in my software i see that PAL TV restitution is 768x576 and not as you say - 720x576
so i checked on google what is 720x576 and sow that 720x576 is a 16:9 widescreen restitution..
Or am I wrong?
So what do you say? Record on the resolution 720x576 or 768x576 ?
and as you can see, i can't change the frame rate in the software...
+ Reply to Thread
Results 31 to 60 of 114
PAL D1 video is 720x576. PAL DVD uses 720x576 for both 4:3 and 16:9 display aspect ratio. 768x576 is the equivalent display resolution for 4:3 PAL video (when retaining the 576 line height). If your software doesn't allow PAL capture at 25 fps it is severely defective. Analog PAL has only one field rate: 50 fields per second. When digitized those fields are packaged in pairs to make 25 frames per second digital video.
so what resolution is better?
i do not going to play this on DVD..
it is for players like Xtreamer..
and i chaked again and sow that if i change in the software to record to mp4 file then i can change the frame rate:
but what settigs is the best for mp4 file?
this is my settings:
i want to record with compressor because i no space in my hardisk for big files...
i can set H264 or MPEG4 in the compressor...
if i record to avi file with compressor than i have a very high CPU use..
the H264 compressor is verry havy for the CPU.
it takes 99% of all my 4 cores..
this is the settings that working best:
what do you thing?
i need to change something in the settings?
Last edited by gil900; 1st Feb 2013 at 09:58.
If you're serious about cleaning the video you should capture YUY2 video with lossless compression (HuffYUV, Lagarith, UT, etc.). That way you will avoid all the problems of high compression codecs -- loss of detail (not that VHS has much detail), blocky artifacts, temporal noise artifacts, etc.
If you can't afford that much disk space (~40 GB/hr) and have to use high compression codecs, use the highest bitrates you can. But it's absolutely critical that you capture PAL video with 525 scan lines and 25 fps.
Capture at 720x576 and 25 fps interlaced. That's the international standard (ITU/BT 601) for analog to digital PAL capture. If you can't do that capture at 768x576 25 fps interlaced instead. But you should look for software that conforms to the ITU specs.
Last edited by gil900; 1st Feb 2013 at 11:40.
That's not a problem. It's what's on the tape. You can crop or mask it later while filtering.
Do not deinterlace while capturing. The card will probably do it poorly. And you will lose half the temporal resolution, maybe even half the spacial resolution.
Your goal while capturing is to get as close as possible to what's on the tape.
Thank you very much!
I start recording to avi with MPEG4 compressor on 720x576,25FPS, 20,000 bitrates
At first I thought recording to AVI is very much slows down the computer and high use CPU.
but i learned that it is the compressor and not the file type..
Yes, and 23.976 could be thought of as a slightly-adjusted version of 24fps (the FILM standard).
Unlike what Sanlyn said, 24p (or 23.976p) telecine to 25i FPS PAL is usually done with a framerate speedup, so no pulldown/pull-up, no interpolation/blending. Constant cadence, just an adjustment of the timebase (which will make things SHORTER/QUICKER). Often makes them higher pitched (though this can be adjusted with good audio software). The great majority of 24->25 conversions use this method.
Now as far as your audio is concerned. I don't know what you did with it, nor why, nor which app you used, but let me first say that "Audio has NO framerate". To be perfectly clear: Not counting data packetizing, THERE IS NO DIGITAL AUDIO FORMAT EVER THAT HAS FRAMERATE. Samplerate, yes, Framerate, No.
So "needing to match framerates" is a red herring, probably due to bad nomenclature on the part of the app(s) you were using. Actual good software will allow you to use standard samplerates, with standard codecs at standard bitrates, and enable you to get perfectly compliant (for DVD or BD purposes) audio that should easily mathc the length of your existing/original/intended video.
Personally, regardless of what time & effort you may have already put into this, I think it would be worth your while to go back to "square-one" and re-digitize (capture) your originals all over again. This time, using a capture software that doesn't mangle the recording at the first stage. Your tape is either PAL-based (more likely) or NTSC-based (less likely). If PAL, the settings should be: 720x576 @ 25 FPS interlaced. If NTSC, the settings should be: 720x480 @ 29.97 FPS interlaced (=30000/1001). Once you get them into digital format, then you can address the possible problems of whether the ORIGINAL (prior to it getting to your VHS tape) was Film-based, PAL video-based or NTSC video-based.
edit: WHOOPS! this post was meant for earlier in the thread (my browser didn't take me to the end!) If you've already worked it out, never mind. I'll catch up the thread later and see if my suggestions need updating...
Last edited by sanlyn; 25th Mar 2014 at 12:04.
I just wanted to point out that the last screenshots from the OP show the default or native capture settings of the Webcam/Capture Device are at 640x480@30fps YUV and all of the other screenshots are the output settings from the software. Is this possibly the source of interpolation and resizing problems?PB
Ok I finished recording to an AVI file.
But there is a small problem with DGIndex.
DGIndex not support avi files..
so i load the avi file(45 min, 2.65GB, 25FPS,720x576) on VirtualDub an then i got a wonning massage about the audio:
but i dont have that problam with the audio that described in the massge.. so i think that it is ok..
Anyway, I encoded small part from the video without compression (PCM UNCOMPRESSED)
to make the bast sample i can..
I have attached the sample video.
I think the quality is much better now
I just wanted to point out that the last screenshots from the OP show the default or native capture settings of the Webcam/Capture Device are at 640x480@30fps YUV and all of the other screenshots are the output settings from the software. Is this possibly the source of interpolation and resizing problems?
but i Identified that problem and change the settigs to this:
and in the video options i changed the FPS to 25
the first part was very important
Last edited by gil900; 1st Feb 2013 at 13:49.
The new sample is getting closer. But there are still problems with the handling of the chroma channels. The colors from the two fields are blended together:
The blotches of color at the right are from the next shot (the other field in the frame). And notice how the rest of the colors (the red shirt for example) are very muted because they are mixed with the color of the next shot.
I don't know if this happened during the capture or during your conversion to RGB. You should upload a sample without any conversion. Use VirtualDub in Video -> Direct Stream Copy mode.
Cropping in AviSynth would be something like Crop(4,2,-20,-10).
Last edited by jagabo; 1st Feb 2013 at 15:39.
AviSynth site explains a couple ways to do it. However, since it's interlaced you have to be careful, and interlaced cropping and resizing does throw a monkey wrench into it. Even if you did the filtering in VDub (not a good idea, especially for simple stuff such as cropping), you'd have the same problem. In AviSynth, maybe making slight adjustments, and be careful of the field order:
Unfortunately, the problem is in the captured video. Interlaced frames have been encoded as if they are progressive. That has caused the color blending. Can you force the encoder to use interlaced mode?
These same articles are installed on your computer in the Avisynth program folder.
Your previous avi sample had been converted to RGB, and nearly doubled the size of the original cut. To cut a piece of video in VirtualDub and maintain the same color space, don't save the AVI using "full processing mode". Use "Direct stream copy".
I still had only the original sample1.m2v you posted earlier, so I spent some time with it. It's an example of how much grief is caused by capturing damaged and noisy analog source to lossy codecs such as MPEG. Do yourself a favor and keep learning to capture properly. These sample images are from the 640x480 video, which I didn't try to resize. The number of filters used to clean the capture .m2v is pathetic. Most of the work was spent trying to repair compression artifacts, some of which couldn't be fixed.
The image below is frame 529 from the original clip. It's too bright, and highlights and red are too "hot". There is coarse grain and banding in shadows and contours from the MPEG encode, seen in the face and arms. The halos are obvious, but oversharp artifacts are also evident on the right and left of edges.
[Attachment 16076 - Click to enlarge]
The image below is frame 443 from the mkv posted below, after cleanup (frame 529 in the original m2v). Encodes can be cleaned somewhat, but you can see how much very fine detail is lost after running so many filters.
[Attachment 16077 - Click to enlarge]
The mks encode required more bitrate than would have been necessary for a cleaner capture. In the mkv look at the nosie and mottling that remains in the dark an d black areas. This isn';t going away. It's the result of what the original encoding did with tape noise, which could have been cleaned up considerably with a proper AVI capture.
A file of ideas for the filters I used in cleanup is also attached.
Last edited by sanlyn; 25th Mar 2014 at 12:04.
this is my new script on the new recording:
loadplugin("C:\Program Files (x86)\AviSynth 2.5\plugins\FFT3DFilter.dll") loadplugin("C:\Program Files (x86)\AviSynth 2.5\plugins\aWarpSharp.dll") Import("C:\Program Files (x86)\AviSynth 2.5\plugins\QTGMC-3.32.avs") AviSource("D:\*******\1.avi") Crop(4,4,-44,-8) #MergeChroma(McTemporalDenoise(settings="high",interlaced=true)) FFT3DFilter(bt=-1, plane=0, dehalo=1, hr=2, ht=50, svr=0) QTGMC( Preset="Slower", Sharpness=1 ) SelectEven() mergechroma(aWarpSharp(depth=40))
1. In VirtualDub, with a video loaded, go to :Video..." -> then "copy output frame to clipboard".
2. Open a photo app and copy that captured photo into a new photo.
We can seldom judge much from screen captures. Your script does not correct invalid luma or chroma levels. After you go from YUV to RGB, it's too late for that.
Last edited by sanlyn; 25th Mar 2014 at 12:05.
Things start a bit confused ..
Ok so if too late then i can start again from 0.
Someone said that the size of the file will be similar with Lagarith or HuffYUV.
I did not know that.
So I'm willing to spend another 45 minutes in the recording for the best output.
sanlyn, thanks for the help but I think you're working too hard.
I did what jagabo said - re-record on correct settings.
And I did it. And you are trying to improve an old and not good recording
I have over 30 tapes to convert.
I prefer to dwell on the first tape to do everything else right.
this is the test file:
there is no motion on the first frames(frame 0 to frame 24) because the i just start to record..
this is not a problem
And I have a question.
This is not really related, but that is the question -
Is it better to remove the tape from the VCR when not recording?
Because I thought about it. The VCR just warms the tape without doing anything.
Last edited by sanlyn; 25th Mar 2014 at 12:05.
If you really can't afford the disk space required by HuffYUV you might try using x264vfw. Set it to interlaced encoding, max GOP size of 1, with a CRF value around 12.
Last edited by jagabo; 1st Feb 2013 at 20:12.