View Full Version : 2-Pass Mainconcept MPEG2 rendering question


Bryan Quarrie
February 13th, 2018, 07:06 AM
Hi all - just a quick question (maybe needs a technical answer, who knows...).

When 2-pass is selected when rendering as an MPEG2 file, what does vegas do in the 2nd pass?
Is it as simple as passing over the timeline twice before writing the rendered file to its destination drive?
Or are there other things going on related to video detail as it does so?

Just curious, as up until now I’ve not really paid that much attention to this feature.

Regards,

Bryan

David Stoneburner
February 13th, 2018, 08:33 AM
I'm not 100% sure, but I always thought that pass one was a detailed analyzing and then pass 2 was the encoding plus additional analyzing. While one pass analyses and encodes at the same time.

Adam Stanislav
February 13th, 2018, 09:10 AM
It seems to me that using 2 passes allows the encoder to compress the data more efficiently, as MPEG-2 (and later) allows to subtract the data in a frame from the data of another frame, whether that other frame has appeared in the data stream or is yet to appear in the near future (by future, I mean it will be shown several frames later).

So, if you only use one pass, you only know about the several past frames before you send them out to the mpg file, but if you use 2 passes, you can also learn that perhaps subtracting this frame from one that comes in later and make a note of it, so you can use that information in the second pass to compress the frame more efficiently.

At least that is the general impression I have about how the process works.

Adam

Bryan Quarrie
February 13th, 2018, 10:15 AM
Thanks guys.

I’ve tried it with some long form clips today. Normally without 2-pass, a single 1hr clip (HD source with only a single NewblueFX colour plugin) would take about 30 mins to render. Now with 2-pass it is twice that obviously, but there’s little change in the result - unless it may be more noticeable in high-detail scenes like crowds etc. (I’m just speculating but I’ll try it with more dense detail material).

The compression efficiency factor sounds interesting (especially with frame analysis in both passes). Let’s say I have a clip that has some pretty dense transition effects like crossfading between a talking head scene to a scene featuring an audience applauding. Could this improve artifacts like macroblocking between frames etc?

Thanks,

Bryan

Seth Bloombaum
February 13th, 2018, 11:47 AM
The compression efficiency factor sounds interesting (especially with frame analysis in both passes). Let’s say I have a clip that has some pretty dense transition effects like crossfading between a talking head scene to a scene featuring an audience applauding. Could this improve artifacts like macroblocking between frames etc?

What’s written above is correct - first pass analysis, second pass encoding.

The compression efficiency of 2-pass encoding is mostly about motion. It works best with variable bitrate (VBR) settings. Less data on low-motion frames, more data on high-motion frames.

The transition you describe *would* be considered “high-motion”, that is, all pixels are changing on each frame of the transition, and is a good place to apply 2-pass VBR compression.

The MC MPEG-2 codec in Vegas is really quite good, but why MPEG-2? It’s become a specialized encode for DVD and sometimes BluRay. h.264/MP4 is better for BluRay... use it if you can, 2-pass, VBR.

Bryan Quarrie
February 13th, 2018, 01:11 PM
Hi Seth,

Unfortunately I’m still using MPEG2 because my clients still insist on having DVDs 😕.

There was a time when I was torn between having smooth motion but less detail (interlaced footage), or creating “Universal” discs (which are really NTSC progressive videos, but can be played on any DVD player), that restricts the franerate to 29.97fps, but the detail is better than interlaced footage (I live in PAL land but create DVDs for loved ones of clients in the West Indies (NTSC) when I film funerals that happen in UK etc).

I can do HD renders, but these guys are more concerned about having DVDs because they are still common.

Seems like DVDs are still not dead yet lol!

Of course give me 1080p50 anyday!!

Pete Cofrancesco
February 13th, 2018, 01:44 PM
I can’t see the small improvement of a 2 pass and don’t find it worth double the encode time. If you posted two short clips I doubt anyone could pick out the 2 pass. If you upload to youtube or the like they recode it anyways undoing your 2 pass hard work. I often catch things I don’t like in the video once I build a dvd and watch it on my tv and then go back and re encode.

Jeff Pulera
February 13th, 2018, 02:23 PM
The benefits of 2-Pass VBR will be more noticeable when encoding longer programs = lower bitrates.

If the program is like an hour or less, I would just use CBR at 8.0 rate and it will ALL be good, no dipping the quality up and down.

Thanks