Problem with corrupt AVIs (Gspot overly sensitive?)

[Spirogyra]

I attempted to make this post once already, but it appears that I timed out before submitting and the post was lost. Hence, if it appears twice, please forgive me.

I am experiencing an odd problem with corrupt AVIs. I am trying to move some VHS video onto a DVD. I captured the video by converting the analog VHS signal using a DV camcorder with a “passthrough” feature and then running the converted DV signal to the computer via firewire. I captured the DV signal using Pinnacle Studio 7 and the DV capture card which came with it. I used Studio 7 to edit the video and then rendered a new AVI of the edited video. The edited video appeared to work fine – it played in Windows Media Player and could be reopened and manipulated by Studio 7. Since the edited AVI looked fine, I deleted the raw unedited AVI and the Studio 7 project file which contained my edits in order to reopen needed space on my harddrive.

I don’t remember what caused me to do this, but prior to rendering the MPEG2 with TMPGenc, I opened the edited AVI in Gspot and discovered, much to my frustration, that the AVI header was corrupt. According to Gspot, the header was corrupt, it was a “serious” problem and the file would not likely be playable. However, as I said, the file did appear to play fine in WMP and Studio, though I admit I did not watch it from beginning to end (its a little over 40 minutes in length), but I did skip around using the slider bar in WMP and had no problems.

Anyway, now I faced a dilemma. I could go ahead and encode the MPEG with the “corrupt” AVI and risk wasting a lot of compression hours for a bad MPEG. Or I could throw out what was potentially a perfectly good AVI and go through the whole capture, edit and render business again in hopes of getting an AVI which Gspot would sign off on.

I rooted around a bit on this website and around the web generally and couldn’t find any helpful guidance. Since I couldn’t post on the forum until a day had passed following my registration, I did some further experiments. I recaptured the video from the VHS and discovered, much to my surprise, that Gspot said the captured raw AVI – call it “A” – had a corrupt header. I then opened A in Studio and used the trimming tool to render two new AVIs – call them A1 and A2 – of the first and second halves of the corrupt A. Gspot reported that A1 was okay, but A2 was corrupt. So I took A2 and broke it up into three 5-8 minute segments (5, 5, and 8 minutes actually) – call these segments A2a, A2b and A2c. Surprisingly, Gspot now reported that A2a, A2b and A2c were all fine (i.e., no corruption). Confused, I took these three files and recombined them in Studio to create a new AVI file which should be visually identical to A2 – call this recombined file A2*. Gspot said A2* was a compliant AVI, even though it should be substantially the same as A2, which Gspot said was corrupt. Weird.

Note that in checking the various AVIs with Gspot, sometimes it would report that a file was okay, except that it had some additional data at the end of the file which Gspot identified as “garbage” or “junk.” The gist seemed to be that the file was longer than necessary given the header information, but that it shouldn’t be a problem.

Now my questions:

- First, what are people’s experiences with Gspot’s ability to detect corruption in AVI files? Is it good at this? Is it overly sensitive? If not Gspot, what freeware utilities would you recommend for use in quickly checking the integrity of an AVI (something I intend to always do from now on upon creation of any AVIs)?

- Second, given that my first edited AVI seemed to play fine in WMP, notwithstanding Gspot’s suggestion of corruption, and assuming TMPGenc would accept the file without question, should I just go ahead and render an MPEG2 using this first “corrupt” file? Will TMPGenc create a clean MPEG provided it can read the AVI notwithstanding a minor imperfection in the AVI, or is there a risk of getting a crap MPEG here?

- Third, is it possible that a portion of VHS video can have some problem that would cause the capture software to reliably create AVIs with corrupt headers? This wouldn’t seem to make sense.

- Fourth, does anybody have any idea why an AVI file which Gspot reports is corrupt can be rendered non-corrupt simply by breaking it down into parts and then recombining those parts to create a new AVI which is visually identical?

- Fifth, is there any concern that taking an initial AVI and using Studio 7 to render new AVIs using the initial AVI and then building yet more AVIs using the rendered AVIs can result in a degradation of video quality? My intuition tells me that there shouldn’t be a problem since there is no compression occurring and the digital quality of the data should allow perfect copying without loss of data.

Thanks in advance for any tips/advice.

[erratic]

I think it's a GSpot bug. When I capture a big (MJPEG) AVI file with VirtualDubMod, GSpot sometimes tells me that it has a corrupt AVI header, but I can open the file just fine with VirtualDubMod, TMPGEnc or Vegas. I can edit it, re-render it, convert it to MPEG-2, etc. Nothing wrong with the file at all, it would seem. So I don't think your AVI files are really corrupt.

[Spirogyra]

Thanks for the post Erratic. I went ahead and coded the MPEG with the supposedly corrupt AVI and it appears to work fine. I already have it on a playable DVD. I am still curious though what freeware utility is best for checking the integrity of video files.

[erratic]

With VirtualDub or VirtualDubMod (both free) you can scan a video stream for errors. I don't know if there's something better.

I always assume there's nothing wrong with an AVI file if I can open it with an NLE (Vegas 4.0 in my case).