Best way to improve raw DV captures?

[danielscreed]

I recently bought a working Canopus ADVC-110 for a very good price ($55), and I am loving the results that this device gives. However, I don't know if this is really the best quality possible. All of my family's home videos were recorded on a Handycam (CCD-TRV308 model).

I captured the tapes using Premiere Pro and then deinterlaced the AVI files in QTGMC with a few other filters, such as Deblock QED and MCTemporalDenoise. I really don't know if this improved the quality since I am still new to capturing. Anyways, I am really just trying to find out if there is a better method for restoring DV captures. I have attached some stills from one of my captures down below.

[lollo2]

Difficult to judge on 2 images, but I guess clamped blacks (trees in the bottom right of the first pictures) and whites ("ball building in the first picture, t-shirt in the second picture).
If you post your original captured video we can help you better.

[danielscreed]

I've uploaded some snippets from one of the tapes last night on my second YouTube channel. Here is the link.

[lollo2]

No please, stop with youtube videos, it was a mess on the other post. Upload original DV capture of the second image

[lollo2]

On top of that, if we want to see if there is any potential improvement we have to start from the original file, not from a file already restored ;-)

[danielscreed]

Quote:

Originally Posted by lollo2

On top of that, if we want to see if there is any potential improvement we have to start from the original file, not from a file already restored ;-)

Here are some sample files. Don't ask why they are different formats. These have been deinterlaced with QTGMC, no other filters used. All I want to know is if it looks good or not. I really think it's the camera that was used, but that's just my guess.

Hardware Specifications

Operating system: Microsoft Windows 10 Pro
Processor: Intel(R) Core(TM) i5-7600K CPU @ 3.80GHz
Video card: AMD Radeon(TM) RX 460 Graphics
Motherboard: MSI B250 GAMING M3 (MS-7A62)
RAM: CORSAIR Vengeance LPX DDR4 16.0 GB

[lollo2]

Quote:

All I want to know is if it looks good or not

To my eyes they are not too bad, considering the problem with under/over exposure. I am not familiar with NTSC videos, so maybe lordsmurf, latreche34 and msgohan will judge better than me the potential DV 4:1:0 problems...

Quote:

These have been deinterlaced with QTGMC

QTGMC is also a denoiser and a sharpener. Not knowing the used parameters makes a little sense to further blindly process. You really shouls start from raw DV capture.
The histograms show problems, ADVC should capture full 0-255 range so the clamping is in the source or in the way the YUV range has been applied to filters and codec (QTGMC, ...).
Using compressed h264 x265 files also means we'll decompress and compress several times, adding each time a loss of quality.

However, just a quick and dirt eperiment trying to fix dark areas and bright areas. I did not fully check the final result, cause I am in a rush, but attached is the result. It is not a serious restoration, just a share of impressions.

cfr1.jpg

You must be logged in to view this content; either login or register for the forum. The attached screen shots, before/after images, photos and graphics are created/posted for the benefit of site members. And you are invited to join our digital media community.

[danielscreed]

Quote:

Originally Posted by lollo2

To my eyes they are not too bad, considering the problem with under/over exposure. I am not familiar with NTSC videos, so maybe lordsmurf, latreche34 and msgohan will judge better than me the potential DV 4:1:0 problems...



QTGMC is also a denoiser and a sharpener. Not knowing the used parameters makes a little sense to further blindly process. You really shouls start from raw DV capture.
The histograms show problems, ADVC should capture full 0-255 range so the clamping is in the source or in the way the YUV range has been applied to filters and codec (QTGMC, ...).
Using compressed h264 x265 files also means we'll decompress and compress several times, adding each time a loss of quality.

However, just a quick and dirt eperiment trying to fix dark areas and bright areas. I did not fully check the final result, cause I am in a rush, but attached is the result. It is not a serious restoration, just a share of impressions.

Attachment 13542

Attachment 13543

I don't really see much of a difference in the restored version other than the brightness being increased a bit and the overscan/tracking lines being cropped out, which does help improve it. I also thought it looked pretty good after the conversion process, but I wanted the videos to be looked at by someone with more experience in video capturing and restoration, and I guess you do seem like one. Somehow I still feel that it can be improved just a bit more. Overall, though, this is absolutely fine in my opinion. I'm still open for any feedback from others.

[danielscreed]

By the way, what method did you use to restore the clips?

[lollo2]

Quote:

I don't really see much of a difference in the restored version

Let me repeat myself:

It is not a serious restoration, just a share of impressions.

Whites and blacks are not good.

QTGMC is also a denoiser and a sharpener.

You really should start from raw DV capture.

Quote:

and I guess you do seem like one.

Not really, I just work on my own captured tapes, so my experience is narrow and limited. But I learned a lot from very experienced people here and in videohelp and I am happy to share part of what I learned if it can help, now that most of them are not around anymore