ac3 vs HD or Uncompressed audio Is there a Difference?

[waynezo]

When I first started using BDRB I kept HD audio. Now I let BDRB convert everything to ac3 640kbs. I read an article that said the human ear can't tell the difference from HD, uncompressed, or compressed audio.

What are your thoughts on this?

[k-c-ksum]

use your ears

[jdobbs]

I can only give my opinion -- and I'd be surprised if you could tell the difference, and even then I'd suspect it was only because you knew that it had been converted.

There are lots of people who are enthralled with HD audio -- but to me it is more of a marketing gimmick than anything else. They had to use that extra Blu-ray BD-50 storage space for something -- and it helps sell players.

Again, that's just my opinion.

[Capsbackup]

For me, it is a visual thing... I see my receiver says DTS HD Master or True HD verse Dolby Digital. It's for sure my ears can't tell! :p
For BD-25 backups that have 4 hours or less total content, I usually just keep the HD audio, since space and bitrate is no concern. ;)

[waynezo]

I can only give my opinion -- and I'd be surprised if you could tell the difference, and even then I'd suspect it was only because you knew that it had been converted.

There are lots of people who are enthralled with HD audio -- but to me it is more of a marketing gimmick than anything else. They had to use that extra Blu-ray BD-50 storage space for something -- and it helps sell players.

Again, that's just my opinion.

I agree! It is Great American Marketing. I almost upgraded my Kenwood reciever (.007 total harmonic distortion) since it doesn't support HD Audio. Now I Think I will keep it and not miss a thing.

[Video Dude]

What would you guys suggest for movie only BD25 backups?

Keep DTS-HD MA / True HD and reencode video
or
Select AC3 640 Kbs / DTS core and do not reencode video

[Sharc]

Not re-encoding video saves ou a lot of time.
AC3 640kbps is excellent (IMHO).

[soneca]

Low frequency response more articulate ...
Response medium that makes the sound more transparent ...
Better separation between channels ...
What bullshit, just listen carefully!
Compression ...
Of course, Dolby wins clearly using his algorithm that achieves the same sound quality for less than half the space occupied by a track in DTS.
My poor English ...:eek:

[SquallMX]

AC3 at 640 Kbps is good enough... if you use an Dolby Certificated Encoder (Vegas, Soft Encoder, etc), Aften is a little worse and still has decoding issues with some decoders.

[JJB]

AC3 at 640 Kbps is good enough... if you use an Dolby Certificated Encoder (Vegas, Soft Encoder, etc), Aften is a little worse and still has decoding issues with some decoders.

Can somebody tell me what this means, I use BD-Rebuilder and and do not care what other encoder's do. What does this mean if BD-Rebuilder is my source of BD backups and uses Aften.

[SquallMX]

Can somebody tell me what this means, I use BD-Rebuilder and and do not care what other encoder's do. What does this mean if BD-Rebuilder is my source of BD backups and uses Aften.

Is nothing important, Aften (the ac3 encoder used in Rebuilder) is not as good as an official ($$$) encoder, the difference is minuscule, so don't worry, I was just nitpicking :rolleyes:.

[JJB]

Is nothing important, Aften (the ac3 encoder used in Rebuilder) is not as good as an official ($$$) encoder, the difference is minuscule, so don't worry, I was just nitpicking :rolleyes:.

Then my question is ...why bring it up at all if most of us can't afford the $$$$ encoders? Should not we want the best?;)

[jdobbs]

Aften is an excellent encoder and follows the standard, certified or not. I would recommend encoding at 640Kbs (for 5 channel), though. 448Kbs sounds good (and is the standard for DVD) but 640Kbs appears to be the point where all the tests and reviews I've read say that it can't be distinguished from any of the HD sources or DTS. Lots of people say they can hear a difference when they know which source is which -- but none of them can seem to tell HD from DD when a volume compensated source is presented in the blind.

Honestly, though, it's been a while so my reading is a little dated -- but I doubt that the human ear has evolved much in that time.

[A.Fenderson]

AC3 at 640 Kbps is good enough... if you use an Dolby Certificated Encoder (Vegas, Soft Encoder, etc), Aften is a little worse and still has decoding issues with some decoders.

Is nothing important, Aften (the ac3 encoder used in Rebuilder) is not as good as an official ($$$) encoder, the difference is minuscule, so don't worry, I was just nitpicking :rolleyes:.

Have you noticed audible quality differences in a source encoded 640 kbps Aften vs a certified encoder, same source?

Also, what "decoding issues" are you referring to? I'm trying to decide on a standard lossy audio solution myself when using BD-RB (http://forum.doom9.org/showthread.php?t=158960), so as to maximize available space for video, but I don't want to have to worry about having decoding issues if I change receivers (not necessarily "upgrade," as I'm looking for an older Yamaha DSP-A1 (http://usa.yamaha.com/products/audio-visual/hifi-components/amps/dsp-a1_black__u/?mode=model)or similar).

If I decide to use only AC3 640 Kbps for all sources (including those with <640 Kbps AC3 tracks in the original), do I just deselect all the current audio encoding options, as below?
OFF: Do not convert DTS to AC3
OFF: Do not reencode AC3
OFF: Use 448
OFF: Use 192
OFF: Keep HD Audio

If there's already an AC3 640 stream present in the source, under these settings, does BD-RB discard it and reencode to 640 Kbps AC3 when certain lossless options are available? If so, I guess I'll just have to manually toggle the "Do not reencode AC3" setting for those sources to save the trouble, in that the Aften-produced files are generally considered equal or perhaps slightly inferior (quality and/or compatibility) to studio-produced 640 Kbps AC3 streams.

Also, when reencoding AC3 640 from other sources, can BD-RB make use of DD-THD and LPCM? When DTS-HD (MA/HRA) is encountered, it's just decoding the DTS core and then encoding AC3 from that?

[scotthulbs]

I wondered this myself, so rather than listening to what I read online I tested. I used the first Transformers movie. I had the movie on HDDVD and my buddy had it on Bluray. The HDDVD version has the audio in DD+ and the Bluray is TrueHD. I had both the HDVD player and BD player hooked up to my receiver and played the same few scenes a few times switching back and forth between the same scenes on the different players. I came to the conclusion that neither sounded better than the other, but there was definitely more to hear in the TrueHD track. I specifically remember the scene in the beginning where the helicopter thrashes the base when the glass in the air control tower blows you could hear a lot more. So in my experience its not necessarily better sound quality but just more sound information. I keep the HD tracks on movies like Transformers, Ironman, Dark Knight, ect. Most movies though I am more than satisfied with DD. I think you should give it a try you would be surprised at what you might be missing.

[jdobbs]

I wondered this myself, so rather than listening to what I read online I tested. I used the first Transformers movie. I had the movie on HDDVD and my buddy had it on Bluray. The HDDVD version has the audio in DD+ and the Bluray is TrueHD. I had both the HDVD player and BD player hooked up to my receiver and played the same few scenes a few times switching back and forth between the same scenes on the different players. I came to the conclusion that neither sounded better than the other, but there was definitely more to hear in the TrueHD track. I specifically remember the scene in the beginning where the helicopter thrashes the base when the glass in the air control tower blows you could hear a lot more. So in my experience its not necessarily better sound quality but just more sound information. I keep the HD tracks on movies like Transformers, Ironman, Dark Knight, ect. Most movies though I am more than satisfied with DD. I think you should give it a try you would be surprised at what you might be missing. The only thing you can get with HD that isn't in DD would be the 7.1. The DD encodes are 5.1. To me 7.1 is marketing gimmick as well. Fact is that anything over 4 speakers (quadraphonic) is marketing. You can balance 4 speakers to make a sound avalable at 360 degrees -- and about all you might add would be a subwoofer (so all the speakers don't have to be large enough to handle very low signals).

[Ghitulescu]

Most people (those having regular setups) won't feel the difference between [correctly encoded] DD and TrueHD. Then comes the question of money: whether it's worth spending 2000+€ for a very tiny difference (remember, people that could afford spending such an amount of money do not have the hear of those who do not have this money ;)).

It's [almost] like the issue of 24b vs. 16b - very few "regular" gear have the noise floor so low that 16b were not enough to cover it. Most of them don't even manage 12b.

[scotthulbs]

Ghitulescu: I don't see why it would cost 2000+ to hear the difference. Like I said I was hearing more info from the trueHD. I guess if you have bose cube speakers you might not hear the extra info, but if you had even a modest setup you should hear the extra. Its not a huge difference I'll agree with that. It only cost me pennies to keep the HD audio. I mean how cheap are hard drives now 3-4 cents per GB. So for me its worth the extra few cents to know I am getting the best sound I can. To each their own

[juhok]

scotthulbs: It's quite possible that the source for HDDVD and Bluray was mixed differently. So it might not be a very valid test to begin with. This happens all the time with audio CDs for example. Different releases of the same record have different mixing/mastering used.

[waynezo]

The mix can make a huge difference.

Check out this article, you may be happy with ac3.

http://www.hemagazine.com/node/Dolby_TrueHD_DTS-MA_versus_Uncompressed_PCM

"Between the 640 kbps Dolby Digital and the uncompressed, the difference was even less noticeable. Enough so that most people, even those trained to listen for it, probably won’t be able to hear the difference."


"If you’ve been listening at home and are sure you can hear a difference on your favorite discs, be wary. There is absolutely no way to tell that compressed and uncompressed tracks on any disc have anything to do with each other. They could come from different masters, they could be mixed differently, or any number of other variables that makes an in-home test, unfortunately, impossible. That said, trust your ears, and go with the one that sounds best to you."

[Ghitulescu]

Ghitulescu: I don't see why it would cost 2000+ to hear the difference. Like I said I was hearing more info from the trueHD. I guess if you have bose cube speakers you might not hear the extra info, but if you had even a modest setup you should hear the extra. Its not a huge difference I'll agree with that. It only cost me pennies to keep the HD audio. I mean how cheap are hard drives now 3-4 cents per GB. So for me its worth the extra few cents to know I am getting the best sound I can. To each their own

Maybe I should have said 20000+. I can tell that there are other factors more important than lossless audio, the most important one being the speakers. What importance has that an AVR has a THD of 0.001% when the speakers have 5% (and each 0 after the comma adds 00 before the comma to the price ;)). Only people with absolute hearing can discern HD lossless from well encoded DTS or DD with enough bitrates (because they spot the errors not the content). Jitter is the second one.

To your suprise (maybe) you need to find out that Bose is not a profi firma, it's something for "audiophiles" with thick wallets, maily because they use tricks (reflexions) to recreate a "sound space" instead of providing unaltered sound. There's no Bose in the recording studios, the most expensive Bose system doesn't hold a bit against the cheapest near-field studio monitors (there are no reflections in a recording studio) at 1/10 of its price.

For the second item: I would also keep the lossless track if provided. No doubt about it. I said that the difference towards a good lossy audio is minimal (should it not be there) and mainly not sensed.

[SquallMX]

I always keep the Lossless track... after reducing its bit depth to 16 bits, in most cases the bitrate is halved, recovering around 2 Mbits for the Video. To my surprise dialog driven movies usually needs bitrates in the ranges of 1500 - 1700 Kbps (Just a little more than a Full bitrate DTS track 1536 Kbps), action ones 2000 - 2300 Kbps.

Still, I pretty sure that I can not hear the differences between AC3 and lossless.

[A.Fenderson]

I always keep the Lossless track... after reducing its bit depth to 16 bits, in most cases the bitrate is halved, recovering around 2 Mbits for the Video. To my surprise dialog driven movies usually needs bitrates in the ranges of 1500 - 1700 Kbps (Just a little more than a Full bitrate DTS track 1536 Kbps), action ones 2000 - 2300 Kbps.

Still, I pretty sure that I can not hear the differences between AC3 and lossless.

So what's your procedure? You decode to LPCM if necessary, truncate the least significant bits down to 16 (or do some kind of dithering), then reencode to DTS-HD MA or DD-THD? Are you cutting it down to fewer channels too?

[SquallMX]

So what's your procedure? You decode to LPCM if necessary, truncate the least significant bits down to 16 (or do some kind of dithering), then reencode to DTS-HD MA or DD-THD? Are you cutting it down to fewer channels too?

Almost, First i create a normal backup using BD-Rebuilder with the "Keep HD Tracks" option disabled, so i get a backup with a Full Bitrate DTS Lossy File, using the default settings the final size is around 22.5 GBs.

After that I use eac3to and the command "-down16" to create 6 or 8 WAV files from the original DTS-MA (or TrueHD) track, then I encode the files to DTS-HD MA using a core of 768 Kbps, and finally use MultiAVCHD to replace the lossy DTS with the lossless one, ending with a final size of 22.8-23.1 GB. If the output size is above a BD-R capacity i can simply delete a Bonus Feature.

Obviously this only works for non-multiangle movies, for multiangle movies is a hassle so I just keep the DTS track.

[djmasturbeat]

The only thing you can get with HD that isn't in DD would be the 7.1. The DD encodes are 5.1. To me 7.1 is marketing gimmick as well. Fact is that anything over 4 speakers (quadraphonic) is marketing. You can balance 4 speakers to make a sound avalable at 360 degrees -- and about all you might add would be a subwoofer (so all the speakers don't have to be large enough to handle very low signals).
I am not so sure i agree with this, at least the part of quad vs 5.1 ( certainly my rooms aren't big enough for 7.1 so, for me , at least it is unnecessary).
My setup these days isn't as nice as before i had kids (actually comparitively, it is garbage -- when i was single i used to own a few sets of B&W speakers matched w/ high end Yamaha and Sony ES AVR), so some of it may be due to the fact that the inferior speakers and amp i have now require the center channel and sub in order to actually produce a decent separation for dialog and other elements for movies. But i have tried unhooking cntr/sub as well as tried the quad setting on the amp; for music I like quad + sub or 2 x stereo +sub better than PLII and settings like that, but for movies it is really lacking clarity and separation. Again, this could just be the really low end consumer junk I use now.

As for double blind test, i guess the only way to do it is get someone who knows what they are doing and to re-encode using things like SurCode and DTSHDMA suite and the Dolby Suite on Mac, and compress an LPCM track to DTSHDMA, DolbyTrueHD, DD+, DD640, DTS1.5 and blind test on all of these and play each at exactly the same decibels on same system (taking into account the 4dB dialnorm boost of dts, etc). I would guess that some people will detect differences, but even among them, they may be surprised by some of the outcomes of preferences. They will sound different to some people, mostly due to channel separation, clarity, and depth, etc, but most people won't even hear that for a normal film viewing. More folks can hear the difference when the dts is cut in half (or is a 768 core), and ac3 is 448 (or even more especially so below). But often some people can be fooled by just playing even a slightly lower bitrate at a higher volume and/or mixed and mastered to overcompensate with a bit more punch and sizzle (or at least that was my experience with music audio, since i have never done any audio for a film).

In then end, in regards to the OP, i have to agree with the first commenter: use your ears. what do you hear? how good is your system -- will it ? what are you satisfied with? that is all that matters. screw others' opinions or some academic postulations, do what is good for you. do a test with same source, 1 with re-encoded or core audio and (if possible) untouched video, and one with re-encode video and untouched lossless audio, and play them back from your PC --- keep in mind that in TMT3 and 5 the HD audio is downsampled to 48/16, so if it is higher than that it won't be a true test of audio output (been awhile since using PowerDVD so i don't know if it bitstreams untouched or not, but you could download a trial and try it), anyhow it will be a bit different from many standalones that won't downsample, but it can save you burning both off first-- then you decide for yourself if there really is or isn't any need to keep the HD audio, or if you are fine with the core, etc (just do several tests with untouched video and only resampled audio at different rates -- these will take a lot less time to rebuild, you just need the drive space for multiple builds).

[Ghitulescu]

I am not so sure i agree with this, at least the part of quad vs 5.1 ( certainly my rooms aren't big enough for 7.1 so, for me , at least it is unnecessary).
My setup these days isn't as nice as before i had kids (actually comparitively, it is garbage -- when i was single i used to own a few sets of B&W speakers matched w/ high end Yamaha and Sony ES AVR), so some of it may be due to the fact that the inferior speakers and amp i have now require the center channel and sub in order to actually produce a decent separation for dialog and other elements for movies. But i have tried unhooking cntr/sub as well as tried the quad setting on the amp; for music I like quad + sub or 2 x stereo +sub better than PLII and settings like that, but for movies it is really lacking clarity and separation. Again, this could just be the really low end consumer junk I use now.

As for double blind test, i guess the only way to do it is get someone who knows what they are doing and to re-encode using things like SurCode and DTSHDMA suite and the Dolby Suite on Mac, and compress an LPCM track to DTSHDMA, DolbyTrueHD, DD+, DD640, DTS1.5 and blind test on all of these and play each at exactly the same decibels on same system (taking into account the 4dB dialnorm boost of dts, etc). I would guess that some people will detect differences, but even among them, they may be surprised by some of the outcomes of preferences. They will sound different to some people, mostly due to channel separation, clarity, and depth, etc, but most people won't even hear that for a normal film viewing. More folks can hear the difference when the dts is cut in half (or is a 768 core), and ac3 is 448 (or even more especially so below). But often some people can be fooled by just playing even a slightly lower bitrate at a higher volume and/or mixed and mastered to overcompensate with a bit more punch and sizzle (or at least that was my experience with music audio, since i have never done any audio for a film).

5.1 requires a different mixing procedure than 2.0 stereo. One cannot simply remove the center, the LFE/subwoofer from a 5.1 source and say: Now I have stereo!

Who wants stereo, s/he must have a stereo setup as well, separately from the 5.1/7.1 one.

As the proliferation of portable devices has driven the sound engineer to modify the mixes accordingly, I am pretty much convinced that a 5.1 sound engineer would mix the 6/8 channels to accomodate 4/6 small/tiny boxes (of the size of the good old stereo tweeters), a center (sometimes of the same size, but with two tweeters) and a subwoofer - chances that 4 or 6 full-range (stereo) loudspeakers are used in the average home is practically null. I've seen one full-range 5.1 system (within my purchasing power) from Onkyo and it simply "killed" the super-advertised Bose systems at 1/3 of the price (and the AVR was included, too).