Impact of lossy compression when streaming Atmos

[riskylogic]

I only have the "4" album on SACD and did a comparison between Urgent from this one and Tidal. Definitely different mixes. Sad to say that the new atmos mix in every way sound worse than the old, at least in my ears. The old mix has got a much better and precise low end. I think the eq on all the instruments and lead vocal sounds better. It feels like they really did work hard to get all the details as good as possible. As for the Atmos mix it sounds more like it has been done in a hurry. Even the surround effects are more creative in the 5.1 mix. In the atmos mix' defense you could say that the snare sounds more like the original stereo mix, but I prefer it sounding a little updated. The 5.1 mix is great, the Atmos is good. The music is great anyhow so keep 'em coming. Hoping for more full best of albums like Elton's rather than just teasers like Foreigner, Billy Idol and Queen.

That pretty much seems to be a universal problem with streaming Atmos. The dynamic range and depth just isn't there. I have Yello - Point and Steven Wilson - The Future Bites on bluray and they sound much better than streaming from Tidal. Furthermore, whenever I switch from streaming to my hard drive just about everything sounds like a sonic masterpiece. I'm sure it's because the bitrate just can't handle it. Streaming has about the same bitrate as Dolby Digital and there's more channels. It's very lossy.

 

[himey]

That pretty much seems to be a universal problem with streaming Atmos. The dynamic range and depth just isn't there. I have Yello - Point and Steven Wilson - The Future Bites on bluray and they sound much better than streaming from Tidal. Furthermore, whenever I switch from streaming to my hard drive just about everything sounds like a sonic masterpiece. I'm sure it's because the bitrate just can't handle it. Streaming has about the same bitrate as Dolby Digital and there's more channels. It's very lossy.

Thankfully I don't get hung up on formats. There are plenty of great sounding mixes only available in lossy.

I do however get hung up on the quality of the mix, and the quality of the original production. These vary much more than a lossy conversion. I will take a quality mix and production, like the Elton Diamonds, in lossy, over a crappy mix or low budget production, in any high resolution, lossless format.

 

[Oldschoolsoundengineer]

Thankfully I don't get hung up on formats. There are plenty of great sounding mixes only available in lossy.

I do however get hung up on the quality of the mix, and the quality of the original production. These vary much more than a lossy conversion. I will take a quality mix and production, like the Elton Diamonds, in lossy, over a crappy mix or low budget production, in any high resolution, lossless format.

Agreed. Totally. Lossy does not make the mix turn bad. You might loose details in the treble area and bass might not be as well defined as in high res, but the quality of the mix is the most important part. In Elton's case it is also impressive how well the mixes of songs from several decades are mixed and mastered to sound uniform throughout a compilation and still sound excellent.

 

[riskylogic]

Thankfully I don't get hung up on formats. There are plenty of great sounding mixes only available in lossy.

I do however get hung up on the quality of the mix, and the quality of the original production. These vary much more than a lossy conversion. I will take a quality mix and production, like the Elton Diamonds, in lossy, over a crappy mix or low budget production, in any high resolution, lossless format.

Yes, but there are a bunch of streaming only Atmos titles where I really want the nonlossy version. For starters:

Tori Amos - Ocean to Ocean
Beatles - Sgt. Pepper
Crown Lands - White Buffalo
Tom Petty - Wildflowers +

Elton Diamonds would be mighty nice too.

 

[MrSmithers]

The low bitrate and its compression is not at all kind with this song… Listened to it with my headphones and I find the Spotify version much clearer.

Wouldn't Spotify (Ogg Vorbis) offer an even lower bitrate and more compression? The Atmos mix might be a bit lacking and that might explain more of a gap compared to the lossy stereo Spotify stream?

 

[perzon57]

The low bitrate and its compression is not at all kind with this song… Listened to it with my headphones and I find the Spotify version much clearer.

I think they sound good in Atmos from Apple Music on my setup. No headphones, no Spotify.

 

[riskylogic]

Wouldn't Spotify (Ogg Vorbis) offer an even lower bitrate and more compression? The Atmos mix might be a bit lacking and that might explain more of a gap compared to the lossy stereo Spotify stream?

If you are bit rate constrained then you will get better sound quality with fewer channels. I think 600 kbs is usually fine for stereo - the Tidal stereo steams usually sound as good as a CD to me. But if you've got 8 channels instead of two then it's 75 kbs per channel instead of 300. With quad you've got 150 kbs per channel, which might be a better tradeoff is you are trying to stream surround.

 

[Albert Dagger]

If you are bit rate constrained then you will get better sound quality with fewer channels. I think 600 kbs is usually fine for stereo - the Tidal stereo steams usually sound as good as a CD to me. But if you've got 8 channels instead of two then it's 75 kbs per channel instead of 300. With quad you've got 150 kbs per channel, which might be a better tradeoff is you are trying to stream surround.

That’s not how it works for multichannel encoders. They can be more efficient than simply dividing the bit rate per number of channels.

For stereo files, I’m not sure how you got to 600kbs. It’s either lossy, 320 or less like Spotify, or lossless. If lossless, the bitrate is irrelevant. Flac 600, 900 or 1000 is still a lossless version of the 1411 wav file. So the tidal stream should sound exactly like the CD, unless it’s labelled as Master and it is MQA encoded. Even in that case, I doubt the difference is audible.

 

[riskylogic]

That’s not how it works for multichannel encoders. They can be more efficient than simply dividing the bit rate per number of channels.

For stereo files, I’m not sure how you got to 600kbs. It’s either lossy, 320 or less like Spotify, or lossless. If lossless, the bitrate is irrelevant. Flac 600, 900 or 1000 is still a lossless version of the 1411 wav file. So the tidal stream should sound exactly like the CD, unless it’s labelled as Master and it is MQA encoded. Even in that case, I doubt the difference is audible.

• Dolby Atmos audio in Dolby Digital Plus is typically encoded at bitrates between 384 and 768 kbps.

It's not lossless or even close. I'm sure the encoders can be more efficient when not all the channels are used. But with "Personal Shopper" they can't keep up. It's obvious.

 

[Albert Dagger]

• Dolby Atmos audio in Dolby Digital Plus is typically encoded at bitrates between 384 and 768 kbps.

It's not lossless or even close. I'm sure the encoders can be more efficient when not all the channels are used. But with "Personal Shopper" they can't keep up. It's obvious.

I didn’t say it’s lossless, and I didn’t say it sounds as good as lossless. I just said your bit rate calculation is not correct and is misleading. The quality you get from a 768kbps e-ac3 stream is much higher than that bitrate divided by the number of channels, and the result is perfectly listenable.

 

[riskylogic]

I didn’t say it’s lossless, and I didn’t say it sounds as good as lossless. I just said your bit rate calculation is not correct and is misleading. The quality you get from a 768kbps e-ac3 stream is much higher than that bitrate divided by the number of channels, and the result is perfectly listenable.

I think I read somewhere than 600 kbps is the average, but that's trivial. 768 kbps is about the same as a stereo flac, and it's lossless. A 320 kbps mp3 is usually almost as good, but I think you're asking for trouble when you go lower than that. 768 kbps for 8 channels is the same as 192 kbps for stereo.

It does seem to me that fidelity varies between recordings. The ones with lots of dynamic range and channel usage suffer from compression the most. I also think there's a negative correlation between how much you have to turn it up to get to normal listening levels and fidelity.

 

[Doug G.]

I don't understand, at all, sacrificing sound quality for what is judged to be a superior mix. What is the point? Is this what Atmos is all about?

Doug

 

[LuvMyQuad]

I don't understand, at all, sacrificing sound quality for what is judged to be a superior mix. What is the point? Is this what Atmos is all about?

Doug

I don't think that's really the case. The few Atmos albums I have on disc are stellar sounding. On something like Harrison's ATMP the artificial ceiling effects are kept to a bare minimum. Abbey Road is mostly like that too. The Atmos channels provide ambiance. and they do a good job at it. It can envelop like 5.1 cant. There is no loss in fidelity at all that I can detect, in spite of Atmos topping out at 48kHz. All that is what Atmos is about.

The Atmos that is being streamed (Apple in my case) is something different, but not by much. I have compared selections of the two titles above. This was not a scientific exercise by any means. A/B testing with a 5 second gap when inputs are switched is not ideal. But i did try to get the volumes close. and the timing reasonable. Each time I preferred the disc version. Try not to laugh at my stab at an audiophool term, but the stream sounds more grainy, especially through the midrange. There is more of what sounds like electronic haze on the top end. Its not difficult to hear.

That being said, its also easy for me to not listen so critically and just enjoy it. Stuff that I can't compare like the Atmos EJ stuff sounds just amazing. And there is even a great version of Rush's Moving Pictures in Atmos. A mix that finally does the album justice. Its so good it makes me want to buy the ridiculously priced box.

 

[ar surround]

Try not to laugh at my stab at an audiophool term, but the stream sounds more grainy, especially through the midrange. There is more of what sounds like electronic haze on the top end. Its not difficult to hear.

If I'm driving or doing something around the house...casual listening...then this level of Atmos fidelity is OK. However, if I'm going to sit down and listen to music, I want reasonable audiophool quality otherwise it's not worth my time. Others may feel differently and that's reasonable, but I guess I'm becoming even more of an audio snob in my old age.

(A friend of mine said that I've also turned into a skiing snob because I won't take the time to drive somewhere if the snow conditions aren't up to my standards.)

 

[kfbkfb]

https://www.avanquestusa.com/products/enhancer/index.htm^^^
...MP3s, Windows Media, Real audio, CDs or Internet radio, DFX dramatically enhances your music listening experience.

(maybe this tech could be adapted to enhance [low data rate] Dolby Atmos surround sound)


Kirk Bayne

 

[Pablo]

I think they sound good in Atmos from Apple Music on my setup. No headphones, no Spotify.

Towards the end 4:00 I find the song quite compressed. I’ll make a few more tests…

 

[perzon57]

Towards the end 4:00 I find the song quite compressed. I’ll make a few more tests…

Is that in Atmos?

 

[Strilo]

Yes, but there are a bunch of streaming only Atmos titles where I really want the nonlossy version. For starters:

Tori Amos - Ocean to Ocean
Beatles - Sgt. Pepper
Crown Lands - White Buffalo
Tom Petty - Wildflowers +

Elton Diamonds would be mighty nice too.

I would add all three Pearl Jam Atmos mixes to this list as well.

 

[himey]

I don't understand, at all, sacrificing sound quality for what is judged to be a superior mix. What is the point? Is this what Atmos is all about?

Doug

If you can't get any enjoyment from these slews of new Atmos mixes, you are doing it wrong.

 

[Pablo]

Is that in Atmos?

Yes, the Atmos version of Zeit on Apple Music. Will compare soon to the lossless version and see if it is a mixing thing or a compression thing.