stereo vs surround, your views.

[futon88]

Thank you for this, I will try it!

More questions:

1. If the entire centre channel will be muted, is it still necessary to import it?
2. Do I still need to manually pan the channels left/right before exporting? I'm guessing yes, as I hear the track from both sides when I preview it in Audacity.
3. What format do I use for export? Does .wav still work, or should I use one of the ffmpeg options (M4A or AC3)?

 

[Hal9000]

Thanks, Possessed, for that breakdown. Let me breakdown a hypothetical scenario to see if I understand.

Let’s say I want to edit from a 2.0 DLII mix of Star Wars, so I rip it and do what is needed with it. Right there within the stereo source material’s audio is all the intrinsic (non-metadata) information necessary to make its way back into DLII output. AND, if I even had a 5.1 source I wanted to make use of, I can go ahead and edit the 2.0 and 5.1 tracks (and monitor it in stereo). Then, when it comes time to encode the final output, targeting DLII stereo ought to deliver a pretty decent product. (And the new portions from a 5.1 source ought to be encoded fine by the algorithm on its way down to stereo.)

Pretty nifty; I almost wish I had a non-arbitrary task that would prompt me to make use of it.

 

[DigModiFicaTion]

I could use a bit of help in this area. Sorry if this feels like a thread hijack, but maybe it will help others that are new to editing with 5.1...

There are a few areas in my current edit-in-progress where I've wanted to maintain all audio except the centre track. I'm thinking I might have gone about this the wrong way. Here's what I did:

1. Open all five mono tracks in Audacity (all but centre): FR, FL, BR, BL, LFE
2. Pan the tracks either left or right according to name (but leave LFE alone).
3. Export as 16 bit WAV

I'm concerned this process has essentially created a stereo mix, and it will be obvious to the viewer when the movie shifts from this audio to the original 5.1 DTS, even if the ultimate output from my NLE is still 5.1. What effect will this have for the listener?

I noticed today Audacity has the option of exporting as M4A (AAC). Is it just a matter of repeating what I did but exporting as that instead of .wav, or will this approach always leave me with a two channel mix?

Audacity would only be used to drop your source audio into, then encode as pcm wav files for editing. Because they would export as individual mono files, you don't need to pan them (that happens in your NLE).

Some NLE's such as Vegas don't have compatibility with AC3 or dts audio so it's either a lossy aac 5.1 mix or individual lossless wav files. I've also seen that se NLE's, while having the ability/compatibility for 5.1, don't actually offer an easy interface for editing 5.1 audio. I've often seen/experienced 5.1 audio showing up as a single stereo audio channel.

I think there are two different ideas being discussed.
1) listening to stereo vs surround
2) editing in stereo vs surround

I watching everything on a stereo setup, but I personally think it a waste of time to edit stereo. Stereo is extremely limiting when it comes to editing.

 

[Possessed]

Thanks, Possessed, for that breakdown. Let me breakdown a hypothetical scenario to see if I understand.

Let’s say I want to edit from a 2.0 DLII mix of Star Wars, so I rip it and do what is needed with it. Right there within the stereo source material’s audio is all the intrinsic (non-metadata) information necessary to make its way back into DLII output. AND, if I even had a 5.1 source I wanted to make use of, I can go ahead and edit the 2.0 and 5.1 tracks (and monitor it in stereo). Then, when it comes time to encode the final output, targeting DLII stereo ought to deliver a pretty decent product. (And the new portions from a 5.1 source ought to be encoded fine by the algorithm on its way down to stereo.)

Pretty nifty; I almost wish I had a non-arbitrary task that would prompt me to make use of it.

Yes, as long as it's all being encoded to pl2 at the same stage. I'm not sure how to word this. So basically as a matter of source material yes they would be compatible, I would just be careful about editing with both 2 channel and 5.1 channel audio together in the same timeline unless your editor has a built in pl2 encoder, because what I would worry about is If you are rendering out at 5.1 to downmix later the nle might do some weird upmix to the 2 channel audio to get it to be 5.1 for the export that might ruin it. If that's the case you could always pre-mixdown the 5.1 track before editing. Since the data is in the audio itself you could use a lossless format so no quality would be lost.

However if your editor would simply contain the audio in the 2.0 portions to the front left and right and doesn't try to expand it or anything, *Hypothetically* yes that should also be fine, although I have never tried such a scenario. But I don't see why it wouldn't work...

 

[Hal9000]

Weird hypothetical, agreed. But thanks for clarifying, and I think I understand.

 

[Possessed]

Weird hypothetical, agreed. But thanks for clarifying, and I think I understand.

Well, not all that weird. When I made my Episode IV fanedit (well more of a hybrid of different versions with a unique color correction and some enhanced vfx and all blasters and sabers redone... Sort of a diet adywan but in 1080p if you will) I was basically in this scenario across a broad variety of sources. The main base was the 1997 mix which was 5.1, but it also revolved based on the scene and which sounded base. Some was the original 77 mix, some was the 93 mix (which was the go to for OOT content because it by far has the best quality but there were one or two instances that I liked the 77 mix better artistically), and some was adywans mix which is 5.1 (mostly just scenes with Vader dialog since adywan had enhanced it to sound more like the other two movies. I could have done that myself but why fix what isn't broken, he did a great job so might as well save time since we're a community right?). Since there weren't any like actual "edit" edits I didn't need to edit in 5.1 so I just did the downmixing of the 1997 mix and adywans mix (which is based on the 2004 mix) beforehand and eq'd them all to match and then applied some slight processing in post to make it pop a bit.

And then for esb it's a hybrid of the 2011 sound mix and the 93 mix and again there weren't any deep cuts so I just pre downmixed the 2011 mix and matched them up and went for it.

However for my return of the Jedi edit it was actually much more of an "edit" with significant cuts so I needed the surround channels so I edited that in surround and mixed down after. But since it was going to be mixed down I didn't really have to worry about the surround channels as long as there wasn't anything sticking out that you could hear in stereo, because the downmix and then re upmix would automatically smooth that over, so to speak. There is exactly one cut in that I wish I would have been more careful about but it's not a big deal and nobody has pointed it out yet so I won't ruin it for anybody and then I downmixed after.

 

[Adabisi]

Just my opinion on this: Unless you absolutely, 100% know what you're doing, stay with stereo and don't try 5.1 audio. There are an annoying number of fully approved edits from experienced editors out there that advertise "5.1 audio" but are really just expanded stereo.

 

[DigModiFicaTion]

Just my opinion on this: Unless you absolutely, 100% know what you're doing, stay with stereo and don't try 5.1 audio. There are an annoying number of fully approved edits from experienced editors out there that advertise "5.1 audio" but are really just expanded stereo.

Do you leave reviews for this large quantity so people know?

 

[Adabisi]

Do you leave reviews for this large quantity so people know?

Of course.

 

[Possessed]

Care to name some examples? I haven't encountered any yet, which is odd if the number of them is annoyingly high, but it would still be good to know because it would probably annoy me if I was watching them too.

 

[Adabisi]

Care to name some examples? I haven't encountered any yet, which is odd if the number of them is annoyingly high, but it would still be good to know because it would probably annoy me if I was watching them too.

Specific to 5.1 presented as expanded stereo:

Review for Split Unbreakable Glass, - Fanedit.org

I can't say I'm a big fan of the trilogy this edit is based on, but it definitely had some neat ideas. Overall, I enjoyed this edit, but it's not without flaws. 3.5 hours is a lot to ask of your audience. I think this story could have been cut down even more - not every scene felt vital to me...

ifdb.fanedit.org

Review for Se7en: The Sinful Cut, - Fanedit.org

I was really excited for this edit because the almost identical one I made over a decade ago would be great with an HD upgrade. I was glad to see Bobson Dugnutt took the reigns and I'd love to recommend this as a replacement for my original edit, but there are some serious technical issues with...

ifdb.fanedit.org

General issues with 5.1 mixes:

Review for Star Wars: Episode VIII - The Last Jedi: Anti-Cringe-Cut, - Fanedit.org

I hated The Last Jedi. The whole thing was a mess. I was hoping for a fanedit that fixed it and this is it. I loved all the narrative choices - the movie is improved tenfold. It does get a little clunky during the final battle, but the editor did a good job with what he had. The only additional...

ifdb.fanedit.org

Review for American Gangster: Two Tales, - Fanedit.org

I'm a huge fan of removing extraneous elements from gangster films. Years ago, I tried to edit American Gangster to remove the police angle and focus on Frank. Every which way I tried, I couldn't get it to play in a way that I liked it. I even tried something similar to what L8wrtr pulled off...

ifdb.fanedit.org

 

[Possessed]

I see. Interesting. So by expanded stereo you're saying that instead of it either being an actual discrete mix (or as some of those reviews say it is discrete but the channels aren't assigned properly) or when it has to use a 2.0 source it just mirrors the front channels into the back instead of upmixing them properly? Because yeah that would ruin an experience for me as well

 

[Adabisi]

I see. Interesting. So by expanded stereo you're saying that instead of it either being an actual discrete mix (or as some of those reviews say it is discrete but the channels aren't assigned properly) or when it has to use a 2.0 source it just mirrors the front channels into the back instead of upmixing them properly? Because yeah that would ruin an experience for me as well

Yes, that's essentially it. A dead giveaway is when you hear dialogue coming out of the rear speakers. Unless a character is off screen, pretty much all dialogue in a proper 5.1 mix should be coming from the centre speaker. Depending on the scene, or how it was originally mixed, this can carry over to front L and R, but they shouldn't be talking behind you.

 

[Possessed]

I would hope most people know what it should be, and just goofd.

 

[GelflingHand]

If the entire centre channel will be muted, is it still necessary to import it?

It's necessary to have a silent channel in that position; you could create it manually, but I don't see any reason to do that. The source should be a single 5.1 file, not six separate mono files; Audacity can probably extract it directly from the video file.

Do I still need to manually pan the channels left/right before exporting? I'm guessing yes, as I hear the track from both sides when I preview it in Audacity.

No. The order of the channels detemines which speaker each goes to. You need to export it in order to listen to it as 5.1; within Audacity it's just six mono tracks.

What format do I use for export? Does .wav still work, or should I use one of the ffmpeg options (M4A or AC3)?

Not sure; give .wav a try and see what happens.

 

[futon88]

It's necessary to have a silent channel in that position; you could create it manually, but I don't see any reason to do that. The source should be a single 5.1 file, not six separate mono files; Audacity can probably extract it directly from the video file.


No. The order of the channels detemines which speaker each goes to. You need to export it in order to listen to it as 5.1; within Audacity it's just six mono tracks.


Not sure; give .wav a try and see what happens.

.wav worked, but the 4GB limit produced a cut-off, so I exported it as an .m4a file instead (AAC); six mono channels (no panning) with nothing mapped to #3 (centre). I added that to my NLE in place of my two channel soundtrack, and when I export the entire thing as 5.1 audio it sounds...perfect? I don't have a surround setup to confirm, but I can't tell the difference between the bluray and my exported file, so I think I'm on track. Thanks for your guidance.

 

[addiesin]

.wav worked, but the 4GB limit produced a cut-off, so I exported it as an .m4a file instead (AAC); six mono channels (no panning) with nothing mapped to #3 (centre). I added that to my NLE in place of my two channel soundtrack, and when I export the entire thing as 5.1 audio it sounds...perfect? I don't have a surround setup to confirm, but I can't tell the difference between the bluray and my exported file, so I think I'm on track. Thanks for your guidance.

It sounds like you achieved what you were trying to, but if you want to keep something in the center channel like music or sound effects, but get rid of dialogue, you can try running the center channel as an individual WAV through an online AI voice isolator/separator and see if the result is useable. Then swap that one in instead of a muted one, if it works to your liking.

 

[Hal9000]

If I rip a DLII stereo source and edit it, would encoding to an AC3 and checking the box that it has been surround encoded be sufficient for a DLII-compatible output?

 

[Eyepainter]

After reading all of this talk, I decided to try out Possessed's idea on a project that I've been working on the past few weeks and see how Dolby Pro Logic II went. The project was just over three hours in length, and for that reason, I needed to cut corners and look for ways to make the file size smaller, making it a perfect candidate for this sort of idea.

Here's the boring stuff. Since I use DaVinci Resolve, I exported the edit in GoPro Cineform with a 16-bit 5.1 Linear PCM Audio track. Then, I launched Handbrake, converted the Cineform file to an .m4v video file, and I changed the 5.1 LPCM track to an AAC track with a mixdown to Dolby Pro Logic II, a bitrate of 448kbps, and a 48khz sample rate. I then took the converted file, put it on a USB flash drive, and watched the edit via my blu-ray player on a 1080p HDTV. I have no home theatre setup, so the audio is just from the TV.

As I watched the edit, I thought I heard the sound of horses whinnying behind the room. I live in a city where horses are uncommon. I rewound the scene again, and sure enough, the sound of the horses actually came from the edit itself! I was also stunned when I heard the sound of wooden boards over on the right side of the room clunking down to the ground after a big explosion. It was the first time I ever heard those boards in the film and an extra detail I never noticed before.

So, in a nutshell, 10/10 would convert again.

 

[Hal9000]

If the sound came from the TV, it’d be playing back from two speakers, right? I don’t think you’d benefit from DLII or 5.1 with only two speakers.