56 hours of Hi8 to digital project- two burning questions

[m_00]

Hi everyone, I hope that what I am posting is not redundant at all. I've spent the best part of a year reading these forums and have learnt a huge amount already, but just have a few extra questions if anyone can help.



I am trying to convert old family videos from Hi8 to digital. There are 56 in total, ranging from 1985 to 2004. I am doing this in the UK, with the hardware below:
Camera: Sony Handycam CCD TRV 91E PAL, which was the camera used to record much of the footage.



Capture card: I eventually settled on the Blackmagic Intensity Pro, recording via S-video. Whilst I understand that many people on this forum feel that recording to DV from Hi8 is all the quality you need, I couldn’t quite bring myself to settle for that, not least considering that I would be spending what seems to be a similar amount of money on potentially inferior quality.
The rest of my setup is as follows:
Win7 x64, Intel i7, Gigabyte EX58 UD3R motherboard, 6GB RAM, Seagate barracuda 7200RPM HDD


My current plan is to capture in 8bit 4:2:2 YUY2 (I think, if not then UYVY) via Vdub, to huffyuv AVI purely to give me the option of editing/ remastering in the future. I’ll save this and then create MPEG-2 or similar files for day to day use.
The intensity pro did come with the software ‘Blackmagic media express’, which supports MJPEG capture or 10bit uncompressed AVI. I think choosing uncompressed AVI option may be too much data for my system to handle, as recording in this results in the video lagging, then speeding up, then lagging again.


My first question refers to the quality of the images. From the photo below, there are a few things I am having trouble with.
Firstly, I am assuming that the artefact at the bottom of the picture is due to overscan so I have so far ignored that. I presume that the green/ pink lines on the right and top borders of the picture are due to dirty video heads (if that is the right word) on the camera. I have used a cleaning tape with no benefit. Buying a new camera, let alone a Hi8 deck would be quite expensive, and I could potentially live with these lines. Moreover, I have read that the best quality to be gained (within reason) was from playing the tapes on the camera they were recorded on, and that DV8 cameras that can play Hi8 cameras aren’t as good as the original camera which was designed to record and play only in Hi8).


Secondly, the camera has a timebase corrector enabled. I understand that this is likely to be a ‘line’ TBC and not a full frame TBC and therefore not provide much benefit. However from all the examples I can find online, the amount of tearing in the picture from the VHS tapes used is quite significant and not quite what I am experiencing here. For all intents and purposes, attempting to ignore the fact that I am watching the video on a 1900x1200 monitor (deinterlacing using VLC makes no difference) makes me believe that I have a reasonably good quality copy of the Hi8 tape. So I suppose the question has to be whether getting a TBC (my eyes on the AVT-8710) would correct the ‘jagginess’ of the video, or whether the picture I have now is essentially as good as it is going to get.

I appreciate that some of this may be repetition but all the answers I find relate either to VHS specifically, or hardware that is more than five years old. I suppose therefore my main questions are whether it is worth buying a) a new camera and b) a timebase corrector.


Any other suggestions are of course welcome.


Apologies for length (and if any of this has been answered in other posts but I have missed) but I figure its easier to give any potential responders as much info as possible.


Thanks in advance for any help,

Malek

[Cornucopia]

Well, I'm one of those that feel that DV is more than capable of fully capping Hi8's quality (even using the best of cameras, which yours isn't), so I think all the other stuff is overkill. The only thing missing in the DV spec compared to Hi8 is the reduced colorspace, though if you're going to be ending up with MPEG2, and you aren't doing compositing, you won't notice or miss it anyway.

AFA TBC is concerned, you need to zoom in on your image and verify whether that jagginess is just standard interlacing or not. If it's just alternating lines, it's likely normal interlacing with maybe the field order being wrong; if the jaggies have more than 1 line thickness, you've got other problems and probably will need a TBC (good to have anyway).

Scott

[Brad]

Are the shots from the MPEGs instead of the AVI captures? I'm seeing significant blocking.

[lordsmurf]

I see significant "DV checkerboard" pattern (moire).

VLC deinterlacers are not perfect. You'd have to watch on TV for true test.

DV introduces extra layer of colorspace compression.
Generally speaking, multiple colorspace changes are not a good thing.

Even former DV lovers have recanted in the age of ProRes, DNxHD and Cineform.
I was never a DV fan, outside of the cameras.
For workspace and conversion codecs, DV was simply awful.

[edDV]

The intensity pro did come with the software ‘Blackmagic media express’, which supports MJPEG capture or 10bit uncompressed AVI. I think choosing uncompressed AVI option may be too much data for my system to handle, as recording in this results in the video lagging, then speeding up, then lagging again.

Issues:

DV - reasonable quality at 13GB/hr (fast editing)

10 bit uncompressed - on order of 130 GB/hr ... needs RAID

BM MJPEG - 75 GB/hr (fast editing)

Huffyuv ~ 40-60 GB/hr (laggy for editing but i7 compensates)

,,, With the BM Intensity, I'd go MJPEG. Approx 4.2 TB for 56 hours.
,,, Or a digital intermediate like Cineform.

My first question refers to the quality of the images. From the photo below, there are a few things I am having trouble with.
Firstly, I am assuming that the artefact at the bottom of the picture is due to overscan so I have so far ignored that. I presume that the green/ pink lines on the right and top borders of the picture are due to dirty video heads (if that is the right word) on the camera. I have used a cleaning tape with no benefit. Buying a new camera, let alone a Hi8 deck would be quite expensive, and I could potentially live with these lines. Moreover, I have read that the best quality to be gained (within reason) was from playing the tapes on the camera they were recorded on, and that DV8 cameras that can play Hi8 cameras aren’t as good as the original camera which was designed to record and play only in Hi8).

Lines at bottom are head switch. Edge color issues seem related to high luma. Not sure of cause. All fall into TV overscan area.

Secondly, the camera has a timebase corrector enabled. I understand that this is likely to be a ‘line’ TBC and not a full frame TBC and therefore not provide much benefit. However from all the examples I can find online, the amount of tearing in the picture from the VHS tapes used is quite significant and not quite what I am experiencing here. For all intents and purposes, attempting to ignore the fact that I am watching the video on a 1900x1200 monitor (deinterlacing using VLC makes no difference) makes me believe that I have a reasonably good quality copy of the Hi8 tape. So I suppose the question has to be whether getting a TBC (my eyes on the AVT-8710) would correct the ‘jagginess’ of the video, or whether the picture I have now is essentially as good as it is going to get.

I keep mine interlace and let the HDTV hardware handle the deinterlace.

I appreciate that some of this may be repetition but all the answers I find relate either to VHS specifically, or hardware that is more than five years old. I suppose therefore my main questions are whether it is worth buying a) a new camera and b) a timebase corrector.

I'd want to have the camera evaluated or test compare it to a known good one.

[2Bdecided]

My first question refers to the quality of the images. From the photo below, there are a few things I am having trouble with.
Firstly, I am assuming that the artefact at the bottom of the picture is due to overscan so I have so far ignored that.

It's head switching. Just blank it out. Sometimes a TBC lines it up better than you've shown, but sometimes not. Even with the best TBC it still comes out unusable.

I presume that the green/ pink lines on the right and top borders of the picture are due to dirty video heads (if that is the right word) on the camera. I have used a cleaning tape with no benefit.

It might be the way the original camera recorded the tapes. Those edges were invisible on a CRT due to overscan - consumer cameras often hid a lot of evils in the overscan (e.g. like the head switching!).

I have read that the best quality to be gained (within reason) was from playing the tapes on the camera they were recorded on

Only if it's still working properly (though a random purchase off eBay might have issues too). It's only significant if there's some misalignment between the original camera and whatever you use to play the tapes. That's quite common though (to varying degrees).

Secondly, the camera has a timebase corrector enabled. I understand that this is likely to be a ‘line’ TBC and not a full frame TBC and therefore not provide much benefit.

A decent line TBC can provide a huge amount of benefit - and often all the benefit you need (unless your video capture device chokes when there's no signal). A dumb frame TBC without a decent line TBC often leaves wiggles in the video.

So I suppose the question has to be whether getting a TBC (my eyes on the AVT-8710) would correct the ‘jagginess’ of the video, or whether the picture I have now is essentially as good as it is going to get.

That's interlacing.

Here are your two screen shots deinterlaced (crudely):
http://www.mediafire.com/?327h9kq19iw83xq
http://www.mediafire.com/?sgrgetaydcmrdyx

If VLC didn't deinterlace these at least that well when watching it live, then you didn't actually select deinterlacing. Right click, video, deinterlacing, linear.

Or set it in the options if you want it to remember this for each video.

With AVI, VLC might pick the wrong field order, which will cause horrible judderiness/jumpiness on movement and pans. 50/50 chance!

I appreciate that some of this may be repetition but all the answers I find relate either to VHS specifically, or hardware that is more than five years old. I suppose therefore my main questions are whether it is worth buying a) a new camera and b) a timebase corrector.

There's nothing wrong with what you have AFAICT. Though I'd have bought a DV capture device and saved myself the headache of dropping frames and huge lossless files. But what you have is fine if the TBC is really working. I can't see any dramatic problems on those stills. Maybe the levels need a tweak.

Cheers,
David.

[Cornucopia]

I see significant "DV checkerboard" pattern (moire).

VLC deinterlacers are not perfect. You'd have to watch on TV for true test.

DV introduces extra layer of colorspace compression.
Generally speaking, multiple colorspace changes are not a good thing.

Even former DV lovers have recanted in the age of ProRes, DNxHD and Cineform.
I was never a DV fan, outside of the cameras.
For workspace and conversion codecs, DV was simply awful.

Interesting you see DV checkerboard moire when the OP stated he didn't use DV as a codec...

In his case, PAL DV, with it's 4:2:0 color is better matched to MPEG's 4:2:0 (although there is still the minor difference between co-sited and non-co-sited). So color "recompression" would be minimal.

I'm not singing the praises of DV, particularly over ProRes,DNxHD or Cineform, but I strongly believe that for HIS/HER situation, it's more than good enough, plus makes easy capping, editing, archiving & conversion.
No, it's not a 8 or 9 out of 10 in quality, but a 6 or 7 is still fully capable of handling 4 or 5 material (just ballparking the numbers for easier understanding to the OP).

Scott

[m_00]

Hello everyone and thanks for the quick replies.


Firstly, Vaporeon800- the shots are from the AVI- I haven’t converted them to MPEG2 yet. The plan for MPEG2 was more to make them easier to move around/ have an informal copy my parents can watch whilst keeping the originals for remastering/ editing.
By blocking do you mean the pixelated effect? Is that the same as the DV checkerboard pattern Lordsmurf is referring to? Regarding this, I found a thread on this forum which found that the checkerboard pattern was due to a standalone TBC (incidentally the AVT-8710 I was considering), as opposed to the line TBC/ other parts of my current setup. Is that possible in a non DV setup?


Cornucopia- I’ve zoomed in on the photos (example posted) and am I right in presuming this is what you mean by single alternating lines?


As for the interlacing- I think what I was doing was setting it to blend instead of linear. 2bdecided your deinterlaced pictures do look quite good- I think I would definitely be happy with that quality of video. The only problem is enabling it in VLC. When playing a clip captured in uncompressed 8bit YUV, turning on linear deinterlacing results in the screen turning black (sound carries on fine). When playing a clip captured in MJPEG, I get the juddery image you described.


All other ‘doubler’ deinterlace modes (which is the type I want, right?) that I have tried in VLC such as Bob and Yadif (2x) give me the same result (i.e. black screen for the uncompressed, juddery image on the MJPEG). It seems like one is not able to change the field order in VLC either.


If my main problem here is related to the interlacing, would I be right in assuming that I can either
a) capture the video (without the TBC) and then sort out the VLC issues after, or
b) capture the video deinterlaced?
Is option b) even possible (it doesn’t seem to be using blackmagic media express) and if so, would the resulting video would look ‘wrong’ if I were to play it on a normal TV instead of a monitor/ digital TV?



Thanks again,


Malek

[m_00]

Cornucopia- I just saw your post and you've answered one of my questions. It does clarify things a bit more, so thanks. I'll read up on the other formats you and lordsmurf described (cineform etc) tonight.

I guess now that I have the blackmagic card the best bet would be to continue with the AVI plan and capture in MJPEG like edDV suggested (or potentially capture in the uncompressed 8bit YUV and then compress it to huffyuv for extra hard drive space) though if it looks like I will have to spend more DV may look like a very attractive option.

Cheers,

Malek

[Brad]

Not sure how you don't see the blocks in those blown-up photos. Look near the bottom right of the first one in post #8. Unless those were caused by your JPEG compression, your captures must not have been truly lossless.

[Cornucopia]

Yeah, there's got to be DCT blocking going on there...

OP, Your stuff is Hi8 analog. That's SD material. ALL interlaced.
No matter which route you choose, for best In-to-Out, you should be capping and saving ALWAYS in interlace, unless you were to be doing something different like editing with other (progressive, HD, web) kind of material. Leave it in Interlace throughout including final MPEG2/DVD. Then let your DVD/BD player and/or TV do the deinterlacing/upconverting.

Scott

[2Bdecided]

Not sure how you don't see the blocks in those blown-up photos. Look near the bottom right of the first one in post #8. Unless those were caused by your JPEG compression, your captures must not have been truly lossless.

Agreed - and I don't think it's the final JPEG compression - look at the file size vs the non-existent detail! So something is very wrong there, unless these captures are mode from lossy copies.

Cheers,
David.

[2Bdecided]

All other ‘doubler’ deinterlace modes (which is the type I want, right?) that I have tried in VLC such as Bob and Yadif (2x) give me the same result (i.e. black screen for the uncompressed, juddery image on the MJPEG).

It sounds like your PC can't keep up. Not surprising with uncompressed video in VLC.

There are players (like Media Player Classic) that require less horsepower, and degrade more gracefully when they eventually run out of steam (e.g. dropping frames on playback rather than giving a black screen!).

Using a different video display/overlay mode, or enabling deinterlacing by default, can help in VLC.


Then again, when I have used lossless video (for intermediate stages when processing/filtering), it's never even occurred to me to try and play it because I know my system (old P4 2.8GHz!) just can't cope. I always make a DV or MPEG-2 proxy to check what it actually looks like.


You need to keep the interlacing. Lossless capture, and then either throwing half the info away (same rate deinterlacing), or being stuck with the output of a particular deinterlace (double-rate deinterlacing) is insane.

Likely output candidates are 704x576i50 ~8Mbps MPEG-2 for DVD, 768x576p50 constant quality H.264 for PC, and/or 640x480p25 or 960x720p25 ~4Mbps H.264 for YouTube. The latter two require deinterlacing - the 50p version requires very good deinterlacing IMO. Do it properly, and if you know what you're doing, then the 50p version will be a perfectly good enough master for re-encoding to anything else you require.

Cheers,
David.

[m_00]

Ah yes I see what you mean now. I think I was looking at the pictures in my first post. I took the pictures by using the 'snapshot' button in VLC, and the zoomed in ones by literally just zooming in in windows photo viewer and then print screening. Not the most high tech of methods I admit. Repeating the process into jpeg or png didn't really improve things. Is there a better way to be doing this?

As for the interlacing, this may be a stupid question, but how would I NOT capture it interlaced? From the fact that I am capturing analogue interlaced video, is the digital capture not interlaced by default? It's just that I can't find any settings either in media express or vdub to change/ turn on or off interlacing during capture.

I'll try converting the captured video to 50p like you suggest and see if VLC can deinterlace that. Or do you mean you deinterlace the video when encoding it? Would the 50p then be good enough quality to delete the original uncompressed/ MJPEG AVIs and use the 50p for potential mastering?
I'm not entirely sure why the PC can't keep up the deinterlacing- CPU usage barely goes over 10% when trying to deinterlace the uncompressed video in VLC. Unless it is being throttled by the hard drive/ somewhere else on the motherboard which would make sense..

Thanks as always,

Malek

[edDV]

As for the interlacing, this may be a stupid question, but how would I NOT capture it interlaced? From the fact that I am capturing analogue interlaced video, is the digital capture not interlaced by default? It's just that I can't find any settings either in media express or vdub to change/ turn on or off interlacing during capture.

I'll try converting the captured video to 50p like you suggest and see if VLC can deinterlace that.

Read again Cornucopia's post #11. ... For best results do not deinterlace. Goal is an interlace DVD. Deinterlace is destructive and should only* be done separately for web posting needs. Of course your computer player will need to deinterlace for computer playback

The great advantage of DV format is the workflow simplicity for capturing and encoding to interlace DVD MPeg2.

The BMI MJPEG codec maintains interlace but the rest of the workflow risks deinterlace and/or color space conversion unless carefully managed.


* There is an super-geek alternative to convert 50i to 50p by vertical field resize in avisynth. Vertical field resize is lossy but is necessary if you need 50p for some reason. As said above, this is done better in advanced player hardware.

[2Bdecided]

Goal is an interlace DVD.

Is it? I don't know, and I'm not sure if the OP has said.

Cheers,
David.

[2Bdecided]

I'll try converting the captured video to 50p like you suggest and see if VLC can deinterlace that.

50p (50 progressive frames per second) is the result of double-rate deinterlacing. You can't deinteralce 50p - it isn't interlaced.

Would the 50p then be good enough quality to delete the original uncompressed/ MJPEG AVIs and use the 50p for potential mastering?

It could be. Someone who has chosen to capture lossless will usually choose to keep lossless. Most people would keep the interlaced original and work from that when they needed an interlaced version.

I'm not entirely sure why the PC can't keep up the deinterlacing- CPU usage barely goes over 10% when trying to deinterlace the uncompressed video in VLC.

Maybe it's 10% of a multicore CPU and the important process can only use one core?

Could be HDD access, but then, you managed to capture it without dropping frames.

Cheers,
David.

[edDV]

Goal is an interlace DVD.

Is it? I don't know, and I'm not sure if the OP has said.

Cheers,
David.

Yes the goal has been poorly defined. For me archive should be closest to source which is 702x576 25i.

Given that, the question is codec and amount of compression.

A bob 50p deinterlace is lossy. It could be done better in playback hardware which can use multi-frame interpolation. In the future, hardware will improve.

[2Bdecided]

A bob 50p deinterlace is lossy.

Plenty of bob deinterlaces preserve all the original lines, and add in new ones. Therefore it's not lossy at all. The original lines are all there and can be recovered with two lines of AVIsynth code! Though if you're not careful the same code could throw away all the original lines and keep only the interpolated ones!

It could be done better in playback hardware which can use multi-frame interpolation.

I haven't seen any playback hardware that's a patch on the best AVIsynth deinterlacing. That's painfully slow though.

In the future, hardware will improve.

Hopefully. Though with HD, it's slightly less important.


I absolutely agree that you should keep the interlaced original.

The only "problem" is when the restoration process involves bob-deinterlacing (e.g. to do optimal denoising and sharpening) - you can always re-interlace afterwards, but here the progressive version is probably a more useful thing to keep. As well as the lossless interlaced original.

Cheers,
David.

[m_00]

Thanks guys- this is clearing things up a huge amount. I've also managed to figure out what was going on in VLC, or at least how to fix it to an extent. In the advanced settings under video>filters>deinterlace there are settings for "display" and "streaming". Setting those to the same mode seems to have more or less fixed the problem. It now can play (and deinterlace) the uncompressed 8bit YUV clips as well as the MJPEGs, however I do still get juddering. The right click>video>deinterlace mode seems to have the final say in the juddering matter.


To properly define my goal, I want to capture the videos to as high a quality as possible to give me the option of mastering/ restoring them in the future. Unfortunately I don’t have the time or understanding at the moment to restore them so this is going to be a slow process. So it seems that what I want is either huffyuv via vdub (less hard drive space) or the BMI MJPEG codec. Either of which should give me the 702x576 25i, right?


The DVD is more of an aside goal. It was more to create a ‘portable’ version of the videos that can either be on DVD, or another format (h.264/ mpeg2) that can be played on a laptop/ HTPC, whilst preserving the originals for the future plan of mastering. It’s looking like most of the money for this project is going to go on storage; HDDs for the originals, their backup and the ‘portable’ version..


I’ve read that multiframe interpolation is possible using ffdshow though I haven’t quite made it round to giving it a try. As far as I can see VLC only does linear interpolation, so this would be a purely restoring issue, unless this ‘advanced player hardware’ (TV? CPU?) is developed. Isn’t multi-frame interpolation less for smoothing out the actual image and more to improve the framerate/ fluidity of the video?


Cheers,


Malek

[edDV]

A bob 50p deinterlace is lossy.

Plenty of bob deinterlaces preserve all the original lines, and add in new ones. Therefore it's not lossy at all. The original lines are all there and can be recovered with two lines of AVIsynth code! Though if you're not careful the same code could throw away all the original lines and keep only the interpolated ones!

This requires an advanced understanding of AVISynth deinterlacers. The more advanced modes will interpolate new fields and not leave a reverse path to the original fields. This seems risky for a newb as the only archive. Also requires a processing load for all 56 hours of source.

[edDV]

Thanks guys- this is clearing things up a huge amount. I've also managed to figure out what was going on in VLC, or at least how to fix it to an extent. In the advanced settings under video>filters>deinterlace there are settings for "display" and "streaming". Setting those to the same mode seems to have more or less fixed the problem. It now can play (and deinterlace) the uncompressed 8bit YUV clips as well as the MJPEGs, however I do still get juddering. The right click>video>deinterlace mode seems to have the final say in the juddering matter.

These are all playback issues. VLC defaults to blend deinterlace. With sufficient CPU speed you will get better results with Yadif or Yadif (2x). A simple Bob results in low motion flicker.

To properly define my goal, I want to capture the videos to as high a quality as possible to give me the option of mastering/ restoring them in the future. Unfortunately I don’t have the time or understanding at the moment to restore them so this is going to be a slow process. So it seems that what I want is either huffyuv via vdub (less hard drive space) or the BMI MJPEG codec. Either of which should give me the 702x576 25i, right?

They will give you 720x576 25i with 9 pixels of black blanking left and right. The Hi8 source is contained in the center 702x576 frame.

The DVD is more of an aside goal. It was more to create a ‘portable’ version of the videos that can either be on DVD, or another format (h.264/ mpeg2) that can be played on a laptop/ HTPC, whilst preserving the originals for the future plan of mastering. It’s looking like most of the money for this project is going to go on storage; HDDs for the originals, their backup and the ‘portable’ version..

Since 2005 I've archived Hi8 to DV-AVI @ ~13GB/hr using the original Sony CCD-V5000 Hi8 camcorder with internal line TBC to a Canopus ADVC-100 for DV conversion. I then make an interlace MPeg2 DVD for general use.

If I was starting today, I'd look at a digital intermediate (huffyuv, Cineform or maybe MJPEG) because CPUs are fast and hard drives are so much cheaper. I haven't closely evaluated the pro/con of a particular codec but I have full confidence in Cineform.

I've been looking at various interlace h.264 codecs (x264, Sony, Mainconcept) as an alternate to MPeg2 DVD but so far for standard def none offer a compelling reason to change. They all need near the same 8+ Mb/s bit rate for equivalent performance. DVD remains a universal distribution format.

You will need to do your own evaluation.

I’ve read that multiframe interpolation is possible using ffdshow though I haven’t quite made it round to giving it a try. As far as I can see VLC only does linear interpolation, so this would be a purely restoring issue, unless this ‘advanced player hardware’ (TV? CPU?) is developed. Isn’t multi-frame interpolation less for smoothing out the actual image and more to improve the framerate/ fluidity of the video?

Again, these are all playback issues. Advanced deinterlacers evaluate motion over several frames, then apply different deinterlace strategies for fast vs slow moving areas of the frame. The more advanced techniques must be done non-realtime or in hardware. Sophisticated hardware processors require several frames of memory. Deinterlace filters like Yadif are intended for realtime CPU computation so are more limited in performance.

As CPU/GPU performance improves, realtime deinterlacers will improve. As memory prices drop and patents become generic, hardware deinterlacers in media players and HDTV sets will improve.

[2Bdecided]

For PAL, with DVD as the end goal, I can't believe there are many cases where there's a visible difference between capturing DV and capturing lossless. Horrendous amounts of noise, which you subsequently remove in software, may create a visible difference, but that's about it. For something better than DVD, you might see a difference more often.

I wonder if Cineform works out cheaper than the larger HDD to do YUV? But then you have the worry about still being able to get the codec to work on your Windows 27 PC when you finally do something with these videos in 15 years time!

I have 300+ hours of tape. I am using DV. There's no way I can store 300 hours lossless. I've satisfied myself that even decent real-time 8Mbps MPEG-2 is good enough as a source 95% of the time! My source is all S-VHS though, so maybe slightly better to start with - and the really noisy parts are low light and will look crap whatever I do with them. FWIW I intend to edit the raw DV, and then run the edited versions through AVIsynth for restoration. Applyrange() is a great thing. Obviously I'll have to avoid crossfades between very different pieces of footage, but it's worth this hassle to avoid processing the whole lot, or figuring out which 5% I need to process for the final edit. I doubt I'll ever need more than 5% of the raw footage.

Cheers,
David.

[m_00]

Haha after all of this I think it may have been much easier and cheaper to just go for DV from the start instead! However there is always that nagging feeling that I will be able to do something with the extra quality. I'm thinking that in the not too distant future I will have a worthwhile HTPC setup which could negate the need for DVD entirely. I'll have to look into the Cineform codec and make my decision as to what to capture to. Thankfully I don't have as much as 300 hours, though perhaps if I did it would make the decision a bit easier!

Apart from that I think I'm more or less ready to go ahead. Thankyou all again for your help- I highly doubt there is much I can offer in return but if I can let me know! I'm sure I'll be back on this board before too long trying to figure out how the hell to use AVIsynth, and what this impressive sounding 'applyrange()' function is..

Cheers,

Malek