Stereo with Side Reflections vs Absorption and Upmixing/multi-channel - Does a Best Practice Exist?

[Danaxus]

A lot of you are way deeper down the rabbit hole than I, so I'm really curious what you guys have read and experienced.

Ignoring budget (obviously, or this is over before it starts), is there an optimum room/speaker configuration for music-listening? On the one hand, you can have a 2.x setup, and manage the decay times with eq and optionally, some light treatment, assuming your speakers have good directivity/off-axis response, and let your room reflections give you the sound stage and comfort. On the other hand, you can kill the room with treatment, then simulate it with extra channels and upmixing.

My amateur intuition tells me that scenario 1's benefit is that it's really easy to get working, and will make all music sound good. The downside is you'll never get that feeling of being in a concert hall - the room will always feel narrow. Plus it's perhaps not great for actual multi-channel listening or home theatre.

Scenario 2 on the other hand, seems much more difficult to get right. Upmixing will sound unnatural, especially if you're aware of the sounds you yourself are supposed to be making (but not hearing right, even with diffusion). It should work fine for music that doesn't need a wide soundstage, and should be great for spatial audio / multi-channel audio recordings, as well as home theatre, but maybe not so great for traditional 2 channel wide soundstage tracks.


Is one better than the other? Is the world slowly moving towards a more multi-channel world? Does Upmixing/multi-channel stuff work well even in a minimally/untreated room? Would love to know what people think,

 

[DVDdoug]

On the other hand, you can kill the room with treatment, then simulate it with extra channels and upmixing.

I'd say "mostly unrelated".

With up-mixing you might want lots of treatment/absorption to minimize the short small-room reflections, and then generate artificial reflections of a larger room.

But personally, I do use up-mixing with a "theater" or "hall" soundfield setting and I don't have any treatment.

Scenario 2 on the other hand, seems much more difficult to get right. Upmixing will sound unnatural,

It depends on what you mean by "natural". If you are "hi-fidelity purist" the sound should be reproduced in stereo as intended, and probably with minimal reflections (similar to how it sounded in the recording/mastering studio). But maybe the simulated sound of a large music hall sounds more like live music and "more natural" to you.

 

[ernestcarl]

I don't think you'll find a complete "best practice" guide here that most everyone will agree on all points, but there does appear to be some general agreements out there like avoiding too strong/peaky early reflections at least up to 10 to 15 ms to improve imaging and clarity. This can be achieved in more than one way e.g. using directivity control to steer sound away from nearby boundaries and then optimizing speaker-listener positioning.

I'm in the killing the room and then upmixing to suit one's taste camp -- since that seems easier to implement and integrate in conjunction with stereo-MCH setups -- but there is also no clear one-way best approach here since records (not to mention rooms and speakers) vary. There will be some songs or tracks that just sound wrong with the upmixing algorithm... or you have to modify the DSP so much that it will only work best for that particular record/track. Which is kind of already the same case for EQ tone controls anyway -- you want to be able to control/modify the level of the effect depending on the played media content. Not all signal processing effects will work best for all available stereo music. I usually just try to find a mostly satisfying (yet not perfect) medium setting. And only on certain occasions modify the DSP to taste.

 

[Danaxus]

With up-mixing you might want lots of treatment/absorption to minimize the short small-room reflections, and then generate artificial reflections of a larger room.

But personally, I do use up-mixing with a "theater" or "hall" soundfield setting and I don't have any treatment.

That's what I was thinking. I got the feeling that treatment would make spatial audio (or anything mixed in surround), sound "better", or at least give a clearer 3D soundstage, but would make stereo music sound simulated - i.e. the artificial reverb won't be convincing.

It depends on what you mean by "natural". If you are "hi-fidelity purist" the sound should be reproduced in stereo as intended, and probably with minimal reflections (similar to how it sounded in the recording/mastering studio). But maybe the simulated sound of a large music hall sounds more like live music and "more natural" to you.

I'm not a purist and don't really care about the "right" or "wrong" way to reproduce music. I "just" want it to envelope me, feel like I can reach out and touch the singer, place the individual performers of a quartet in front of me, or feel like in a lively hall, listening in a full-on orchestra. I don't want be reminded that the music is coming from a bunch of speakers placed around the room, or hear a digital echo instead of natural reverb. I'm honestly not sure if heavy treatment will give me that with a clear phantom centre due to controlled reflections, with surround filling in for the room, or if tamed modes and untamed reflections will give me the feeling of music surrounding me, placing me in the soundstage. I don't mind sitting in a quiet room - if anything I think I'd find it peaceful - I just don't want the music to sound thin, distant, like a wall of sound in front of me, with dead silence everywhere else, or worse, sound coming from discrete locations wherever speakers happen to be, with big silent gaps in between.

I'm in the killing the room and then upmixing to suit one's taste camp -- since that seems easier to implement and integrate in conjunction with stereo-MCH setups -- but there is also no clear one-way best approach here since records (not to mention rooms and speakers) vary. There will be some songs or tracks that just sound wrong with the upmixing algorithm... or you have to modify the DSP so much that it will only work best for that particular record/track. Which is kind of already the same case for EQ tone controls anyway -- you want to be able to control/modify the level of the effect depending on the played media content. Not all signal processing effects will work best for all available stereo music. I usually just try to find a mostly satisfying (yet not perfect) medium setting. And only on certain occasions modify the DSP to taste.

That makes sense - I'm guessing the stuff mixed for surround sounds better to you, right? Btw, what is an "MCH" setup? Do you mean main channel (centre speaker)? Also, out of curiosity, what led you decide on room-killing route instead of a more natural-sounding room?

 

[ernestcarl]

I'm guessing the stuff mixed for surround sounds better to you, right?

The surround audio content I consume mostly comes from video media e.g. Blu-ray and streaming Netflix. I don't listen to a lot of dedicated multichannel surround records -- so mostly just upmixing the usual stereo recordings to surround. However, I also have not had too many issues with the few surround music mixes that I've heard -- other than the occasional mix where there's some LFE+BM Sub summing levels imbalance.

what is an "MCH" setup? Do you mean main channel (centre speaker)?

Uhhh, MCH is simply short for multichannel. "Mains channels" for me would refer to the front left, center, and right (LCR) channels.

what led you decide on room-killing route instead of a more natural-sounding room?

More "natural sounding room" could possibly mean anything... since rooms vary so much e.g. wall-to-wall carpeting and insulated drywall construction ~vs~ minimalist living room space and furnishings with reflective leather chairs, lots of glass and hard concrete wall surfaces.

Hmmmn... Perhaps a room with maybe something like the ff. characteristics fits your idea of a "normal" room?

Wharfedale Linton Heritage with Audiophonics HPA-S250NC first REW meauserements

Mdat file.

www.audiosciencereview.com

That's a room where reflections largely dominate what's being heard. The speakers' directivity and bare walls with EQ still produces an acceptably flattish response. However, imaging focus and transients will be effectively "blurred". Is that what an ideal "natural sounding room" should look like? I dunno... My family room which is also an also open-plan space is much, much dryer than the linked example despite having no acoustic treatments installed at all.

My own main listening room is not, of course, entirely anechoic (i.e. literally "room-killed) despite being somewhat dryer than most spaces. Generally the RT and reflection-decay is quite evenly flat. Also, the most egregious low frequency modal resonance peak(s) -- around 25 Hz -- is suppressed with EQ. It would be very difficult to impossible to correct that with passive treatment alone.

The "natural" RT levels in my listening room probably should be around 250-300 ms, but with (excessive?) absorption it's mostly between 120-130 ms and then rising steadily in the lower frequencies.

RTx without upmixing:


RTx with upmixing:



One can also visualize the after-effect objectively in other plot views such as the spectral decay:



The wavelet spectrogram would show even finer details...

The actual bandwidth range, levels, and FR shape of the introduced surrounds and center channels can be individually modified and refined to taste. At least that is what I do using JRiver in combination with convolution filtering.

With additional third-party VST plugins one can apply even more tailored reverb effects and match their strength or "wetness" per channel to taste...

While the implementation may involve quite a steep learning curve and lots of tinkering or experimentation -- plus, there may be hardware and software limitations e.g. commercial AVR vs home theater PC -- overall, I still believe a "dryish" MCH music playback room setup has quite the advantage, despite its naysers.

 

[Danaxus]

*A lot of great info!*

Thank you for sharing your room setup - it's really interesting and helpful being able to see how others have thought out their rooms.
I'm nearing the end of my room design journey, and I'm getting the pre-committal jitters. I keep telling myself that front/back wall treatments will only do good (that at least, seems to be universally accepted), and allow me to help tame some of the room modes. I can always replace the rear BAD panels with 3D quadratic diffusers, or even just a wood panel, if I feel the room is uncomfortable. The side wall treatment can be torn down at any point - it'll be made of a frame attached to the wall, so a little filler and a coat of paint, and poof - it's gone.

I think my fears are brought about by the fact that I've never been in a room that's been "properly" treated before. In my current room, sound is just there - you kinda get a phantom centre, sorta, if you sit just right, but otherwise everything sounds clear, but not defined. I assume it's like enjoying the effects in an 480p movie, without realising that 4K exists and just how mind-blowing that is by comparison. The one time I was in a treated room, was an audio shop that had covered the entire basement in treatment. I demoed a bunch of speakers in there, but every single one of them sounded muffled, like there was a thick blanket covering the tweeter. Knowing what I do now, I can only assume that treatment itself wasn't to blame, rather it was probably thin treatment, the extensive absorption of the highs, allowing the mid-lows and bass to run rampant. The experience clearly traumatised me though, and I'm terrified of doing the same thing to the room I'm designing for my dad. But I want to try to build the best I possibly can for him.

So yeah, I think I'll just commit to the treatment plan I laid out in my other thread, and fix things later if it goes to hell.

I think I'll set up the speakers as laid out in the Auro-3D guide (skip the centre channel, but include a single overhead speaker) - I really like what I'm hearing about it, and if its upmixing is as good as I hope it will be, hopefully it will let me have whatever room sound I want. Your room seems to be in a really good position to transition over to Auro-3D as well - was that by design?

 

[ernestcarl]

Your room seems to be in a really good position to transition over to Auro-3D as well - was that by design?

No -- maybe coincidence? -- since this basement space was already "finished" by the previous home owner before I bought the place -- he also snaked cabling through the walls for a MCH setup, but the outlets are all in unsuitable places for me personally. I would have added a lot more features if it were me originally doing the whole basement space e.g. soundproofing the HVAC system and door to its utility room -- as well as isolating the ceiling better where when kids are running/jumping above the family room it shakes my dropdown projection screen and projector itself. Minor annoyances, really, in the grand scheme of things.

 

[Philbo King]

A lot of you are way deeper down the rabbit hole than I, so I'm really curious what you guys have read and experienced.

Ignoring budget (obviously, or this is over before it starts), is there an optimum room/speaker configuration for music-listening? On the one hand, you can have a 2.x setup, and manage the decay times with eq and optionally, some light treatment, assuming your speakers have good directivity/off-axis response, and let your room reflections give you the sound stage and comfort. On the other hand, you can kill the room with treatment, then simulate it with extra channels and upmixing.

My amateur intuition tells me that scenario 1's benefit is that it's really easy to get working, and will make all music sound good. The downside is you'll never get that feeling of being in a concert hall - the room will always feel narrow. Plus it's perhaps not great for actual multi-channel listening or home theatre.

Scenario 2 on the other hand, seems much more difficult to get right. Upmixing will sound unnatural, especially if you're aware of the sounds you yourself are supposed to be making (but not hearing right, even with diffusion). It should work fine for music that doesn't need a wide soundstage, and should be great for spatial audio / multi-channel audio recordings, as well as home theatre, but maybe not so great for traditional 2 channel wide soundstage tracks.


Is one better than the other? Is the world slowly moving towards a more multi-channel world? Does Upmixing/multi-channel stuff work well even in a minimally/untreated room? Would love to know what people think,

Our living room surround system (4.1) occupies about a 15x15 foot area with a vaulted ceiling from 12 feet to 15 feet at the peak, though the room itself is far larger (roughly 30x40 foot). It has no treatment other than carpet, couch and loveseat. So we listen in a more or less midfield distance.

I was a little concerned that the lack of treatment would be an issue, but the direct sound far outpowers the room reflections/reverberations.
My wife loves it, except for the occasional time in a movie where someone knocks on the door in the surround speakers and she gets up to find no one at the front door.

For a smaller room some absorbers might be useful, at least at SBIR points.