Reverberant room vs "dead" room+dsp

[VietnamRadioEnjoyer]

The problem is - The short reflections-reverb in your living room are nothing like the wonderful natural reverb that you hear from all around in a larger music hall or concert hall.

Maybe a surround system where L&R play music and the surround speakers play the reverbs? a simple script could do that and since reverb settings are in our hands we should be able to get close to hall experience (or better?) with trial and error (& science)

Just to add that this is at least in part due to the fact that large halls contain diffuse reverberant sound energy which simply isn't possible in small rooms.

Further, when we record from a seat in a hall and reproduce the recording with two loudspeakers in a small room the result is that both the original direct sound (which we can normally localize) and the diffuse/enveloping reverberant field is reproduced from the same distinct sound sources (loudspeakers), and both follow the same early reflection pattern in the small room, with no reverberant field to speak of. So naturally we no longer have the means to distinguish the two, the original hall suddenly becomes part of the direct sound in our "small" room.

Maybe a ML Model can remove baked reverb sounds from the recording & add them back in the form of surround or binaural

But concert organizers should get a binaural recording device like this. its 2023

 

[GXAlan]

Yamaha CinemaDSP HD3 with YPAO RSC is actually really, really good for this. If Yamaha came out with a 16 channel processor, I would be first in line. I still have their CX-A5100 for my music room even though I use the Monolith HTP-1 for my movie theater.

The copycat reverb modes from other vendors is horrible. Yamaha is great, based upon its knowledge of the room based upon room EQ, knowledge of the speaker position including height, based upon triangulated measurements like Trinnov and the Sony ZA series, and accounting for volume and room size is tuning the amount of reverb and how the reverb is timed.

User control is quick and easy enough that I can switch it on and off depending on the source recording. There is very little fiddling around needed, though it is there if you want it.

 

[kemmler3D]

So wouldn't it make sense for audiophiles to aim for a well treated "dead" room & only add reverberations when needed using DSP filters?

Regardless of any other points to be made here - Adding reverb via DSP will basically never play the role of room reflections. Even if you get the RT60 and spectrum identical to your room, all spatial / directional cues will be completely different.

It's like bolting a sun lamp to your ceiling and hoping you mistake it for a skylight.

 

[dominikz]

Maybe a surround system where L&R play music and the surround speakers play the reverbs? a simple script could do that and since reverb settings are in our hands we should be able to get close to hall experience (or better?) with trial and error (& science)

I believe multichannel (e.g. Atmos) mixes of music try to achieve that, as do various stereo to multichannel upmixers.

Maybe a ML Model can remove baked reverb sounds from the recording

There are some plugins that do reverb removal (e.g. WavwsClarity Vx DeReverb) but I'm not sure how well they'd work on a full mix.
Cool ideas though, I'd love to see more implementations!

Also concert organizers should get a binaural recording device like this. its 2023

I'm sure you can get binaurally recorded music already, though maybe not as much as one would like.
Note that there are some drawbacks to dummy head recordings as well: not a lot of flexibility in shaping the mix, no head-tracking and no personalized HRFT support.
Binaural technology is however evolving beyond dummy heads to address these limitations and I believe it might have a pretty bright future in the XR world of tomorrow!

 

[kemmler3D]

if we take random peoples random room interactions out of the equation we can have standards & guide lines & customizations based on individual taste.

Perhaps yes, but it's generally impossible to construct an anechoic chamber at home. Even getting a dry-ish room below 100hz is unreasonably difficult / expensive for most people.

It might even improve the sound of headphones & IEMs because they should be the most dry

I've tried IR-based crossfeed on my headphones (basically, a room simulation for headphones) and eventually went back to plain crossfeed. The changes to frequency response and phase smearing were eventually too much to justify, although it does sound cool.

Give it a try: https://github.com/ShanonPearce/ASH-Listening-Set you can use the True Stereo IRs (4 channel) files in EQAPO using this free VST: https://impulserecord.com/convology-xt/

Ultimately, I think it's easy to underestimate how much our brains/ears compensate for / adjust to the room you're in. It's very hard to fake it or trick the brain into 1) not "hearing past" the room or 2) hearing a different room.

 

[GXAlan]

Regardless of any other points to be made here - Adding reverb via DSP will basically never play the role of room reflections. Even if you get the RT60 and spectrum identical to your room, all spatial / directional cues will be completely different.

It's like bolting a sun lamp to your ceiling and hoping you mistake it for a skylight.

+100

This is why the regular reverb stuff doesn’t work well.

BUT this is where Yamaha’s DSP actually works. Built during Japan’s bubble economy, the use of actual measurements taken across the globe to characterize venues is yet to be replicated by any other consumer product.

@restorer-john knows a lot more about these original Yamaha DSP efforts.

 

[VietnamRadioEnjoyer]

I've tried IR-based crossfeed on my headphones (basically, a room simulation for headphones) and eventually went back to plain crossfeed. The changes to frequency response and phase smearing were eventually too much to justify, although it does sound cool.

Give it a try: https://github.com/ShanonPearce/ASH-Listening-Set you can use the True Stereo IRs (4 channel) files in EQAPO using this free VST: https://impulserecord.com/convology-xt/

Ultimately, I think it's easy to underestimate how much our brains/ears compensate for / adjust to the room you're in. It's very hard to fake it or trick the brain into 1) not "hearing past" the room or 2) hearing a different room.

If you want the sound of your speaker+room in your headphone, best solution is the impulcifer+HeSuVi. Impulcifer measures your room with in-ear mic & creates the profile for HeSuVi to use. Since it is your speaker + your room + your ears + your headphones, it should sound natural

 

[kemmler3D]

If you want the sound of your speaker+room in your headphone, best solution is the impulcifer+HeSuVi. Impulcifer measures your room with in-ear mic & creates the profile for HeSuVi to use. Since it is your speaker + your room + your ears + your headphones, it should sound natural

This makes sense for single-position headphone listening. If you don't want to buy in-ear mics the Shannon-pierce responses are edited to have head but not ear filtering, as I understand it. So they work well for headphones.

 

[kemmler3D]

the use of actual measurements taken across the globe to characterize venues is yet to be replicated by any other consumer product.

How do they work? Unless your room is extremely dead, even very nice IRs are still just special effects on top of the original audio...

Even a well-recorded multichannel (5 or 7) convolution reverb is still going to be fighting with your room. It would be (IMO) mighty hard to trick the ear into thinking you were actually in e.g. a cathedral or concert hall without extreme acoustic treatments.


And perhaps it isn't a "consumer" product, but Waves has a plugin that will do this for $30... supposedly with recordings of 2000 spaces. https://www.waves.com/plugins/ir360-convolution-reverb

 

[GXAlan]

How do they work? Unless your room is extremely dead, even very nice IRs are still just special effects on top of the original audio...

Well stereo is a special effect too? Of course, it is a special effect over the audio.

I don’t know all of the technical details although seeing the user configurable options and how they apply them differently to the bed layer versus presence/Atmos layer and Yamaha’s angle measurement on top of distance (“triangulation” but with 4 points) gives it a scaling factor that again, works well.

It focuses on having different parameters for your speakers, and different rates of early vs late reflections and the timing and decay and FR of it.

I never used DSP from any vendor before Yamaha, but kept it on all the time when I ran Yamaha.

It would be (IMO) mighty hard to trick the ear into thinking you were actually in e.g. a cathedral or concert hall without extreme acoustic treatments.

I guess the question is how much your have a suspension of disbelief. I think it is better but even binaural recordings don’t fully capture environments.

And perhaps it isn't a "consumer" product, but Waves has a plugin that will do this for $30... supposedly with recordings of 2000 spaces. https://www.waves.com/plugins/ir360-convolution-reverb

There’s definitely a very different approach to saying that we have reverb that matches a venue versus capturing it almost ambisonic-like.

Steinberg lets you use a pair of stereo mics to do your own measurements like this

True Stereo

Impulse responses recorded as true-stereo files allow you to create a very realistic impression of the corresponding room.

steinberg.help

Anyway, getting back on topic, I do like the Yamaha DSP’s ability to add pleasant reverb that is sophisticated and in a consumer product for playback of music and movies in real-time.

I think the OP’s approach is valid.

 

[kemmler3D]

Well stereo is a special effect too? Of course, it is a special effect over the audio.

I don’t know all of the technical details although seeing the user configurable options and how they apply them differently to the bed layer versus presence/Atmos layer and Yamaha’s angle measurement on top of distance (“triangulation” but with 4 points) gives it a scaling factor that again, works well.

It focuses on having different parameters for your speakers, and different rates of early vs late reflections and the timing and decay and FR of it.

View attachment 322775

I never used DSP from any vendor before Yamaha, but kept it on all the time when I ran Yamaha.



I guess the question is how much your have a suspension of disbelief. I think it is better but even binaural recordings don’t fully capture environments.


View attachment 322771

There’s definitely a very different approach to saying that we have reverb that matches a venue versus capturing it almost ambisonic-like.

Steinberg lets you use a pair of stereo mics to do your own measurements like this

True Stereo

Impulse responses recorded as true-stereo files allow you to create a very realistic impression of the corresponding room.

steinberg.help

Anyway, getting back on topic, I do like the Yamaha DSP’s ability to add pleasant reverb that is sophisticated and in a consumer product for playback of music and movies in real-time.

I think the OP’s approach is valid.

Interesting, this seems as sophisticated as you could hope for from a consumer product. And I guess there's a reason they're selling that plugin for $30, ha.

Anyway, I still am not really comfortable with the idea. It's easy enough to overdo reverb at the mixing stage. To me, adding reverb at the listening stage is like adding more garlic to your food at the table, not just salt. If the producers thought the music needed more space, they would have added it. Why not add a flanger or chorus for fun while you're at it?

If your room is so dead that artificial room acoustics are both viable and desirable... maybe just remove some treatment?

 

[fpitas]

maybe just remove some treatment?

Some small rooms can be pretty obnoxious. I can see adding judicious amounts of reverb in recordings.

 

[GXAlan]

To me, adding reverb at the listening stage is like adding more garlic to your food at the table, not just salt. If the producers thought the music needed more space, they would have added it.

No disagreement from me there. I think it is essential that these types of tweaks are user defeat-able.

But maybe it is closer to the idea of adding gravy to turkey, or ketchup or mayonnaise to fries. Yes, adding sauce to your fries in advance will make them soggy, but have you noticed that different people have different amounts of sauce they prefer? Even your choice of ketchup vs. mayo is a preference.

I will say that for movies, Cinema DSP really convinced me after watching Top Gun: Maverick with the “Sci-Fi” mode enabled. That is Atmos-aware and is called “Sci-Fi” because it maintains precise special effects positioning, so the reverb is subtle. It makes your home feel like a larger room than it actually is without affecting speech intelligibility. Yamaha’s naming scheme are a legacy of 80’s/90’s marketing but the underlying math is good.

For music, showing my kids some of the classic Balanchine Nutcracker ballets (both the 1993 one and NYC ballet movie) also sounded better in the soundfield modes.

For a lot of classic well-recorded audiophile 2 channel recordings from typical labels like Chesky Records, I was happy to have the DSP off or to use Dolby Pro Logic II.

 

[kemmler3D]

maybe it is closer to the idea of adding gravy to turkey, or ketchup or mayonnaise to fries.

Well, I don't know if the analogy holds, because when you cook turkey or fries, you probably serve it with that condiment and expect people to use it.

I fully believe there are settings where the extra reverb sounds great. It's just my philosophical discomfort with the whole thing that throws me off.

 

[Puddingbuks]

As I have a heavily absorbed, nearly “dead” listening room, I think that all the reverb and spatial info is already in the recording.

In my room, some songs sound small and compact between de speakers (little room by Norah Jones), while other songs sound broad, wide beyond the speakers (love’s a stranger - warhaus).

In a lively room with lots of reverb, songs sound less defined and the room info of the recording is less clear.

In a song like creep - radioheads you can almost hear the room itself of the recording.

 

[goat76]

Adding reverb post-production to a full mix will depend highly on the content. It will almost always work pretty well for productions made with just a few microphones where the instruments already share some "acoustic partnership" with each other, but for multi-mono recordings, as most modern recordings are, it will likely be a "hit or miss" situation if the reverb is not reasonably subtle.

It's way better to be able to apply different types of reverbs to different sound objects, or at least apply different pre-delay times to the different instruments in the mix if the same type of reverb is used, which will make them appear closer or farther away to add some depth. Things like that must (obviously) be made during the mixing stage of the production.

I think most recordings already have the "right" amount of reverb. Some music productions simply sound better fairly dry while others sound the best with a more roomy kind of sound, and the best-sounding ones are often the ones that contain the real acoustics of a good-sounding studio room, the ones where you can hear the natural and real depth of the venue and where the instruments were placed in this space.
All of that recorded information is of course best heard in an acoustically well-treated listening room.

 

[NIN]

Reading through some threads here, it seems that the general advise experts give is to not go after room treatment because we crave reflections & reverberations specially in orchestral music,
But for edm & pop music, gaming audio & surround speaker setups, a well treated "dead" room is better. Even for orchestral music some recordings have reverberations baked into the recording itself making a well treated room ideal

Some might prefer no or few room treatment but there are many here that prefer room treatment. I prefer the recordings "reverberation" over my rooms same reverberation on all music + the recordings reverberation. So I would suggest do room treatment.

 

[Justdafactsmaam]

Reading through some threads here, it seems that the general advise experts give is to not go after room treatment because we crave reflections & reverberations specially in orchestral music,
But for edm & pop music, gaming audio & surround speaker setups, a well treated "dead" room is better. Even for orchestral music some recordings have reverberations baked into the recording itself making a well treated room ideal

So wouldn't it make sense for audiophiles to aim for a well treated "dead" room & only add reverberations when needed using DSP filters?
I use easyeffects on linux. Here is its Reverberation filter settings:

View attachment 322671

Shouldn't we be able to fine tune the settings and only have reverberations when we want them the way we want them?
I think "dsp + well treated room" is a more controlled setting compared to relying on room interactions & if some scientific research is done in this area we can have better guide lines

(i don't have experties or the equipment to say anything confidently so this is just an idea after reading some forum threads)

Personally I have zero interest in hearing either the listening room reverb or added DSP reverb. I want to hear the unobstructed reverb and spatial cues on the recording. If some DSP can enhance *that* reverb and spatial information then I’m interested. I want my room as dead as possible.