Upmixing - where are we at? Have people compared upmixers?

[Alex42]

Как это сравнивать? В этой ветке много разных мнений, но они верны только для автора сообщения.

I did a few variations of upmix little wing Jimi Hendrix - logic7 h/k avr, halo nugen VST, DTS neural VST, posted on my google drive and gave the link so everyone could download and vote but got banned))). Thank you Amir for restoring my account. But I don't know of another way to objectively compare and evaluate upmixes.

 

[krabapple]

I do find it unfortunate that the current DS on my AV20 only has the sole option turning center spread on or off. I would think there could be some areas where a little tweaking might give a better over outcome? I consider Speaker Virtualizer a useless option, trying to create a height dimension when no height speakers exist in the system.

Center Spread (on/off) is the only user option Dolby Surround (aka DSU, 'Dolby Surround Upmixer") has ever offered. And even that option, they took away for awhile (complaints brought it back).

Ah, Mr Negativity again,

Yeah, I never like any surround mixes. U mad because I don't worship your hero.

funny how most everyone in the multich world love 90% of their engineering work.
YMMV.

LOL. You haven't polled 'most everyone in the multich world'; don't pretend you have.
Not everyone who listens to surround sound posts on QuadraphonicQuad (where 99.99% of surround releases get 10/10 stars from *somebody* ..and where 5 stars somehow doesn't mean 'average'. Like Camelot, 'tis a silly place')
Anecdotally, I've read plenty of posts on internet forums by people who aren't enamoured of Mr. Wizard's work

(But I don't pretend that's anything like a scientific sampling)

 

[krabapple]

Qutoed for emphasis.

When original stems are not available, only finished masters, then a 'Penteoed' remix it is mostly. That is a compliment. There are myriad ways to mess up the immersive mix (ages ago we even had a discussion where 'Sean Olive was not impressed with MC mixes'). There are (were) proper solution for the 'artist' to not mess up panning+reverb, channel decorrelation and rendering. Sadly that venture went belly up. So bass guitar in the LFE it is. Lets even give an award for that, just for laughs.

Sometimes the whole band is in LFE. I am not kidding. And full-bandwidth LFE is also a not uncommon thing.

Bass *only* in the LFE, that would be a funny one. Luckily I haven't come across that stupidity yet.

(personally as a bass guitarist, who needs to transcribe and learn parts, I love when the mix engineer puts an essentially isolated full-range bass part in one of the channels)

It seems loudness war is at least toned down with multichannel mixes.

DVD-A and SACD have been hit or miss on that score (check out the Elton John mixes on both fomats) Atmos spec seems to disallow loudness war mastering, so BluRay is doing better...but the last BluRay disc I bought, released just this month, offers a a full-range Atmos/7.1 mix and a smashed 5.1 version of the same mix. To which one can only respond, what the f*cking f*ck are they thinking?

DSU does not collapse stage without centre spread, as PL2 with 'centre width 0' does (neither does Logic7). DSU with centre spread on is mostly PL2 'width 3'.

As I recall, that's exactly what it is

But if DSU/centerspread OFF does not = PL2 centerwidth 0, what is it?

We are lucky to even have this option. Remember how Dolby forced the firmware 'upgrade' couple years back?

Yes, goodbye centerspread! Then Gene from Audioholics got everyone mad about it and Dolby reinstated it. ;>

What I personally would like to see is something that transforms the 2ch material to a time-intensity panned multi-speaker render with a sensible primary sound steering and 'leaky' objects, with a user adjustable slider. I really think 'Neo:X Music' was it. So was Logic 7 Music with '5 speaker enhance'. Both deprecated.

We live in a diminished upmixing world.

 

[krabapple]

Okay, that's interesting. Even somewhat encouraging.

How do you think it would go if we tell the AVR it doesn't even have a centre speaker, and save that system setting, and use it specifically for playback of 2-channel recordings with upmixing? Dumb idea?

cheers

The funny ting about 'spreading' dedicated center content into L and R is that it kind of defeats the anti-comb filtering benefit of having a dedicated center in the first place.

 

[Chr1]

Yeah, I never like any surround mixes. U mad because I don't worship your hero.

Have to say that I like very few surround mixes too. I usually find fault with some of the chosen surround channel levels, or some other part of the mix.
Having spent some time setting up my preferred channel levels for DTS Neo X: Music, I generally far prefer the results of upmixed 2.0.

Definitely easier on my wallet too!

 

[krabapple]

I didn't say I like very few surround mixes. There are plenty I like.

Sal and I are arguing about Grand Wizard Steven Wilson's work, particularly. He's got a lot of online fanboys.

 

[Chr1]

Ah, ok. Apologies.

Have to say that I do actually like quite a few too actually. Some of the Talking Heads, The Beatles: Love, Kraftwerk and several obscure electronic mixes spring to mind. Unfortunately, with the majority, I often find something that niggles. Often simply the chosen volume of surround channels. (Kind of annoying when the disc in question can be quite costly.)

Anyhoo, all good... horses for courses...each to their own and all that!

 

[Newman]

The funny ting about 'spreading' dedicated center content into L and R is that it kind of defeats the anti-comb filtering benefit of having a dedicated center in the first place.

That would depend on whether it preserves phase when spreading. Plus how much of the original centre channel allocation (without centre spread) is left in the centre (with centre spread).

We are told so little.

 

[goat76]

Ah, ok. Apologies.

Have to say that I do actually like quite a few too actually. Some of the Talking Heads, Kraftwerk and several obscure electronic mixes spring to mind. Unfortunately, with the majority, I often find something that niggles. Usually simply the chosen volume of surround channels. (Kind of annoying when the disc in question can be quite costly.)

Anyhoo, all good... horses for courses...each to their own and all that!

I often find the volume level in the surround channels in Atmos mixes to be mixed too loud compared to the front channels, which, depending on the type of music, sometimes leads to a strange balance that can be pretty distracting. I doubt that was the intention by the mixing engineer for most of these mixes, and I suspect a possible reason may be that the speaker systems were not very well calibrated in some of these studios.

Of course, it’s possible that my surround speakers are set too loud, but when fine-tuning the level of all the loudspeakers using the pink noise calibration in the receiver, I think they sound equally as loud to each other. I can add that I never hear any strange balance between the front and surround speakers when watching movies, and only when listening to music in Atmos.

 

[Chr1]

I suspect that this is one of the main reasons I prefer upmixing 2.0. Having spent the time getting the channel levels set to my preference, there's no random level variation occurring. Consequently I get consistent results.

This is not the case when listening to actual discrete surround mixes, as whoever did the mix dictates the channel levels. And there's a lot of variation.
Some you'll like. Some you won't.

With an upmixer you at least get consistency. Whether you actually like it better than a particular discrete multichannel mix is a different matter of course.
Is one better than the other? Again, subjective I would suggest. My 2c.

 

[krabapple]

I often find the volume level in the surround channels in Atmos mixes to be mixed too loud compared to the front channels, which, depending on the type of music, sometimes leads to a strange balance that can be pretty distracting. I doubt that was the intention by the mixing engineer for most of these mixes, and I suspect a possible reason may be that the speaker systems were not very well calibrated in some of these studios.

If it happens that you use Audyssey room EQ, and use its Dynamic EQ (DEQ, which most people should unless they listen at close to reference level) , it is known that DEQ automatically boosts surround channels by a few dBs, based on some esoteric research Audyssey says they did..

 

[goat76]

If it happens that you use Audyssey room EQ, and use its Dynamic EQ (DEQ, which most people should unless they listen at close to reference level) , it is known that DEQ automatically boosts surround channels by a few dBs, based on some esoteric research Audyssey says they did..

Good information, but no, I don’t use Audyssey or any other automatic room EQ program.

 

[dannut]

But if DSU/centerspread OFF does not = PL2 centerwidth 0, what is it?

PL2 - a classic, broadband active-matrix upmixer. Does gain-riding at speaker level. With CW=0 it just clamps down the broadband L-R speaker feeds for central signals.
DSU adds fancy analysis (STFT filterbank with crosscorrelation analysis) and ERB based steering. Less pumping, less stage collapse.

I suspect that this is one of the main reasons I prefer upmixing 2.0. Having spent the time getting the channel levels set to my preference, there's no random level variation occurring. Consequently I get consistent results.

What 'center gain' setting are you using? How would you describe the difference between 'Music' and 'Cinema' mode, if you set 'center gain' to 0.0 (or was it 1.0 for pinpoint images?) in Music mode? Asking because with Neural:X we are stuck in the 'Cinema' mode... There should be an automatic switch-over for 'music-like signals', but like any automation, it's a crapshoot.

 

[bmc0]

I did a few variations of upmix little wing Jimi Hendrix - logic7 h/k avr, halo nugen VST, DTS neural VST, posted on my google drive and gave the link so everyone could download and vote but got banned))).

That would be a good way to compare, but the source material used is problematic due to copyright. There are some recordings available under a permissive license allowing derivative works; a comparison using one of those should be OK to post.

 

[Sal1950]

Sal and I are arguing about Grand Wizard Steven Wilson's work, particularly. He's got a lot of online fanboys.

Mr Negativity, also self named "Krabapple" again.
I named 3 of the more popular surround engineers of which Steven was only one.
Now of course your entitled to your opinion but it certainly doesn't match the greater number of 5.1 and Atmos listeners.
The fact that Steven is chosen by so many artists and labels to handle the surround mastering chores must say a little something about his reputation in the industry.
Then there is the fact that you can please some of the people all of the time, and all of the people some of the time.
You know the rest.

If it happens that you use Audyssey room EQ, and use its Dynamic EQ (DEQ, which most people should unless they listen at close to reference level) , it is known that DEQ automatically boosts surround channels by a few dBs, based on some esoteric research Audyssey says they did..

Newer processors like my Marantz AV20 offer listeners using Audyssey DEQ the ability to either defeat that feature or set it to 3 different levels surround channel boost.
I've also heard that updates maybe coming for older units.

 

[Chr1]

What 'center gain' setting are you using? How would you describe the difference between 'Music' and 'Cinema' mode, if you set 'center gain' to 0.0 (or was it 1.0 for pinpoint images?) in Music mode?

Can't answer these questions at the moment as I am away from home. It's been a couple of years since I adjusted any of these, but I will have a look when I'm back in about a week.

 

[dannut]

Have to say that I like very few surround mixes too. I usually find fault with some of the chosen surround channel levels, or some other part of the mix.
Having spent some time setting up my preferred channel levels for DTS Neo X: Music, I generally far prefer the results of upmixed 2.0.

Definitely easier on my wallet too!

https://www.audiosciencereview.com/...h-how-dolby-atmos-is-mixed.44905/post-1609056

Griesinger had a nice saying almost 30 years ago: "Comparing the original discrete mix to the same mix after downmixing and upmixing can be quite revealing, as the down-mixed and up_mixed mix is often superior to the original. Analysis of the reasons for this observation can lead to better mixing technique.."

Sadly knowledge of proper mixing technique for the center channel is lacking, still. Tonmeister courses are churning out professionals with knowledge in basic psychophysics, but many mixing 'gurus' haven't got the memo it seems.

 

[Miker 1102]

Mr Negativity, also self named "Krabapple" again.
I named 3 of the more popular surround engineers of which Steven was only one.
Now of course your entitled to your opinion but it certainly doesn't match the greater number of 5.1 and Atmos listeners.
The fact that Steven is chosen by so many artists and labels to handle the surround mastering chores must say a little something about his reputation in the industry.
Then there is the fact that you can please some of the people all of the time, and all of the people some of the time.
You know the rest.

Newer processors like my Marantz AV20 offer listeners using Audyssey DEQ the ability to either defeat that feature or set it to 3 different levels surround channel boost.
I've also heard that updates maybe coming for older units.

You cannot listen to surround music and objectively not like Stevem Wilson. The guy is brilliant at both mixing and playing. You cannot please everyone but you also cannot fix bad taste.

 

[goat76]

You cannot listen to surround music and objectively not like Stevem Wilson. The guy is brilliant at both mixing and playing. You cannot please everyone but you also cannot fix bad taste.

What is it in particular you like about Steven Wilson’s Atmos mixes? Can you describe in detail what it is that makes you think his Atmos mixes stand above other engineers' Atmos mixes?

 

[krabapple]

PL2 - a classic, broadband active-matrix upmixer. Does gain-riding at speaker level. With CW=0 it just clamps down the broadband L-R speaker feeds for central signals.

Sorry, I'm not clear what you mean here.

DSU adds fancy analysis (STFT filterbank with crosscorrelation analysis) and ERB based steering. Less pumping, less stage collapse.

What 'center gain' setting are you using? How would you describe the difference between 'Music' and 'Cinema' mode, if you set 'center gain' to 0.0 (or was it 1.0 for pinpoint images?) in Music mode? Asking because with Neural:X we are stuck in the 'Cinema' mode... There should be an automatic switch-over for 'music-like signals', but like any automation, it's a crapshoot.

I'm not using Dolby PL2 now, alas, because modern Denon/MArantz AVRs don't offer it except at the 'budget' level, where other DSP I want (e.g., advanced Audyssey) is not offered..

I pretty much only used DPL2 Music mode, when I did have a DPL2 AVR. I was happy with its default Center Width, which IIRC was 3. My recollection is that DPL2 Movie mode did not let you adjust Center Width at all, and sounded like CW=0, but I could be misremembering.

There was an article decades ago that went into the details of Music Mode vs Movie Mode but not at hand right now, and I'd have to do some googling to find it/them....

ok, here's what I was remembering , from Secrets of Home Theater . It reports that the three user adjustable parameters -- Center Width, Panorama, Dimenson -- were onl,y available in Music mode:

Product Review - Dolby Surround Pro Logic II - The Technology and the Sound - March, 2001​