Proper way to reproduce 5.1 channel Pro Logic II?

[Seraphic-]

Hi,

I have several Playstation 2 games and I wanted to reproduce their Pro Logic II 5.1 audio features with my video captures. My system has the hardware to capture both RCA and Digital Optical. I've heard you should upmix stereo to 5.1 for something like this. MeGUI has a audio option for "upmix 2 to 5.1 via SuperEQ". Would that reproduce Pro Logic II? My hardware allows for multichannel recording over digital optical, but from "what you hear", so it isn't a direct feed. Only stereo RCA/Optical is a direct source feed.

I'm not sure how to proceed?

[FlimsyFeet]

How will you be playing back these captures? Because it you're going to put them on DVD, then play back through a sytem that has Pro-logic II decoding, I would say you are better leaving the 2-channel audio as it is.

If you want to convert to 5.1, and properly emulate Pro-logic II decoding, then there is a method invloving Graphedit and a DPL II decoder built into either WinDVD or PowerDVD. Sorry I don't know the specifics, but if you search this forum you should find the guide.

[tebasuna51]

I have several Playstation 2 games and I wanted to reproduce their Pro Logic II 5.1 audio features with my video captures. My system has the hardware to capture both RCA and Digital Optical. I've heard you should upmix stereo to 5.1 for something like this. MeGUI has a audio option for "upmix 2 to 5.1 via SuperEQ". Would that reproduce Pro Logic II?
Nope. Like FlimsyFeet say you there are GraphEdit methods to do a correct upmix from dpl II to 5.0 (LFE empty)
You never can obtain 5.1 because the LFE channel isn't in dpl mix. But don't worry, your audio hardware equipment select the low frequencies to play in subwoofer if exist.

Maybe latest PowerDVD versions don't permit use the necessary filters in GraphEdit. But exist a free method to do this upmix using Foobar2000 and this plugin (http://www.hydrogenaudio.org/forums/index.php?showtopic=52235).

My hardware allows for multichannel recording over digital optical, but from "what you hear", so it isn't a direct feed. Only stereo RCA/Optical is a direct source feed.

I'm not sure how to proceed?
To capture you only need stereo signal.
To play you have two options:

1) If you have an audio equipment with dpl II decoder you can safely mix the stereo capture (mp3 format at least 160 kb/s, for compatibility) with the video and decode to 5.1 at play time.

2) If your audio equipment can't decode dpl II, you need decode by soft (GraphEdit, Foobar2000) to 5.0 (or 5.1 with LFE empty) and mix with the video (ac3 format for compatibility at least 384 kb/s)

The preferred method is 1) for size and quality (transcoding always lose quality)

[Seraphic-]

Nope. Like FlimsyFeet say you there are GraphEdit methods to do a correct upmix from dpl II to 5.0 (LFE empty)
You never can obtain 5.1 because the LFE channel isn't in dpl mix. But don't worry, your audio hardware equipment select the low frequencies to play in subwoofer if exist.

Maybe latest PowerDVD versions don't permit use the necessary filters in GraphEdit. But exist a free method to do this upmix using Foobar2000 and this plugin (http://www.hydrogenaudio.org/forums/index.php?showtopic=52235).


To capture you only need stereo signal.
To play you have two options:

1) If you have an audio equipment with dpl II decoder you can safely mix the stereo capture (mp3 format at least 160 kb/s, for compatibility) with the video and decode to 5.1 at play time.

2) If your audio equipment can't decode dpl II, you need decode by soft (GraphEdit, Foobar2000) to 5.0 (or 5.1 with LFE empty) and mix with the video (ac3 format for compatibility at least 384 kb/s)

The preferred method is 1) for size and quality (transcoding always lose quality)

Thanks for the reply.

In this case, these 720p/480p videos would be for internet distribution v.i download then play.

I have the hardware to decode DPLII, however I'm sure about all other users who will access these videos. But, I still wanted to offer some kind of 5.1 audio for the higher bit-rate videos, rather then just stereo two channel.

Another thing is, my hardware supports 48kHz/16Bit (insures audio/video sync) or 96kHz/24Bit (would have to sync to video manually). I'm sure 96kHz/24Bit would be the best bet, but my speakers support a direct feed of 88.2kHz and I would just have to match the audio to the video by hand.

Also, why would the LFE be empty? Doesn't the Playstation 2 just upmix two channel to 5.1 when channel DPLII is active? So wouldn't it be the same thing?

[tebasuna51]

I have the hardware to decode DPLII, however I'm sure about all other users who will access these videos. But, I still wanted to offer some kind of 5.1 audio for the higher bit-rate videos, rather then just stereo two channel.

Of course you can choose. My recommendation is let without changes dpl II stereo mix. If the audio is encoded in ac3, there are a flag to set the receivers in dpl II upmix mode.

Another thing is, my hardware supports 48kHz/16Bit (insures audio/video sync) or 96kHz/24Bit (would have to sync to video manually). I'm sure 96kHz/24Bit would be the best bet, but my speakers support a direct feed of 88.2kHz and I would just have to match the audio to the video by hand.

If you doubt about dpl II decoder forget other frequencies/bitdepth sophistication, let 48/16.

Also, why would the LFE be empty? Doesn't the Playstation 2 just upmix two channel to 5.1 when channel DPLII is active? So wouldn't it be the same thing?

I don't understand very well your question but:

1) A certified Dolby 5.1 downmix DPL II never use the LFE channel, is ignored.

2) If somebody use the LFE channel in a DPL II downmix, this LFE channel can't be separated from the Central channel.

3) Don't worry with a LFE empty, all the lower frequency info is in the others channels and the audio hardware player can redirect them to the subwoofer speaker.

4) Don't mistake subwoofer speaker with LFE channel.
An audio equipment can redirect all low frequencies, from all channels, to the subwoofer speaker (typical in low-cost 5.1 audio)
An audio equipment without subwoofer can redirect the LFE info to front (or all) speakers.

[Seraphic-]

Of course you can choose. My recommendation is let without changes dpl II stereo mix. If the audio is encoded in ac3, there are a flag to set the receivers in dpl II upmix mode.



If you doubt about dpl II decoder forget other frequencies/bitdepth sophistication, let 48/16.



I don't understand very well your question but:

1) A certified Dolby 5.1 downmix DPL II never use the LFE channel, is ignored.

2) If somebody use the LFE channel in a DPL II downmix, this LFE channel can't be separated from the Central channel.

3) Don't worry with a LFE empty, all the lower frequency info is in the others channels and the audio hardware player can redirect them to the subwoofer speaker.

4) Don't mistake subwoofer speaker with LFE channel.
An audio equipment can redirect all low frequencies, from all channels, to the subwoofer speaker (typical in low-cost 5.1 audio)
An audio equipment without subwoofer can redirect the LFE info to front (or all) speakers.

Was just trying to do a test encode using MeGUI with AC3 set to upmix 2 to 5.1 SuperEQ. But it returned as error. So I'll have to find what the problem is or try another program.

And wouldn't it be optimal to record as 96kHz/24Bit? After all, you could always downmix to 48kHz/16Bit if needed. PS3 seems to output 19Bit, so 3Bit is being lost with 16Bit.

[tebasuna51]

Was just trying to do a test encode using MeGUI with AC3 set to upmix 2 to 5.1 SuperEQ. But it returned as error. So I'll have to find what the problem is or try another program.

I can't support MeGUI. If you want experiment about stereo to 5.1 upmix you can read this thread GUIDE LIST: Stereo-to-Surround Conversion Guides (http://forum.doom9.org/showthread.php?t=83752), also BeHappy.
If the source is DPL II never use these methods.

Also, wouldn't it be optimal to record as 96kHz/24Bit? After all, you could always downmix to 48kHz/16Bit if needed.

Maybe, if you can listen the difference.

[Seraphic-]

I can't support MeGUI. If you want experiment about stereo to 5.1 upmix you can read this thread GUIDE LIST: Stereo-to-Surround Conversion Guides (http://forum.doom9.org/showthread.php?t=83752), also BeHappy.
If the source is DPL II never use these methods.



Maybe, if you can listen the difference.

I was just testing out the upmix to see it would work as it offered the AC3 that you were talking about.
The source would never be DPL II, as it would be recorded as stereo right? Or did you mean I should set the source as DPL II and record as stereo or set to stereo and record as stereo?

Also, can you ever use too high of a bit-rate for audio encoding? For example, say you record using 48kHz/16Bit PCM and you wanted to encode with an AAC with your choice being between 16 kbit/s and 640 kbit/s. Would using the 640 kbit/s max always be suggested or is there a point where using too much won't produce higher quality? I guess it would also be different for stereo 2 channel vs 5.1 channel as well. Multichannel would work well with maxed bit-rate I'm sure, but is that the same for stereo?

[tebasuna51]

The source would never be DPL II, as it would be recorded as stereo right? Or did you mean I should set the source as DPL II and record as stereo or set to stereo and record as stereo?
Sorry my english is so bad and I can't understand you.
I say, if the source was, previously, encoded like stereo DPL II (only ac3 have a flag to say: this stereo file is encoded DPL), must be decoded by a hard/soft DPL II decoder to obtain the original 5.0 multichannel.

If you have a standard stereo file you can try several methods to obtain 5 channels.

Also, can you ever use too high of a bit-rate for audio encoding? For example, say you record using 48kHz/16Bit PCM and you wanted to encode with an AAC with your choice being between 16 kbit/s and 640 kbit/s. Would using the 640 kbit/s max always be suggested or is there a point where using too much won't produce higher quality? I guess it would also be different for stereo 2 channel vs 5.1 channel as well. Multichannel would work well with maxed bit-rate I'm sure, but is that the same for stereo?

More channels more bitrate needed.
More bitrate, more quality.
The limits are the encoder range, the size allowed, the audio equipment, speakers and, the more important, your ears.

[Seraphic-]

This program looks like it would work, too bad it costs $500...
http://www.cycling74.com/products/upmix

Edit: My speakers do support 24bit/96khz =D

[Seraphic-]

tebasuna51 (or anyone),

Does ac3/DTS even support 96khz/24bit encoding? Or is 48khz/16bit the max for both?

[tebasuna51]

tebasuna51 (or anyone),

Does ac3/DTS even support 96khz/24bit encoding? Or is 48khz/16bit the max for both?
The max for standard ac3/dts is 48KHz, dts can be Extended to 96KHz but you need special encoder/decoder.

The bitdepth 16/24 bit can't be applied to ac3/dts because the samples are in frequency domain. Two related questions:

1) An ac3/dts encoder accept input wav files with 16/24 bits?
This is a encoder question not related with ac3/dts format.
For instance Aften accept 16/24/32 bits int and also 32 float samples.
Dts format have a field in the header to store the bitdepth of source wav file.


2) The max Dynamic Range of an ac3/dts is equivalent to 16 or 24 bits wav files?
Is accepted to be equivalent to 24 bits.

[n0mag!c]

The limits are the encoder range, the size allowed, the audio equipment, speakers and, the more important, your ears.
Seraphic, real narrow place is, as Tebasuna says, "the encoder range". If you'll decode encoded audio then you can analize of what maximum frequency the audio is. And you'll be sadly surprised with result. And if your source signal is 44kHz/16bit then upsampling it to 96kHz/24bit is pointless because it didn't give more frequency range.

[Seraphic-]

Seraphic, real narrow place is, as Tebasuna says, "the encoder range". If you'll decode encoded audio then you can analize of what maximum frequency the audio is. And you'll be sadly surprised with result. And if your source signal is 44kHz/16bit then upsampling it to 96kHz/24bit is pointless because it didn't give more frequency range.

Well, the source signal output from PS3 for example is 48khz at 19Bits (from what I've seen so far).
So recording at 48hkz with 16Bits removes three of the extra source Bits.
Recording 48hkz at 24Bits would cover this (can only select 8/16/24bit).

From what tebasuna51 said, it seems 48KHz at 24bit (19Bit True) would be allowable.

[Seraphic-]

Hi,

Could an experianced Audiophile please download this raw test recording and convert it from stereo to 5.1 channels correctly?
I've tried a few guides, but mine comes out extremely muffed sounding and the channels don't seem to be defined correctly.
I would just like to get an idea as to what it should sound like when done correctly. Thank you.

70MB - RAW PCM Stereo WAV - 48kHz/24Bit (http://www.temp.seraphicgate.com/48kHz24Bit.rar)

Also, is DTS "ranked" over AC3? What I mean is, is one less compression for higher quality? Or would they both be the same if the same bit-rate was used?

[tebasuna51]

Could an experianced Audiophile please download this raw test recording and convert it from stereo to 5.1 channels correctly?
I've tried a few guides, but mine comes out extremely muffed sounding and the channels don't seem to be defined correctly.
I would just like to get an idea as to what it should sound like when done correctly.
What is correctly?
Don't exist correct method, there are different approaches.
Select your desired one.

My choice is preserve the original source and play with dpl II decoder or other DSP functions in my 5.1 system.

Also, is DTS "ranked" over AC3? What I mean is, is one less compression for higher quality? Or would they both be the same if the same bit-rate was used?

Can't be compared at same bitrate, the ac3 max. is 640 Kb/s and the dts min. is 768 Kb/s. My personal selection is ac3.

[Seraphic-]

What is correctly?
Don't exist correct method, there are different approaches.
Select your desired one.

My choice is preserve the original source and play with dpl II decoder or other DSP functions in my 5.1 system.



Can't be compared at same bitrate, the ac3 max. is 640 Kb/s and the dts min. is 768 Kb/s. My personal selection is ac3.

Hi,

Yes, I understand there are different methods. I've tried a few guides but again, it comes out extremely muffed sounding. Most of the guides I'm finding are from around 2003/2004 on doom9.

But using one of the newer guides I found from 2006, I managed to pull it off for the most part.
http://www.scribd.com/doc/2153277/Converting-Stereo-to-DTS

One problem was I had to do some guessing on a few settings, as the guide was done with voice in mind for the center channel.
Two other problems were my audio levels are maxing out most of the time and I might have messed up the LFE channel.

Test 5.1 Encode (http://temp.seraphicgate.com/51test.rar) - Still sounds kind of muffed, but overall, I'd say it doesn't sound too bad.

It is supposed to sound muffed or is it just part of the upmix process?

http://www.temp.seraphicgate.com/51upmix.jpg

[Seraphic-]

Using another two minute test encode, I did an upmix encode using both ac3 (640kbps) and dts (768kbps).
To be honest, dts sounds much richer in the highs and lows, just better overall.
So I guess I'll go with that, even though it works out to about double the size of ac3.
For two mintues, ac3 was 10.1MB and dts was 24.4MB.

[n0mag!c]

Using another two minute test encode, I did an upmix encode using both ac3 (640kbps) and dts (768kbps).
To be honest, dts sounds much richer in the highs and lows, just better overall.
So I guess I'll go with that, even though it works out to about double the size of ac3.
For two mintues, ac3 was 10.1MB and dts was 24.4MB.
If you did DTS encoding with "surcode" then I suggest you to pass result through "eac3to". This will reduce 768kbps DTS files size nearly twice. (just "eac3to input.dts output.dts")

[n0mag!c]

But when you decide that you've playing with upmixing enough, then you'll conclude the same:
My choice is preserve the original source and play with dpl II decoder or other DSP functions in my 5.1 system.

btw, here is my way (http://forum.doom9.org/showthread.php?p=1136779#post1136779) to playing with upmixing

[Seraphic-]

If you did DTS encoding with "surcode" then I suggest you to pass result through "eac3to". This will reduce 768kbps DTS files size nearly twice. (just "eac3to input.dts output.dts")

Yes, but wouldn't that reduce the quality as well? Isn't it compression the file?

But when you decide that you've playing with upmixing enough, then you'll conclude the same:

btw, here is my way (http://forum.doom9.org/showthread.php?p=1136779#post1136779) to playing with upmixing

In my case, this isn't for DVD playback. These are for HD videos for download over internet distribution.
So the audio track from the video would already need to be up-mixed to 5.1 and encoded to the file, no?

[Seraphic-]

So in the end, when making HD videos for download over internet distribution. Is splitting the stereo source into six mono waves for encoding to 5.1 my only real choice here?

tebasuna51 suggested a dpl II decoder or other DSP functions, but for something that is to be download over internet distribution workable? Is that flag that could be set in the audio stream?

This is the guide that helped me produce the best results in my case.
http://www.scribd.com/doc/2153277/Converting-Stereo-to-DTS

[tebasuna51]

So in the end, when making HD videos for download over internet distribution. Is splitting the stereo source into six mono waves for encoding to 5.1 my only real choice here?
Nope, you can send the original audio stereo, mono or multichannel you don't need upmix.

tebasuna51 suggested a dpl II decoder or other DSP functions, but for something that is to be download over internet distribution workable? Is that flag that could be set in the audio stream?
I suggested use a dpl decoder when the source was a multichannel encoded to stereo Dolby ProLogic. If you know this you can use ac3 and set the appropriate flag to automatically put the receivers in this mode.

[Seraphic-]

If you did DTS encoding with "surcode" then I suggest you to pass result through "eac3to". This will reduce 768kbps DTS files size nearly twice. (just "eac3to input.dts output.dts")

Tried EAC3toGUI, doesn't seem to create any file though.

Nope, you can send the original audio stereo, mono or multichannel you don't need upmix.

I suggested use a dpl decoder when the source was a multichannel encoded to stereo Dolby ProLogic. If you know this you can use ac3 and set the appropriate flag to automatically put the receivers in this mode.

So a program like Sound Forge would allow you to flag a stereo stream as 5.1 when you output as AC3? What other program can do this?

[tebasuna51]

So a program like Sound Forge would allow you to flag a stereo stream as 5.1 when you output as AC3? What other program can do this?

Nope. We can flag the file like stereo DPL encoded. A receiver with dpl II decoder can extract 5 channels (more or less).