DJRumpy, thanks for your detailed post. I will try your alternative suggestion to the Virtualdub/TMPGEnc path I have been using.

Regarding the encoding speed of CCE, I have tryed to encode and Avi file frameserving from Virtualdub to CCE (with resize of Virtualdub active from 480x288 to 720x576) and tryed also to produce first a new resized Avi to disk and only after that encoding with CCE directly from disk (thus no frameserving). Both options took the same time, around 5 minutes for each minute of film (2 pass vbr). Is there any special version of CCE for Pentium IV processors? Is CCE 2.66 faster than 2.5?

[Luis Neves]

DJRumpy, one more point I would like your oppinion. I have encoded to SCVD some time ago many hours of my home videos (Hi8), using standard templates of TMPGEnc (CB 2520). Now I bought a Sony DVD burner and I have two options:

1) Using the header trick, which results ok with my DVD table player, I can fit 6 CD´s in one DVD disk, just reencoding audio. Quality is ok for my eyes.

2) I can capture again all the Hi8 tapes and encode to DVD format, putting between 1 and 2 hours with regularly used bitrates.

The question is: considering the quality of the original film (Hi8), aren´t we just wasting disk when encoding to DVD resolution, which is much higher than the original source? Is it really option 1 the way to go?

[bbeiler]

DJRumpy, Let me see if I understand this correctly. I did a little playing around on the AVISynth.org site to try and understand this after reading your latest post.

Sample situation:
I capture a file from a Hi8 source with virtual dub at a 704x480 resolution with Huffy compression, audio is captured at 48K. I call the file CAPTURE and I save it in C:\mydocuments\ folder. After capture, I open notepad and type

Video="C:\mydocuments\CAPTURE.avi"
Audio="C:\mydocuments\wavfile.wav"
newwidth=352
newheight=480
samplerate=48000
vid=AVISource(video)
aud=WAVSource(Audio)
AudioDub(vid,aud)
ResampleAudio(samplerate)
ConvertToYUY2()
BicubicResize(newwidth,newheight)

and then I save it as CAPTURE.avs. I then open tmpg or cce and instead of opening the original source avi that I captured, I open the CAPTURE.avs file and it will open my original avi only resized to 352x480? Assuming I have installed AVISynth. Which, by the way, where do I get that?

Notice my changes as compared to the script you posted. I made the changes necessay to fit the situation I described. I'm not exactly sure where the audio fits in. Line two I'm a little unsure of, since there is not a seperate audio file for the avi I captured.

Thanks for any input. All of you have been a big help.

[DJRumpy]

Luis Neves: It depends on how much work you want to put into this. If your source material is Hi8, but the SVCD header trick looks fine to your eyes, then I wouldn't bother with the re-encode. Your going to end up using MANY more DVD-R's, as opposed to simply dropping your SVCD's on a DVD-R. As you should have found out by now, DVD-R media isn't cheap. You should consider this only if the quality your getting is not what you expect, or need. Regarding your performance problem, I would guess that the AVI's your working with in VirtualDub are using RGB colorspace, and not YUY2. When you saved your AVI, did you select the HUFFY codec, and configure it to convert to YUY2? Then try encoding a sample of that AVI, and watch your performance skyrocket. As to frame serving from VirtualDub, I've never tried it. I always serve from AVISynth, with steller performance. It could be that VirtualDub is simply slower than AVISynth. Someone with more experience Frame Serving from VirtualDub may be able to answer this one.

bbeiler: Your script looks fine. You can get the newest versions of AVISynth here:
http://sourceforge.net/projects/avisynth2/
The audio is optional. It is intended for those who want to use a seperate audio file. If you do not have a seperate audio file, then you can comment those lines out. AVISynth will read audio from any avi that contains simple PCM/WAV, or MP3 audio. It does not natively read DivX files with AC3 audio, but there are plugins to enable that as well. If your avi has simple audio (pcm/wav, or mp3), then your script should look like the one below. Since your not working with seperate video and audio, it isn't really necessary to assign your video source to a variable. When you do not do this, it removes the CLIP requiremnt from all of your commands. For example, the syntax for ResampleAudio is actaully:

ResampleAudio(clip,samplerate)

If we simply import our video without assigning it to a variable, the CLIP component becomes optional. Try this script instead:

AVISource("C:\mydocuments\CAPTURE.avi")
newwidth=352
newheight=480
samplerate=48000
ResampleAudio(samplerate)
ConvertToYUY2()
BicubicResize(newwidth,newheight)

[bbeiler]

DJRumpy: Wow, the script worked. Thanks . What are the advantanges doing the resizing this way? My test proved that in TMPG at CBR (Highest Quality) using AVISynth to resize it took 8 min 10 sec to encode one minute of video from a 704x480 AVI. Without using AVISynth to resize (I used tmpg to resize), it was actually a few seconds faster (about 30). Maybe BiCubic resize does a better job. I didn't test anyting on the TV. How do I apply filters using AVISynth like temperol smoother. You mentioned that earlier, and I'd like to try it. Is that just a command line in the script? The filters in tmpg just add even more time to encode, so if there is a quick way to do it, that would be awesome!

Everyone: I tried using CCE to do my encoding. How do I keep that DVD compliant? If I check the DVD compliant box it resizes the video to 720x480 (I believe it Mattes the video, putting black bars on the left and right sides with the viewable video at 352x480). There is no other setting to set the max. GOP to 18 like in TMPG. Which, I believe, needs to be set to stay DVD compliant (amoung a few other things).


Speed results:

I tested a one minute AVI (704x480) with CCE using CBR at 3800, resized to 352x480 and I got a .443 speed. Which took a little over 2 minutes. TMPG at highest with same settings took 8:10. In fact it didn't matter in CCE if I used VBR or One Pass VBR it all came in at around a .443 speed. I think this is rather slow. When I use DVD2SVCD with CCE as my encoder at get around these speeds.

DVD as source

DVD Resolution: .980
SVCD Resoultion: 1.220
CVD Resolution: 1.450

Why are my home encodings so much slower?

[DJRumpy]

You have to download MPEG2DEC. You can find it in the tools section. It comes with TemporalSmoother. To use it, just add it to your script somewhere after your video has been imported.

Somewhere before the TemporalSmoother command, you must load your MPEG2DEC plugin like so:

LoadPlugin("mpeg2dec.dll")

TemporalSmoother(2,3)

If memory serves, the first number is the radius, and the second number is the strength of the filter, but I probably have it backwards. In either case, this is the suggested default to use. Give it a try.

Try to avoid using the DVD Compliant box. It can resize when you don't want it to. You won't need to worry about your GOP setting. The default is fine, and will produce compliant video for you.

You haven't filled out your profile, so I don't know what sort of hardware your running. Can you either fill out your hardware profile, or just give me a quick rundown of what you have?

[bbeiler]

DJRumpy: Thanks, you have been a HUGE help. I really appreciate it.

You said:

To use it, just add it to your script somewhere after your video has been imported.

Does this mean I write and save the script like before and then AFTER I add it to, e.g. CCE, I then change the script and resave. OR, does it mean I just need to be selective as to what line in the script to place these two additional lines. e.g. AFTER the "newwidth & newheight" lines. Sorry, I'm such an idiot .

That's odd, what you said about my Profile. It is actually filled out, and I tried clicking on it and it showed everything. I did update a couple of things. Anyway, if it still can't be shown, my hardware is as follows.

OS: Windows XP Home Edition
CPU: 1.6 P4
RAM: 512 SDRAM
CAPTURE CARD: All-in-Wonder RADEON 32 DDR (AGP)
DVD-Burner: SONY DRU 500a 1.f Firmware
Harddrive: 40g 7200 RPM (Pretty fast compared to some)

I know other people have mentioned dropping frames at 702x480 with similar configs. However, I drop about 20-30 frames in one hour of capture using Huffy. Which I think is rather good.

I have a Western Digital 120g 8MB cache on the way. I'm so excited.

DJRumpy, thanks again for your help!

[mikeinfwa]

DJRumpy,

I am unable to find any how-to's on the 2d cleaners.
(I am using regular 8/HI8/Digital8 to DVD with both S-video and firewire connectors. 2 sony camcorders)

I use Vegas video 3, TMPGenc with toolame audio, Nero for burn

Definately use a 2d cleaner, either during capture, or post capture, as noise from your source will waste bitrate

thanks.

mike

[DJRumpy]

bbieler, you would usually add the temporalsmoother line before any resizing, simply because your processing needs will be higher after you resize.

Take the script with the additional lines, and drop that into CCE for encoding.

mikeinfwa, try doing a search of this forum for "dgraft", or try starting here for a few good links: http://shelob.mordor.net/dgraft/