Any difference between D1 & Half D1 when source is VHS?

[ahfairley]

I'm transferring 8mm home videos to DVD using capture via DV cam analog passthrough.

Should I see a difference between encoding and burning to NTSC D1 (740x480) or half D1 (352x480), assuming I used the same bitrate for each?

How about it I encode D1 at bitrate 4000 and half D1 at bitrate 8000 (same lenght of video). Should I see better quality at the lower res and higher bitrate?

Thanks

Alan

[BJ_M]

1/2 D1 works very well for vhs/8mm analog source and always - the higher the bit rate - the higher the quality ... the higher resolution of D1 would be marginally better , but try it and I dont think you would see much diff....

[tvtuna]

Everyone sees the same thing differently. I keep reading how 1/2 D1 is the way to go for VHS, but I can see the lower quality and it bugs me.
I prefer D1, max bit rate 8000 but using variable bit rate.

It depends upon how many tapes you want to archive on DVD. If it is a lot, then it gets expensive if you don't compress. If it is a reasonable amount, you might want to squeeze out the most quality you can, so it looks its best in 50 years.

[BJ_M]

in 50 years .... you will not be able to find a DVD player anyway ...

disks will be the size of a fruitloop and hold 200tB's using Molecular or Atomic bonding storage

[tvtuna]

Great - another opportunity - the DVD to Fruit Loop transfer business

[Bel]

mmmm, fruit loops....

[drdy111956]

Well, Zed, Guess I'll have to buy the Beatles White album ...again...on Fruit Loop

[Thorn]

Just don't drop your fruit loop disc into your breakfast cereal.

Anyway... if you EVER intend to convert these to another format in the future (like High Definition Fruit Loop [HDFL]) you want to keep the best possible source material. So if you're ditching the old tapes, you might want to keep the resolution and bitrate high to keep the quality as high as possible (even if you can't visually see the quality difference now). I encoded some DV caps to mpeg a few years ago and now I wish I still had the source material.

Despite some views that 1/2 D1 is good enough (and in many cases it most likely is) the full res and high bitrate encode will store more information and be a better source. It's a personal choice as to how much that extra bit of quality is worth (doubling your filesize for a 2% increase in quality may be overkill for you... but we really don't have hard numbers).

As for your last question I'm curious myself, but for now I'm playing it safe and doing high bitrate full res encodes for my home-video.

[SatStorm]

For VHS, even 1/4 D1 is enough, with the right bitrate and interlace output. Because that is -x- for any optical media, 1/2 D1 is the closest frame size mach. Ideal for VHS and SVHS.

There are 2 parties in this world: One use only what it needs and the other use more for no reason.
The problem is that the second party always find "explanations" to justify whatever they support.

[Nelson37]

Why would you double the resolution and half the bitrate? If you want to compare the two, use D1 at 8000 and 1/2 D1 at 4000.

Second, I don't know the specs on 8mm other than it is NOT VHS. As I see you are recording to DVD, ergo you have space to burn, use the higher res AND higher bitrate.

If you are comparing "more than enough" and "just slightly less than enough", more is better.

[vhelp]

Evening all.
Also, to add...

It is BEST to capture it in AVI format (not DV, despite what others
may say) because DV is LOSSY too!
But it may be more convenience for people w/ DV cams etc. to utilitze
this approach, cause they have a) enought experience and prefer this
route, and/or b) this route may be the only alternative to the
old analog capture route, which leads to many frame drop issues.
c) they truely believe that DV is a better source of quality, etc.

In any case. . .
But, before you (or anyone else) embark on this adventure, BE SURE
YOU HAVE THE FOLLOWING ISSUE AT ZERO LEVEL:
* a) FRAME DROPS: 0
* b) RESOLUTION: 720 x 480
* c) AUDIO: a good handle on, ie 44.1k else for DVD or higher, 48k

All three are important, but line a) is MOST important.

Ok, so why line b)'s value?? cause you want to have as much area
to work with as possible. If you go w/ the 352x480 (though good)
later on's technology will only improve and be better, and if you
only have 352x480, you're stuck w/ regrit - "ah, shucks, should've
capped at 720x480!! ...and got THAT much better quality!!"

Yes, 720x480 for VHS capture is pushing it, but would be effective
later on, when you have better tools to work with, ADD to that,
that hopefully, you'll'of gotten better knowledge of capturing and
encoding or whatever they have a few years down the road.

Oh, and of course, you have to have some means of archiving these
LARGE files (AVI's) to disk. Most likely, DVDRs. But, then again,
is it worth all this?? You be the judge. Cause only YOU will
have the regrets later on, after you've destroyed your tapes.
As Thorn said, "I encoded some DV caps to mpeg a few years ago
and now I wish I still had the source material"

But, all in all, all is not lost. that is, your tapes will last a several
more years. So, taking on this embarkment would only tax you out, and quite
possibly cause headakes and issues in the process. IMO, better to wait
till you:
* have a grasp/good knowledge of capturing and encoding (in other words, NAILED)
* better media and cost effective (lower priceses)
* cost effetive methods and processes, and hardware and video devices of
such
That is, if you have no criticle reason to archive them OFF the tapes, so
that you can re-use the tapes. Else, I say, leave them on the tapes, and
just BUY new blank ones, and just build up a library, till you have truelly
MASTERED the process from START to FINISH!!

Give that some thought!
and, Good luck!
-vhelp

[mdm]

Many of the arguments for full D1 seem to talk about the "future", but your source material (VHS) is never going to improve. Is the actual captured AVI (lossless) really better at full D1? And what does this have to do with future technologies? If improved encoding methods are developed, you'll still be working from same capture quality. If full D1 was, say, 5% better in quality than half D1 (for VHS cap), I doubt future encoding methods will increase this difference much (since the half D1 encode would improve also).

Awhile back I read here (in these forums) suggestions to capture at hi-res but then encode to half D1. Is this still a reasonable option?

[BJ_M]

Why would you double the resolution and half the bitrate? If you want to compare the two, use D1 at 8000 and 1/2 D1 at 4000.

Second, I don't know the specs on 8mm other than it is NOT VHS. As I see you are recording to DVD, ergo you have space to burn, use the higher res AND higher bitrate.

If you are comparing "more than enough" and "just slightly less than enough", more is better.

8mm (he didnt say Hi8) is the same resolution as VHS exactly (some systems add 10 lines) and the same S/N also - 43-45

Hi8 and SVHS are also about the same resolution (but there are other differances )

DV stores video in component format, which produces better color accuracy than the S-Video format used by S-VHS/Hi-8 and especially the composite format used by 8mm/VHS. DV is lossy though.

i suspect many people when encoding don't change the default video setting from NTSC to component when encoding DV and also "output YUV data as YCbCr not CCiR601" (for DV) ... also if you CAP in component YUV you also may want to change the video format setting.

[Thorn]

VHelp: good point, for optimal quality I always capture analog signals in Huffyuv. The reason I think we jumped to DV for the storage format is 1) probably pretty easy to get the video into DV with the bridge. 2) DV is the easiest way to store the highest amount of data (at about 6 minutes per DVD-R I won't use Huffyuv for semi-permanent storage unless the footage was very important). 3) I assume the analog tapes will degrade over time, and will degrade faster every time they are viewed.

Essentially from that, I would think you want to capture just once. I don't keep the Huffyuv files, but perhaps this is a more complete solution than what I proposed (a bit more work, but there's always a tradeoff):
Capture to Huffyuv at full res. Then encode to the currently selected format (D1, 1/2 D1 or whatever). Then take that huffyuv file and A) Split the huffyuv avi into <4.3 GB chunks and archive to DVD-R (highest quality solution) or B) encode it to DV, and save that DV file to tape. (I'm assuming that the quality will be as good as the bridge's encoding to DV). My assumption is that the DV tape will weather better than the analog tape in the long run. Comments anyone?

MDM: If your final format (this time around) is 1/2 D1 or whatever, yes you will benefit from capturing at full resolution. I did a lot of VCD encodes that essentially went: cap at full D1 resolution->filter/edit in virtualdub->frameserve to TMPGenc and encode to VCD. The reason that you want to do this (among others) is that the analog signal is not in 1/2 D1 resolution, so it is being rescaled (in a sense) anyway. It is much better to downscale in software where it can compensate rather than letting your capture card decide how to scale the picture down (most likely the card won't be as good).

On the subject of the future: If the source material is still fully intact, recapturing from it would be a good option. However, will your analog camcorder work in 10 years? Would you buy another one just to read the tapes? I work on the assumption that when I capture from analog sources that I won't want to (or won't be able) to capture again.

Having the highest resolution and quality source material is important. My DV->DVD encodes today look SO much better than my DV->VCD encodes a few years ago (and therefore DV->VCD->DVD as well). It's not just the encoding technology, it's the newer formats that become available (and sometimes it's just the higher bitrate).

Now, for analog sources this becomes more of an issue. The noisier the source, the more high-resolution and high bitrate can help the quality.

BJ_M: I see the "YCbCr not CCiR601" option in TMPGenc for DV source, but where else in the endode/decode process does this come in to play?

[vhelp]

thorn,

The YCbCR comes into play (or not) given below:

I have found that if you tick the [x] Output YUV... in tmpg under the
[Quantize Matrix] tab for DV source materials, it helps a little bit more on
the quality side, w/out resort to adding the Color filter in tmpg.

But, I'm not too sure that it really matters, so you should/could try it out
and see how it goes for your Source's final encode. Take into consideration
that ALL the different version of TMPG that are out, have bugs in various
functions. So, while one function may work in one version, this same
function may not work in another (try all, or whatever you have) and while
this same function didn't work in one version, but now works in another
version, may not work yet, in ANOTHER version. I got so dizzy by Tsumai's
TMPG version development, that ended up leaving all the version but v2.53
I'm pretty somewhat steady w/ v2.53 at the moment. There are post
beyond numbers on these issues w/ TMPG versions floating around.

Currently, the latest (handful) issues w/ TMPGenc are:
* TMPGEnc 2.58 compressing times.
* Anybody used TMPGenc's latest version 2.58 yet?
* Buying TMPG
* TMPGEnc color problems
* 2.56 or 2.57?

So, bare the above in mind, in your endeavors.

v2.57 is where we were sold out to M$S$ and Licensee legalities w/ the
MPEG consortium, if you dare to believe - it's blanketed by the License
issue (even if it really is one of the issues) Look'it how the color has
changed for some. An good issue to NOT have.

Don't get me wrong. I Love using TMPG, but it's just consistently full of
bugs in EVERY version put out. They just NEVER, NEVER cease!! Nor will
they ever! But, I still love using it. So, I can't complain, can I?

-vhelp

[BJ_M]

i have to agree w/ vhelp on the TMPGenc version issue ... each version has produced different results .. some hardly noticable and some major ...

I use 1.x and 2.56 on various machines and so far going to stick with them ..

vhelp's info on YCbCR i agree with also.. always best to run your own tests though ..

the other setting for componet is in video format (which is the first tab shown on "settings") .. the location of this setting is somewhat misplaced because it is grouped with output settings but it is really the video format of the source.
all source settings should be grouped together and all output settings should be grouped together .. but they are all over the place.

i use most of the time frames as my input source (targa files) so my results are going to be different than someone who is using DV and someone who is doing vhs (for example)...

[vhelp]

Afternoon,

I forgot to mention, that when you source is DV, ie, Home-Made movies shot
through your lens, these can be fet directly into TMPG. But, the penalty
is the color washout issue. Luckely, you don't have to suffer this, as
again, turn on the [x] Output YUV... and it should re-sample (to some
extent) the colors, as when you bring into VDUB.
Note, when you bring your DV source into vdub, it loads in the appropriate
DV codec (one that is installed on your PC) and colors will be re-sampled
per DV codec. Frameserving this into TMPG doesn't require the need to
tick the [x] Output YUV... in TMPG.

Ok, to recap:
* if source is DV and from home-made videos via your camera's lens,
a) --> drag your source directly into tmpg, and tick [x] Output YUV... on.

* if source is DV and from pre-recorded sources ie miniDV of sources like
Cable, Satalite or DVD, and your source is not Telecined, but pure 29.970,
b) --> drag your source directly into tmpg, and tick [x] Output YUV... on.
no need to bring into vdub first for frameserving into tmpg. But, some may
find that they can convert to 23.976 and benefit. If you are one of those,
and you bring into vdub first, you don't kneed to tick [x] Output YUV... on.

* if source is DV and from home-made videos via your camera's lens, and you
bring into vdub first...
c) --> no need to tick [x] Output YUV... in tmpg (you can test though)

I think I got the above layed out correctly.
The benefit of dragging your source directly into TMPG is simple. Speed.
Your encoding will go a lot faster, thanks in part, of NOT framevering into
another source ie vdub. If your not IVTCing your DV or variatoin of such, you
don't need to frameserve the DV footage, unless you have other reasons to *.AVS
or vdub it first. But, if you know for sure that your source has ben Telecined
in some way, then don't tick [x] Output YUV... on.

-vhelp