|
|
|
|
#1 |
|
Junior Member
Join Date: Jul 2003
Posts: 39
|
A whole bunch of sound quality questions...
First off, why hasn't winamp made the switch to MAD mp3 decoder? It is supposed to be much better. There is a plug-in for it but it doesnt support dsp so I have only played around with it a little bit. And, is Wav output really better than directsound? What problems can i expect by switching? Right now I have just deleted the directsound plug-in from my plug-ins folder so i assume winamp is now using wav output by default? Am I wrong in this assumption? What kind of input plug-in, output plug-in settings should i use for highest sound quality?
Thanks for reading this ,Greg |
|
|
|
|
|
#2 |
|
Techorator
Winamp & Shoutcast Team Join Date: Jun 2000
Posts: 35,894
|
MAD is a 3rd-party plugin. It will NEVER be included with the default Winamp setup.
Report all problems / direct all queries to the plugin author. You won't catch me using it, that's for sure ![]() Note: We don't officially support 3rd-party plugins here. Output plugins: The active output plugin is whatever is hi-lighted in Prefs -> Plugins -> Output. There's no need to physically delete anything. If WaveOut is hi-lighted, then it is the active Output plugin. The other (non-highlighted) output plugins are not used. DirectSound Output is recommended under Windows 2k/XP The difference between WaveOut & DirectSound http://forums.winamp.com/showthread.php?threadid=100492 If you want any further advice, please include the following info: mobo, cpu, ram, soundcard, speakers, Windows OS, DirectX version. |
|
|
|
|
|
#3 |
|
Junior Member
Join Date: Jul 2003
Posts: 39
|
Thanks for clearing up directsound for me. While there is a MAD plug-in for Winamp, MAD is really another mp3 decoder that doesnt have alot to do with Winamp (which is unfortunate). It was not developed with Winamp in mind. Someone just coded a plug-in allowing the replacement of the Fraunhofer mp3 decoder (Winamp standard mp3 decoder) with the MAD decoder. The Fraunhofer decoder is VERY old and, though it is still quite good compared to other competition, I think Winamp should make the switch to the much more up-to-date MAD, so people like me dont have to settle with poor plug-in implementation of the superior decoder (because, even though MAD is good, there are no supported dsp plug-ins for it and it breaks the equalizer because it was made for the Fraunhofer decoder).
|
|
|
|
|
|
#4 |
|
Guest
Posts: n/a
|
Old doesn't mean bad though. It is the decoder that AOL has licensed, and it isn't worth the effort to go through the product of licensing another encoder because it is newer, besides newer doesn't always mean better.
http://mp3decoders.mp3-tech.org/overall.html The Winamp 2.7 decoder seems to stack up better then MAD. There is no reason to justify any decoder change. If a user want to use MAD, then they can download the third party plug-in then. But the current decoder in Winamp is arguably the best available. |
|
|
|
#5 |
|
Major Dude
Join Date: Apr 2003
Location: where the llamas come from!
Posts: 952
|
the reason is that the fraunhoffer codec works flawlessly. there are some "features" of the MAD decoder that make it superior in THEORY (dithering, 24 bit support), but in practice you can't hear any difference with 16 bit music.
and the fraunhoffer codec is permanently worked on. last time they added support for wa 2.9x's new library. I think that's a feature MAD currently lacks and this is a feature you WILL notice in practice. eeeee eeeeeee eeeee eeeee eeeee 8 8 8 8 8 8 8 8 8 8 88 8eee8 8e 8 8 8eee8 8e 8 8 8 88 8 88 8 8 88 8 88 8 8 8 88 8 88 8 8 88 8 88 8 8eee8 |
|
|
|
|
|
#6 |
|
Junior Member
Join Date: Jul 2003
Posts: 39
|
Though dithering does seem a little iffy, there are no ifs, ands, or buts when it comes to 24 and 32 bit support. That IS superior to the 16-bit support offered in the Fraunhofer decoder (in theory and in practice). Especially if you have hardware with 24-bit capability. The proof is in the pudding, and I can hear a definite difference between the two. On testing the two (I used coolplayer for the MAD decoder part of the test) I find that MAD sound much more realistic.
Well, I guess its just a matter of opinion and that it will probably not change... T |
|
|
|
|
|
#7 |
|
Member
Join Date: Jul 2002
Posts: 62
|
Running Windows2000 on a Celeron 466 Mhz machine, I've found that the Winamp 2.91 decoder shows 0-1% system resource use, and the Mad decoder demands much more from the CPU (can't recall now, but probably 7% or more), as does the Thompson MP3Pro ver. 1.2 decoder. My soundcard is an old Soundblaster Live!, so I can't comment on 24-bit soundcard results. For me, resource use is an important factor.
|
|
|
|
|
|
#8 |
|
Member
Join Date: Jul 2002
Posts: 62
|
Oh -- and I don't think I was able to get the MAD decoder (latest version) to decode streams reliably. That's a critical function for me.
|
|
|
|
|
|
#9 |
|
Junior Member
Join Date: Jul 2003
Posts: 39
|
For me, its all about quality of sound. Extra resource use doesnt bother me (unless it is something that is always memory resident like antivirus software and such). And the MAD plug-in for Winamp doesnt support streams. If you want to do a good comparison of the differences in winamps default decoder and the MAD decoder download Coolplayer. It implements the MAD decoder much better than the Winamp plug-in. The Winamp plug-in is a very poor example of how good MAD is.
|
|
|
|
|
|
#10 | |
|
Member
Join Date: Oct 2001
Location: Singapore
Posts: 90
|
Quote:
And in any case, most people's audio files are encoded in 16-bit, so 24-bit playback is useless, and won't even be activated in the Winamp plugin. Even if you manage to force 24-bit decoding, the soundcard will resample, and most consumer soundcards suck at resampling (all Creative cards except Audigy2, for example). There seems to be this grave misconception that because 24-bit has 8 bits more depth than 16-bit, it MUST be audible. Likewise, 32-bit is only there so that you can pad 24-bit to it for software compatibility with goodness-know-what. And 96-bit is marketing hype, designed for those with money to spare. Dithering would be more prudent to use than 24-bit decoding, because it adds about 2 more bits of depth, which you can more likely hear (relatively) than the 6 bits of depth after it. I am no audio expert though, all the above info came from months of reading on www.hydrogenaudio.org . |
|
|
|
|
|
|
#11 | ||
|
Member
Join Date: Oct 2001
Location: Singapore
Posts: 90
|
Re: A whole bunch of sound quality questions...
Quote:
Quote:
|
||
|
|
|
|
|
#12 |
|
Major Dude
|
I personally use DirectSound with Winamp, having Hardware Acceleration enabled on my Terratec DMX Xfire 1024 soundcard under Win98SE, and it's working nicely for me. Much better than WaveOut (CPU usage difference between WaveOut vs. DirectSound barely are ~2% on my Celery 433
).A mate of mine with a Creative Audigy card can't have Hardware Acceleration enabled at all, he'd hear artifacts, crackling and popping .It highly depends on your soundcard and its drivers if WaveOut or DirectSound is better, whereas it's true WaveOut is not recommended under 2K/XP. As for the MAD plugin, I could never hear any difference in sound quality. Maybe it's just because I have a 16bit soundcard, but even with dithering enabled, I couldn't notice a difference. What always made me not touching the MAD plugin is its incompatibility with streams, and at times it does not seem to decode right at all . It uses too much resources anyway for the job it does, the Fraunhofer decoder that comes with Winamp is heaps better.
|
|
|
|
|
|
#13 |
|
Member
Join Date: Jul 2002
Posts: 62
|
Concerning sound quality, a DSP with Winamp can certainly help. Personally, I've tried a bunch of 'em, including paid versions of DFX, and I've settled on Enhancer version 0.17. With ambience turned off, Enhancer seems to add only 1% resource use, if that. The value and sound quality of Enhancer has been promoted by many, many Winamp users in these forums.
|
|
|
|
|
|
#14 | |
|
Major Dude
Join Date: Jan 2001
Posts: 1,668
|
Re: Re: A whole bunch of sound quality questions...
Quote:
|
|
|
|
|
|
|
#15 |
|
Junior Member
Join Date: Jul 2003
Posts: 39
|
Yes, I did a completely blind test. I had someone else run the controls and I sat back and guessed. I didnt have a very high success rate with pop music, but my high success rate with jazz and classical music more than made up for it. But even having done a blind test i understand thay my opinion is still just that, an opinion. If people havent been encoding 24-bit wav files to .mp3 that is their problem (Most people probably encode mp3's with musicmatch or something anyway). They should be
Below are some visual comparisons of MAD at 24-bit dithering, 16-bit dithering, and Fraunhofer at its limit of 16-bit). Also,below are the results to someone elses MAD blind test. http://www.ff123.net/dither.html http://www.ff123.net/madchallenge.html |
|
|
|
|
|
#16 |
|
Major Dude
|
If you think MAD is so much better, then just use the 3rd party plugin. The Fraunhofer decoder is more compatible, and less resource hungry.
Moreover, Nullsoft/AOL most definitely won't go through yet another major licencing hassle to make a tech like you can find in the MAD decoder as a default piece of code that distributes with Winamp. Future will show how it's gonna be when 24bit soundcards are the standard. Until then, the current decoder does its job just fine, no need to toss it. Like I said, if you wanna use the MAD decoding tech, just use the plugin ... |
|
|
|
|
|
#17 |
|
Junior Member
Join Date: Jul 2003
Posts: 39
|
I dont quite understand how you think the Fraunhofer decoder is more compatible than MAD... They are both compatible with all mp3s. And I have already explained that the plug-in for MAD on Winamp is a poor implementation of MAD and its features.
|
|
|
|
|
|
#18 |
|
Major Dude
|
With compatibility I was referring to crackling and noise during playback. I noticed crackling on 4 different soundcard of mine, and no, I'm not the only one noticing that
.The MAD plugin might be quite poor for Winamp, but trying CoolPlayer with MAD built-in, it's quite ok (shame CoolPlayer's EQ sucks galore ), but nothing I would prefer towards the Winamp Franhofer decoder.Like I said, future will show when 24bit soundcards are standard. |
|
|
|
|
|
#19 |
|
Junior Member
Join Date: Jul 2003
Posts: 39
|
I agree wholeheartedly with you on Coolplayer. Its saving grace is the MAD decoder. And that alone doesnt make it worth the switch.
|
|
|
|
|
|
#20 |
|
Major Dude
|
Coolplayer is really nice, I don't understand why the MAD plugin for Winamp needs so much more CPU power than CoolPlayer does
|
|
|
|
|
|
#21 |
|
Junior Member
Join Date: Jul 2003
Posts: 39
|
Winner2, thanks for the enhancer reccomendation. It is great. So far I have tried DFX, Dee2, and Enhancer (the ones I remember). DFX was pretty good, then I saw Dee2 and it was even better (a little too much goin on in that plug-in though (the presets are kind of overbearing). And now I find Enhancer which is better than both of them (It really adds a lot of depth without added hiss and distortion and it isnt as heavy on the processing as Dee2). Thanks again!
|
|
|
|
|
|
#22 | |
|
Member
Join Date: Oct 2001
Location: Singapore
Posts: 90
|
Quote:
But anyway, I sort of forgot to mention that I'm not using Winamp for my main audio playback anymore, though I always come back to look out for the latest version of Winamp, and debate stuff in the forums. PS: Kudos to you Pearljam for blindtesting. Well, that means you have better hearing and better equipment than me (I don't even use headphones). And yes, I've seen those graphs before. And I go to the same forum as the guy who created that website
|
|
|
|
|
|
|
#23 |
|
Junior Member
Join Date: Jul 2003
Posts: 39
|
I wouldnt say that about the equipment. We probably just hear things a little differently. I would switch to coolplayer, but I wont because Winamp has a much better team and amazing functionality and second and third party plug-ins out the bum.
|
|
|
|
![]() |
|
|||||||
| Thread Tools | Search this Thread |
| Display Modes | |
|
|