Creating a simple still image movie for DVD-Audio disc

[DRP]

Looking for advice on how to author a DVD video disc for the primary purpose of the audio on it rather than the pictures. My DVD player does not support playing DVD-Audio directly, but it says it can play 24bit/96kHz audio files when included as the soundtrack of a DVD video disc.

I have a collection of 24bit, 96kHz 2-channel stereo audio files in uncompressed WAV format that I would like to be able to play on my DVD player in this manner. The problem I have is how to author such a disc with just token (or blank) video files for the picture part that I understand is required to make a compliant DVD video disc.

I am familiar with authoring DVDs using DVD-lab Pro. I also have a collection of JPEGs that I can convert into compliant MPEG-2 files to use as the video component if necessary. I have TMPGEnc Plus for the making of MPEG-2s and I have Nero for burning the end result to DVD disc. I also have discWelder Chrome II although this is useless I think without a DVD player that supports DVD-Audio discs.

Is there a guide out there somewhere for what I'm trying to do? I'm only really interested in the audio tracks, not at all in the video content, but I understand that I need to have the video in some form or another in order for my DVD player to recognise the disc and actually play it. What's the quickest and easiest way of accomplishing this preferably using the software I already have and have learnt how to use?

From the discWelder documentation this is ultimately what I want to do:

Universal DVD discs

A "Universal" DVD disc (sometimes referred to as a "Hybrid" DVD disc) is a DVD that contains both a video zone and an audio zone. That means that the replicated Universal DVD disc will play in any DVD player, whether it is a DVD-V or a DVD-A player.

Note: The two zones must be authored in separate programs, and the video zone authoring program's output, which is a set of "VIDEO_TS" files, is imported into discWelder CHROME to make a Universal disc (CHROME produces the "AUDIO_TS" file set for the audio zone). "Import VIDEO_TS" can be found by clicking on Menu item "File", then "Album Properties", then "Import VIDEO_TS".

Note: If the soundfiles in the video zone are in 5.1 surround format, they MUST be encoded into Dolby Digital (AC-3) files, as well as encoded into DTS Surround files if desired, to be included in the DVD-Video zone (stereo soundfiles can be AC-3 encoded for the video zone as well). The encoded soundfiles (called "streams") are multiplexed into the final "VIDEO_TS" file set that will be imported into CHROME. The audio content is typically encoded from the same source files as those in CHROME, although they may be encoded from different soundfiles if desired. The "VIDEO_TS" file set MUST include a Linear PCM (16-bit, 48kHz) track OR a stereo or surround Dolby Digital (AC-3) track, or both, depending on the amount of space available and the wishes of the producer.

Obviously I know my DVD player won't be able to play the DVD-A portion of the disc, but it should be able to play the DVD-V section. The audio files in question do not fill the disc or even come close to that, so there will be more than enough room to have both a DVD-V + DVD-A section, even in 24bit 96kHz resolution in both cases. I can incorporate a second 16bit, 48kHz audio track section on the DVD-V section if necessary to meet compliance requirements if what the discWelder documentation says is correct with regard PCM 16/48 or AC-3 must be included somewhere.

[racer-x]

I did this many years ago with DVD-Lab. you essentially are creating a new tittle for each song with an image. I believe that it is limited to 99 songs in that configuration, if memory serves me. It doesn't require much extra space at all though.

[DB83]

You do not, as you say, want to create a dvd-audio or even a hybrid disk.

Actually, depending on your version, Nero should be able to create a dvd-player compatable audio disk straight out of the box.

In any event, the audio HAS to be dvd-video compliant else it will have to be converted. So that means 1536 kbps, 48 MHz PCM maximum. IIRC.

[DRP]

Fantastic! A little test of concept I just did with just one track works! Mind you, that one track is only 4:26 long and recorded as both a DVD-A & DVD-V section on the same hybrid disc in 24/96 resolution has taken up some ~900MB of disc space, so at that level of data usage, I might be foregoing the whole DVD-A section entirely in order to fit the whole collection on one disc!

I found out that DVD-lab PRO does actually have the option of creating audio-only "Movies" right there built in, so I created the VIDEO_TS folder using that and then inported that into discWelder CHROME's authoring process for the DVD-Audio creation of the AUDIO_TS folder. Output as a complete ISO file and burn with Nero and the thing plays perfectly exactly as designed.

And no, you don't need any audio to be in 16/48 resolution at all. I created this test disc with full 24/96 audio in both the DVD-A & DVD-V sections with no secondary 16/48 tracks elsewhere for compliance with specs and it plays just fine on my Sony DVP-NS728H upscaling DVD player which does not support DVD-Audio playback. The display even shows on screen that it's playing 24 bit 96kHz audio as it does so.

Now all I have to do is make the screen images a bit more classy than the basic white lettering on black background I just did up quickly for the test and figure out how to import a background image into DVD-lab PRO, but very pleased I've got the audio to play without going to the expense of buying an obsolete DVD-Audio player.

[DB83]

Just check the specs using mediainfo for the vob inside the video_ts folder. I will bet my last dollar that the audio is not higher than 48 khz.

Better still. Post that mediainfo (text mode) report for both folders.

900 meg for under 5 mins is just crazy - even dvd-video at max with actual video would not go that big.

And if you do not have these folders and vob(s) you do not have a dvd. Your player might play it but you got lucky.

[DRP]

Just check the specs using mediainfo for the vob inside the video_ts folder. I will bet my last dollar that the audio is not higher than 48 khz.

Better still. Post that mediainfo (text mode) report for both folders.

MediaInfo clearly doesn't support or isn't capable of reading the audio data correctly when it's up to 96kHz because here's what it's reporting.

General
Complete name : R:\VIDEO_TS\VTS_01_1.VOB
Format : MPEG-PS
File size : 150 MiB
Duration : 4mn 26s
Overall bit rate mode : Variable
Overall bit rate : 4 743 Kbps

Video
ID : 224 (0xE0)
Format : MPEG Video
Format version : Version 2
Format profile : Main@Main
Format settings, BVOP : No
Format settings, Matrix : Default
Duration : 4mn 26s
Bit rate mode : Variable
Bit rate : 3 625 Kbps
Maximum bit rate : 9 000 Kbps
Width : 720 pixels
Height : 576 pixels
Display aspect ratio : 4:3
Frame rate : 25.000 fps
Standard : PAL
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Progressive
Scan order : Top Field First
Compression mode : Lossy
Bits/(Pixel*Frame) : 0.350
Time code of first frame : 00:00:00:00
Time code source : Group of pictures header
Stream size : 115 MiB (76%)
Color primaries : BT.601 PAL
Transfer characteristics : BT.470 System B, BT.470 System G
Matrix coefficients : BT.601

Audio
ID : 189 (0xBD)-160 (0xA0)
Format : PCM
Format settings, Endianness : Big
Format settings, Sign : Signed
Muxing mode : DVD-Video
Duration : 4mn 26s
Bit rate mode : Constant
Bit rate : 1 024 Kbps
Channel(s) : 2 channels
Sampling rate : 32.0 KHz
Bit depth : 24 bits
Stream size : 32.5 MiB (22%)

I am absolutely convinced that is not correct because the size of the video VOB is 150.4MB whilst the size of the corresponding AOB in the DVD-Audio section on the same disc in the AUDIO_TS folder is 149.4MB and the only difference between the two is a single frame still image picture with the Track Name in white text on a black background. The size of the uncompressed 24 bit 96kHz audio WAV file used as source for both is 146.2MB @ 4:26 duration as you can see above. I have no idea where MediaInfo gets 32kHz from, but it definitely isn't accurate.

I'm using MediaInfo 0.7.76 but it can't even decode anything at all about the corresponding AOB file made by discWelder CHROME II, so it probably isn't even coded to bother analysing those files at all since they're probably considered obsolete and extinct these days.

discWelder, DVD-lab PRO & my Sony DVP-NS728H player are all reporting audio at 24bit and 96kHz, so with 3 out of 4 reporting what I expect and the file sizes matching up and the source file being reported as 24bit 96kHz, I'm quite convinced that MediaInfo is just plain wrong on the VOB. Here's what it says about the source file:

<Edit> To be completely fair, DVD-lab PRO does warn you when importing the Audio assets at 24bit 96kHz that they are incompatible and non-compliant with DVD-specs and that continuing may result in an unplayable disc. However, it is not a show stopping critical error that prevents the program continuing if you choose to do so. DVD-lab Pro is good like that. It will let you continue doing what you want to do even if it's out of spec. It treats the user like an adult who knows what they're doing and gives them responsibility for their own actions. Completely unlike Apple Corp. for example.

General
Complete name : C:\01.wav
Format : Wave
File size : 146 MiB
Duration : 4mn 26s
Overall bit rate mode : Constant
Overall bit rate : 4 608 Kbps

Audio
Format : PCM
Format settings, Endianness : Little
Format settings, Sign : Signed
Codec ID : 1
Duration : 4mn 26s
Bit rate mode : Constant
Bit rate : 4 608 Kbps
Channel(s) : 2 channels
Sampling rate : 96.0 KHz
Bit depth : 24 bits
Stream size : 146 MiB (100%)

900 meg for under 5 mins is just crazy - even dvd-video at max with actual video would not go that big.

To be fair I was only looking at the resultant ISO file that came out of the discWelder compile of the combined DVD-Audio section with the imported DVD-Video section I imported from DVD-lab Pro. That ISO did come to a total 902.9MB, however the total resultant files burnt to disc has only ended up being 300MB. I have no idea why discWelder created an ISO 3x the size of the files contained within and since that software is not supported or even available anymore, I doubt I'll ever find out either.

And if you do not have these folders and vob(s) you do not have a dvd. Your player might play it but you got lucky.

Here's the relevant page from my DVD player manual about what it does with audio of various resolution:
https://dl.dropbox.com/s/gu5qyahshpwqq4a/Page%2037%20Operating%20Instructions.pdf

[DB83]

Mediainfo is not perfect but it does not make too many mistakes.

As it stands, that audio is not dvd-compliant but if your player can handle it then who gives........

[Cornucopia]

The DVD-Video & DVD-Audio specs BOTH allow for 24/96/2 in their streams, since this 4.608 Mbps easily falls under both per-stream audio bitrate limits (6.144Mbps and 9.6Mbps respectively). - (just checked the spec)

Whether or not it is truly compliant to the spec (which in this case, it ought to be), it may NOT be compliant with DVDLab Pro's authoring capabilities, and so it might very well be that DVDLP is downsampling/downconverting - yes, to 32kHz - in order to maintain its authoring procedure, when given 24/96. This is not uncommon for Authoring apps that have generating/re-encoding features, nor uncommon for apps to limit the audio quality to "common" settings, and lots of times it is more convenient for an app to say that something is "not compliant" than to say "we aren't designed to handle more complicated exceptions". The highest end pro apps (such as Scenarist) won't have that restriction.

Reason this might go unnoticed by your playback: in order to combat "piracy", and in order to keep costs down, many manufacturers use only stock D->A audio decoders and so many of the available machines CANNOT produce true 24/96 on their output. They will downsample to 24/48 (or possibly even 20/48 or 16/48), prior to D->A conversion.
Those better machines that DO allow true 24/96 output, are NOT ALLOWED to output 24/96 if the content is CSS-encrypted (part of Hollywoods CP rules), although this might be allowed now for fully end-to-end protected output such as HDMI to the receiver. This is the most cited reasoning behind saying it's "not compliant", even though it is.

If you look at the MI output, you can see that the combined filesize of the DVDLP title is equivalent to the audio input, but if you subtract the video stream size (which ought to be QUITE accurate, as MI has thorough understanding of that portion), it is impossible for the audio to be any better than 24/32/2 or 16/48/2 (which BTW are both 1536kbps) as it's stream size is only 32.5MB. It's just doing the math.

Have you tried downsampling your source WAV file to 16/48/2 or 24/48/2 and seeing if DVDLP gives you warnings/errors, and what MI says about the output there? It might shed some light on what is going on behind the scenes with DVDLP.

Scott

[DB83]

Scott,

I accept what you say about dvd-audio(that can go as high, in some cases, to 192 kHz) but dvd-video. Surely that has to be 48 kHz ?

[El Heggunte]

http://stnsoft.com/DVD/LimPcmAud.html

[DB83]

Ok. I was referring to the info on a little site called videohelp.com. You may have heard of it

Like I said above, methinks there is a distinction between pure dvd-audio which supports higher(and lower) sampling rates than pure dvd-video.

But I have not had the best of weeks with my arguments so maybe I should just retire on this one to my cave.

[racer-x]

I'm pretty sure DVD-Lab is incapable of re-encoding anything. It doesn't even have an encoder, aside from the menu image encoder. That was one of it's main selling points, it expects you to import compliant streams. It's been years since I've used it though. It's been years since I've authored a dvd for that matter....

[Cornucopia]

You could be right, however the code to downconvert 96kHz to 32kHz can be as simplistic as dropping every 2 out of 3 sample points. Since it's LPCM, no real "reencoding" necessary.

Couple of alternate tests ought to sort it out.

Scott

[DRP]

it might very well be that DVDLP is downsampling/downconverting - yes, to 32kHz - in order to maintain its authoring procedure, when given 24/96.

I don't think this is happening. From watching the authoring process on the screen there is absolutely no indication by dialog or time taken that DVDLP is doing this. In fact, whilst DVDLP does have a few basic transcoding tools inside it for adjusting both video & audio files, transcoding down to a different sampling rate does not appear to be one of them. Conversion to MP2 or AC-3: yes, but not conversion from 96 to 48 or 32kHz.

Reason this might go unnoticed by your playback: in order to combat "piracy", and in order to keep costs down, many manufacturers use only stock D->A audio decoders and so many of the available machines CANNOT produce true 24/96 on their output. They will downsample to 24/48 (or possibly even 20/48 or 16/48), prior to D->A conversion.

Yes this is why I included the page from my DVD player manual. It specifies what the DVD will and won't do with respect to the audio output. Right now I fully expect that I am not hearing true 24/96 sound because just temporarily while my amplifier is being repaired I'm only hearing the audio through a 1/8" headphone socket from the TV into a couple of compositie RCA plugs due to input limitations on a spare amplifier. I think that to keep the signal as is, it will need to be transferred to the amp via either HDMI or co-ox cable as a bare minimum ultimately. For now, I'm just proving the concept and it doesn't matter because no-one can actually hear the difference between 24/96 and 16/44.1 sound anyway.

If you look at the MI output, you can see that the combined filesize of the DVDLP title is equivalent to the audio input, but if you subtract the video stream size (which ought to be QUITE accurate, as MI has thorough understanding of that portion), it is impossible for the audio to be any better than 24/32/2 or 16/48/2 (which BTW are both 1536kbps) as it's stream size is only 32.5MB. It's just doing the math.

The video size is also quite obviously wrong. There is no video at all. The only video of sorts is a plain black background single frame slide with a few words of white text on it. That's it. I'd be amazed if it takes more than 1MB to store. Certainly nothing like the 115MB MediaInfo is claiming. What I think is going on here is that DVDLP is putting false information in the header of the VOB file to trick DVD players into playing a VOB file that effectively contains only audio and no video at all. This is a "new" feature of their DVD-lab PRO2 product I'm using as advertised here from

http://www.mediachance.com/dvdlab/dvdlabpro.html

:

Audio-Only Track
A 'Movie' can now have also audio track(s) only. This is a good option for distributing large amount of music on a DVD. As addition you can also edit the still screen that will be displayed during the audio playback. (For example a song title)
The new version adds second channel and the ability to insert new still image on any audio mark.

I think this is a designed feature of DVD-lab PRO to enable this functionality and have as little player protest as possible. The players will just see a fully compliant disc structure by analysing the header information and go ahead and play it anyway, even though it's 'weird' with all audio and no proper video. I think MediaInfo goes no deeper than analysing the VOB header information too and comes up with the same incorrect analysis. DVD-lab PRO is effectively 'tricking' both the DVD player and MediaInfo too with its authoring process to hide the true status of the video & audio files contained within the VOB

Have you tried downsampling your source WAV file to 16/48/2 or 24/48/2 and seeing if DVDLP gives you warnings/errors, and what MI says about the output there? It might shed some light on what is going on behind the scenes with DVDLP.

Yes, I have downsampled the original 24/96 WAV file to 16/48 with foobar2000/SoX and both DVDLP & MI report it as being exactly as such and DVDLP accepts it as an input for a DVD asset without any protest or warning at all.

Here is the warning I get from DVDLP when importing the original 24/96 WAV file into assets:


Note how it's not complaining about the sampling rate as such. Just the bit depth.

<Edit> Just tested DVD in much older and less sophisticated Sony DVP-NS415 player and it works there too no problem. I think my theory about DVDLP tricking the VOB header is correct.

Now, can anyone tell me if it's possible to add a background image to the simple audio title frame in DVDLP when wishing to make audio-only movies, or is it as basic as it looks only and just limited to simple text and basic Microsoft Paint type quality graphic art only?

[theewizard]

most DVD players play CD music discs
is there some reason NOT to create Normal CD's

[DRP]

most DVD players play CD music discs

No kidding. You obviously haven't read this thread through before replying. I am not making "CD music discs" as you like to call them.

is there some reason NOT to create Normal CD's

How about my source of music files is not CDDA red book compliant and converting them to be so is not a lossless process?

[Cornucopia]

DVDLP is fudging: there is NO SUCH THING as a "Audio Only" title in the DVD-Video spec. In fact, the spec expressly forbids it. ALL titles must have video. Even stills/menus are not truly still photos, but single I-frame-only videos.

Your understanding of the DVD-Video spec is flawed.
The proof is in the pudding, or in the eating, and is quite empirical - just demux your VOB to elementary M2V video and LPCM/WAV audio (you can use TMPGEnc freely for this, or many other apps).
Then try playing the M2V in VLC or similar. If it plays for the said amount of time, you've surely got video. Plus you could check the demuxed M2V and the LPCM WAV themselves in MI.
My guess is that what you've got is an MPEG-Still-As-Video (aka I-frame-only), which doesn't use up nearly the amount of space as full-blown video, but still is more than you're thinking (and more than Zero). If I were economizing on space, that's how I would have done it. MPEG stills usually hang up VLC, though.

My "theory" would also exhibit the same playback you are experiencing.

Can't tell you about DVDLP's menu/art capabilities, as I mainly use DVDArchitect or Encore for personal projects, and DVDMaestro or Scenarist for pro projects (which have their own MPEG Stills Encoders) and I create nearly all my menus/photos in Photoshop.

BTW, the thing about "normal DVD players only support 16bit files" is either an outright lie or a bit of severely antiquated/out-of-date deflecting jargon (my guess is the latter). The last part about "24bit only playable on DVD-Audio players" was NEVER true. Sounds like somebody needs to update their software design or their documentation or both.

You could of course counter with some other argument, but as I said, it's easy enough to get to the bottom of this by trying what I suggested. Or if you aren't sure of the steps, just upload a short clip and we'll do it.

Scott

<edit>Most "MPEG Stills" in VOBs are listed as stream # 0xE2. What you have is # 0xE0, which is what is used for NORMAL video, so it probably is NOT an MPEG Still. Shame and likely a major waste of bitrate.</edit>

[DRP]

DVDLP is fudging: there is NO SUCH THING as a "Audio Only" title in the DVD-Video spec. In fact, the spec expressly forbids it. ALL titles must have video. Even stills/menus are not truly still photos, but single I-frame-only videos.

You're getting too hung up on semantics and your interpretation of strict adherence to written specifications. DVDLP are clearly meaning that most users would consider a DVD that contains 99.99% audio and 0.01% "video" which consists of nothing more than a slideshow of still images to be an "audio only" movie. I would agree with that layman's description. The fact that DVDLP makes the 'video' portion itself to make the end result playable in a conventional DVD player is a true credit to the program IMHO.

Your understanding of the DVD-Video spec is flawed.

Quite the contrary. Everything I've done so far has confirmed I'm right in my theories.

just demux your VOB to elementary M2V video and LPCM/WAV audio (you can use TMPGEnc freely for this, or many other apps).
Then try playing the M2V in VLC or similar. If it plays for the said amount of time, you've surely got video. Plus you could check the demuxed M2V and the LPCM WAV themselves in MI.
My guess is that what you've got is an MPEG-Still-As-Video (aka I-frame-only), which doesn't use up nearly the amount of space as full-blown, but still is more than you're thinking. If I were economizing on space, that's how I would have done it. MPEG stills usually hang up VLC, though.

Done that. Here's the 'video' you can download and see for yourself. All 253KB of it ripped out of a 1.7GB VOB file, all of the rest of it is PCM audio at 24/96.

https://dl.dropbox.com/s/epdeho5856nzjtt/VTS_01_1.m2v

AFAIC that's proof that MediaInfo is wrong and DVDLP is doing no resampling down to 32kHz as suggested. I have effectively an Audio only movie (as the vast majority would describe it) on a conventional DVD that plays high res music on a conventional DVD player, which is precisely what I was wanting to achieve. Somebody owes me a dollar.

[DB83]

^^ I will readily admit, having read the fuller spec, that I was wrong about the 96 kHz sampling for audio on dvd-video (even tho it is stated that many players sample that back down to 48 kHz which defeats the purpose in the higher setting).

But the dvd-video - and that is what we are talking about here - spec mentions nothing about 32 kHz audio. Gee I would like to try that but my software only creates audio that the dvd-video spec normally accepts ie PCM 1536 (not 1024) 48 (not 32).

Maybe mediainfo has been fooled by the header of that vob. So even if it is reporting that, your player is not playing it.

[DRP]

OK new problem now. I've got the DVD working and playing in my DVD player as I want it to be. All except for one small thing. How do I get gapless playback on a DVD? I've joined all the separate WAV files together with WavCat so there are no gaps at all and then inserted the whole contiguous WAV file into the DVD as one 'movie' and entered chapter points at the track change locations. I thought this would play perfectly seemlessly through the chapters one after the other, but it doesn't. There is still a momentary stutter at each chapter point I think as the next slide for the next track is read and loaded for display. It's only very brief, maybe half a second or less, much better than a previous test where I had each track as a separate 'movie', but still not perfectly gapless as it's supposed to be.

Any ideas? If not, then I think I'll have to test it with chapters but with no slide changes at the change of track points to see if that makes a difference.

[racer-x]

If memory serves, DVDLP has a slideshow tool. You import images and add your audio track. You'll need to do this for each track. You can set playback to be sequentially or by button link only.

It's been years, but I did this for my Led Zeppelin, Pink Floyd and Jimi Hendrix collections. Although I only used 16-bit PCM or AC3 in mine.

[manono]

It probably doesn't matter since your players play it, but if you're curious whether or not the DVD is within spec, load a 720x576 BMP and the audio into Muxman and see if it authors it for DVD. It's quite strict about following the rules.

[DRP]

Well that didn't work. Deleting all the custom slides at each track/chapter change made no difference. It still stutters between tracks/chapters. So it would seem there is no way to make true gapless audio playback on DVD whilst still retaining the ability to individually track seek. I can only see one other option to try and that's removing all the chapters, which effectively just means one long contiguous audio file with no ability to seek to individual tracks at all.

[manono]

So it would seem there is no way to make true gapless audio playback on DVD whilst still retaining the ability to individually track seek.

That's not true. If you really have individual chapters within the same title, as opposed to each song being its own title, then there shouldn't be any pauses.

[DB83]

Well if you are using DLP for this, keep the tracks separate. Use the playlist for 'Play All' and a simple menu for track-select.

[DRP]

So it would seem there is no way to make true gapless audio playback on DVD whilst still retaining the ability to individually track seek.

That's not true. If you really have individual chapters within the same title, as opposed to each song being its own title, then there shouldn't be any pauses.

I agree, that was my thinking too. That's why I joined all the files together with WavCat first to one big long contiguous WAV file. Surely that can't then possibly have any gap playing problems, but sure enough, on playback via DVD player, it introduces gaps at the chapter points. I have no idea why. What plays seemlessly via foobar2000 on the computer, suddently gets short stutter gaps played on a DVD player.

[manono]

I don't use DVDLab so I can't help there. Follow DB83's suggestion to see if you can get it to play smoothly that way. I know that, using the authoring programs with which I'm familiar, the playback is seamless at the chapter points.

You're sure the whole thing consists of a single title? You could open the final DVD in PGCEdit to make sure.

[pandy]

You could be right, however the code to downconvert 96kHz to 32kHz can be as simplistic as dropping every 2 out of 3 sample points. Since it's LPCM, no real "reencoding" necessary.

Couple of alternate tests ought to sort it out.

Scott

Only if low pass (brick at 16kHz) precede decimating as you described - if no low pass then you may and (probably) you will introduce aliasing.

[racer-x]

Well I just played my Led Zeppelin dvd audio (hybrid) disk I made back in 2004 using DVDLP with slideshow images. It played smoothly from track to track. I have no idea what you did wrong, but clearly you did something wrong.

I don't have DVDLP installed and don't feel like installing it, so I can't really help. Keep at it till you figure it out.....

[Cornucopia]

You could be right, however the code to downconvert 96kHz to 32kHz can be as simplistic as dropping every 2 out of 3 sample points. Since it's LPCM, no real "reencoding" necessary.

Couple of alternate tests ought to sort it out.

Scott

Only if low pass (brick at 16kHz) precede decimating as you described - if no low pass then you may and (probably) you will introduce aliasing.

THEORETICALLY, that is very true. But in real life, if the upper harmonics (>16kHz) are of very low level/power to begin with (quite common), many people MAY not even notice enough the foldover aliasing artifacts even though they would exist. Maybe DVDLP is banking on this and ignore LP/AA needs, maybe they do include an LP/AA filter, maybe they delegate that algorithm to an external process, or maybe no downsampling was done and I'm wrong - hard to say.

Scott