Audio Reference Levels

[pojibonzai]

Does anyone know if Cyan has audio reference guidelines for age creation?
I was thinking there must be some standard to insure that music levels and sfx levels would be consistent from age to age.
I've noticed that the audio levels in some fan ages are inconsistent and was wondering if there were published standards for reference levels.

Also, if there is anyone with expertise in game engine audio processing, please PM me if you have the time to answer a couple of questions.

This is a wonderful forum with so many knowledgeable and thoughtful members, I thank you for your time.

Poji

[Branan]

There are none published, but it's something we might be able to contact them about. I can't make any promises we'll actually get an answer back, but we can try.

[pojibonzai]

After digging around a bit, I found this document from AES (Audio Engineering Society). http://tinyurl.com/7ggnelh
Although It's from 2004 and quite technical, it does address the issue.
I found it to be informative, others may as well.
It seems to me that most games are mixed to be as loud as possible (more like a pop song than a film). Which makes it difficult to create a dynamic and realistic audio environment due to the enormous amounts of compression needed to achieve these louder levels.

I'd be interested to know if any other age creators have wrestled with this issue, and what they use as a reference level.

Thanks Branan, for the speedy response. If you do hear from Cyan on this, I'd appreciate any info they might provide.

Thanks to all

[Deledrius]

Given the amount of control and authority the musician and sound designer had (from what I read in the CWE Plugin docs, anyway) I wouldn't be surprised if Cyan leaned more toward film attitudes in Uru.

At any rate, it would be nice to have more audiophiles around to put down some guidelines for this. Especially if they're also coders who can help with the EAX/EFX code.

[dendwaler]

Because Pojibonzai is a sound " scaper" and not a builder, it is well possible he did not see that there are some possibilities to at least control the volume for every object inside the gameengine.
You can control the volume from " 0 " to "1".
So most important is that the sounds when they are mixed use the full dynamic space and will not clip in any case.(nor been compressed)
Then in game the sound will set standard on a volume level of 0.5.
From that point the builder has the possibility to tweak it due to the circumstances.
A bit louder or abit less loud.

Besides the volume parameter we have the " minimal distance" and " maximal distance" setting.
where the minimal distance is the distance of the avatar where the sound will start to fade from its default setting to a lower volume.
the max distance is the point where the sound reached volume 0.
if you have a bumblebee with an overall low volume.
This will be very loud if the bee is nearby your ear.
In that case the minimal distance where the setted volume is 0.5 is a small distance.
After that it will fade out in a very short distance.
The opposit is the case with a bomb, the volume setting is the same but its min and max setting will be very far from each other.
In RL the bomb will be heard as loud very nearby as from a bit distance because your ears have reached the end of its dynamic range.
You can compare it a bit with radiosignals of your mobile phone.
The loudnes of what you hear through the phone is always the same, it has to do with the modulation index.

How far it reaches is not dependent of the audio loudness, but is dependent of the carrier.
The carrier defines the cell , in which it can be received and on which point the sensivity of the receiver can't longer demodulate the signal.
There is a analogy in this.

In game , you have to set up " rings" (Like planning radiocells)
Inside this ring , (the circle of the cell of the mobile phone) there is full loudness.
outside the ring is a fade away area where hearing (or receiving for the mobiles) will become more difficult.

I hope this adds some value on the understanding of how to deal with volume control.

At the end you see, that what you hear in game is mostly dependent of a correct fading setting, because the source of every sound has in general the same value. its totally dependent of the distant from the source.
In real live every sound is loud if you are only close enough to the source.

[pojibonzai]

Thanks dendwaler!

Good metaphor. Very helpful for understanding source-point audio. (in blender,yes?)
I think that there is still the matter of environmental ambiances and scoring. These elements exist independently with no source point and need to (hopefully) blend in with the various localized audio events to achieve a pleasurable audio experience. A widely used standard in film is 85db. This insures a proper reference level for being friendly to most consumer electronic devices. It also helps to make sure that the dialog isn't overwhelmed by a booming score or that the subtle and nuanced aspects of a score are not trampled by loud effects at lower listening levels, among other examples.

Ultimately the game designer has the final say in the "mix", but as audio content providers, we could better serve your needs (less work for you) if there were some guidelines to follow. (at least, for this composer )

Thank you

[dendwaler]

yes , of coarse you are completely right that every electronic device has a max input level.
Which varies, but follows standards.
I don't know the entry levels no experience with that but for a microphone input its different then for the auxilary entrance of a device.
The microphone entrance is folowed by a preamplifier, which has a weighted gain following a correctioncurve and at the output of the amplifier it has the same level as the entrance of the auxilary.
I think the levels will be about the same as specificated as intrance levels for audiocards.1Veff? , not sure
You have to stay within the dynamic range of the card , but on the the other side as high as possible to get best signal to noise ratio, because the card will also add some noise to a perfect clean intrance signal.
That added noise has a constant value, so is relative higher by low inputs then it is by high inputs.

[pojibonzai]

Line input levels for analog devices are -10dbV for consumer electronics. +4dbU for professional.
digital input levels are usually -20dbV.
Microphone input levels are usually more sensitive,(as you mention) usually around -50dbV