Did you use Composite between the ES10 and blackmagic? Or did you just mean composite from the VCR.
The Nokia VCRs have an interesting "ASO Plus" function that does some processing on the raw rf signal before demodulation, is it on or off? I have a 2-Head Nokia VCR but missing remote and display is broken so I don't know if it's on or off by default, but I found the image to get a bit sharper when pressing the ASO High button.
It's better to use S-Video or Component(3x RCA + audio) between the ES10 and the capture card, otherwise you may get some extra noise due to Y/C crosstalk. Component may be slightly better than S-Video if available.
this test did start when i did see diagonal noise when i did connect my old Nokia 3786 to my Studio 1 Productions PA-200 with composite to my Ati all in wonder 2006 PCI-e
i did try like this
Nokia 3786 composite > PA-200 composite > Ati all in wonder 2006 PCI-e =Diagonal Noise
Nokia 3786 composite > PA-200 Svideo> Ati all in wonder 2006 PCI-e =Even worse sharper Diagonal Noise
i did the same thing with blackmagic_intensity_shuttle and Ati Tv wonder 600 usb no diagonal noise perfect picture
hmm blackmagic_intensity_shuttle did give me the best quality or i did like it the most on this old Nokia vhs test i did
Pretty ugly stuff. I don't know how you guys watch it, much less praise it. But, then we live now in the advent of the Video Dark Ages when visual literacy is down the toilet. I use Neat Video when needed. It can work miracles with noise. But it can't do anything with ugly.
It's not bad. Not even "not good". But it (probably) would not satisfy me. It did knock out the patterning, but it left problems, and actually made one worse.
The main issues is that chroma noise has now been baked in, and should be removed pre-(grain)NR. Some offset as well. Now it's a mess, impossible to fix.
The next problem is that it lacks a good median filter, so you still have errant noise that distracts.
The temporal NR is really pushing the envelope at times. Sometimes it's unavoidable with lots of NR, but Avisynth tends to do better here. I've seen worse examples, but this is part of why NeatVideo is a "hatchet job" by comparison.
Overall, yes, it's much better, but could be better yet with some practice in scripting.
I'd give the effort a B+.
Do the pre-grain chroma NR, and it'd be a solid A- attempt.
I'm not anti-NeatVideo at all, it's in my video toolbox/arsenal. In this exact case, a blend of Avisynth, VirtualDub, and NeatVideo would probably work well.
By comparison, something like this is almost good quality compared to the nightmare tapes I get.
lordsmurf Thanks for the info i have to remember that fix next time i try to fix bad video
i know this was lazy attempt to fix but it was more i wanted to show how Neat video 5 did with lazy and fast setting and for me it did look ok not so plastic clay look i did not even try to fix other things
and of course i have better super vhs players if i want to get little better quality from this vhs tape
The use of NeatVideo aside, which I use myself when it's the right tool for the job, there's as lot in the unfiltered sample that doesn't make sense. It was filmed in the USA at 23.976fps but the sample is PAL speeded up to 25fps. The original film was made at 1.85:1 aspect ratio but the PAL tape is obviously pan and scan 4:3 with the sides and part of the top and bottom cropped off. As analog tape it would play at 25fps interlaced but the AVI is progressive. So is your tape source really a progressive PAL video or was it deinterlaced during during or after capture? In any case, no matter whose denoiser is used the results look eviscerated and plastic. Darks are crushed and unnaturally and grim (for some reason darks always look that way with BM gear) and something has made a complete mess of the color. Did it really look that bad when it was broadcast? The noise itself isn't just tape grain, it appears to be a mix of tape noise and RF interference. In any case, any resemblance in detail or texture to a film original disappears after denoisers are used, whether it's NeatVideo or something else.
Nevertheless the sample and the capture gear info is appreciated and very informative. They demonstrate, again, that inferior players and BM gear should be avoided for analog source capture. I've been through quite a few bad tapes myself and have a few more awaiting capture. IMO a tape in poor condition doesn't justify the use of poor players or capture cards not designed for analog input.
Last edited by sanlyn; 12-06-2019 at 10:54 AM.
The following users thank sanlyn for this useful post:
jjdd (12-06-2019),
lordsmurf (12-07-2019)
i did not do any deinterlace but i did fix the error when itīs Pal film sometimes i donīt need to do it but this is the code that works when that error happens
Yes, colors are good now. My earlier comments only addressed the grain/noise. But color did need work.
themaster1 sample was good, and with more script work/testing it could probably be refined closer to NeatVideo ... but without the NV artifacts. Or again a mix of methods, which may include NV. NV has narrow uses, and this may be one of them, it's mostly good at patterning noise.
To me, it's always fun/magical to see before/after restored video.