I agree with this. There is some average going on, I just don't know where it is.
You must be my twin brother, because this is what I believe as well.
it probably compensates for silent portions of a song and skips it, thats not difficult to do in programming.
someone feed it a silent song and see if it processes it faster than another song of the same size and length but with audio in it, if the result is a very high number on the silent song then we have problems.
someone feed it a silent song and see if it processes it faster than another song of the same size and length but with audio in it, if the result is a very high number on the silent song then we have problems.
I don't know about this because songs start out lower in level and end with a fade out or on a reverb decay.
I don't think the program could skip that?
there is one way to find out, do the above test, usually there is a 3 second delay for skipping silent portions of a track, it waits before it hears 3 seconds then it starts to skip ahead.
though averaging the score does have the same effect, I'm just afraid of it losing accuracy of an albums true dynamic range, say for example an eminem album which you would expect to have very nasty compression going on being weighted in a favourable light simply because the track has 2 songs that are partially silent and one hidden track that is silent for 15 minutes........
Like on hip hop tracks that have women crying in boots of cars or rappers talking for 3 minutes at a time.
if they are going to implement an automatic database system then this might need to be adjusted for by including a silence detector.
if it has that silence detector then I can believe the DR score more.
though averaging the score does have the same effect, I'm just afraid of it losing accuracy of an albums true dynamic range, say for example an eminem album which you would expect to have very nasty compression going on being weighted in a favourable light simply because the track has 2 songs that are partially silent and one hidden track that is silent for 15 minutes........
Like on hip hop tracks that have women crying in boots of cars or rappers talking for 3 minutes at a time.
if they are going to implement an automatic database system then this might need to be adjusted for by including a silence detector.
if it has that silence detector then I can believe the DR score more.
Last edited:
The RMS value of a track doesn't skip the quiet parts, it counts them too. That would mean that the RMS value of a track is lower than just the average parts of the music - minus the silence.
However, in practice, if I remove the silences the value does not change much. Most popular music has very little silence in it. So in practice it's all averaged out.
That leaves me still not understanding the numbers in the DR database, unless they simply don't relate to decibels. That's possible.
However, in practice, if I remove the silences the value does not change much. Most popular music has very little silence in it. So in practice it's all averaged out.
That leaves me still not understanding the numbers in the DR database, unless they simply don't relate to decibels. That's possible.
Dec. 13 guys.
Commercial Advertisement Loudness Mitigation Act - Wikipedia, the free encyclopedia
First TV, then radio.
Then maybe they can start mastering digital audio the way it was intended....for incredible dynamic range!
This is an example of why capitalism and the free market isn't always the best philosophy.
Commercial Advertisement Loudness Mitigation Act - Wikipedia, the free encyclopedia
First TV, then radio.
Then maybe they can start mastering digital audio the way it was intended....for incredible dynamic range!
This is an example of why capitalism and the free market isn't always the best philosophy.
The DR test is somewhat useless to judge the actual/experienced quality of compression because it doesn't model psychoacoustics. Wether overly compressed material sounds crushed is a matter of time constants used, and clever producers use the ones which make the sound still punchy in the short time scales while it may be leveleled out "flat" in the longer time scales (2++ seconds).
I think while your overall comment is correct, the use of compression should be replaced with limiting. It is the overactive use of limiting that has caused the loudness war, not compression as many believe.
This article is much too optimistic...
https://community.klipsch.com/index.php?/topic/143320-loudness-war-and-the-dynamic-range-dr-database-some-observations/
From page four of this thread:
"Will the loudness wars result in quieter CDs?
"One [mastering engineer] once turned to me after I made a request for more dynamics and said, 'I have a reputation to uphold, I can't make it that quiet.'"
Will the loudness wars result in quieter CDs? | Technology | The Guardian"
Radio in the U.S. has been regulating loudness for many years already using their own algorithms. So if you subtract out radio from this argument, the only pieces left that I see are mp3's, CDs/SACDs/DVD-As, and the various non-mp3 download services, such as HDTracks, if you discount the effects of streaming online music on the total recording industry marketplace (as I do).
Apparently, the big record companies are the source of the problem. I recommend reading about the cultural issues surrounding recording/ mixing/ mastering engineers practices. Until that culture changes (and good luck with that), the problem will persist and even leak into the next generation of "engineers".
I pay attention to the DR Database site when buying, and it's made a huge difference in my listening library makeup (for popular recordings only).
Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.