Frage von area2051:Dear All,
I've already read several posts and tried s.diverse Programs, while some questions have arisen that I would like to clarify.
Starting point:
-I have an old analog Cam of SonyHi8.
Hauppauge WinTV-type unknown (old but works)
-capture software I use VirtualDUB
-coding software TmpegEnc
1.Frage:
In the "set custom format" of VirtualDUB are there so many adjustment options (size) and-I believe-color format (such as RGB, UYUV etc.)
-What format because I have to actually set it exactly and what color?
It depends I suppose off of my cam, right?
kssagts me
2.Frage:
After VirtualDUB I set a filter 480x576 with Lancazor3 and start a FrameMaker Server and then hand over the s.TmpgEnc.
- Since I SVCD's behind ever find out the size of the filter may well be ok, right?
3.Frage
- What is the meaning Lancazor3 or so ähnllich?
've Found instructions that only one of the ZIG, but so far not found a solution!
4.Frage
Since I'm running the server and the frame encoder on the same computer, it takes forever. If there is a possibility, a "simple design" frameserver somehow - possibly gibts da ja ne software for it, which allows a 2nd Calculator to use to encode?
5.
What settings should I choose to mpeg2, so it still rel. looks good to me-of the computer is synonymous anblieben overnight.
Have been just 2Pass selected and the highest bit rate. Is it that ok? gehts better? , ggf. ner with more bessersern quality?
6.
The result is flickering light to play with the DVD player. Film can look at, but somehow I got the impression as if he was there trying to balance a couple of frames that are not available.
- Has already had a time? Lösungsansetze?
VIEEEELEN YOU for each answer now!
Greeting
Mark
Antwort von Stefan:
1 / This is "no preference".
Here you can alter what is to deliver the card preferably s.Virtualdub. What can you actually set off the map and depends of the drivers. What can not the card gives an error message. Since you later, eh one post (Resize doing), you can just take an RGB format. A YUV format would have a faster without post-processing.
2 / Yes. Size is right for SVCD.
3 / Lanczos3
If a calculation method for the resizing of images.
A Comparison of different methods for photos, brought to you such as:
http://www.americaswonderlands.com/digital_photo_interpolation.htm
Here are a few "species" s.diskutieren on the Basics
http://www.avsforum.com/avs-vb/archive/index.php/t-460922.html
4 / No clue. If you find that, let me know ;-)
5 / My personal opinion:
It is not worth these days do not SVCD. Make a DVD. The burners cost "almost nothing" and the blanks are not synonymous. The quantum leap in image quality you bring through years of working meticulously with the Encodingeinstellungen not go outside.
6 / it flickers or jerky it?
As a flicker, I know a certain amount of unrest in areas with lots of details. The reason is that some small structures are presented in one field and not in another. The place for a long time s.einem interferes with viewing. This can be enhanced by a previous resize when the process is slipping just under the problematic structures Flimmergrenze. Remedy drüberzulegen is a slight blur filter (or make deinterlacing, and thus all fields equal to).
When bucking, I know missing frames (dropped frames) and exchanged frames (field order). Dropped frames VD you would register with the recording.
Swap regularly and especially with movement (pans, zooms) to see. Workaround is to tell the encoder to the correct field order from the source material and reinzubasteln in the production of the target material does not spin.
7 / Oops, we already finished ;-)
Good luck
The fat Stefan
Antwort von MiXMaster:
1 / This is "no preference".
... Since you later, eh one post (Resize doing), you can just take an RGB format. A YUV format would have a faster without post-processing ...
As flicker ... I know a certain amount of unrest in areas with lots of details. The reason is that some small structures are presented in one field and not in another. The place for a long time s.einem interferes with viewing. This can be enhanced by a previous resize when the process is slipping just under the problematic structures Flimmergrenze. Remedy drüberzulegen is a slight blur filter (or make deinterlacing, and thus all fields equal to).
First of thanks for the long answer.
to the first:
That is, if I ere afterward on SVCD rüberconvertiere I need not adjust these other formats, ie directly RGB?
Is it less memory intensive? IMHO is used in RGB 8bit per pixel then. what is stored in UYUV etc? .. I uUkönnte the codec can compress a bit less and so get even better picture. ;-)
the second:
Jap. it is no flicker, but this unrest, which you've described. How / what do I then turn s.besten s.in VD? What do you mean with blur filter?
Greeting
Mark
Antwort von MiXMaster:
or I can do that even with Tmpegenc??
Greeting
Mark
Antwort von Stefan:
RGB and YUV (UYUV is a subspecies of it) in the memory requirement to pay anything. They differ in the so-called color space (had recently written about it) and in the installation of the data in memory (ditto).
The setting of the codec, of course, depends on how much space you can sacrifice for the recording file / want. I usually take MJPEG with Quality 18-19, was approximately 10:1 compression or lossless Huffyuv what about 2.5:1 compression.
A blur blur-alias filter of the Picture "makes a soft". In VD, there is an internal filter to
blur, the
blur more and the
box blur filter. I would first try the simple blur filter s.einer notorious scene. Thou can turn the frameserving.
Whether TMPGEnc has) a blur filter (blur, I do not know at the moment does not look synonymous. TMPGEnc can resize. If you do not necessarily want to use Lanczos, can you save the frameserving. Synonymous Then you can think about how you can distribute Recording and encoding two calculator on the net.
Good luck
The fat Stefan