Posted by PTravel on 02/18/07 07:09

"Gene" <genes@wildblue.net> wrote in message
news:W7qBh.36$EP6.44076@news.sisna.com...
>
> "PTravel" <ptravel@travelersvideo.com> wrote in message
> news:53mkkjF1susrbU1@mid.individual.net...
>>
>> "Gene" <genes@wildblue.net> wrote in message
>> news:DFnBh.33$gP6.19476@news.sisna.com...
>>
>>> 2. I was very disappointed with the speed that it took to get from
>>> AVI to the 19.6GB MPEG2/VOB file. I had hoped for an improvement,
>>> but I suspect that there is only so much low level coding that can
>>> be done ... just a lot of arithmetic to do.
>>>
>>> Here's the times:
>>>
>>> a. Copy 1.5hr from camcorder to disk ~ 1.5hr :-)
>>> (Tape was 1:37:32 long, it correctly saw EOT & stopped the
>>> capture automatically.)
>>>
>>> b. Convert AVI to VOBs = 5 hours, 38 minutes, and 17 seconds,
>>> which is probably not that bad, given my old 2GHz P4.
>>> If there was a way to SKIP this step, my life would be much better:-)
>>
>> Of course, there is not. This is the transcode step and, as you note, it
>> is very processor-intensive. Unfortunately, you can't have it both ways.
>> If you want good quality video, the transcode will take time. If you
>> want a speedy transcode, the video quality will not be good. For what
>> it's worth, a 2-hour transcode at the highest quality settings takes
>> 12-24 hours on my 3.2 Ghz P4 running tmpgenc.
>>
>
>
>
> "If you do not need to edit", and all you want to do is burn the best
> DVD-R
> from a MiniDV camcorder tape, then there in "NO" better , or faster, way
> than
> using good quality HARDWARE. A well inplemented DSP, talking to another
> DSP is
> by far the better (and faster) solution, IMHO. Am I wrong about this?

I think so. All other things being equal, multiple-pass analysis will
always yield a better transcode. Unlimited time for analysis will always be
more thorough than real-time.


> I assumed that
> the PC capture programs do nothing more than transfer the bits from the
> tape to the PC disk.

That's right.

> Do the capture programs actually enhance the quality
> of the resulting A/V file in the PC?

It's a straight file copy -- no data is changed.

> If you just want to caputure the tape
> to a best quality single DVD-R - then why would you want to spend an
> extra five and a half hours on a PC?

To get a better transcode which, in turn, will determine the final video
quality.

> Looks like plugging a camcorder into a
> DVR is a MUCH faster, and from what I have seen, gives as good as, if not
> better
> quality than DVD-Rs generated from a PC program.

Then I'm sure you'll be happy doing it that way. First of all, you're
wrong -- a DVR will not produce as good a DVD as _properly_ transcoded and
authored DVD produced on a computer. I don't know what you've seen.
Obviously, it's either not produced by good software, or whoever produced it
didn't know what they were doing. However, everyone's standards are
different. If you're happy with a DVR, then more power to you.

> I'm sure I do not
> undersdtand, so what am I missing? As I stated above, I have zero interest
> in
> editing, or using a PC to in any way to enhance the quality of my tape.

I explained at some length how video quality is effected by the type of
mpeg2 transcoding. I'm not explaining it again.

> All I want is
> a good conversion to video DVD in the form of a fully stuffed DVD-R x 2.
> 1.5hr to copy, and 10 minutes to burn seems like a MUCH better solution
> than hours on a PC - especially when the final DVD-R is comparible.

I'd strongly urge you to use your DVR. It's bullet-proof, requires no
understanding of the underlying processes that are involved and, evidently,
produces results that you're happy with. Since, clearly, that was the
answer you wanted, I'm not quite sure why you went through this exercise in
the first place.

Have fun.

>
>
>
>
>>
>>> As for the filming that I am currently doing, I think I will just keep
>>> shooting
>>> MiniDV & D8, rewind the tapes & use them over & over again,
>>
>> That's not a good idea. Repeated reuse increases the likelihood that the
>> magnetic coating will flake off the substrate, causing drop-outs and
>> increased camera wear. Tape is cheap. Use them once, capture them once
>> and then store them.
>>
>
>
>
> No big deal, when the tape starts going bad you will start seeing
> artifacts,
> typically at the very beginning of the tape. That tape is tossed into the
> trash can. MiniDV tape is not cheap, DVD-Rs are cheap:-) I have been
> doing
> this for over 20 years, and have never lost an important tape. I began by
> shooting
> expensive 8mm and copying over to cheap VHS. You lose a little quality,
> but
> was no big deal to me.
>
>
>
>
>>
>>> and simply
>>> copy the 1.5hr LP digital tapes onto two DVD-Rs with a set-top DVR at
>>> ~ 95%+ full. If there is a VERY special tape, then I will burn the two
>>> DVD-Rs & put said tape in coolers.
>>
>> D8 and miniDV tapes will last far, far, far longer than a DVD-R. They
>> are the cheapest and most reliable archive solution. Additionally, if
>> you ever want to edit your videos, you'll want to preserve them in the
>> highest quality. Mpeg2 transcodes are lossy -- you will lose detail and
>> quality. Mpeg2 is also difficult to edit and not well supported by the
>> better editing packages.
>>
>
>
>
> LOL - how do you "know" that tapes will last longer than DVDs - we have
> not had time to tell? My gut feeling is that they will both be around in
> 20 years.
> We never plan on editing our data, except for cutting out a section and
> pasting
> for an occasional video DVD, so we could care less about editing. The
> bottom
> line is "we are happy with the video DVD quality of the DVD-Rs".
> They look just fine on our 1080i TV.
>
>
>
>
>
>
>>>
>>> I initially thought that all the freeze-ups that I was having was
>>> related to the
>>> DSP/firmware implementation inside my three Sony D8 & MiniDV camcorders.
>>> However, after replacing the VIA chip with the TI chip, all three
>>> camcorders
>>> are about as bullet-proof as they can be. That is, I can copy from
>>> camcorder
>>> to PC hard drive with ANY cheap firewire cable & have never had a
>>> freeze. With
>>> the VIA chip, it was unlikely you could copy a tape without freezing.
>>> ALL of
>>> the capture programs worked without incident, as this application is
>>> mostly
>>> hardware talking to each other.
>>
>> I've edited on PCs with VIA chips and never had trouble capturing.
>> Capture stability varies depending on the capture software that you use
>> (I use Scenealyzer Live, which is rock-solid), and what you have running
>> in the background.
>>
>
>
> In my specific case, the VIA chip was the problem, 100% for sure.
> I can recreate the problem by simply replacing the VIA board & the Sonys
> crash.
> My Canon ZR65 did not have a problem with the VIA, as I remember - just
> the Sonys.
>
>
>
>
>>>
>>> So my challenge will be to find a DVR with DV in & a hard drive that has
>>> a
>>> very well implemented firewire specification.
>>
>> All that is needed is OHCI-complaint 1394. Virtually everything is these
>> days.
>>
>
>
>
>
>
> Aparently my VIA PCI card, or the Sony camcorders were not compliant -
> else why does
> this combination crash, and the TI chip board run hours on end without a
> problem?
> The ONLY change that I made was swapping the PCI cards. Your statement
> that all are
> virtually compliant is simply wrong. You give the erronous impression that
> every PCI
> firewire card from China will function without a problem. That
> conventional wisdom cost me
> a week in time, I assumed the VIA card was implemented properly for the
> Sonys. I can
> swithch cards and crash at will - 100% of the time. That's a pretty good
> indication that
> this specific VIA card did not have what it takes to capture from all 3 of
> my Sonys, and the
> Belkin with the TI chip did. My VIA card could just be defective, and all
> other VIA cards
> from China are OK, but I really doubt it. I once part-owned a hardware
> company that made
> state-of-the-art cards for the PC, I know how easy it is to have a logic
> error, especially in DSPs
> implemented in custom chips:-)
>
>
>
>>
>>
>
>

[Back to original message]