Keep DTS or convert?

[hogger129]

When you guys are making backups, do you keep the lossy DTS tracks off of your discs, or are you converting the lossless tracks to AC3? I'm just curious which I should do. If I am giving myself a ~8.5GB limit, should I just keep the lossy DTS 5.1 tracks or should I convert the DTS 5.1 to AC3 5.1? For titles that have Dolby TRUEHD, I am keeping the AC3 5.1 core track. My playback hardware can support both lossy DTS and AC3. What do you guys think?

[newpball]

I wonder how you think the lossy/lossless part is relevant?

Both DTS and AC-3 are lossy.

[hogger129]

Well... I'm just curious if you would keep the DTS 5.1 or AC3 5.1 tracks that are included, or if you would convert the lossless to lossy and keep it as AC3 5.1 across the board.

[newpball]

... convert the lossless to lossy...

There seems to be a disconnect here.

With respect to surround sound, I am seldom impressed!

Most of the time the movie is shot with mono/stereo tracks and the surround engineer proddles something together that is supposed to make it into full fledged surround.

[Cornucopia]

Again, @newpball, you don't know what you are talking about. Surround sound production is a mature field, pros having experience in doing it since ~'80s. Most movies these days that SAY they are in surround are fully mixed in surround (often with binaural/HRTF/psychoacoustic assist). Just like most music recordings that say they are in stereo are fully mixed in stereo. Even though RARELY are they recorded that way - for reasons that would be obvious to you if you did it often enough (probs with: phase cancellation, wet ambience conflagration, directional steering). Maybe you are confused in how all those processes work - it's known as multi-mic recording & mixdown. Regardless, you are again moving things off topic. And there is no such word as "proddle" (except as a stupid recent mashup having no relation to this discussion).

Back ON TOPIC, @hogger129, you have to decide what is your priority - (small) size? (high) quality? consistency? If it were me, I would leave the tracks alone (keeping the original rip format) and who cares what the final size ends up. If you HAVE to keep the size down (to fit onto a DVD9 limit for example, which BTW is 7.95GB not 8.5), you should convert to the format(s) that makes the most sense to your selection of players, and you should only convert once and at the highest bitrate allowable within that limit.

Lossless to lossy might be a more natural and obvious degradation choice. No matter: Avoid reconversion to lossy formats if at all possible. Is it possible? Only you can answer that.

Scott

[Kerry56]

hogger129
If you are starting with commercial Blu-ray with DTS HD MA audio and compressing down to BD-9 size, your video quality is going to suffer if you keep the lossless audio. It takes an enormous percentage of that space. Same thing for Dolby TrueHD.

I only keep the HD audio streams intact if I am making backups to fit 25gb size recordable discs. And not always then. It depends on the importance of the movie soundtrack to some extent. The HD sound in Interstellar is worth keeping...the talk, talk, talk of Before Midnight, not so much.

[hogger129]

I guess what I should ask is should I keep the core tracks (DTS, AC3) or always convert the losless tracks (DTS-HD MA, Dolby TrueHD, LPCM) to AC3?

[newpball]

Surround sound production is a mature field...

I remain to be unimpressed!

Running panned mono tracks through some high quality convolution reverb would give far less worse surround than your run off the mill (throw in a fumbly paper prop in the back channel once in awhile and boost that bass) 'surround' production.

[Kerry56]

I guess what I should ask is should I keep the core tracks (DTS, AC3) or always convert the losless tracks (DTS-HD MA, Dolby TrueHD, LPCM) to AC3?

Cornucopia has already answered that for the most part. It really depends on the capabilities of your equipment and how much space you are willing to use for the audio. The core audio will be fairly high bitrate...probably 640kbps for the AC3.

[Cornucopia]

@newpball, Why don't you for once stick to the topic and be helpful to the OP!!??

@hogger129, I already said: that is ultimately YOUR CALL.

Backups of what? - DVDs? BDs?
Is this size limit clear-cut? Why?
Lossless tracks are, by their very nature, VBR - do you need CBR (as most lossy tracks are)?
What apps/devices are you going to be playing these with (both now and in the near future)? This helps determine what you should keep.
Figure out WHY you are making your choices, prioritize them, and then follow those based on confidence in your own decision.

Scott

[newpball]

@newpball, Why don't you for once stick to the topic and be helpful to the OP!!??

I did, he asked what other posters think, I answered!

I gave him my opinion that surround is mostly a waste of time!

Apparently you do not like that opinion, fine no problem with that! But it seems you want to go further and like to prevent me from even stating it!

I suppose I must apologize I did not know that the North Korean Oversight Committee on Sound Engineering Speech was objecting.

Sorry, sound engineers are the best people in the world, even their farts smell good!

Better?

[fritzi93]

Well, to answer the question directly, I always convert to DD 5.1 at 640 kb/s.

1) Although I have a reasonably good 5.1 setup in my HT, I can't tell the difference.
2) Compatibility. I sometimes play movies from hard drive on the living room TV's onboard player. The TV will pass through DD 5.1 intact to soundbar and sub. The subwoofer is the thing there.
3) It's smaller, which is not very important to me, as I do crf encodes. But if you encode to target size it may be.

Other preferences are just as valid. Whatever suits you and your hardware.

[hello_hello]

There was a TV with a built in media player in this house without support for DTS audio (it's optional for Divx certified HD players) and due to licensing issues my Android video player (MXPlayer) dropped DTS support a while back, so that's two reasons for converting it for me.

If you have a receiver that decodes DTS and AC3 you might want to stick to AC3, but I don't worry about that sort of thing as I'm not a fan of surround sound myself anyway, although after agreeing with newpall about something, albeit for different reasons (I simply prefer stereo) I might have to re-evaluate my view of the universe and everything in it.

So I convert DTS to AAC. I use variable bitrate, quality based encoding, but I do keep the multi channel audio as multi-channel. You never know, If I can agree with newpall about something there's a chance one day I mightn't dislike surround sound too, so I keep it that way. I mainly use AAC rather than AC3 as it should offer the same quality at lower bitrates. For an average movie with 5.1ch audio you can probably save around 1GB in file seize compared to keeping the DTS. If there's only AC3 (no DTS) I just keep the AC3. At the quality settings I use for AAC it doesn't reduce the bitrate enough to warrant re-encoding, in my opinion. Not unless you always mix multi=-channel down to stereo for encoding. You might save a couple of hundred MBs but when the average 720p encode is probably close to 4GB anyway it's doesn't seem enough to matter. Of course lower quality AAC settings might change that.

I you're interested, the different AAC encoders have different ways of setting a variable bitrate quality (different command line options) and these are the "qualities" I use with the various AAC encoders. NeroAAC -q 0.50 (it's the default), Apple/Quicktime AAC -V 91 (foobar2000 says the default is -V 63) Winamo/FhGAAC --VBR 5 (the default is 4) and for FDKAAC I use -m 5 (the default is 4).

I used the NeroAAC encoder elusively for years, and to be honest the main reason I settled on those quality settings is they result in roughly the same bitrates as Nero's Q0.50 At least for stereo. Multi-channel seems to vary a bit more. For FDKAAC it was hard to find the right quality setting to match bitrates with the other encoders, but I discovered after some messing around that it seems to be due to some odd choices for low pass filter frequency. By default, it's 15kHz for -m 4 (I'm pretty sure) and there's effectively no low pass filter for -m 5, so I had a look at what LAME does with it's VBR presets, and it uses a low pass filter of 18.5kHz, which seems reasonable to me given I doubt I've been able to hear anything higher than that for many years (if I ever could) so I tried using that and it did the trick. The bitrate for quality 4 increased while for quality 5 it decreased and there was no longer such a huge bitrate difference between the two quality settings. My quality 5 FDKAAC command line for foobar2000 is this:
--ignorelength -S -w 18500 -m 5 -o %d -

Also FYI, I switched from NeroAAC to AppleAAC when the "no delay" option was introduced. All lossy encoders add a delay (it's effectively silence called "padding"). For NeroAAC it's around 50ms. Audio players know to ignore it but most muxing programs are oblivious (MKVMergeGUI compensates for it) so it changes the audio/video sync a little. AppleAAC has a "no delay" option with zero padding. The help file says it might result in a pop or click at the beginning, but I suspect that might only apply to music tracks where there mightn't be any silence at the beginning. Movie/TV soundtracks invariable have a second or more of silence and I'm yet to hear a pop or click.
--ignorelength -s --no-optimize --no-delay -V 91 -o %d -

Don't use ffmpeg's AAC encoder. Apparently it's not very good.

If you're using an encoder GUI that doesn't supply all the necessary files, here's a post explaining where to get the various AAC encoders and any files required to use them:
https://forum.videohelp.com/threads/370412-Looking-for-an-AAC-LC-encoder-from-Frauenhof...=1#post2376348

[hogger129]

I've just decided to stay with AC3 5.1 at 640kbps for the best space savings balanced with quality. Question. I'm using BDtoAVCHD to do my re-encodes. Why would the program tell me that the video bitrate is higher if I leave the 24-bit FLAC track (which I re-encoded from the lossless TrueHD) as opposed to re-encoding the FLAC track to AC3 5.1 @ 640? Both are 5.1 so it's not like BDtoAVCHD is converting it to stereo. I noticed similar issue with Vidcoder in its video bitrate estimates. That can't be right.

[lordsmurf]

I hate the dynamic range of DTS, and it's useless to me.
I actually prefer stereo sound, coming from the TV, and do not want to be immersed in boom-booms all around me.

[hech54]

I hate the dynamic range of DTS, and it's useless to me.
I actually prefer stereo sound, coming from the TV, and do not want to be immersed in boom-booms all around me.

I agree. Never had any interest in 5.1 in general and probably never will.....until I'm forced to....hopefully I'll be dead by the time TVs no longer come with speakers.

[Cornucopia]

Well, I think you are both missing out on a whole lot of cool stuff, but I guess we can agree to disagree.

Regardless, that's a bit OT, since the OP isn't asking about whether surround is good or not, but which 5.1 file format makes the most sense to keep.

Scott

[fritzi93]

I shouldn't have said I can't tell the difference between DTS and AC3, because sometimes I can. But only when repeatedly comparing the two on the same movie, and adjusting volume to match. Explosions and the like are, dunno how to describe it... more dramatic? Or take the opening fanfare from the Star Wars movies; in the DTS sound track the horns sound, hmm, a little "brassier" to me. Is this the difference in dynamic range?

There is one difference easily detected between DTS and AC3. I've noticed DTS always sounds louder than an alternate AC3 track, or a re-encoded AC3 audio track, with the receiver at the same volume.

What's the deal with that anyway? At any rate, I don't think DTS sounds better. JMO.

[El Heggunte]

But today it's very easy to compare AC3 with DTS or vice-versa:

a) copy an audio-CD track to your HDD;
b) encode to AC3 with Aften (or FFmpeg, or a Dolby-certified encoder) at 256kbps;
c) encode to DTS with ffdcaenc at 468.5625kbps;
d) LISTEN;

[Cornucopia]

But to do that comparison correctly (apples to apples), one needs to use similar bitrates and nearly equal output volumes.

The difference in AC3 and DTS volume can probably be attributed to the AC3 use of DialNorm, which when used correctly, restores & levels the audio to an acceptable average. Unfortunately, PC players often ignore it or interpret it incorrectly, or apply it ONLY to AC3 without a similar feature in DTS, So that makes DTS sound louder and AC3 sound quieter. And as we all know, LOUDER is almost universally interpreted as better.

Scott

[newpball]

And as we all know, LOUDER is almost universally interpreted as better.

Especially, and shamefully, with mastering engineers!

And no, louder is not better.

[El Heggunte]

FWIW: Aften, by default, sets "dnorm" to -31dB, and "dynrng" to none;
the user may prefer to set the bandwidth to -w 60 though;

as for ffdcaenc: it knows nothing about dialog normalization and dynamic range;

regarding bitrates: yes, I've already done the math for the brave guinea pigs

[Cornucopia]

And as we all know, LOUDER is almost universally interpreted as better.

Especially, and shamefully, with mastering engineers!

And no, louder is not better.

Dude! Did you not read what I wrote?!!

INTERPRETED
I never said it WAS better.

This is a documented and impirical fundamental of psychoacoustics. Look it up.

And stop shooting the messenger. Mastering engineers are mostly just facilitators. The shame is that it is being demanded of them by artists, producers, and mostly by a*s*s*h...* record executives (who pay the bills only when they get what they want).

I know this from multiple personal experiences. Unless you have direct experience yourself, you have no truthful basis in spouting such harassing drivel.

Scott

[newpball]

I never said it WAS better.

And nothing I said implied that.

Mastering engineers are mostly just facilitators. The shame is that it is being demanded of them by artists, producers, and mostly by a*s*s*h...* record executives (who pay the bills only when they get what they want).

I do not deny there is sometimes pressure from those groups.

But on the other hand mastering engineers en masse didn't protest much either the last 15 years and neither did broadcasters. It was only after some mastering engineers and artists brought it to mainstream attention that most mastering engineers became "concerned".

And no, I disagree that they are just facilitators, they have a responsibility and a reputation. Doing a botched job and hiding behind "I just did what I was told" is of questionable value to me.

Who is responsible? I think it is a mix but definitely the mastering engineers are not totally without blame either.

[lordsmurf]

"Louder is better" is not any more/less stupid than "more resolution is better".

(And yes, both are stupid.)

[hello_hello]

That was a reasonably impressive back-pedal. From mastering engineers almost universally interpreting loudness as better, to mastering engineers bringing the problem to mainstream attention, to mastering engineers becoming "concerned", to mastering engineers not being totally without blame....

And reversing such a large distance within the space of two posts without any admission the original statement was a ridiculous generalisation..... that must take some practice.