• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Can dialog be improved ?

BN1

Active Member
Joined
Feb 9, 2020
Messages
121
Likes
66
New to multi-channel HT, old ears, damaged hearing, not a discriminating listener but improving. How to approach the problem of difficulty in hearing movie dialog ? Yes, I use sub-titles, working on room acoustics, turned up the center channel, used YPAO, put a rug down, absorbing some first reflections, played with EQ, etc, etc. I have a 5.1.2 system using Elac Debut 5.2 L/C/R (BS, CC, BS) driven by Yamaha Aventage RX-A660 (80 wpc w/2 driven). TV dialog (news) is fine but movie mix is the primary problem, even with BD. Willing to upgrade if necessary but is there a "clear" solution out there in the hardware or do I continue working on the finer details ?
 

Kal Rubinson

Master Contributor
Industry Insider
Forum Donor
Joined
Mar 23, 2016
Messages
5,270
Likes
9,767
Location
NYC
I wonder if it is the CC and/or its placement. That is a typical inexpensive 2way CC with an MTM driver arrangement. These are prone to irregular horizontal FR variations. If possible, you should consider replacing it with another Debut 5.2 bookshelf, if possible, and placing it at ear level.
 
OP
BN1

BN1

Active Member
Joined
Feb 9, 2020
Messages
121
Likes
66
I wonder if it is the CC and/or its placement. That is a typical inexpensive 2way CC with an MTM driver arrangement. These are prone to irregular horizontal FR variations. If possible, you should consider replacing it with another Debut 5.2 bookshelf, if possible, and placing it at ear level.
L/C/R tweeters are pretty close to ear level and cc is pulled out to edge of cabinet plus only two of us so mlp is approx. on-axis. The Elacs do very well with 2-channel music but, my impression, only "ok" with HT. It just seems that center channels get very little love from speaker designers. Most reviewers spend a lot of time talking about music impressions but limited time is spent on the cc. Terms like "clear" dialog or "intelligible" voices are about all you get. It seems that a 3-way design might be the best for dialog, something that crosses over the drivers outside of the normal dialog range. I realize that specific gains will be small and likely an accumulation of many things rather than a single hardware solution.
 

Fluffy

Addicted to Fun and Learning
Joined
Sep 14, 2019
Messages
856
Likes
1,423
I do not have a HT system, but I can speak from a point of view of someone who helped mix several short films, and recorded dialog for feature films.

In the context of spectrum, dialog inhabits a certain bandwidth. You can cut frequencies below 100 hz and above 6khz and have perfectly sounding and intelligible dialog. Some applications like telephone lines push this even further. But if you are hard of hearing inside of this range, you could have problems understanding dialog, because different syllables inhabit different parts of the spectrum, and different people have a different spectral signatures to their voice. So it could be beneficial to have yourself tested to see exactly which parts of the spectrum are damaged, and use EQ to elevate those parts in your system. Though I don't think a simple high/low boost will solve it, it will probably require custom graphic or parametric equalizing.

It could also be a problem of balance between speakers, when the middle channel is too low in the mix. I see that you already tried to turn up the center channel, so I don't think this is the issue. I think though it's important to remember that the center channel is not exclusively for dialog, but contain certain sound effects and music. So boosting it will sometimes boost other sounds that will still mask the dialog.

I would also like to mention that English for me is a second language, so even with crystal-clear dialog and good hearing, I also sometimes struggle to understand what people are saying in movies. Most dialog is recorded on location and not in a studio, and acting performance is dominantly preferred over pronunciation in the editing room. Weird accents and actors without diction can make understanding dialog difficult. Using subtitles is definitely acceptable, and a lot of countries outside of the US\UK use them regularly.
 
OP
BN1

BN1

Active Member
Joined
Feb 9, 2020
Messages
121
Likes
66
I do not have a HT system, but I can speak from a point of view of someone who helped mix several short films, and recorded dialog for feature films.

In the context of spectrum, dialog inhabits a certain bandwidth. You can cut frequencies below 100 hz and above 6khz and have perfectly sounding and intelligible dialog. Some applications like telephone lines push this even further. But if you are hard of hearing inside of this range, you could have problems understanding dialog, because different syllables inhabit different parts of the spectrum, and different people have a different spectral signatures to their voice. So it could be beneficial to have yourself tested to see exactly which parts of the spectrum are damaged, and use EQ to elevate those parts in your system. Though I don't think a simple high/low boost will solve it, it will probably require custom graphic or parametric equalizing.

It could also be a problem of balance between speakers, when the middle channel is too low in the mix. I see that you already tried to turn up the center channel, so I don't think this is the issue. I think though it's important to remember that the center channel is not exclusively for dialog, but contain certain sound effects and music. So boosting it will sometimes boost other sounds that will still mask the dialog.

I would also like to mention that English for me is a second language, so even with crystal-clear dialog and good hearing, I also sometimes struggle to understand what people are saying in movies. Most dialog is recorded on location and not in a studio, and acting performance is dominantly preferred over pronunciation in the editing room. Weird accents and actors without diction can make understanding dialog difficult. Using subtitles is definitely acceptable, and a lot of countries outside of the US\UK use them regularly.
Since you have some experience and I note that you mentioned "cut", which is more effective, to cut surrounding f or raise specific f ?
 

Kvalsvoll

Addicted to Fun and Learning
Audio Company
Joined
Apr 25, 2019
Messages
878
Likes
1,643
Location
Norway
This vocal-dialog problem is caused by the loudspeakers in combination with room acoustics. With controlled directivity speakers the problem is fixed, and you can understand the dialog - at least, it will be much more intelligible.

If the center speaker is old-school with no directivity control, but otherwise of decent design, this can be improved by absorbing all early reflections acoustically. But it will never be as good as a better center speaker.

This question keeps coming frequently, and indicates there are potential for significant improvements in sound quality in typical old-school home theater systems. Sound quality in the movie of course is part of this, but eventually it is the speakers that has limitations due to how they perform in a typical home environment. Dialog clarity and intelligibility is in my opinion kind of a starting point for good sound quality, meaning, if dialog is a problem, there is huge potential for improvements.

The typical low placement of the center also makes the problem worse.

This is now a solved problem. Good controlled directivity speakers, with decent output capacity, placed properly.
 

bigx5murf

Addicted to Fun and Learning
Joined
Nov 6, 2018
Messages
522
Likes
343
Do your sources have dolby/dts mutli channel mixes, or are you upmixing stereo content to multi channels?

Have you ran YPAO? Do you know what adjustments it made?
 
OP
BN1

BN1

Active Member
Joined
Feb 9, 2020
Messages
121
Likes
66
This vocal-dialog problem is caused by the loudspeakers in combination with room acoustics. With controlled directivity speakers the problem is fixed, and you can understand the dialog - at least, it will be much more intelligible.

If the center speaker is old-school with no directivity control, but otherwise of decent design, this can be improved by absorbing all early reflections acoustically. But it will never be as good as a better center speaker.

This question keeps coming frequently, and indicates there are potential for significant improvements in sound quality in typical old-school home theater systems. Sound quality in the movie of course is part of this, but eventually it is the speakers that has limitations due to how they perform in a typical home environment. Dialog clarity and intelligibility is in my opinion kind of a starting point for good sound quality, meaning, if dialog is a problem, there is huge potential for improvements.

The typical low placement of the center also makes the problem worse.

This is now a solved problem. Good controlled directivity speakers, with decent output capacity, placed properly.

Ok, I need the 101 basic course here.

This is now a solved problem. Good controlled directivity speakers, with decent output capacity, placed properly.[/QUOTE]

Can you give some examples or, at least, some specific characteristics of such speakers ?
 
OP
BN1

BN1

Active Member
Joined
Feb 9, 2020
Messages
121
Likes
66
Do your sources have dolby/dts mutli channel mixes, or are you upmixing stereo content to multi channels?

Have you ran YPAO? Do you know what adjustments it made?
Mostly HT, either streamed or BD and dolby/dts capable. Music is typically 2 channel. Have run YPAO (basic, single pt), distances look ok, I typically tweek the volumes based upon listening to the test tones and balance to my ears. Center channel is up about 5 dB. Experimenting with EQ but results seem to be minimal.
 

Fluffy

Addicted to Fun and Learning
Joined
Sep 14, 2019
Messages
856
Likes
1,423
Since you have some experience and I note that you mentioned "cut", which is more effective, to cut surrounding f or raise specific f ?
I didn't mean that you should cut frequencies, I meant that most dialog exists between 100hz and 6khz. You shouldn't outright cut frequencies in your system.

If you already measured your system using a calibration microphone, I guess there shouldn't be missing bands or over excited ones. Therefore, the problem may be your hearing. So boosting up the problematic frequency ranges in your hearing could help (if they are not gone completely…).
 

Kvalsvoll

Addicted to Fun and Learning
Audio Company
Joined
Apr 25, 2019
Messages
878
Likes
1,643
Location
Norway
Ok, I need the 101 basic course here.

This is now a solved problem. Good controlled directivity speakers, with decent output capacity, placed properly.

Can you give some examples or, at least, some specific characteristics of such speakers ?[/QUOTE]

I can try to show which properties affect speech clarity, and then how a speaker affects this.

Speech clarity - in the meaning how easy it is to understand words and sentences - is important in acoustics, for design of auditoriums and such, there are defined standards for acoustic properties of such rooms, standards that defines how to measure and how to interpret and compare what was measured, and how the numbers relate to speech clarity. Decay rate is an important parameter, and this is the one that is important for hifi/home theater reproduction, because frequency range, noise and distortion can be assumed to be within acceptable limits.

Early reflections and too much early energy can be seen as noise that obscures and disturbs the original signal, and this noise makes it more difficult to understand speech. So we want to keep early reflection level below certain limits.

We can look at C50 in REW, from REW description:
"The early to late energy ratio in dB, using sound energy in the first 50 ms as the 'early' part. C50 is most often used as an indicator of speech clarity."

One speaker is like this:
anitkk clarity C50.png


A different speaker in the same room is like this:
F205 C50 h.png


And this last speaker in a different room:
F205 C50 Rom2 h.png


So we see that both speakers and room can create different results.

Now, this is a quite lean and crude analysis, if we look at the IR - impulse response, we can see a better imsage of what is going on very early in time. The first speaker:
antikk IR.png


The other speaker, in the same room:
F205 IR h.png


This other speaker in the other room:
F205 IR Rom2 h.png


Now we see that the speaker has more influence early in time, compared to the room.

This second speaker has a much better directivity control, which gives a huge reduction in early reflected sound. This again, gives this speaker better speech clarity.

From this one may conclude that the speaker with very narrow radiation is the best. But a speaker with laser-like radiation will not sound nice, and you would need to sit exactly on-axis. So the best radiation pattern is something that covers as much of the room as possible, and at the same time avoids sending sound into boundaries that causes early reflections.
 
OP
BN1

BN1

Active Member
Joined
Feb 9, 2020
Messages
121
Likes
66
Ok, I'm beginning to understand your point. Allow me to jump to a conclusion - typically, the LF and RF speakers have some degree of flexibility in positioning and toe-in but the center channel does not. Would this imply that a virtual center might, in fact, produce a better vocal clarity than an poorly designed center speaker ?
 

Kvalsvoll

Addicted to Fun and Learning
Audio Company
Joined
Apr 25, 2019
Messages
878
Likes
1,643
Location
Norway
Ok, I'm beginning to understand your point. Allow me to jump to a conclusion - typically, the LF and RF speakers have some degree of flexibility in positioning and toe-in but the center channel does not. Would this imply that a virtual center might, in fact, produce a better vocal clarity than an poorly designed center speaker ?

Yes, that may be. Partly because the L-R speakers are better - more smooth freq response, smoother off-axis, but also because center speaker position often becomes very compromised - low close to floor with table between center speaker and listener.

This is easy to test in your own system, simply configure with "no center" option.

For one listener - the one sitting exactly centered between the speakers, the center image will be equal, very solid and real, exactly how depends mostly on the speakers and also on room acoustics. For listeners off-center the image will collapse to the nearest speaker, but that has little to do with speech clarity and ability to understand dialog.
 
Top Bottom