Sharpening during capture in procamp

[Rahneshin]

Hello all,

Just wondering what the general consensus is regarding sharpness settings in procamp during capture: on (50/100) or off? Or somewhere in between?

Haven't tried comparing results with both ways yet, partially because there are so many ways of sharpening in conversion/editing that it's hard to tell what's best.

Actually hoping this is one of those "iron rule" things with a clear yes or no answer.

My set up: USB-Live2 -> Amarec (Lagarith) -> Hybrid or Avidemux

Cheers.

Edit: haven't seen anything discussing this directly in the forums.

[lollo]

For USB-Live 2 I use middle value (50), and noticed no added halo or strange artifacts. However you should check with your videos (mine are 90% recorded from TV channels)

[Rahneshin]

Thanks, lollo.

Mine are from Video8 captures. I noticed that sharpening - unsharp mask - in avisynth (via Hybrid) works much better for me when procamp sharpness is set to 10 or even 0. When it is set to 50 and sharpening is done in Hybrid after, the result is an over-sharpened image, even if the values are set quite low.

Cheers.

[lollo]

OK, but I do not understand if your video is over-sharpened before the filtering. You should distinguish between the impact of the card and the impact of the post-processing.

If the captured video is ok, you can use a different AviSynth sharpener, such as LSFmod (with default overshoot prameters) or CAS.

[Rahneshin]

The impact of the card on its own is minimal/OK in sharpening at 50. Only when I do additional unsharp mask in avisynth do I get an over-sharpened look.

So, I thought better to leave procamp sharpening at 0 (or maybe 10), to keep the captures as 'untouched' as possible, and then do all sharpening in post-processing.
I have tried this and the result is quite good. But I have not compared it directly side by side with procamp sharp at 50 (and no post-process sharpening).

Cheers.

[lollo]

It appears to me then that the oversharping is in the post-processing.

keep the captures as 'untouched' as possible

Not sure that 0/10 values is "untouched"; it may be that the default 50 value is "untouched" and lower values blur a little bit. Just speculating, because a reference is needed to judge the impact of the setting. However, I did some comparison between captured videos and dvb-s stream of the same program (in this case I do have a "reference"), and at default 50 sharpness I did not notice any added sharpening in the captured video. But YMMV.

As usual, the best is to do all the necessary experiments and find the appropriate choice for your specific case

[jagabo]

The sharpness setting means different things with different captures devices. With some neutral is at the far left and moving it to higher values sharpens the video. With some, neutral is somewhere in the middle, moving to the left blurs the video, to the right sharpens. I recommend leaving it at neutral. Increasing the proc amp sharpness with low res sources like VHS only sharpens and increases the noise.

[Rahneshin]

Thanks lollo for taking the time to do a test I would be unable to, given my set up.

It looks like 50 value might indeed be the "untouched" one, then. (why would anyone want a purposely blurred image?)

I will do further experiments and see what happens using LSFmod and CAS with captures at procamp sharpness 50. I had not good result with procamp->50, LS->Range Sharpening. And good results with procamp->0, LS->UnsharpMask->110.

As usual, the best is to do all the necessary experiments and find the appropriate choice for your specific case

Indeed. Just had hoped this would be one of those questions for which there is a clear firm answer. But, as I'm discovering, nothing in the field of video tech is clear and firm.

Cheers.

[Rahneshin]

I recommend leaving it at neutral. Increasing the proc amp sharpness with low res sources like VHS only sharpens and increases the noise.

Thanks jagabo,

Any idea what the "neutral" sharpness setting might be with the USB-Live2?

Cheers.

[Brad]

0 is probably neutral for Conexant Polaris devices in general, based on the Diamond VC500, but who knows. The drivers for Hauppauge USB-Live2 are completely different, so it's possible that this differs.

You can check for yourself using test patterns, if you have the ability to burn and play back a DVD: https://forum.videohelp.com/threads/360704-2013-my-video-capture-device-comparison-screenshots