w64 or wav for B-R, 48kHz/96kHz & video render

Gabonviper wrote on 9/1/2012, 2:23 AM

Hi,

I am creating my first blu-ray music video (MainConcept MPEG-2, 1440x1080-60i, 25Mbps video stream) in Vegas Pro 11 (a multiaudio four-track mix from various audio sources) and was wondering if it makes any difference whether I use the Sony Wave64 template or the Wave (Microsoft) one for the audio. Both sound the same to me, but are there any compatibility issues?

Also, should I render it as 48kHz or 96kHz? 16, 24 or 32 bit?

I will finish it in DVDA 5.2 and with the audio file rendered as 48kHz the four-hour 15 minute project now fits on two 25GB discs, with about two hundred megs of free space on each after menus and all.

Another question: The multicam video mix is a compilation of altogether 20 video sources (varying quality and frame rates) and the default Pro 11 gives for render is HDV 720-30p as those were the settings I chose for the project in the properties (1280x720, 29,970 NTSC).

Of course this template is no good for blu-ray, so I have already rendered the whole project with the MainConcept MPEG-2, 1440x1080-60i, 25Mbps video stream template and it looks fine on the computer screen and projected with a Full-HD projector onto a large screen. I haven't tested it on regular TVs yet, so I don't know if it this would be the right setting for regular TV-viewing in Finland.

The only thing is the project is now so huge I won't be able to add animated buttons to the menus in DVDA, as both discs would then exceed the space by about 500 megs.
Then the question: If I lower the output width and height of the template from 1440x1080 (pixel aspect ratio 1,3333 to 1280x720 (pixel aspect ratio 1,000), would this

a) affect video quality (I made test comparisons and they both look the same on the computer screen qualitywise, though the 1440 one opens to full screen whereas the 1280 one doesn't.)

b) reduce the size of the rendered file? (the 15-second test files vary surprisingly: the 1440 is 43 megs and the 1280 is 45,6 megs, so maybe I shouldn't change it, but then, would changing the resolution

c) affect playability in various devices (I watch it on PS3)?

Lastly, should I choose 1440x1080 or 1280x720 in DVDA for resolution in properties?

Any help appreciated. Bear with me and be gentle, as I am a novice ;-)

Cheers,

Marko

Back to post

Comments

Chienworks wrote on 9/3/2012, 7:10 AM

Laurence, i'd be very skeptical of the frequency response of the audio transducers in your iPad. Chances are that it's got a huge drop in response to frequencies over 10K anyway, so it's a very poor test of your hearing in that range.

Gabonviper wrote on 9/3/2012, 8:55 AM

Thanks all you guys for all the info and suggestions. Not particularly a manual buff, I usually first resort to trial and error, and then the manual. However, in order to make use of a manual one needs to know what to look for, and that's where forums like these and guys like you who are willing to invest time and (a lot of ) patience in novices' questions really come in handy :).

"When everything else fails, read the manual" - Murphy's Law.

But I still hate manuals.

Laurence wrote on 9/3/2012, 9:03 AM

>

Another story: I went to my kid's annual science fair a few years ago and one boy's project was on hearing range. He had a cd with test tones on a portable CD . He had done tests and graphed out the frequencies. Every single father that had taken the test had very little hearing over 10k. Moms did a little better up into maybe the 12k range and most of the children could hear tones that their parents couldn't. This experience is what inspired me to look for the iPad test.

As far as the iPad test goes, every child I try it with can hear the 15 to 16k, I have one adult male friend who can hear 19k. I'm sure of this. I had him look the other way as I turned it off and on and he could tell repeatedly exactly when the tones started and stopped. I am sure the iPad can do at least 19k because of this.

A little background on myself. I worked as an audio engineer for years at Disney world shows. In the process I did quite a bit of damage to my hearing. My ears constantly ring at several pitches. None the less I still do audio at two churches which I love because the praise bands sound awesome and nobody minds the lower spl levels that I now strive for.

John_Cline wrote on 9/3/2012, 9:11 AM

It makes me cringe when someone defiantly says they never read manuals.

john_dennis wrote on 9/3/2012, 9:13 AM

"But I still hate manuals."

So do I. I've read a few in my life.

John_Cline wrote on 9/3/2012, 9:14 AM

Typically, women and children can hear higher frequencies than men. High frequency sensitivity in men starts to drop off at around age 25.

Laurence wrote on 9/3/2012, 11:04 AM

Check out this hearing test:

http://www.phys.unsw.edu.au/jw/hearing.html

If you can hear 12k, your ears are better than mine.

mudsmith wrote on 9/3/2012, 1:58 PM

Some audio-only input:

Dolby Digital is definitely a lossy algorithm as stated by the poster above. It ALWAYS sounds worse than PCM.

Even if the original material is only at 16 bits (PCM), the mixed final product can certainly sound way better using the 24bit output of the program because of the internal 32bit mixing engine.....which is to say that the mix, including effects, has more going on than the original audio.

In fact, if one is going to print the mix at 16 bits, even if the original was also at 16bits and even if you are going to Dolby Digital or MP3 quality, it is important to BOTH add a little mastering compression and do some careful dithering.....otherwise, you will be truncating what you are hearing during playback and potentially reducing the apparent "size" of your mix.

As I have spent years trying to figure out the best way to accomplish this last bit of processing in a variety of situation, it might take a while to give you recommendations. This would also require a knowledge of your audio plugin packages.

If you have the space to keep the final at 24bits PCM, and if you have your project set for that, then there is an audible difference. If not, play with dithering and Master Buss compression.

mudsmith wrote on 9/3/2012, 2:01 PM

Also, I agree with the poster above that it is highly unlikely in your situation that printing a mix at 96k will buy you anything but less space on your BluRay. 96k can be startlingly good sounding, but requires more going on in the path than just a sudden SRC at the last stage.

Gabonviper wrote on 9/9/2012, 3:54 AM

I tried several modifications to the AVC and MPEG-2 templates, but had to discard the AVC options, as they tended to intensify any camera shake in the footage, rendering the video partly unwatchable.

Of the various Main Concept MPEG-2 25mbps 60i templates, only the default versions of 1440 60i (upper field first) and 1920 60i (both upper and progressive) with no modifications of the bit rates were accepted by DVDA as such, so no recompression was required. If I tried to modifiy the bitrates in the template, DVDA would require recompression. The only modifications DVDA allowed were changing the max and average bitrates to 25 (from 30 and 20) and the lowest bitrate from 20 to 16. Any other combination (and I tried a few) would result in DVDA stating the need for recompression. This of course meant that I could not really lower the bitrate of the 25 mbps 60i templates very much.

Interestingly though, although in the above it was firmly stated that resolution does not affect the size of the rendered file, this did not appear to be the case to some extent. Whether the contributing factor was resolution or something else, is of course another thing, but here goes:

If I rendered a 1280x720 clip as both 1280x720 and 1440x1080, there would not be any change in output file size. But, if I took a larger part of the video containing various sources, the size of the file varied between the 1280, 1440 and 1920 resolutions (bit sizes used 25, 25, 16 for max, average and minimum respectively). Additionally, an upper field first render was invariably larger than a progressive one. (This was not the case when there was only one source in the rendering test. )

I tested this by rendering the exact same 2:48 sequence in all three resolutions, upper and progressive. In the sequence, 1280x720 progressive sources predominated, with about 30 % being from a 1920x1080 interlaced source.

It turned out that the 1920x1080 60i upper field first template rendered the smallest file, allowing me to fit the material with animated buttons and 24 bit audio onto two 25GB discs. With the 1440 template, I would have had to give up on the animated buttons. (And DVDA did not accept a 1280 render (with the 1440 template modified) without recompression.)

Another interesting finding is that Windows shows different lengths for these renders, even though the loop area rendered remained the same: the 1280 render (504 megs) is 2:49, the 1440 render (503 megs) is 2:48 and the 1920 render (492 megs) is 2:45 minutes, though really they cover the same sequence.

Anybody any ideas why this might be the case?