Preface: As a newcomer when I Slashcam synonymous to Search no longer find an answer to a problem which Many videographers have to actually relate to: I try to time just to my many old video -8 To bring to digitize analog recordings and on DVD. (As a software I use Nero7, but that's another story.) On playback, it always comes back to image defects, short dropouts of individual lines or line segments for 1-2 frames (If we call the "drop-outs"?). As an example, see Attached Picture. I Play the head repeatedly cleaned without noticeable improvement, so I think the errors are already incurred in recording or by long storage.
Is there a filter, with whom I wegrechnen such errors can "? If so, what software / how do I merge them with Nero and ULead? Since each only 1-2 frames are concerned, it ought to be be an error correction by pixel replacement from the previous or to make the next frame. Or is that a market niche?
For any help I would be grateful Holger
PS: supplementary s.The professionals: What can you do to the quality to improve these shots?
Antwort von Markus:
You've seen either you or not really the wrong posts look for it. There is a simple trick, drop-outs during digitizing of Video8 and Hi8 recordings almost completely eliminated and not to Herumrechnen long.
Antwort von Holger Schlingloff:
First of all, thanks for the answer!
Does this mean that the recommendation is to provide a specialized hardware (SonyTRV480) to buy that has the drop-out filter HW moderately installed already? I need the part, but only once to create the copies. Somehow it seems to me exaggerated. Or you can rent something?
Antwort von prem:
Question which captures you now? graphics card, TV card or what?? When I helped another program to use. So I had as many drops with VirtualDub VirtualVCR with no more ... What matters is synonymous the codec. Where there is much to be compressed (eg DivX, Xvid and others), it is usually wrong. Unless man uses a special capture card, and therefore more expensive. Huffyuv zb is a fast codec, which works almost kompriemiert and virtually lossless. But beware, the files are then synonymous entsprechnd great .....
Antwort von miciculi:
It is about 30 V-8 tapes a 90 minutes, which I the Ravages of Time snatch wants (ie extra) hardware only in extreme emergencies. I have a feeling that the image quality due to storage worse (or is it just subjective, because today's recordings always better?)
Currently I have the old lion-Video8 Camera (Model 1988) to the A / V input in the new SonyPC330 (Model 2003) connected. The changing digital / analog carried by the Sonyand is reliable. From there it via FireWire s.das notebook, with Nero and ULead's is captured, cut and recodes for DVD.
The problems appear already s.Anfang on the chain, ie, the dropouts are already visible on the screen of the video camera -8. At the beginning of a band they emerge strengthened, but synonymous in the middle, so that watching is no fun. As I said, brushing Video Head is pointless, the errors are on the tape. And they almost always concern only one single frame and single lines.
My brother pointed me to a program called "avidemux". There is a filter called "drop", which goes in the direction of that what I need: it reduces the problem somewhat, but not eliminated it.
Grateful for any help Holger
Antwort von AndyZZ:
As you have already ruled that the drop-outs before already on the O-material. At those places is your material, and the magnetization of the tapes, already partially destroyed. Since then helps filter no longer synonymous. I had this problem for some synonymous Hi8 cassettes. If it does not take too much upper hand, then you can track drop-outs to eliminate tedious manual work using Photoshop:
Herausrenden The affected image area with some good frames on the right and left of the broken frames from Premiere, and indeed as a filmstrip. Import of the filmstrip in Photoshop and drop-outs there with a brush or stamp disguise. Ggf. before you have to deinterlace, ie, the join fields in Premiere when exporting. Picture synonymous with small changes you can picture the destroyed part of a good frame and superimpose herauskopieren s.The destroyed body. Most of the time works very well: good body frame with select, copy, paste () as a new layer, transfomieren (and to a certain number of pixels were I think 580 pixels) to move up or down accordingly. then the copied piece is s.genau the right place. Just a little experiment, it funzt quite well.
Antwort von coconut:
Well, the transformation is almost optimal.
The quality can be slowly indeed, but it takes quite a while. I guess rather that in comparison to today's cams of the quali. differed so much is obvious.
But if the drops already visible on the Cam are similar, you can improve s.der "chain" anything, then he would in fact only help a prog that the missing frames "artificial" dazurechnet. This should, however, be quite complicated, because the know prog yes "must be used where there is something missing ... Is he sure the frames are missing, or jerky, the tape and thereby comes into this effect. Also possible by changing lane position of the band. As far as I know, there would only help a TBC, but whether it's worth ....
I would, if it is possible for you, first let's see if you can improve the source, so try another camcorder or recorder.
However, if the tapes but have suffered so that it is definitely in it, then I see no (cost) the possibility to reconstruct perfectly.
Perhaps one of our species still has a hot stock tip ...
Antwort von AndyZZ:
But if the drops already visible on the Cam are similar, you can improve s.der "chain" anything, then he would in fact only help a prog that the missing frames "artificial" dazurechnet. This should, however, be quite complicated, because the know prog yes "must be used where there is something missing ... Is he sure the frames are missing, or jerky, the tape and thereby comes into this effect. Also possible by changing lane position of the band.
If I have correctly grasped it, then he does not complain about dropped frames on small special bolts s.verschiedenen points of the analog tape. Just drop-outs. For digital tape that would be a formation of blocks or the like. When you drop out, mostly tiny white flash, just missing a part of the image information s.dieser certain point in the Picture.
Antwort von Holger Schlingloff:
"If I have correctly grasped it, then he does not complain about missed > Frames special about small parts of the lightning s.verschiedenen > Analog tape. Exactly! Thank you, one understands me! See example picture on the O-posting.
The solution with Photoshop (missing points from adjacent frames manual override) seems to be feasible, but I will while old and gray before I just synonymous with a single tape'm done. (even 1 'processing time per frame add up to 90' strip material more than 2000 h working hours ...)
> This should, however, be quite complicated, because the know prog yes "must > Where something is missing ...
Basically it seems to me the problem computationally not so difficult: If for a frame n is the pixel (x, y) is white, and both in frame (n-1) as synonymous (+1) s do not know but, for example green, replace the white By a green pixel. It could be synonymous even before last and the consider the second frame, etc.
I assume that the proposed hardware solution proceeds as follows. At least I think I read somewhere that the DV error correction work that way. But: I can not even program ...