- From: David MacDonald <befree@magma.ca>
- Date: Thu, 18 Nov 2004 13:55:08 -0500
- To: <w3c-wai-gl@w3.org>
- Message-Id: <200411181855.iAIIt9wN017111@mail3.magma.ca>
On last week's call I provided my findings and suggestions on Guideline 1.4, particularly as it applied to audio contrast. The question of testability came up. I mentioned that it would be difficult to test it. Gregg suggested a technique of sampling the track as follows: 1) Measure the volume of the track (in DB's) at a point where no one is speaking (background only) 2) Measure the volume when someone is speaking over the background 3) Compare the two measurements to ensure there is at least a 20db difference between the two samples I have no problem with us presenting this as a technique for measuring audio contrast. There are however some serious considerations. I think we would have to set the following conditions: 1) The audio background would need to be at a similar volume in both samples. 2) There cannot be any compression/expansion applied to the track. (Currently, media uses compression which would skew the results of the measurements.) The practical effect of compression is that the background is actually louder when there is no talking than when there is talking. When there is talking the entire recording (background and talking) is squashed down. So under these conditions the background would rarely be the same volume for the sample taken when there is no speaking and the sample taken when there is speaking. However, the recording may be perfectly accessible because when the person is talking the background is squashed and when the person is silent the background is expanded. Perhaps I could work out an algorithm which compensates for compression and would not skew results. Additional info: Here's a simplified explanation of the way compression works. (I'll leave out peak limiting for now which is a special case of compression.) Every radio station, TV station, (and musician) wants their signal to be as loud as possible over the airwaves. They want to use up all the headroom. The way they make that happen is to apply compression to the final mastering of their recordings and then boost the overall compressed signal. Compression marks a threshold, say of -15db. Every part of the recording above the threshold will be squashed at a predetermined ratio, say 2:1. In this example every part of the signal over -15db is half as quiet as it would normally be. This squashes the loud parts of the recording. Then the overall signal is boosted so the entire track fills up the headroom (meaning the signal is now as louder). Excerpt from Spinal Tap the movie (thx Wendy) Nigel: This is a top to a-you know, what we use on stage, but it's very, very special because if you can see... Marty: Yeah... Nigel: The numbers all go to eleven. Look...right across the board. Marty: Ahh...oh, I see.... Nigel: Eleven...eleven...eleven.... Marty: ..and most of these amps go up to ten.... Nigel: Exactly. Marty: Does that mean it's...louder? Is it any louder? Nigel: Well, it's one louder, isn't it? It's not ten. You see, most, most blokes, you know, will be playing at ten. You're on ten here...all the way up...all the way up.... Marty: Yeah.... Nigel: ...all the way up. You're on ten on your guitar.. where can you go from there? Where? Marty: I don't know.... Nigel: Nowhere. Exactly. What we do is if we need that extra push over the cliff, you know what we do? Marty: Put it up to eleven. Nigel: Eleven. Exactly. One louder. Marty: Why don't you just make ten louder and make ten be the top number and make that a little louder? [pause] Nigel: These go to eleven.
Received on Thursday, 18 November 2004 18:55:18 UTC