Video Editors

[Happy Expat]

Hi there,
Anybody know of, or feel like writing a utility to scan video files MPEGs, AVIs etc for the gaps between "scenes".
The target market would be all amateur video photographers who tend to use their new toy more like a video snapshot camera and produce 90 minute tapes with loads of dross between the scenes.
That dross needs to be edited out and finding the stuff - almost frame by frame, takes an eternity.
These gaps are "normally" a few frames - or seconds, with the snowy interference pattern you get on a screen when the transmission ends but the TV is left on, or sometimes just black frames.
If the entire file could be scanned and an index built giving the position to within a couple of seconds, or frames - then it would make life so much easier.
I imagine it would be "simple" enough for the user to edit a few frames from an existing video containing nothing but the type of frame being sought and then saving this as a sample.
The sample would then contain a template of the data that triggers an entry into an index.
There are obviously dozens of CODECS but I think they are identifiable by the header and the saved sample would contain both the header and frame data.
I checked some commercial products with no success so it could be either too complex to incorporate into a sophisticated editing package or, hopefully, something nobody thought of writing as a simple standalone utility.
I would imagine a product like that could generate quite a lot of exposure for the originator.

[superboyac]

This exists, I just forgot the name of the software that can do it.  Sorry, useless!!

[Vurbal]

Your best bet is probably to focus on finding a more general purpose editor that has such a feature (most of them do) and figuring out whether it can be tuned to do what you want.

The general feature you're looking for is typically called scene detection. Based on your description, though, your best bet would probably be finding an editor with specialized scene detection for identifying commercials. For example there's a standalone tool (only for MPEG 1 and 2 video IIRC and maybe only TS files) called ComSkip. It's also used (or was at one time) in at least one MPEG editor called VideoReDo.

If you can't find a simple solution there's always AviSynth. It's definitely the most capable tool for the job but the learning curve is steeper since it's script rather than GUI based. It does have the advantage of supporting nearly any video and/or audio format there is. If you end up going that route my advice would be to post a question over at Doom9's AviSynth user forum and you should be able to figure it out relatively quickly. When you setup an account there you have to wait 5 days before posting so plan accordingly.

Also you would probably want the 2.6.0 alpha 5 version rather than 2.5.8 stable. Despite the alpha designation it's actually considered more stable and also has a significant number of fixes and new features.

[4wd]

For AVISynth there's MSU TV Commercial Detector which will make using AVISynth (reasonably) easy.

Besides the normal black frame detection it can also analyse for the station logo, which tends to disappear during commercials so as not to annoy the sponsors.

The resulting .avs file, (which contains sections to keep/discard), can be used by any program that can make use of AVISynth or understand it's files, eg. VirtualDub, AVIDemux.

NOTE: VirtualDub is geared towards AVI media not using MPEG2, (due to licensing), but AVIDemux can handle most media codecs - there's also the offshoot VirtualDubMod which can handle MPEG2.

Don't know how accurate it's going to be, (I use VideoRedo, as Vurbal mentioned).

[Vurbal]

It's generally best to avoid VirtualDubMod now since it hasn't been updated in over a decade and there are plugins available to open pretty much everything under the sun in VirtualDub. That includes 1 or 2 formats that still give AviSynth fits. There's a list on the VirtualDub forums which also includes information about, or at least links to, some third party filters. I haven't heard of any commercial or scene detection filters but I use VirtualDub almost exclusively for previewing AviSynth scripts so I'm far from an expert.

[Happy Expat]

I think I need to become more familiar with the subject before I venture too far into the unknown. As a rank amateur I was rather hoping there was an easier solution. Still, I have plenty of time to learn.
I was trying to preserve my wife's "video snapshots" for posterity but it appears that "point and shoot" really means "point and shoot and something dies" - the dreams you thought you were capturing :-).
Thank you all for your time and endeavours.

[Vurbal]

I think I need to become more familiar with the subject before I venture too far into the unknown. As a rank amateur I was rather hoping there was an easier solution. Still, I have plenty of time to learn.

-Happy Expat (March 05, 2014, 11:59 PM)

Here's a post I made some time back that might be a good starting point.

The bad news is it doesn't include information about the container used by almost all point and shoot cameras which is Apple's QuickTime, easily identified by the .MOV extension. It's also missing an entry for the video compression used for all video on older still cameras and still typically used today at lower resolutions which is Motion JPEG.

On the good side you're already ahead of a lot of people who seem determined to find a simple solution even after you explain the complexity of the problem. At least you've got that going for you.

[Happy Expat]

Just read most of your post.
If that's where I've got to start, can you hold my hand a little more along the way :-)
Thanks!

[Vurbal]

No problem. Here's my first tip. Hold off on wading through all that for a bit and we'll focus on figuring out what kind of files you have. Download MediaInfo and open one of your video files with it. Take a screenshot and post it here and I can to give you a significantly shorter list of what you should know before starting.

[Happy Expat]

Vurbal, I was in the process of typing this as your earlier response came through, I suspect that some of what the mediainfo screenshots convey will also contribute to the answer for this query.
As I was copying Hi8 tapes with two as ancient as 20 years and possibly already transcribed from cine film, and the rest dating back as much as 10 years and all probably saved in whatever passed for high res in those days, I thought it would be overkill to use the highest resolution digital format available from the capture device. Consequently, files are all under four Gigs for each 90 minute tape. I actually captured them on an old XP machine and intended to transfer them via FAT32 USB sticks, so the 4Gig file limit was a consideration. Clearly, I can re-format the Sticks to NTFS or even utilise a pluggable NTFS hard disk to transfer them if I reimport them into MPEG files.
In your opinion would I be wasting my time grabbing a higher resolution?
I shall now go and download mediainfo and provide the screenshots.
Thanks

[Curt]

Download MediaInfo and open one of your video files with it. Take a screenshot ...

-Vurbal (March 06, 2014, 02:32 AM)

1) maybe "and click Save to text file" ... rather than a screenshot.

2) then come here and click "preview", and on the new page click [SP] and post the text (not the file) inside the brackets ]text[

like this:

Spoiler

General
Complete name                            : F:\Video Extern\Amy Macdonald - Don't Tell Me That It's Over.mp4
Format                                   : MPEG-4
Format profile                           : Base Media / Version 2
Codec ID                                 : mp42
File size                                : 17.1 MiB
Duration                                 : 3mn 20s
Overall bit rate mode                    : Variable
Overall bit rate                         : 714 Kbps
Encoded date                             : UTC 2014-01-02 16:50:25
Tagged date                              : UTC 2014-01-02 16:50:25
gsst                                     : 0
gstd                                     : 200851
gssd                                     : B535E79C8HH1393519101767761
gshh                                     : r5---sn-uqj-j2ie.googlevideo.com

Video
ID                                       : 1
Format                                   : AVC
Format/Info                              : Advanced Video Codec
Format profile                           : [email protected]
Format settings, CABAC                   : No
Format settings, ReFrames                : 1 frame
Codec ID                                 : avc1
Codec ID/Info                            : Advanced Video Coding
Duration                                 : 3mn 20s
Bit rate                                 : 615 Kbps
Maximum bit rate                         : 1 397 Kbps
Width                                    : 640 pixels
Height                                   : 360 pixels
Display aspect ratio                     : 16:9
Frame rate mode                          : Constant
Frame rate                               : 25.000 fps
Color space                              : YUV
Chroma subsampling                       : 4:2:0
Bit depth                                : 8 bits
Scan type                                : Progressive
Bits/(Pixel*Frame)                       : 0.107
Stream size                              : 14.7 MiB (86%)
Tagged date                              : UTC 2014-01-02 16:50:26

Audio
ID                                       : 2
Format                                   : AAC
Format/Info                              : Advanced Audio Codec
Format profile                           : LC
Codec ID                                 : 40
Duration                                 : 3mn 20s
Bit rate mode                            : Variable
Bit rate                                 : 96.0 Kbps
Maximum bit rate                         : 101 Kbps
Channel(s)                               : 2 channels
Channel positions                        : Front: L R
Sampling rate                            : 44.1 KHz
Compression mode                         : Lossy
Stream size                              : 2.30 MiB (13%)
Title                                    : IsoMedia File Produced by Google, 5-11-2011
Encoded date                             : UTC 2014-01-02 16:50:26
Tagged date                              : UTC 2014-01-02 16:50:26

[Happy Expat]

It just occurred to me that the only data files I have are already converted from the native tape format to MPEGs
'They were captured using Arcsoft Showbiz through an EZCAP analogue converter. The bundled early version of ShowBiz 3.5 couldn't process the analogue stream, claiming it was copy protected, so I downloaded a trial version of 5.x and captured the files using that. As the audio in the captured files was a simple mono on one stereo track, I then stripped the audio into a WAV file (Audacity) and got Power Sound Editor to then append a mirror track to make it twin track mono. Much better reproduction but still simply two inherently inferior tracks. Then the original video and the new audio were loaded into power director and the original audi track stripped out and replaced with the twin track mono. Everything was then output again in MPEG4 with all the setting targeting the familiar 4Gig threshold. It occurs to me that as I have up to 4TB of Hard Disk and numerous 16Gig Sticks I could easily accommodate much larger files.
You clearly type and/or think fastr than I do as this post is also overlapping another of your responses.

[Happy Expat]

General
Complete name                            : C:\Users\Colin\Videos\Home Capture\Camcorder\Trimmed but not edited\Tape01_Oct2001-Mar2002.mp4
Format                                   : MPEG-4
Format profile                           : Sony PSP
Codec ID                                 : MSNV
File size                                : 3.96 GiB
Duration                                 : 1h 32mn
Overall bit rate mode                    : Variable
Overall bit rate                         : 6 126 Kbps
Encoded date                             : UTC 2014-03-04 11:08:50
Tagged date                              : UTC 2014-03-04 11:08:50

Video
ID                                       : 1
Format                                   : AVC
Format/Info                              : Advanced Video Codec
Format profile                           : [email protected]
Format settings, CABAC                   : Yes
Format settings, ReFrames                : 2 frames
Codec ID                                 : avc1
Codec ID/Info                            : Advanced Video Coding
Duration                                 : 1h 32mn
Bit rate mode                            : Variable
Bit rate                                 : 6 004 Kbps
Maximum bit rate                         : 8 000 Kbps
Width                                    : 640 pixels
Height                                   : 480 pixels
Display aspect ratio                     : 4:3
Frame rate mode                          : Constant
Frame rate                               : 25.000 fps
Color space                              : YUV
Chroma subsampling                       : 4:2:0
Bit depth                                : 8 bits
Scan type                                : Progressive
Bits/(Pixel*Frame)                       : 0.782
Stream size                              : 3.88 GiB (98%)
Language                                 : English
Encoded date                             : UTC 2014-03-04 11:08:50
Tagged date                              : UTC 2014-03-04 11:08:50

Audio
ID                                       : 2
Format                                   : AAC
Format/Info                              : Advanced Audio Codec
Format profile                           : LC
Codec ID                                 : 40
Duration                                 : 1h 32mn
Source duration                          : 1h 32mn
Source_Duration_FirstFrame               : -21ms
Bit rate mode                            : Constant
Bit rate                                 : 117 Kbps
Nominal bit rate                         : 128 Kbps
Channel(s)                               : 2 channels
Channel positions                        : Front: L R
Sampling rate                            : 48.0 KHz
Compression mode                         : Lossy
Stream size                              : 77.4 MiB (2%)
Source stream size                       : 77.4 MiB (2%)
Language                                 : English
Encoded date                             : UTC 2014-03-04 11:08:50
Tagged date                              : UTC 2014-03-04 11:08:50
mdhd_Duration                            : 5549385

[Happy Expat]

Did I at least get that bit right?

[Vurbal]

Download MediaInfo and open one of your video files with it. Take a screenshot ...

-Vurbal (March 06, 2014, 02:32 AM)

1) maybe "and click Save to text file" ... rather than a screenshot.

2) then come here and click "preview", and on the new page click [SP] and post the text (not the file) inside the brackets ]text[

like this:

-Curt (March 06, 2014, 02:49 AM)

That's crazy talk! There's a reason I never remember the easy way to do that and sooner or later I'll figure out what it is. 

[Curt]

my main concern was that a screenshot most likely would produce a very large picture, and though large pictures work wonderful well on desktop monitors, they doesn't work at all on smart phone screens. The new generation of youngsters all use smart phones.

Anyway, forgive me having made noise, I will now keep silent and enjoy your teaching on video editing.

[Happy Expat]

Well, I think I've spotted one thing from the output.
Showbiz 5.x is stated as having the AAC CODEC slugged for the trial period which is probably why the audio was so poor!
Curt: I hope I'm the "new generation youngster" you're referring to.

[Vurbal]

Did I at least get that bit right?

-Happy Expat (March 06, 2014, 03:06 AM)

Yep - but you should also put spoiler tags around it like Curt mentioned. That will keep things easier to read/find later on. Let's go through that first and I'll get to the analog capture bit in a separate post.

The first thing to look at is the General section which will give you container information.

Format                                   : MPEG-4
Format profile                          : Sony PSP

The Format field says MPEG-4 which you might logically conclude means it refers to the official MPEG-4 container aka a MP4 file. However you also have to look at the Format profile field. If that field said QuickTime instead it would actually be in the QuickTime container aka MOV.  In this case it says Sony PSP so you know it is actually a MP4 file. It could say any number of things but all you really have to remember is if it says QuickTime that's the container and otherwise it should always be an actual MP4 file.

If this were video from a point and shoot camera I would expect this to indicate a QuickTime file. That's significant because QuickTime files aren't particularly well supported on Windows so you have to be careful (or at least thorough) when planning a workflow.

Next is the Video section which obviously provides details about the video stream.  Once again 2 lines will tell you everything you need to know at this point.

Format                                   : AVC
Format/Info                             : Advanced Video Codec

This tells us the video is encoded to the MPEG-4 AVC aka H.264 standard. That's also confirmed by the Codec information:

Codec ID                                 : avc1
Codec ID/Info                           : Advanced Video Coding

Finally there's the audio section:

Format                                   : AAC
Format/Info                              : Advanced Audio Codec
Format profile                           : LC

In this case you've got AAC LC (Low Complexity) audio which is what you would expect from a newer point and shoot camera. If it's an older camera that should be PCM aka Pulse Code Modulation aka uncompressed audio. The most important thing here, particularly if it's anything besides PCM, is to remember you ideally want to avoid re-encoding without a good reason.

[Vurbal]

my main concern was that a screenshot most likely would produce a very large picture, and though large pictures work wonderful well on desktop monitors, they doesn't work at all on smart phone screens. The new generation of youngsters all use smart phones.

Anyway, forgive me having made noise, I will now keep silent and enjoy your teaching on video editing.

-Curt (March 06, 2014, 03:42 AM)

No, you're right. I'm just used to capturing screenshots for publishing guides so I never remember the text output option. 

[Happy Expat]

My apologies, I'm so unused to Forum procedures that I didn't fully understand the shorthand Curt was using.
I did go straight into the [preview] pane and then inserted the text by [ctrl]+v as I couldn't identify any other way to do it. It certainly copied the info so I thought I had "cracked it" - I obviously need yet further instructions.
As I mentioned, the files I have now, are already reprocessed from the original Hi8 tapes and are the result of encoding by Arcsoft Showbiz 5.x and then power director 10. I tried not to change anything apart from ensuring the project was always defined as 16:9 as opposed to 4:3 and that the file size remained fairly consistent but I have no idea whether there is actually a better way to get the info of the Hi8 tape.
Camera comes with 3.5mm AV and s-video out ports but as the signal is analogue the only way to make them PC usable is via a converter. I used the EZCAP USB analogue converter which has provided usable MPEGs but I'm not sure whether the aficionados would consider this more "Mickey Mouse" than "Cecil B. DeMille"
Should I start all over again with a better tape transfer mechanism?

[Vurbal]

On the capture question, it's always best to capture at a higher resolution except when it isn't really a higher resolution. Yeah, it's just that simple.

Actually here's what I mean. Pretty much every (consumer) capture device will have 1 specific resolution that the actual capture (sampling) happens at. It may allow you to select half that resolution which it achieves by simply throwing away half the samples. Obviously that's not what you want. Depending on the hardware, though, it may also offer higher resolutions which involve processing frames beyond what was sampled and that's not good either.

I find the best way to look at it is to distinguish between resolution and definition. This isn't the technical meaning of definition because, in fact, there isn't one. It's just marketing speak but for lack of an alternative it's the word I use.

Any digital image is a collection of samples representing an analog image. Think of each pixel as an individual detail. I refer to that original resolution as the image's definition. In other words definition, as I use the term, refers to the level of captured detail. If you reduce the resolution you also reduce the definition. However if you increase the resolution, the definition remains the same. You haven't added any actual details. All you've done is told your computer to interpolate new details which may or (more likely) may not be accurate.

In other words you should always capture at the highest resolution possible without exceeding the definition of the capture hardware. That's assuming both the hardware driver and capture software allow it, which they often don't. USB capture devices, in particular, are usually very limiting because they tend not to have hardware encoders and USB itself isn't particularly reliable for sustained throughput.

The question is whether there are other realistic options available now that most people have abandoned analog video entirely. Of course at the end of the day the most important thing is how happy you are with the quality. If you aren't happy with it and end up deciding to get better hardware I can only give you general advice. Back in the day Hauppauge capture cards had a good reputation but I don't know if that's still true since they don't use hardware encoding any more. That might be a good starting point at least.

[Happy Expat]

As I made all the original files at the "normal" setting for the capture device I imagine that was It's optimum. Is there a way to verify this?
Would the secondary processing where I modified the audio and then re-saved in the same format and resolution likely have adversely affected the true video data or is that likely to be negligible?  Arcsoft use their own CODEC whereas Power Director (I think) use a more "industry standard?" CODEC. Or am I talking complete rubbish

[40hz]

Any digital image is a collection of samples representing an analog image. Think of each pixel as an individual detail. I refer to that original resolution as the image's definition. In other words definition, as I use the term, refers to the level of captured detail. If you reduce the resolution you also reduce the definition. However if you increase the resolution, the definition remains the same. You haven't added any actual details. All you've done is told your computer to interpolate new details which may or (more likely) may not be accurate.

-Vurbal (March 06, 2014, 05:48 AM)

@V - thank you for that! That was the shortest and clearest illumination of the difference (and one-way interaction) between 'definition' and 'resolution' I've ever read. Next time I need to explain those terms to someone I'm going with your definition and example.

[Vurbal]

As I made all the original files at the "normal" setting for the capture device I imagine that was It's optimum. Is there a way to verify this?
Would the secondary processing where I modified the audio and then re-saved in the same format and resolution likely have adversely affected the true video data or is that likely to be negligible?  Arcsoft use their own CODEC whereas Power Director (I think) use a more "industry standard?" CODEC. Or am I talking complete rubbish

-Happy Expat (March 06, 2014, 06:03 AM)

I did some digging and came up with a reasonable amount of information on your card plus a little information (and some educated guesses) about the company that sells it. First off I would make sure never to buy anything from them again. Between some questionable, not outright false, statements about their products and the supposed knock offs and their requirement to sign up for their forum before even reading any posts they seem slimy and untrustworthy.

The good news, though, is you seem to have a pretty much industry standard USB capture device. It's one of at least 3 made by different Chinese OEMs, and possibly the best of the 3 since it looks like Hauppauge sells (or sold) essentially the same unit under their brand. The key components, in terms of capture software compatibility and video capture specs are as follows:

Video capture chip: Empia EM2861 (WDM capture hardware)
Video processing chip: Philips SAA7113 (Samples full frame SD video to NTSC/PAL,Uncompressed YUV 4:2:2)
Audio processing chip: possibly Empia EMP202 2 Channel AC'97 (Dolby Digital) or perhaps other chip supporting just 8000Hz mono

The captured video will be filtered automatically in the hardware but that's pretty much unavoidable unless you're ready to shell out $200 - $300 for a prosumer level capture device from BlackMagic Design. On the bad side the manufacturer's quality control is about what you'd expect from a low end commodity electronics product. If the sound is bad the only solution may be bypassing it and capturing with your sound card.

That option would also mean going with a different capture program but it doesn't look like ArcSoft's product is really suited to proper capturing anyway. Although it appears to stick with industry standard formats, I question the quality of their encoders, particularly for realtime encoding. Also the standard method for high quality capturing involves using an intermediate lossless codec initially and then encoding to your final format as a separate step. Show Biz 3.5 (or 5) may or may not be able to do that. It mostly depends on whether it gives you access to any VfW or ACM encoders you may have installed.

Back to the good news it looks like VirtualDub should have no problem capturing from it and also supports capturing audio through a sound card at the same time. I've never used it for capturing and I do know some people have horrible problems with audio sync, at least until they spend time tweaking some settings. It would definitely work with some great free lossless codecs which are designed for capture.

For maximum quality given your hardware that's where I would start. If that's what you want I'll do what I can to help but you need to understand up front that it could involve quite a bit of frustration in the beginning. Or it might be a walk in the park. There just isn't any way to tell ahead of time.

[Curt]

Curt: I hope I'm the "new generation youngster" you're referring to.

-Happy Expat (March 06, 2014, 03:44 AM)

^hahaha! Tyvärr!