why is picture quality on dv-cam hooked up to tv better than

[kurtvidal]

Can anyone tell me why the picture quality on dv-cam hooked up to tv is better than the same footage dumped to avi and burnt to dvd?

[adam]

Well are you compressing your footage at all when you encode to avi? I can't imagine fitting much content on a DVD with uncompressed avi. Anytime you compress video you are going to lose some amount of quality. Also, when you capture dv to your pc you are probably capturing using a specific dv codec and some of these do filtering, ie: chrominance adjustments, blurring, etc... which may or may not result a decrease in quality to your eyes.

Another thing, if you are burning avis to a dvd then its unlikely you are playing these on a dvd player unless you have one of the few players that support mpeg4, in which case just refer back to my first paragraph. But if you are comparing dv played directly from the camcorder to the tv, versus the dv footage captured to your pc then it will always look better on the tv because DV is interlaced and pc monitors are progressive. When played on your pc your playback software is deinterlacing so it is just throwing out half of the temporal information, which results in a blurred image.

[Luddite]

one hour of dv avi is like 13gig
ya can convert it to dvd keeping the same quality
bitrate 9000....just squeezes on one dvdr

[adam]

Including audio, DV uses something like 30Mbits per second. There is simply no way to do a lossless conversion from DV to DVD even if you only have 1 second of footage. You are still going to be compressing it by more than %50 because the DVD standard just doesn't support that high of a bitrate, though at 9Mbits it should still look pretty darn good. Regardless of the bitrate, there are still often things being done to your source which you can't control, ie: whatever filtering your codec does.

[Xesdeeni]

It's heavily dependent on your MPEG-2 encoder and your encoder process. Some of the consumer-level DVD authoring software come with very bad encoders. For best results, get CinemaCraft Encoder Basic for $58. IMO, TMPGEnc isn't as good at MPEG-2. I've heard that Canopus & MainConcept's encoders are OK as well. You may also want to feed your DV video through some filters, at least to correct the 4:1:1 chroma issues that most DV codecs seem to introduce, prior to feeding them to your chosen MPEG-2 encoder.

One other thing, many consumer-level DVD authoring software packages use only PCM for the audio. That means the audio takes a constant 1.5 Mbps. For a 1 hr video, this means the audio will take up 691 MB. That leaves 3.6 GB for video, limiting your video to 8 Mbps. Using AC3 (or MP2 for PAL or if you are sure your NTSC DVD player can play it), you could cut that down to 384, 256, 192, or even 128 Kbps without much noticable lose of quality (depends on the quality of your audio at the start). Then you could max out your audio data rate at over 9 Mbps for best quality.

Xesdeeni

[tnc1]

Posted: Aug 07, 2003 10:42

--------------------------------------------------------------------------------

It's heavily dependent on your MPEG-2 encoder and your encoder process. Some of the consumer-level DVD authoring software come with very bad encoders. For best results, get CinemaCraft Encoder Basic for $58. IMO, TMPGEnc isn't as good at MPEG-2. I've heard that Canopus & MainConcept's encoders are OK as well. You may also want to feed your DV video through some filters, at least to correct the 4:1:1 chroma issues that most DV codecs seem to introduce, prior to feeding them to your chosen MPEG-2 encoder.

One other thing, many consumer-level DVD authoring software packages use only PCM for the audio. That means the audio takes a constant 1.5 Mbps. For a 1 hr video, this means the audio will take up 691 MB. That leaves 3.6 GB for video, limiting your video to 8 Mbps. Using AC3 (or MP2 for PAL or if you are sure your NTSC DVD player can play it), you could cut that down to 384, 256, 192, or even 128 Kbps without much noticable lose of quality (depends on the quality of your audio at the start). Then you could max out your audio data rate at over 9 Mbps for best quality.

Xesdeeni


I was wondering just what programs specifically you are using? I am impressed with the depth of knowledge you have on this subject and sounds like you've " been there, done that" already. I am still experimenting with multiple programs trying to figure out just which ones really do a good job and which are just easy and merely turn out alright. I have a dv Camcorder and also want to get the best quality dvd-r as I can . Could you list which programs you are using from capture to burn to get your great results. Thanks for sharing.

[kurtvidal]

I'm using a Sony DCR-TRV22 digital Camcorder to dump via firewire onto my PC (winxp). My Camcorder manual states that the i.link baud rate is appx 100Mbps. I'm not familiar enough w/encoding to know if it is hardware or software driven. The capture & editing software I'm using is photostudio 8.5. I dont know if any of you are familiar with this software but it has option of allowing you to capture "full dv quality", or "mpeg quality". I always use "full dv quality". Here are the video capture settings in photostudio:
compression=dv video encoder
framerate=29.97
kb/sec=3745
After I finish editing the footage I again use photostudio to output/burn to a DVD. The "burn" option I always select is "best video quality". I generally get 60min of video/audio per DVD. Here are the output settings in photostudio:
video=best quality
kb/sec=8000

I also own a canopus ADVC-100. Would any of you recommend I run my DV Camcorder thru the canopus then to PC to get better video quality? I see mr. tnc1 mentioned something about canopus encoders.

[Xesdeeni]

Please don't think this is the absolute best or only way to go about the process. This is just the method I have settled on at present, that appears to give me the best quality, and be the least trouble:

- Capture Hardware: PyroDV Basic - PCI IEE1394 card
- Capture Software: Premiere 6.5 - has batch capture
- Edit Software: Premiere 6.5
- Titling Software: Premiere 6.5
- Other Processing: Avisynth - use for exotic processing (see my script for removing a sports ticker) of some sections of video that are placed back into the Premiere timeline; SoundForge - Used to normalize the audio prior to encoding...BeSweet will probably do the same thing
- DV codec (when necessary): MainConcept
- Video Encoding: CinemaCraft Encoder - has plugin for Premiere 6.5
- Audio Encoding: BeSweet (with BeSweetGUI or AC3Machine)
- DVD Authoring: DVD Maestro - not the easiest, but allows AC3 audio, multiple audio tracks, multiple video files to be grouped as a single one (I sometimes encode sub-sections that I want to play together), and multi-path playback (which I haven't actually experimented with...yet)
- DVD Burning: Nero 5.5.x

When I capture with Premiere, I save type 2 DV AVIs, so I can process them with AVISynth if necessary. The batch capture allows me to preview (log) the video and then enter the areas I want to actually capture. The capture process is then automatic, and I can leave the room while Premiere and my camera work out the details.

Premiere is a decent editing program, but it has a number of quirks. It has some fancy transitions, but I never use anything but crossfades, fades to/from black, and a very occasional diagonal wipe. The earmark of an amateur is use/overuse of fancy wipes, which actually detract from the video.

Premiere isn't great for audio, it can only cut on frame boundaries, and the processing is very coarse. But you can crossfade and adjust volume somewhat.

Before Premiere 6.5, the titler was just passable. I used an animated title on my daughter's first year DVD and it had soft edges that didn't occur if I hadn't added movement. This makes matching up still and moving titles impossible. The titler has been completely redone in 6.5, but I haven't used it enough to say whether it is actually a lot better yet...but there is hope.

When you export to the CCE MPEG-2 encoder, you get no extra loss if you are processing any video (e.g. fades, wipes, or titles), because the uncompressed frames are fed to the encoder. If you export to a DV file, you get a recompression, although the video you have only cut edited will not be recompressed.

This process doesn't allow any overall video processing, like noise reduction, deinterlacing, or 4:1:1 correction, that saving an intermediate file would allow. To do this, you would need to save either in a lossless codec (like Huffyuv) or DV (which gets the above recompression). Then you could process through AVISynth and feed to CCE. The problem is that you are still limited to the 2/4GB limits of AVI files (even with NTFS), so you would have to save out the video in parts, and the rejoin them at some point. DVDMaestro makes this pretty easy, since you can just drop lots of video files on the same line and have them treated as the same movie. SpruceUP didn't do this, and of course DVDIt! and the other cheap titles have even less flexibility.

1 Hour of DVD using AC3 audio (384 Kbps or less) can be encoded at 9 Mbps. CCE does a much better job (IMO) than TMPGEnc, and is faster as a bonus (even with 3-pass VBR). I've heard conflicting reports about comparisons to MainConcept and Canopus, but the MPEG-2 encoder included with Premiere 6.5 (I think it's MainConcept) isn't as fast or as good in my tests.

For home movies, 1.5 hours is acceptable (I'm EXTREMELY picky). I use 6.25 Mbps for the video. Most of the artifacts remaining are directly from the camcorder (a single CCD camera will have lots of resolution and noise issues), the DV compression (4:1:1 isn't pretty), or the interaction between the artifacts in these formats.

Remember, you'll never get Hollywood quality DVDs from a 1-CCD DV camcorder source, but you can get very close.

Xesdeeni

[geek rock]

I also own a canopus ADVC-100. Would any of you recommend I run my DV Camcorder thru the canopus then to PC to get better video quality? I see mr. tnc1 mentioned something about canopus encoders.

running your camera thru the canopus advc-100 would be a pointless task, as it's a DV converter (as is your camera). possibly running the camera through to a canopus mpeg-2 converter would help, but the advc-100 will not do that.

try using TMPGEnc - it's a superb encoder.

-Mark

[powaking]

I have this same camcorder. What I find is you can either have 12bit or 16bit audio. That being said you know that its not DTS quality so you can probably get away with compressing the audio a bit more and try to use less for the video. I don't have a dvd burner yet, just got the camcorder myself.

I did capture a 30second clip off my camcorder with 12bit audio. Raw data came in at 108megs. Using Divx with 2000bit rate for the video the file compressed down to 10megs. Can defenitly tell the difference. I haven't tried it at 9000bit rate. I'll give it a whirl and see what results I get.

[nimco]

I am confused, what makes DV AVI files limitation to 4GB with NTFS, is it file by itself has problems or some peaces of software will not allow to process it properly? Is it really not available to run DV avi from Avisynth to CCE without cutting a file to pieces?

I would appreciated for your help.

[andkiich]

If you've got an NTFS file system and the files keep getting stopped at 4GB, then there is some software somewhere that is arbitrarily doing it. Also, make sure that there is no confusion, like the OS is on a seperate drive with NTFS and the capture drive is FAT32.

I routinely capture DV and analog streams around 60GB with no issues.

[MOVIEGEEK]

@nimco,
Try DVapp or DVIO.

[ronnylov]

I am confused, what makes DV AVI files limitation to 4GB with NTFS, is it file by itself has problems or some peaces of software will not allow to process it properly? Is it really not available to run DV avi from Avisynth to CCE without cutting a file to pieces?

I would appreciated for your help.

Try OpenDMLSource() or AviFileSource() in avisynth if Avisource doesn't work. See http://www.avisynth.org/index.php?page=AviSource

To open segmented avi files you can use SegmentedAviSource() in avisynth.

I'm pretty sure I have opened DV avi files type2 larger than 4 GB in avisynth with opendmlsource()

[nimco]

Thanks for your answers.

Sorry, I did not formulate my question correctly.
I am using AviSource() and it seems to me it is no problem in Avisynth for files larger than 4GB. But when I saw that Xesdeeni has mentioned about this limitation in this post above,

The problem is that you are still limited to the 2/4GB limits of AVI files (even with NTFS), so you would have to save out the video in parts, and the rejoin them at some point.

I thought maybe I just missed something. I am just curious what kind of limitation he was mentioned?

I appreciate for your help.

[handyguy]

I run my camcorder output to a dvd recorder & make a dvd & it looks the same as the original that way. Sony is coming out with a new camcorder that records on dvd-r, although panasonic already has one.

[vhelp]

.
.
I have the TRV22 also, ..great cam choice, by the way..

It's unfortnate that some of you's all are having "quality" issues w/ your
final output (MPEG-2)

Are you all shooting footage in 4:3 mode ??
These tend to produce the most issues w/ quality ..ie, blocks & pixelation

Are you guys encoding w/ Interlace on under TMPG ??
If so, make sure that you follow the steps below:

* VIDEO Tab: Encode Mode [Interlace]
* ADVANCED Tab: Video Source Type [Interlace]

Here are some helpful and useful tips:
* Shoot in widescreen 16:9 mode instead, and encode to 16:9 out for widescreen
...output
Unfortunately, there is a slight side-effect when shooting 16:9 mode w/ the
TRV22 (or any other CAM out there) ..saw-tooth on some diagnols and some
minor glitchs in horizontals.. but only if you look for them.
So, don't go discounting this feature (16:9 mode) use it, and see for your
self.
.
.
This may or may not bring down the bitrate or increase in quality. But, do
make sure you encode w/ Interlace on (observe above note on Interlace)

But, also bare in mind that encoding w/ Interlace on will decrease in your
final output quality. Most of the "quality" hits are due to the user panning
to hard and fast, and also quick distorted movements such as improper grip
or re-positioning of cam in hands, or movement w/ Tripod arm.
Also note, that the faster you pan your CAM, the harder it is on Interlace
(or the more the Interlace separates the fields [aka, motion temporal])
Another "quality" hit is when the user shakes the CAM in his/her hands,
..ie, gripping too tight give CAM vibrations/shakes. ..Instead, hold your CAM
loosely, and when you pan, use your BODY's fullness to turn, instead of your
"hand and cam" turning.
Also, breath in "timmed" increments, not freely. I found this to cause the
CAM to shake, jerk or whatever (you'll see when your encodes to poor in
quality)
Note, every little detail above, helps add to poor quality.
It's really so very important to use a TRIPOD. Prefereby one w/ Oil-filled
for smoohtest/glidest/silky movements. These kinds of TRIPODs will eliminate
those sudden jerky or sticky-jerky movements in your CAM. i'm heading on
getting one myself. I'm currently using a regular, el-cheepo one, and I too,
have issues w/ those jerky or sticky-jerky movements.. though I've learned
a few tricks to help in eliminating these jerkes.
But, it's very important to get a tripod, and learn how to use it, and get
along w/ it, being that it's a bit clunky to setup and use. Get over it,
and you'll very SOO MUCH HAPPIER. Huge difference, when I take outside footage
and in low light, ..ie, shooting footage of baseball leage game at night, when
only some flood lights are on the field. Lots of "low light" noise. But,
when you rest your CAM on a tripod, things get a whole lot better, ..ie, no
more shakes (which cause poorer quaity) These shakes are constant, when CAM
is in users hands, plus add in the user not breaking in a "timmed" set increments.
Anyways... I'm rambling on here..

If you did not pan so much (or eliminated them all together) you'd see a
DRAMATIC difference in quality. Pan only when necessary. I've ben studying
and observing how TV and Video work together. I've found some pretty nifty
techniques that the cause user would not see, and use in THEIR OWN shooting
practices. But, I now know some of the gurus techniques and secrets of the
trade.. thanks to a keen eye on events on TV
.
.
Perhaps I'll incoporate this in another topic I'll start at a later time/date.

If I've missed some things, sorry.

Please do observe some of the above in your next DV CAM projects.

Enjoy,
-vhelp