Possible oversharpening during capture?
 

Had some interesting results with two different SDI capture cards that use Analog Devices ADV7xxx series chips to convert a component signal off of a Rackmount TBC (FA-310) to SDI and then recorded as Apple ProRes422HQ. This really doesn’t follow any of the recommendations in terms of hardware other than the VCR having a line TBC (which is on), but it’ll be part of my grand comparison of what will hopefully become something like 20 capture devices/TBCs tested.

The capture is from a commercially produced Johny Carson regular (non-S) VHS tape. I believe there’s some clipping baked into the recording and the source video they were using wasn’t perfect as their official release of “The Great Flydini” which preceded this screenshot on YouTube seems to also have a rather poor capture and white clipping, guessing transcoded from a later DVD release: https://www.youtube.com/watch?v=s9xKU8eYCFk&t=113s ….Or it could be the 13-year-old YouTube encoding, who knows lol. Of note, I saw there were some identical partial line dropouts on the official YouTube posting as was on this VHS tape, so those must also be baked in the master recording that the VHS was made from and they didn’t bother to correct frame by frame.

The screenshot attached with the side-by-side is a screenshot of both captures open at once side by side and are not deinterlaced, it just happens to be a rather still moment in the video. Deinterlacing would probably sharpen both of them more. I’ve also attached short clips of each of the ProRes files for any further analysis or seeing if any miracles can be worked through avisynth on the less sharp one.

My question is whether that less sharp capture can possibly be brought up to the same sharpness using Avisynth scripts? Seems like there’s some sort of edge enhancement going on, but it’s hard to say which is closer to the true representation of what is actually recorded on the tape which arguably is what you’d want if you were doing archival work, even though I personally think the sharper one looks better (with caveats being an increase in some “shadows” at areas of high contrast edges). Best of both worlds would be capturing as it is and being able to sharpen later if desired I would think.