I have some questions for any audio engineers out there. I keep hearing that lossless is always better and I thought it was more complicated then that. It doesn't look like we have an insider's thread anymore, which is why this is posted here. Mods - please move if you feel it should be elsewhere.

1. Is a lossless audio track identical to a studio master, or does lossless simply mean that no non-audible information was removed?

2. If lossless is a perfect copy of the studio soundtrack, then why are some lossless audio tracks 16bits compared to 24bits, or 48Khz compared to 96Khz?

3. If properly done, could a DD+ soundtrack sound identical to a lossless audio track?

4. Would an 8 bit 24Khz LPCM soundtrack still be lossless? (Just an extreme example, haven't seen any.)

5. Regarding question 4, would you consider that PCM track better than a 20bit 48Khz DD+ soundtrack?

I have some questions for any audio engineers out there. I keep hearing that lossless is always better and I thought it was more complicated then that. It doesn't look like we have an insider's thread anymore, which is why this is posted here. Mods - please move if you feel it should be elsewhere.
I have created an audio thread here: http://www.avsforum.com/avs-vb/showthread.php?t=960826

Maybe we can treat that as poor man's insider thread and make it sticky for future reference?

1. Is a lossless audio track identical to a studio master, or does lossless simply mean that no non-audible information was removed?
It is identical to what was fed to the lossless encoder. Just like the fact that we have 10-bit video masters, but then move to 8-bits before compressing it, audio can be reduces in resolution before losslessly compressed for delivery.

2. If lossless is a perfect copy of the studio soundtrack, then why are some lossless audio tracks 16bits compared to 24bits, or 48Khz compared to 96Khz?
Because many times the original recording is at 48Khz. 16-bit tracks are sometimes downconversion from 24 bit, but other times because the source is that way.

3. If properly done, could a DD+ soundtrack sound identical to a lossless audio track?
It is not a matter of being properly done. The encoding is automatic so there is no room for it being done one way vs another. What makes DD+ transparent to the source depens on data rate used, and the complexity of the source relative to the data rate selected. The #1 enemy of a lossy codec is high-frequency transients. They take more data to encode without distortion. Fortunately, this kind of thing is far more prevelant in music than video. And at data rates such as 1.5mbit/sec DD+, we have ample headroom to reduce their perceptual effect.

4. Would an 8 bit 24Khz LPCM soundtrack still be lossless? (Just an extreme example, haven't seen any.)
Of course. This is used as poor man's compression before we started to use "preceptual" lossy codecs like MP3 and such. Some camcorder audio formats do the same.

In general however, 8-bit sounds very harsh so typically 12-bits or so is used.

5. Regarding question 4, would you consider that PCM track better than a 20bit 48Khz DD+ soundtrack?
Most definitely not. The 8-bit track would only have 48 db of dynamic range. Cassette tape had 70 db! You can add dither to it to make sound decent but it would then be very noisy.

I have some questions for any audio engineers out there. I keep hearing that lossless is always better and I thought it was more complicated then that. It doesn't look like we have an insider's thread anymore, which is why this is posted here. Mods - please move if you feel it should be elsewhere.

1. Is a lossless audio track identical to a studio master, or does lossless simply mean that no non-audible information was removed?

2. If lossless is a perfect copy of the studio soundtrack, then why are some lossless audio tracks 16bits compared to 24bits, or 48Khz compared to 96Khz?

3. If properly done, could a DD+ soundtrack sound identical to a lossless audio track?

4. Would an 8 bit 24Khz LPCM soundtrack still be lossless? (Just an extreme example, haven't seen any.)

5. Regarding question 4, would you consider that PCM track better than a 20bit 48Khz DD+ soundtrack?

Not an engineer but will take a stab.

1. In most cases yes, but there are exceptions. You can have a 16bit track noise shaped to 24bit or a 24bit dithered to 16bit. You could have a 48Khz up sampled to 96Khz or a 96Khz down sampled to 48Khz.

2. Masters can be 16bit or 24bit, 48Khz or 96Khz.

3. If it fools your ears into sounding identical then it does sound identical to you. You could possibly reach a point with perceptual coding where you could fool almost 99% of the ears if it sounding the same. It is still technically missing info so it is NOT the identical.

4. Yes, lossy refers to perceptual coding.

5. Depends. Take a 48/24 Master and dither a 8bit lossless LPCM and a 48/24 DD+ (at 1.5Mbps) and the DD+ will definitely sound better than the lossless LPCM.

Looks like a REAL expert beat me to the enter key.:D

Thanks guys!

Hmmmm

So dithering does effect the sound quality now.

Geeez. I wish the "insiders" here would keep that straight.

Hmmmm

So dithering does effect the sound quality now.

Geeez. I wish the "insiders" here would keep that straight.
Oh, the insiders do keep that straight :). But it seems that my brief answer caused some confusion. So a bit more explanation.

Dither does indeed do away with the steppiness when you convert an analog signal to digital. However, the price you pay for it, is increased noise. When you are talking about 16-bits, the noise is very low. When you are talking about 8-bits, the noise is far, far higher. While you will not hear the noise in 16-bit dithered signals, you will surely hear it in 8-bit and it would be quite bothersome.

So it is all a matter of degree and context. Dither for conversion of 24 to 16 is perfectly fine. But not so if you go down to 8 bits.