It will obviously depend on how wll your source footage is, because the cleaner and smoother that your image is, the better it will compress.

I've gotten up to 3 hours of VHS tape onto a DVD without being drasticly worse than the original footage. But that's because the original was not prisine in the first place so the artifacts blended in with the natural VHS noise.

Basicly I rendered out the video as an AVI and then rendered over and over again untill it would fit on the DVD with menus.

You can probably leave your minimum bit rate at the default of 192,000. Set your maximum bit rate to 6,000,000 (if you have a lot of high speed motion you may have to increase the maximum after you try it and see). The average bit rate is where the main control is, so just set your average to 4,000,000. and start experimenting up or down untill you get a file size that will fit on the free space of your DVD.

Before you start the video renders... render out your audio file so you can group it with your resulting video file to check the size.

JBJones

I've put 3-31/2 hours on a DVD with great quality, it all depends on your source footage and the quality of your encoder.

I just put 2 hrs 30 minutes on a DVD and the quality was pretty darn good. (at least I was impressed) ;-). I’m recording a TV series that’s 1 hr long. Without commercials each episode is only 40 minutes so I get 4 episodes on a single DVD. (4 x 40 = 160 min.) I recorded the TV program on VHS tape at SP speed and then captured that with my Pinnacle Deluxe card as DV25. Then I run the AVI through VirtualDub with the Noise Reduction 2.1 filter, which really cleans up the VHS noise a lot. Less noise means better encoding. Finally, I customized the Vegas MPEG2 preset DVD Architect NTSC video stream by setting the Video VBR Max to 5,000,000 and Average to 3,000,000. The resulting DVD is just as good as the VHS tape. There is an occasional compression artifact but overall it looks great

~jr

The key is to have clean source footage, as the other folks in this thread have mentioned. Now, we don't always have the luxury of cleaning up footage, but click to view this thread on filters that will clean video. The work done by johnmeyer is awesome. I've actually got one client whose old videos end up looking better using this method than the original VHS tape (well, with a little Secondary Color Corrector magic happening too). I owe a considerable amount to john, and if he's watching, I thank him profusely. If you do use those filters, be aware that they're geared to clean up the type of noise present in old VHS - namely, red chroma noise and temporal noise. You have to use some common sense when applying them and understand what the source of noise in your original footage is. Also, if your original camera work is extremely shaky or consistently high motion or high complexity, you'll have problems and will need to encode at a higher bit rate regardless of what filtering you apply.

I guess the other things that go along with capturing old footage are relevant too, such as using a time base corrector if you're capturing analog video. But there's only so much you can do. The MPEG-2 format is itself somewhat limited, and you'll still encounter block noise in certain instances. If you want to encode 2 hours of high quality video, set your bit rate to around 4,500,000 to 4,650,000 bits per second. With VBR encoding it's a bit of a crap shoot, and you may end up overshooting the capacity of a DVD-5 a bit. The bit rate calculator will give you a better feel, but 4.5 Mbps is what I usually use and works just fine. DVD-5 discs were designed to hold 130 minutes of video and audio.

If you told us a bit about your source, it might be easier to direct an answer.

Look at the issue of the Tips, Tricks, and Scripts Newsletter that contains a bitrate chart.

I don't get these types of threads. What is the problem of splitting a 2 hr video into two DVD's? It can't be cost.
Hollywood uses dual layer DVD's and we use single layer DVD's. That to me says that if you want similar quality use two DVD's.

John

First off, to TVCmike, thank you!

As to putting two hours on one DVD, this has been asked a lot on this forum and over in the DVDA forum. Lots of people claim they can set the bitrate really low (which is how you fit two hours on one DVD) and still get good results. I've never had really good luck.

But then again, I am somewhat fanatic about quality and, having run a videoconferencing company a decade ago, I learned a lot about MPEG video artifacts, and spent hours looking at various tests, so unfortunately even subtle blemishes stand out to me. However, I am no snob: if it looks great to you, then do it!

A few notes. As I pointed out in a post several weeks ago, the DVDs that Hollywood produces have several advantages. One is that, as already pointed out, they contain two layers which gives them almost (but not quite) twice the capacity of a DVD-R or DVD+R disk. Thus, they can get almost twice the duration without decreasing quality one pixel. Second, they are mostly from 24 fps film. Compared with 30 fps (nominally) video, this is another 20%. Hollywood also has access to amazing encoding hardware and software, and they pay people very good money to tweak certain frames and scene transitions to eliminate problems.

I do a lot of sports video, and even at 8,000 kbs encoding, I see a lot of artifacts that weren't in the original. High contrast material can be particularly difficult, as can material with lots of dissolves (I shudder every time I see a dissolve on DirecTV -- artifacts everywhere).

I had some ancient VHS footage that I cleaned up, but I had one scene that was taken in winter of Grant Park in Chicago, from a moving car. The "moving" barren trees against the clear blue winter sky created nothing but a pixel storm until I got the encoding rate over 7,500 mbps. Even then, the footage didn't come close the footage that was captured and saved as DV AVI. All this, even though the original was shot on non-HQ VHS on 1981 consumer equipment.

Bottom line: Do some simple tests. Take 10-30 seconds of video from a portion of your project. Pick something with movement, a dissolve, and something with high contrast. Encode at 4, 6, and 8 mbps. Then, author a DVD, burn onto a DVD-RW, and view on the largest TV or monitor you've got. Watch each clip at least three times. If you can't tell any difference, then go with the lowest bitrate, and cram everything onto one DVD. If you CAN tell a difference, then following the earlier suggestions, and use multiple DVDs.

BTW, you can get jewell cases that hold up to four DVDs in one standard size jewell case, so storage of the extra disks really won't be a problem.

I am somewhat mystified as to why one needs a chart or calculator to do this. It's very simple, and takes a few seconds:

Total bitrate = 600 / minutes

That's it. For your example, 600 / 120 minutes = 5Mb/sec. That's the target average bitrate you need to be at. Since a 2-channel Dolby audio stream is .2Mb/sec, it leaves a video bitrate of 4.8Mb/sec.

So you set your minimum to 0, the maximum to about 9, and the average to 5.8. Done.

What I can't speak to is whether the MPEG2 encoder in Vegas is all that accurate at meeting the average bitrate you specify. The encoder I use (Cinemacraft) nails it.

The quality you will get at this low bitrate is completely dependent on how noise-free your source video is. Noise just kills an MPEG2 encoder, which gets a lot of its bitrate reduction by looking for similarities between frames. Video noise implies that identical pixels from adjacent frames are never the same value. The encoder is forced to spend all its effort in encoding the noise. I've done two-hour DVD-5s that look very clean.

Finally, I've had a chance in recent months to do a lot of experimentation with a new program from Nero called "ReCode." It allows one to take a conventional Hollywood DVD and squish it down so that the 8GB data fits on one of our 4.37GB DVD-5 disks. In some of my movies that I've compressed in this fashion, we're talking about a further 40% increase in compression over what Hollywood had already done in making the original disk. The amazing thing is that the video still looks remarkably clean. I attribute it to the fact that, once again, Hollywood has access to very very clean images. 35mm film, done right, is almost noise-free. As mentioned before, it also doesn't hurt that they only have to run at 24 rather than 30 frames per second.

One thing we all forgot to mention, encode the audio to ac3!
That leaves a lot of room for more video.
As for th rest of what's been said, it's absolutely correct. Hollyowwd not only has the very best encoders but also the cleanest source, both of those make a HUGE difference. If got a DVD of footage from a HDW 900 and it looks stunning.
I'd also point out that they don't get these results in a few hours, as much time can be spent producing the DVD of the movie as went into making the thing in the first place. That's why it bugs me when I hear complaints about how long encodes take. I'd love to have the money to buy the top line CineCraft encoder and run 400 pass encodes. It'd take weeks but I bet the results are worth the wait.

Question to riredale and farss (and anyone else):

Do you do anything to DV video to reduce noise prior to doing the MPEG2 encoding?

John:

Nothing as elaborate as you've described in your 10/4/03 thread.

Last year I experimented with the "Smart Smoother" in VirtualDub, and found it to help quite a bit. If taken too far, however, the images took on a sort of "cartoonish" appearance, with subtle detail wiped out in broad areas, which is what one would expect. The CinemaCraft MPEG2 encoder also has an adjustable "antinoise" function, but since it's just a simple filter (as far as I can tell) I leave it off.

When I get some extra time I'd like to experiment with some of your techniques.

One thing I did do this year was to get a new camera. The VX-2000 is famous for its low-light ability, and I think this will help reduce video noise a lot for my indoor shots. When I win the lottery I'll invest in a portable 35mm camera, complete with a Steadicam rig. Jaws will drop when I walk into the next VASST seminar.

Thanks to everyone for the insight on Mpeg compression. I have another question, I rendered out to mpeg2 in V4 with bit rate of max of 4 to a project of 1hr 58min. The resulting files (I spelt the live music video to 4 files at band change breaks.) where a total size of 4.2 gigs, yet when I brought them into DVDA the resulting DVD total size was 7.9 gigs(according to DVDA estimation of output, I included a small nav menu). I went ahead and let DVDA prepare the DVD without recompression the resulting size was 6.9 gigs. I thought that no compression was needed if you total file size plus menu was with in 4.7 gigs.

DVDA has serious bugs when importing files. There are lots of topics on this. Try starting a new DVDA project and bring that file into DVDA again, and immediately look at the Optimize tab. Most of the problems with getting 2x the file size seems to involve using submenus.

"don't get these types of threads. What is the problem of splitting a 2 hr video into two DVD's?"

Becasue there is no need to do that, 2hrs can easily fit on a DVD-R and look great. And why waste the cost of having to master/duplicate 2 DVD's when you can just use one. It's simply a matter of learning compression techniques and what works and what doesn't. I guess that is why they call compression an "art".

"don't get these types of threads. What is the problem of splitting a 2 hr video into two DVD's?"

Becasue there is no need to do that, 2hrs can easily fit on a DVD-R and look great. And why waste the cost of having to master/duplicate 2 DVD's when you can just use one. It's simply a matter of learning compression techniques and what works and what doesn't. I guess that is why they call compression an "art".
============================================
A. "Look great" is a subjective matter. To my eyes they don't look as good at lower bitrates no matter who the "artist" is who did them.
B. "Waste the cost"? What cost?

John

Bottom line is: If putting two hours on a DVD is important to you.....squeeze away!

@ johnmeyer,
of course, DV noise cleaning it's a must before encoding.
DV video needs filtering(cleaning) and it doen't matter if you shoot with CCD or CCCD cam.Also the cleaner the source the better final result as the encoder doesn't have to encode the noise as well+the encoding it's faster.
For my encodings I use Convolution3d,VagueDenoiser,Dust,MipSmooth,PeachSmoother,UnDot....
and I preffer AVIsynth over VirtualDub(Mod) because it's faster.So far I'm more than happy with AVIsynth+Canopus Procoder Express.
Very good post by Spot about rendering the timeline from Vegas,but also depends which codec you use to render to AVI first.PicVideo has lots of options also huyffyuv.Rendering to DV-AVI from Vegas timeline means rendering to MainConcept DV-AVI, meaning version 2.xx
You can set default codec in Vegas if you have the new MainConcept Encoder v.1.4.1 which is v.2.4.9.0(codec version) lots of improvements.

Very good denoisers comparison@ Doom9, have look.
(By the way I'm DSP8000)

AudioIvan

AudioIvan,

Interesting post. I've experimented with Peach and Dust, but am not familiar with the others. They all do similar things. I assume that you don't use all of them, certainly not all at once. Is there one that you use most often when cleaning DV footage, and if so, which one and with what settings?

You mention a post from Spot about rendering from the timeline. Are you saying that you frameserve from Vegas into AVISynth?

You also mention using the Mainconcept codec. I assume you are referring to the Mainconcept DV codec. I have this (latest version), but have still been using the Vegas codec when working within Vegas. Are you saying that I should use the Mainconcept codec within Vegas in lieu of the Vegas DV codec?

Finally, I'll look at Doom9 to see if I can find the denoiser comparison you refer to. I haven't seen it, but would love to read it. If you can give me a clue where to look, let me know

Thanks!

[Edit]
Ah, perhaps this is the link:

Denoiser Comparison

@ johnmeyer,
First, depends on the source(high,low motion,light,cam quality...),for each video different script,you have to experiment to get the best possible result.
Some examples:

LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\aWarpSharp.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\MipSmooth.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\old plugins\ColorYUY2_for_25.dll")
AVISource("D:\video.avi")
ConvertToYV12()
aWarpSharp(depth=16.0, blurlevel=3, thresh=0.5, cm=2)
MipSmooth(spatial=2, temporal=2, method="superstrong", scalefactor=0.80, weigh=true, upsizer="bicubic", show=false)
ConvertToYUY2()
ColorYUY2(Levels="PC->TV")
Letterbox(54,54,0,0)
DoubleWeave.SelectOdd
---------------------------------------------------------------------------------------------

Import("C:\Program Files\AviSynth 2.5\plugins\AddAudio.avsi")
LoadPlugin("C:\LoadPluginEx.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\old plugins\DustV5.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\VagueDenoiser.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\kerneldeint140.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\UnDot.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\UnFilter.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\old plugins\ColorYUY2_for_25.dll")
AviSource("D:\video.avi")
ConvertToYUY2()
SpaceDust(5)
VagueDenoiser(threshold=1.5,method=1,nsteps=6,chroma= true)
KernelDeint(order=0,threshold=0,sharp=true,twoway=false)
Undot()
Unfilter(+5,+5)
ColorYUY2(Levels="PC->TV")
AddAudio()
---------------------------------------------------------------------------------------------

Import("C:\Program Files\AviSynth 2.5\plugins\AddAudio.avsi")
LoadPlugin("C:\LoadPluginEx.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\cnr2.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\eDeen.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\PeachSmoother.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\kerneldeint140.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\UnDot.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\UnFilter.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\old plugins\ColorYUY2_for_25.dll")
AviSource("D:\video.avi")
ConvertToYV12()
Cnr2()
eDeen(5,6,18,1,3,false)
ConvertToYUY2()
PeachSmoother(NoiseReduction=50,Stability=0,DoSpatial=false)
KernelDeint(order=0,threshold=0,sharp=true,twoway=false)
Undot()
Unfilter(+2,+2)
ConvertToYUY2()
ColorYUY2(Levels="PC->TV")
AddAudio()
---------------------------------------------------------------------------------------------

Import("C:\Program Files\AviSynth 2.5\plugins\AddAudio.avsi")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\convolution3d.dll")
AVISource("D:\video.avi")
ConvertToYUY2()
SeparateFields()
odd=SelectOdd.convolution3d(1, 6, 10, 6, 8, 2.8, 0)
evn=SelectEven.convolution3d(1, 6, 10, 6, 8, 2.8, 0)
Interleave(evn,odd)
DoubleWeave.SelectOdd
AddAudio() (this is one of the most common(entry level) scripts for CCE)
---------------------------------------------------------------------------------------------
"You mention a post from Spot about rendering from the timeline. Are you saying that you frameserve from Vegas into AVISynth?" NO, this one
"Rendering to an MPEG from the timeline IS a better option, because you are at that point effectively rendering 4:4:4 to 4:2:0 instead of 4:4:4 to 4:1:1 to 4:2:0. The reason I'm seeing the difference, is that most of our MPEG output is to progressive scan formatting, and not to interlaced formatting. No scaling occurs in interlaced formatting, only in deinterlaced mode. This is why I'm seeing the improvement that I'm seeing. With high motion content even in interlaced mode, you'll see some improvement with the DV to 4:2:0 flow, ONLY because of some anti-aliasing that is taking place, but if you take the direct gradients and render them to MPEG from the timeline with zero movement, you can see a marked improvement over rendering to DV first.

So, to clarify:
Regardless of the origin of timeline content, render straight to MPEG, not to DV first. It's slower, but more accurate to original.

If you are blending fields, there is a visible difference/improvement in MOTION quality, but a loss in COLOR quality. "

Yes, the Mainconcept DV codec, but not the one that you pay for(v.2.4.4),the one from the encoder because it's newer.
As I stated in my previous post,first you have to overwrite the Vegas DV codec or replace with the one from the encoder, otherwise the video gets corrupted and you'll see macroblocks,artifacts...it is some kind of mainconcept bug.It happend to me and I learned the hard way.
Finaly, yes thats the link, Wilbert & Ivo comparison test.

Hope this helps,
AudioIvan