Previous Page | Next Page
- by Lunar at 12:46 PM EDT on March 18, 2008
- EditorArtist: yeah this is something i brought up before, but it fell on deaf ears at the time and was "not a problem." instead, a highly esoteric "replay gain" was thought to be the solution. even though it isn't because.. well, it's esoteric :) and poses absolutely no advantage over normalising the signal. but whatever, i just have to turn my speakers up every time i listen to those mp3s instead, which is a great shame. all that time spent arduously making that set and it isn't normalised...
- by Mouser X at 1:25 PM EDT on March 18, 2008
- Prime Blue did say why he chose the volume he chose. He said that any higher, and it would start clipping. Perhaps there's a fix for this? I don't know about stuff like that (I only recently learned why clipping even happens...), but that's the reason that Prime Blue gave. If it's fixable, perhaps someone can try to fix it? Mouser X over and out.
- by EditorArtist at 3:39 PM EDT on March 18, 2008
- Hmm, so if he used replay gain, the tracks aren't the intended volumes no matter what way you look at it. Which did seem to be unusually low even for my trying to reason that maybe they were that low so they could be mixed directly with SFX with no volume adjustment. Replay gain sounds like a good idea in theory, but either because of wrong settings or because the official soundtracks don't use replay gain, it hurts more than it helps.
Yes, I read about the clipping issue in the readme, and either he doesn't realize how low most of the tracks are due to the plugin he used are or the mp3 release I downloaded via torrent isn't what he thinks it is.
You'll only get clipping if you try to raise the volume for the song's peak above 0db. I analyzed the waveforms, and at the fifteen I looked at, every one of them could safely be raised 3.4dB to about 10dB without clipping, and would have if they were released on a soundtrack.
edited 4:03 PM EDT March 18, 2008
- by valiant at 3:40 PM EDT on March 18, 2008
- Hey, I can see how people have problems with the low volume - because I do so myself.
I can't see however why Lunar still wants those files normalized.
Replay Gain >> Peak normalization
...at least when it comes to game rips.
See the Wikipedia article for more information.
Bottom line: Replay Gain analyzes the whole file's (let's say "semi-average") volume while peak normalization doesn't give a shit about how loud a song actually is.
I don't understand why it bothers you to turn up the volume when listening to the rip, yet you don't care when you have to adjust it between two tracks of the same soundtrack that have a different average volume.
- by EditorArtist at 4:19 PM EDT on March 18, 2008
- Hey Prime Blue, I edited my last post after you posted yours, taking into account previous conversation on the board which I hadn't read yet. I understand replay gain's purpose, but it's not worth it unless the majority of music albums use it. Unfortunately perhaps, most audio engineers nowadays make their songs peak at 0db because everybody else peaks at 0db. People are used to the differences in average volumes from track to track, but are put off by old recordings for instance that have a much lower rms than what is possible when peaking at 0db (not to mention when using dynamic compression, but I'm not advocating that). Anyway, please reconsider your reasoning for the sake of the average/majority listener, and if you want your rip to feel as 'professional' as possible. ;)
- by EditorArtist at 4:55 PM EDT on March 18, 2008
- Also from that wikipedia article you linked to: "Finally, should the audio at its original levels be desired (i.e. for burning back to hard copy), the metadata can simply be ignored.
Replay Gain implementations usually involve adding metadata to the audio without altering the original audio data. While the Replay Gain standard specifies an 8-byte field in the header of any file, many popular audio formats use tags for Replay Gain information. FLAC and Ogg Vorbis use the REPLAYGAIN_* comment fields. MP3 files usually use ID3v2 or APEv2 tags."
It seems your usage of replay gain isn't the standard. Making replay gain optional so that listeners could either choose to hear the volume as it is on the disc or with normalized perceived loudness seems to be the best solution for the widest variety of listeners.
- by Chupperson Weird at 6:45 PM EDT on March 18, 2008
- Clipping is not good. Compression is not good either, whether it's considered "professional" or not. I like the rip as it is.
- by EditorArtist at 6:57 PM EDT on March 18, 2008
- Nobody's recommending compression or increasing the volume to the point of clipping.
- by Lunar at 7:44 PM EDT on March 18, 2008
- "He said that any higher, and it would start clipping"
No, it wouldn't. Normalisation won't cause clipping.
"Clipping is not good. Compression is not good either, whether it's considered "professional" or not. I like the rip as it is."
Normalisation isn't compression! And whether compression is good or not is a whole other debate. Fortunately not one anyone needs to get into here.
As for tracks designed to play a lot quieter, normalising them may boost them too much. In that scenario, manually increasing the signal by a few db instead would be fine. Leaving stuff peaking at -10db is stupid whatever way you look at it though, period. Not everything supports ReplayGain, forcing people to use players that do is silly. I'm damned if Winamp will work for me, even if there's a ReplayGain option (I hear no increase in volume.)
"I don't understand why it bothers you to turn up the volume when listening to the rip, yet you don't care when you have to adjust it between two tracks of the same soundtrack that have a different average volume"
Simple; I don't have to do it with most decently made rips/albums. The tracks are balanced appropriately and finished to an acceptable universal volume that doesn't rely on a proprietary gain technique.
EDIT: just realised EditorArtist pretty much said all this already, but it doesn't hurt to hammer the points home.
edited 7:48 PM EDT March 18, 2008
- by Chupperson Weird at 10:01 PM EDT on March 18, 2008
- For the record, I do realize the difference between normalization and compression.
Anyway, you stated your post more succinctly than EA.
Previous Page | Next Page
Go to Page 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Search this thread
Show all threads
Reply to this thread:
HCS Forum Index
Halley's Comet Software
forum source