Avid Pro Audio Community

Avid Pro Audio Community

How to Join & Post  •  Community Terms of Use  •  Help Us Help You

Knowledge Base Search  •  Community Search  •  Learn & Support


Avid Home Page

Go Back   Avid Pro Audio Community > Pro Tools Post Production > Post - Surround - Video

Reply
 
Thread Tools Search this Thread Display Modes
  #81  
Old 07-11-2007, 01:06 PM
Branko's Avatar
Branko Branko is offline
Member
 
Join Date: Jan 1999
Location: Lisbon, Portugal
Posts: 608
Default Re: Loudness and room-size was: TV-Mix for Theatre

I was not that much worried about perceived loudness, so our tests are trying to achieve a much simpler goal - we're trying to find how wrong is the use of pink noise/RTA system. We believe that it is already proven that RTA measurement is far from perfect and we concentrate our efforts to find a measurable error related to room size or RT60, or to whatever these experiments will reveal.
Essentially, our experiment is to align as many as possible different rooms using the conventional (RTA) method, then measure the differences using different test material. Our test programme consists of pink noise followed by several excerpts from dialog, music and combined film soundtracks. We expect to see differences (at the moment we focus our attention to amplitude differences) in playback of the material and to be able to draw some conclusions. Recordings (of this test material) from various studios will be normalized using the pink noise as reference. Further analysis will bring, we hope, some useful results.
If you want to participate, email me for further instructions.
Branko
__________________
------------------------
Branko Neskov c.a.s.
www.loudness-films.pt
Reply With Quote
  #82  
Old 07-12-2007, 03:25 PM
georgia georgia is offline
Member
 
Join Date: Aug 1999
Location: NY,NY
Posts: 1,859
Default Re: Loudness and room-size was: TV-Mix for Theatre

Hi Branko,

I down loaded the files and tried the experiment... I'll email you the results early next week. But basically we hit almost exactly the same level as the Warner dub stage guys. ( i think it was the warner dub stage..)

cheers
geo
__________________
georgia hilton CAS MPSE MPE

Hilton Media Management

Film Doctors http://www.filmdoctors.com
Me... http://georgiahilton.webs.com/
Stage 32 http://www.stage32.com/profile/6569/georgia-hilton
My Production Company http://www.hiltonmm.com

CREDITS (partial) http://www.imdb.com/name/nm0385255/resume
MEMBER: IATSE LOCAL 700
Reply With Quote
  #83  
Old 07-17-2007, 03:45 AM
Branko's Avatar
Branko Branko is offline
Member
 
Join Date: Jan 1999
Location: Lisbon, Portugal
Posts: 608
Default Re: Loudness and room-size was: TV-Mix for Theatre

Thanks, we´re having 7 recordings so far. Btw, what Warner stage are you referring to? It would be nice to have some tests done there too...
Branko
__________________
------------------------
Branko Neskov c.a.s.
www.loudness-films.pt
Reply With Quote
  #84  
Old 12-30-2010, 06:09 PM
garnoil garnoil is offline
Member
 
Join Date: Jul 2002
Posts: 1,261
Default Re: Loudness and room-size was: TV-Mix for Theatre

Quote:
Originally Posted by Frank Kruse View Post
I thought of starting a new topic instead of hijacking the other.

Speaking of wich there definitely is some connection between roomsize and "felt" loudness. Example: When I listen to a master mixed on a "real" cinema dub-stage in my way smaller edit-room (5.1) wich is also calibrated to 85 even normal dialog almost blows my ears off. Allthough in theory 85 spl at my listening position is EXACTLY the same 85spl on the large scale cinema. When I edit I have found a good level translation at 79spl wich "feels" equivalent to the level I hear when I listen to the exact same sassion 1:1 on the dub-stage. I think Holman has mentioned this problem in one of his books. If I would work at 85spl in my small room and then walk to the dub-stage all my levels would be way to low for someone to have a preview of what I intendet the elements to be.

I always wondered if there are some white-papers that contain investigation on the relationship between the room-size and the "felt" loudness at a given spl level. It makes absolutely no sense to calibrate a 15squaremeter-room to 85dBSPL. The mix will definitely trun out too soft I think.


Of course it´s a totally different cup of tea when it come to mixing and not track-laying but still, there must be a physical reason for this. I doubt that it has to so with hearing psychology. Maybe something with the relation between the listener-speaker-distance in relation to the room-size wich leads to higher direct sound while a large scale stage induced more indirect sound at the listener´s position wich makes program played at 85 seem WAY louder in a small room that in the large one.

Any fundamental info on this problem would be of interest for me.

Frank.
Frank, I just came up this post and though I would share something with you regarding the difference in level "perceived" level in a small room and a big theatre (as you point above).

The main thing is that any number such as 85dbSPLC as read by an SPL meter is in a way "subjective" to the overall frequency response of the speakers reproducing the Pink test tone.

For example, this is what happens and you can try this yourself: if you were to take a very small speaker that has a very limited frequency response at the lower end -for example the Tannoy Reveal: 6.5” with frequency response of 60Hz – 30kHz and you use that speaker to calibrate a room at 85dbSPLC, you would get a "wrong or miss-leading reading of the SPL even it the meter says 85dbsplC.

This is why: The Pink noise being reproduced by the limited response speaker is being affected, in essence the low end of the audio spectrum where a lot of the power exists, is being truncated by the inability of the speaker to reproduce the low end at normalized power. So, when you adjust the level of the amp to read 85dbsplC, the meter is reading/reacting/measuring the sound MINUS the acoustic power of the low end of the spectrum (because the speaker can not reproduce it correctly). Therefore, for the meter that is design to "over all acoustic power metering regardless of frequency distribution", to actually meassure 85dbsplC, you have to deliver a higher sound power. So at this point, once the meter is reading 85dbsplc people thinnk, great now I am aligned like a theatre NOT TRUE. If you then just unplug that speaker and replace it with a full range speaker (frequency response from 40 HZ and up) you will see the SPL meter jump up about 4 dbs. So really, the meter now says that you are at 89dbsplC *which is absolutely true*. In other words, the room when aligned with the small speaker was "miss-aligned" by 4db TOO HOT. Therefore, if you now mix in the misaligned room, your mixes will translate to the theatre about 4db too soft!

There is another very important consideration that does not have to do with the frequency response of the speakers but rather the positioning. It has t do that in a small room, most people use near fields (or mid fields at best). These speakers are located by definition to approximate radiating into full space, meaning they are away from walls and floors and ceilings. Theatre/dub stage speakers are located to radiate into 1/2 space, they are usually either in a soffit or mounted flat on very large walls. When a speaker radiates into 1/2 space (meaning the have a huge wall behind them), they are affected by SBIR (speaker boundary interference response) which essentially re-directs the lower end of the audio spectrum towards the front as an immediate reflection from the rear wall. Because the wall is only a few feet away from the rear of the speaker and low frequencies have very long waveforms, the reflection from the wall become a positive additive to the source wavefront, therefore re-enforcing the low end of the spectrum. The result of this SBIR is very similar to my example above from the little speaker but of course in reverse. The pink noise at -20dbfs played in a theatre or alignment suffers from "extra" bass content as a result of SBIR and therefore the meter reading 85dbsplc is reading a little higher than it should (I don't know how many dbs higher and I really don't want to do the math because it is very complex because it has t be done by frequencies similar to figuring out room modes).

So, the combination of these above 2 examples can yield a very large error in the room alignment. Examlple, the room aligned with the small speaker would sound may be close to 6 db lower in volume in a theatre (I am mentally combining/guessing at these numbers). So this is why, people say: in a small room mix at 79dbsplc.

If the small room has full frequency speakers (same as the theatre), then the missalignemet will occur from the SBIR that occurs at the theatre. There are several other factors which I will only name: the slew rate of the room which is the time it takes for a large room to become fully energize and achieve steady state (one can think of this as the opposite of Reverb time because it is on the onset of sound), the rate of the decay of sound in non-free space which is frequency dependent, air absorption which I would actually think of as "free field sound dispersion" of high frequencies (which behave like more like ray acoustics) and 1/2 field sound dispersion of low frequencies (which behave more like wave acoustics).

To try a test:

1_.-Take a very small speaker (auratone like) and put Pink and meassure SPL.
2.- Replace small speaker with better frequency response unit and meassure SPL.
3.- Replace speaker with full frequency and meassure spl.

To test for SBIR:

1).-Take full frequency response speaker and place flat onto large wall. Run low frequency tone (Limited Pink noise is fine) and meassure SPL.

2).-Move speaker 10 feet ahead of wall (suspend on mid air using thin stand) and meassure SPL.

Now, these SPL variations that you will see are for 1 (one) speaker, now you can just imagine what happens when you multiply the error by 5 speaker + the sub. Again, we are dealing with acoustic power and therefore spl reading as decibels can not be mathematically added, they need to be combined by logs but just by experience I guess thatif you have a 3 db spl error on 1 (one) speaker, the entire system will probably have a 4 to 4.5 db general error.

G
Reply With Quote
  #85  
Old 12-30-2010, 08:02 PM
garnoil garnoil is offline
Member
 
Join Date: Jul 2002
Posts: 1,261
Default Re: Loudness and room-size was: TV-Mix for Theatre

Quote:
Originally Posted by garnoil View Post
Frank, I just came up this post and though I would share something with you regarding the difference in level "perceived" level in a small room and a big theatre (as you point above).

The main thing is that any number such as 85dbSPLC as read by an SPL meter is in a way "subjective" to the overall frequency response of the speakers reproducing the Pink test tone.

For example, this is what happens and you can try this yourself: if you were to take a very small speaker that has a very limited frequency response at the lower end -for example the Tannoy Reveal: 6.5” with frequency response of 60Hz – 30kHz and you use that speaker to calibrate a room at 85dbSPLC, you would get a "wrong or miss-leading reading of the SPL even it the meter says 85dbsplC.

This is why: The Pink noise being reproduced by the limited response speaker is being affected, in essence the low end of the audio spectrum where a lot of the power exists, is being truncated by the inability of the speaker to reproduce the low end at normalized power. So, when you adjust the level of the amp to read 85dbsplC, the meter is reading/reacting/measuring the sound MINUS the acoustic power of the low end of the spectrum (because the speaker can not reproduce it correctly). Therefore, for the meter that is design to "over all acoustic power metering regardless of frequency distribution", to actually meassure 85dbsplC, you have to deliver a higher sound power. So at this point, once the meter is reading 85dbsplc people thinnk, great now I am aligned like a theatre NOT TRUE. If you then just unplug that speaker and replace it with a full range speaker (frequency response from 40 HZ and up) you will see the SPL meter jump up about 4 dbs. So really, the meter now says that you are at 89dbsplC *which is absolutely true*. In other words, the room when aligned with the small speaker was "miss-aligned" by 4db TOO HOT. Therefore, if you now mix in the misaligned room, your mixes will translate to the theatre about 4db too soft!

There is another very important consideration that does not have to do with the frequency response of the speakers but rather the positioning. It has t do that in a small room, most people use near fields (or mid fields at best). These speakers are located by definition to approximate radiating into full space, meaning they are away from walls and floors and ceilings. Theatre/dub stage speakers are located to radiate into 1/2 space, they are usually either in a soffit or mounted flat on very large walls. When a speaker radiates into 1/2 space (meaning the have a huge wall behind them), they are affected by SBIR (speaker boundary interference response) which essentially re-directs the lower end of the audio spectrum towards the front as an immediate reflection from the rear wall. Because the wall is only a few feet away from the rear of the speaker and low frequencies have very long waveforms, the reflection from the wall become a positive additive to the source wavefront, therefore re-enforcing the low end of the spectrum. The result of this SBIR is very similar to my example above from the little speaker but of course in reverse. The pink noise at -20dbfs played in a theatre or alignment suffers from "extra" bass content as a result of SBIR and therefore the meter reading 85dbsplc is reading a little higher than it should (I don't know how many dbs higher and I really don't want to do the math because it is very complex because it has t be done by frequencies similar to figuring out room modes).

So, the combination of these above 2 examples can yield a very large error in the room alignment. Examlple, the room aligned with the small speaker would sound may be close to 6 db lower in volume in a theatre (I am mentally combining/guessing at these numbers). So this is why, people say: in a small room mix at 79dbsplc.

If the small room has full frequency speakers (same as the theatre), then the missalignemet will occur from the SBIR that occurs at the theatre. There are several other factors which I will only name: the slew rate of the room which is the time it takes for a large room to become fully energize and achieve steady state (one can think of this as the opposite of Reverb time because it is on the onset of sound), the rate of the decay of sound in non-free space which is frequency dependent, air absorption which I would actually think of as "free field sound dispersion" of high frequencies (which behave like more like ray acoustics) and 1/2 field sound dispersion of low frequencies (which behave more like wave acoustics).

To try a test:

1_.-Take a very small speaker (auratone like) and put Pink and meassure SPL.
2.- Replace small speaker with better frequency response unit and meassure SPL.
3.- Replace speaker with full frequency and meassure spl.

To test for SBIR:

1).-Take full frequency response speaker and place flat onto large wall. Run low frequency tone (Limited Pink noise is fine) and meassure SPL.

2).-Move speaker 10 feet ahead of wall (suspend on mid air using thin stand) and meassure SPL.

Now, these SPL variations that you will see are for 1 (one) speaker, now you can just imagine what happens when you multiply the error by 5 speaker + the sub. Again, we are dealing with acoustic power and therefore spl reading as decibels can not be mathematically added, they need to be combined by logs but just by experience I guess thatif you have a 3 db spl error on 1 (one) speaker, the entire system will probably have a 4 to 4.5 db general error.

G
I forgot to mention that one of the initial "attempts to a solution" is to have small rooms first professionally aligned for a "flatter" frequency response using an RTA analyzer (much easier said than done). Corrective electronic equalization for a small rooms, especially when dealing with the low end, is very difficult to achieve because the room modal distribution occurs mostly at the octave bands found approximately between 60Hz and 200 Hz (depending on the size of the room). This region of the frequency spectrum is where most of the power of music is contained and even 1/3 octave equalizers are often NOT able to correct modal resonances without negatively affecting its neighbors as well as introducing phase issues as comb filtering occurs. A strong axial mode can exceed a loudness of 12dbs and even higher...on one single frequency and that is just 1 (one) primary axial mode and there are and infinite number of them.

Another problem of aligning a small room for flat response with an RTA before pinking for 85dbsplC, is that in a small room wall reflections "significantly" contribute to the overall response of the room (much, much more that a theatre) and therefore, RT (and especially its relative frequency distribution) becomes of paramount importance. In my opinion, it is often better to "know" your room empirically, and get used to how it sounds, than to try to "correct it" by means that border the laws of physics. An experienced sound person (that knows his/her room) automatically mentally averages and corrects the sound so that the mix will sound good anywhere (of course the room needs to be at least descent and not a marble washroom cube and with good equipment).
Reply With Quote
  #86  
Old 12-31-2010, 07:28 AM
Postman Postman is offline
Member
 
Join Date: Dec 1999
Location: where land meets sky
Posts: 2,375
Default Re: Loudness and room-size was: TV-Mix for Theatre

Hi garnoil, while everything you say is true there is more to the puzzle. Accoustic energy of the low frequencies do indeed affect simple SPL measurements in a big way. This is why you always address room bass trapping, speaker placement, and speaker eq before setting overall speaker level.

Pink noise with restricted bandwidth (high passed at 300hz for example) should mostly overcome the issues you write about, but it does not. Dolby has used a limited bandwidth pink noise for a very long time. In an "ideal" room, the full bandwidth and limited bandwidth noise (with makeup gain applied to produce an equal dbC reading), should measure the same or nearly the same. If the speakers are not capable of the lowest one or two octaves of response, then the limited bandwidth pink alone should be used to set playback SPL. The full bandwidth noise is used with an RTA to check for frequency linearity problems.

"Node" issues, those nasty spikes and dips that are usually invisible on 1/3 octave RTA's, are a non-issue for SPL calibration using pink noise. They are included in the SPL reading. Node issues are much more of a problem with musical tones than with pink noise, and while they do account for poor sound I am not at all certain they explain the problem or even come close to doing so.

We still have a problem of volume perception difference in a small room especially with speakers only a few feet away vs. large room. I've read some theories about energy buildup along the walls, early reflections altering our perception, psychoaccoustic babble about how small rooms increase our awareness through fear, and more. Whatever the causes, we're left with the same problem, which means that either we need a much more accurate way to measure this stuff or for now our standards need to take room size into account. The ATSC's recommendations (soon to be CALM law in the US) do that in a rough way.

Maybe future measurement techniques will take these factors, whatever they are, into account so we don't need to make subjective changes based on room volume. We could all set our dials to "XX" and hear perceive the same volume. That's not how it works today, as you well know!
Reply With Quote
  #87  
Old 12-31-2010, 08:07 AM
garnoil garnoil is offline
Member
 
Join Date: Jul 2002
Posts: 1,261
Default Re: Loudness and room-size was: TV-Mix for Theatre

Hi Postman,

""Node" issues, those nasty spikes and dips that are usually invisible on 1/3 octave RTA's, are a non-issue for SPL calibration using pink noise. They are included in the SPL reading. Node issues are much more of a problem with musical tones than with pink noise, and while they do account for poor sound I am not at all certain they explain the problem or even come close to doing so."

I agree 100% but I was referring to was, what happens in a "typical small size" pre-mix room that has not been RTA flattened (bass traps, electroacustics, geometry) when one simply says play pink noise and set spl to 85 (and then one wonders why the levels do not translate well to a theatre). Specifically I was trying to point out: If one plays non-limited Pink Tone in a room with non controlled modal response, and simply takes an SPL reading, the SPL reading will not be accurate in as much as the entire lower end of the pink noise is being artificially "boosted" by the modal resonances in the room. Therefore, the SPL reading "while it is accurate for the specific room" it is not equal to a reading taken on a large room where no artificial boosting of the lower en occurs (even if both readings say 85dbsplc)
Reply With Quote
  #88  
Old 01-01-2011, 12:57 PM
Postman Postman is offline
Member
 
Join Date: Dec 1999
Location: where land meets sky
Posts: 2,375
Default Re: Loudness and room-size was: TV-Mix for Theatre

Sorry, I thought you were saying that the whole issue of perceived level in small rooms vs. large rooms is due to bad SPL readings due to room nodes and limited bass response of small speakers. If you meant to say that, I do not agree. If you meant to say those factors are important to be aware of when aligning a room, then I completely agree!
Reply With Quote
  #89  
Old 01-05-2011, 08:20 AM
airon airon is offline
Member
 
Join Date: Nov 2000
Location: Germany
Posts: 548
Default Re: Loudness and room-size was: TV-Mix for Theatre

I had that very problem just three weeks ago, as we mixed a short for theatrical presentation. In a test screening, in which the playback chain was set to Dolby standards(7), the whole mix was 2-3 dB more quiet than in our somewhat small room we mixed it in. I'd already tried to compensate for the roomsize by using 82dB as a reference point, but it seems for the bandlimited pink noise (Bluesky file) and even Dolby noise this was not enough. I had to go to 79 dB SPL to get a comparable response in the theater we screened at. Several factors could and probably will have skewed the perceptions. The accoustics in both the theater and mixing room, as well as the speaker placement in the mixing room.

There's always a lot to check.
__________________
"My ego comes pre-shrunk" - Randy Thom
Reply With Quote
  #90  
Old 01-08-2011, 03:49 PM
garnoil garnoil is offline
Member
 
Join Date: Jul 2002
Posts: 1,261
Default Re: Loudness and room-size was: TV-Mix for Theatre

Quote:
Originally Posted by airon View Post
I had that very problem just three weeks ago, as we mixed a short for theatrical presentation. In a test screening, in which the playback chain was set to Dolby standards(7), the whole mix was 2-3 dB more quiet than in our somewhat small room we mixed it in. I'd already tried to compensate for the roomsize by using 82dB as a reference point, but it seems for the bandlimited pink noise (Bluesky file) and even Dolby noise this was not enough. I had to go to 79 dB SPL to get a comparable response in the theater we screened at. Several factors could and probably will have skewed the perceptions. The accoustics in both the theater and mixing room, as well as the speaker placement in the mixing room.

There's always a lot to check.
This is exactly what I referring to. Small rooms by definition will have very inaccurate low end and of course because they are "small", there is hardly ever physical room to install proper bass traps. This will lead to a false SPL reading that will be artificially high because the non-linearity of the low end. One can not say "play pink tone and read 85dbsplc slow" and you are aligned. There is a lot more to aligning a small room to play like a large room, and SPL is not a good measure when comparing both rooms.
Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is On

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Dverb not remembering Room size parameter in 11.0.3 mattrixx Pro Tools HDX & HD Native Systems (Mac) 0 12-09-2013 08:40 PM
post mix theatre vs music console room villagesound Post - Surround - Video 4 10-09-2013 03:40 PM
CALM Acts Audio Loudness, Average Loudness, Program Loudness levels... help needed qr2r Post - Surround - Video 12 11-10-2011 01:41 AM
drum recording room size rems 003, Mbox 2, Digi 002, original Mbox, Digi 001 (Win) 4 04-11-2003 02:45 PM
Control Room Size? seeds Tips & Tricks 24 07-06-2002 03:19 PM


All times are GMT -7. The time now is 11:14 PM.


Powered by: vBulletin, Copyright ©2000 - 2008, Jelsoft Enterprises Limited. Forum Hosted By: URLJet.com