Strange bitrate reading On Transitions

[barneystorm]

I use Womble MPEG Video Wizard for all my MPEG editing, along with TMPGEnc DVD Author (and occasionally Sony DVD Architct).

I have two TV broadcasts -both DVB-S captures with the same bitrate (6500 Variable)- that I've joined together on the Womble timeline. I put a one second video/audio fade on the end of one and the start of the other.

I've done this, literally, thousands of times, and TMPGEnc DVD Author never has a problem with outputted MPEG file.

Someone I sent a finalised disc to has mailed me to say that when he tried to re-author the DVD, with DVD Lab, on the fades, the video bitrate is showing up at almost 15mb sec and won't let him author.

I've checked just the transition in virtualdubMPEG and sure enough, it shows that it spikes up to around 15mb , but only on the fade in/out.

I know that when you capture some DVB-S broadcasts their bitrate flags at 15mb/sec when in fact they are absoluely nowhere near that (Never had a satisfactory answer as to why that happens). In those cases, I just run the whole file through DVD patcher to patch every header with the right bitrate (or once close enough to) before authoring.

Has anyone any idea what could be causing this? Is it a bug in Womble?

[guns1inger]

Most encoders spike occassionally, and if a spike is going to happen, then a cross fade or fade to black is a likely place. Fades require a lot of bitrate because every pixel is affected. The encoder should prevent this from becoming a spike, but obviously womble has barfed this time.

[barneystorm]

Hmm..sounds like a flaw in Womble to me

I've just gone back and checked on several other DVD's I've already authored and burnt and it seems to be the same on every single one. Which begs the question-if every fade in fade out area is going up to 13-14/mb sec, why doesn't TMPGenc DVD Author - or Sony DVD Architect- refuse to accept the MPEG for authoring when DVD Lab does??

[SLK001]

The absolute max bitrate that a DVD can have is 10.08Mbps, of which 9.8Mbps is supposedly the max video rate. Your authoring software may just clamp the spikes in BR to 9.8Mbps, which should result in macroblocking - but in a fade to black, you may not notice it unless you look closely.

It's possible that buffering could hold enough data to sustain a 13Mbps spike for about a quarter of a second, so that although there is the spike, the average BR over a one second interval is only 9.8Mbps.

I believe that an authoring program should throw an error when it encounters a BR above the max - but there may be exceptions to that "max" rule that I am not aware of.

I've done quite a few fade to black transistions and I don't see the spikes that you mention. I think more experimentation is in order.