Virtualdub to TMPGenc in realtime?

[Baron_Samedi]

Hi,

I was wondering if there is a way to capture say from an ADVC-100 using virtualdub, process the capture (noise reduction) and frame serve the output to TMPGenc in one step.

Is this possible, or am I stuck with capturing an AVI and then frameserving the "fixed" file to TMPGenc?

Thanks!

Gary

[feenix]

Yes you can do it by framserving direct with Virtualdub or avisynth. Tmpgenc has it's own noise reduction filter so you could just use that and not bother framseserving at all.

Just try the options and see what's best for you

[Baron_Samedi]

Ahhh, sorry to appear dense but can virtualdub be set to capture, filter and frameserve simultaneously?

If so, how?

[feenix]

If only!!!

Not a chance.

Virtual dub doesn't give you the option to capture and frameserve at the same time and anyway even if it did Tmpgenc is just too slow to handle it.

[pijetro]

Nope, unless you have a super duper, abacadabra system that NASA is secretly working on..

The closest thing that you could achieve is capturing straight to MPEG2, rather than .AVI...

[IJM]

And furthermore, the last time I checked, Virtualdub cannot do Firewire captures.

Ian.

[Baron_Samedi]

Well, firstly I don't think it would need something super duper system to manage it as long as the transition from capture to processing in virtualdub could be handled by frameserving in the same way as TMPGenc is fed....

I agree about capturing straight to MPEG-2 but... that would leave the problem of cropping the head noise off the bottom of my VHS captures

And regarding DV capture into virtualdub, its possible, because I have done it using the Panasonic DV codec as per a guide I found here or a linked page

[feenix]

As an aside - why are you looking to capture from an ADVC-100 when you already have an ATI Radeon card?

You can capture excellent mpeg2 with this AND use cropping to get rid of your vhs 'noise' from the botom of your captures.

Ah - just had a thought. Are you using the ADVC-100 to bypass macrovision?

[pijetro]

Well, firstly I don't think it would need something super duper system to manage it as long as the transition from capture to processing in virtualdub could be handled by frameserving in the same way as TMPGenc is fed....

Sure you do..Look at the CPU usage during capture. Look at your CPU usage during encoding. Add the two, together, and you'll need at least two processors, and a motherboard that's resources don't get bogged down.

Even if that were true, did you ever hear of anybody doing 1:1 encoding in Tmpgenc??

If all you're concerned about is removing the garble from the bottom of your tape, then see if you can capture with the cropping filter included??