I have been for some time in a quest to improve and squeeze the maximum possible quality from my old video8 PAL tapes.
After an evolution from capturing from a a Digital8 cancorder ( Sony DCR-TRV238E to DV/AVI , to lossless Huffyuv
/Avi using a legacy Philips Saa7134 and also an external device Startech USB3HDCAP ( uncompressed video through usb3 ) I am now capturing using an Hauppage usb live 2 usb and an ATI HD 600 usb using virtualdub
to get Huffyuv
/AVI on a improved XP PC (Pentium 4 , 3,4 GHz, 4GB ram, 1TB+2TB hybrid SSHD 7200 disks). I have refined the workflow as much as I could but there are still some details that can eventual provide small additionally improvements and for which I would like to have your opinions:
1- Will it be better to connect the audio from the camcorder to the capture card together with the Svideo signal or can I eventually get best results bypassing the capture card and connecting the audio direct to the PC sound card letting only the video signal through the ATi hd 600 usb. In principle the audio will be in PCM format directly or indirectly .Can I expect any type of problems namely synchronizing problems between audio and video for instance or can I expect the direct connection to provide better sound? For now it seems the direct connection can provide better sound however the audio signal coming from the camcorder is quite high and I need to decrease the record level a lot in the audio sound card mixer and seems easy to get audio clipping. I havenīt detected audio video synchronizing problems for now.
2- Capturing with the ATi Hd 600 usb generates about 40-50 inserted frames per each 90 min. tape, so an external TBC as an AVT-8710 seems the logical option, I insert one in the workflow and it solved the problem. However It seems to me that the images captured with the AVT-8710 inserted are not as good as the ones captured directly. Possible alternatives could be equipment that as some sort of internal tbc used as pass through instead of the AVT device.
I have 3 devices that I could eventually use for this purpose an LG DVD recorder LG RH199HS with some sort of TBC that I canīt identify , a JVC HR-DVS3U ( digital TBC/NR, Digital R3,Video stabilizer and a JVC HR-S9500 with 4 MB TBC/DNR.
I tried all the 3 decks in pass-through for this purpose, however only the LG DVD recorder completely cancels the inserted frames during capture the other two only reduce the number of inserted frames to a value of about 15-20 for a 90 min. tape. However the image I got using these decks seems in any case better than the one I get using the external TBC ( AVT-8710) inserted in the workflow.
To try to get to a conclusion I used the svideo out signal from a DVD player running a DVD with test patterns to get a reference and stable signal to test the alternative workflows. To try to assess the image in more detail I exported identical frames to BMP in virtualdub
using the different alternatives. I īm including this samples in order to get your opinion .
Probably I have only two solutions, to use the AVT-8710 to avoid the inserted frames as it doesnīt seems possible to capture directly without inserted frames, and accept a very small quality degradation if confirmed or use instead the LG RH199HS DVD recorder in pass through that seems to provide better final quality and also avoids the inserted frames. With the other decks even if they provide eventually better quality I still have 15-20 frames inserted and I can only use them if the image is better and this number of inserted frames doesnīt pose a problem situation about which I am not sure.