IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Subjectively pleasing codecs?
julf
post Jan 4 2014, 13:02
Post #1





Group: Members
Posts: 226
Joined: 14-January 12
Member No.: 96426



As we have seen, some people prefer stuff like vinyl and tubes, despite their objectively inferior performance. Apart from perceptual bias, the preference is probably also driven by "pleasant" coloration and distortion. There also seems to be some cases where people have actually preferred low-bit-rate mp3 to uncompressed in listening tests - explained by the way the mp3 encoding removes extarneous stuff, emphasizing the real musical information.

In designing and testing codecs, we naturally try to focus on making the codecs as transparent as possible - the goal being compression that is audibly indistinguishable from the original.

Is anyone aware of any research into codecs and digital processing that would be non-transparent on purpose, instead focusing on emphasizing the "pleasant" aspects of the music? I am not talking about digital versions of teh good old Aphex Aural Exciter, but stuff that is based on the same kind of sophistication in psychoacoustic modelling that the more advanced lossy codecs are based on.

I am aware of a fair bit of work in speech codecs, emphasizing legibility rather than transparency, but I am not aware of similar work in music reproduction/processing.
Go to the top of the page
+Quote Post
probedb
post Jan 4 2014, 15:29
Post #2





Group: Members
Posts: 1214
Joined: 6-September 04
Member No.: 16817



Surely you'd just want EQ to do this? After all there's no point having a codec that basically produces a "warm" sound? You can get the same results by using any codec and applying EQ?
Go to the top of the page
+Quote Post
julf
post Jan 4 2014, 15:43
Post #3





Group: Members
Posts: 226
Joined: 14-January 12
Member No.: 96426



QUOTE (probedb @ Jan 4 2014, 15:29) *
Surely you'd just want EQ to do this? After all there's no point having a codec that basically produces a "warm" sound? You can get the same results by using any codec and applying EQ?


I am curious if someone has gone beyond mere EQ in modelling "pleasant" sound.
Go to the top of the page
+Quote Post
lsn_RU
post Jan 4 2014, 16:30
Post #4





Group: Members
Posts: 17
Joined: 13-December 13
Member No.: 113238



Very interesting! I think to start need to be split all listeners into several age groups, ranging from 10-12 years. I have assume that preference of youth will be given to mp3 and possibly, ogg family codecs.
Go to the top of the page
+Quote Post
saratoga
post Jan 4 2014, 17:23
Post #5





Group: Members
Posts: 4922
Joined: 2-September 02
Member No.: 3264



There are a huge number of DSP effects that are nontransparent but subjectively pleasant at least to some people. You can combine these with a codec like mp3 if you want (and many cheap digital audio players can do this).
Go to the top of the page
+Quote Post
julf
post Jan 4 2014, 17:30
Post #6





Group: Members
Posts: 226
Joined: 14-January 12
Member No.: 96426



QUOTE (saratoga @ Jan 4 2014, 17:23) *
There are a huge number of DSP effects that are nontransparent but subjectively pleasant at least to some people.


Right - but is there any (published) perceptual research underlying them?
Go to the top of the page
+Quote Post
drfisheye
post Jan 4 2014, 21:24
Post #7





Group: Members
Posts: 18
Joined: 4-January 14
Member No.: 113780



If a single algorithm to make music sound better existed, then it would already be mastered in the recording. Since recording engineers are not stupid.

A specific track might sound better with some specific sound effect, but only because the recording/mastering engineers didn't think of using it for that track. But a general trick doesn't exist.

Floyd E. Toole already proved that everyone prefers a neutral frequency response. Young people, old people, Americans, Europeans, Asians, Hard Rock fans, Classic fans, Hip Hop fans etc. So there is no EQ that would make things sound better in general, except if your system isn't neutral, which it shouldn't be. There is no such thing as a subjective preference for certain EQ patterns. There is only subjective preference for musical taste. Not for sound quality. It's why JBL stopped making speakers tuned for Germans.

There is only one way to audio nirvana: get your system to be as neutral as possible. And get the studio's to get listenings rooms as neutral as possible. That way we'll be using the same reference, so that you hear what the recording engineers intended you to hear.
Go to the top of the page
+Quote Post
saratoga
post Jan 4 2014, 21:29
Post #8





Group: Members
Posts: 4922
Joined: 2-September 02
Member No.: 3264



QUOTE (julf @ Jan 4 2014, 11:30) *
QUOTE (saratoga @ Jan 4 2014, 17:23) *
There are a huge number of DSP effects that are nontransparent but subjectively pleasant at least to some people.


Right - but is there any (published) perceptual research underlying them?


Theres a lot of research on various aspects of such DSP effects, things like EQ, compression, stereo expanding, etc. I've looked through a few commercial products, and they tend to use a combination of these effects and others. Presumably they do some tuning to figure out what will sound "good" when put together.
Go to the top of the page
+Quote Post
julf
post Jan 4 2014, 22:57
Post #9





Group: Members
Posts: 226
Joined: 14-January 12
Member No.: 96426



QUOTE (saratoga @ Jan 4 2014, 21:29) *
Theres a lot of research on various aspects of such DSP effects, things like EQ, compression, stereo expanding, etc.


Thanks! Would you have any references?
Go to the top of the page
+Quote Post
item
post Jan 8 2014, 19:01
Post #10





Group: Members
Posts: 33
Joined: 7-August 12
Member No.: 102085



QUOTE (drfisheye @ Jan 4 2014, 21:24) *
If a single algorithm to make music sound better existed, then it would already be mastered in the recording. Since recording engineers are not stupid.

A specific track might sound better with some specific sound effect, but only because the recording/mastering engineers didn't think of using it for that track. But a general trick doesn't exist.

Floyd E. Toole already proved that everyone prefers a neutral frequency response. Young people, old people, Americans, Europeans, Asians, Hard Rock fans, Classic fans, Hip Hop fans etc. So there is no EQ that would make things sound better in general, except if your system isn't neutral, which it shouldn't be. There is no such thing as a subjective preference for certain EQ patterns. There is only subjective preference for musical taste. Not for sound quality. It's why JBL stopped making speakers tuned for Germans.

There is only one way to audio nirvana: get your system to be as neutral as possible. And get the studio's to get listenings rooms as neutral as possible. That way we'll be using the same reference, so that you hear what the recording engineers intended you to hear.


Great first post! Welcome to the forum . . .

There's no accounting for taste universally. I don't think it's too big a stretch to claim that educated listening invariably, broadly leads along a path to the endpoint of neutrality. However, absolute neutrality is a near-impossible goal.

What tends to happen is that individuals come to different conclusions about the kind of a-neutrality that bugs them most: which is why the distortion of vinyl remain pleasant to many, compared to imperfections introduced by the complex apparatus of digital sources.
Go to the top of the page
+Quote Post
DonP
post Jan 8 2014, 21:32
Post #11





Group: Members (Donating)
Posts: 1471
Joined: 11-February 03
From: Vermont
Member No.: 4955



QUOTE (saratoga @ Jan 4 2014, 11:23) *
There are a huge number of DSP effects that are nontransparent but subjectively pleasant at least to some people. You can combine these with a codec like mp3 if you want (and many cheap digital audio players can do this).


Megabass anyone?

Autotune on the supply side.
Go to the top of the page
+Quote Post
Porcus
post Jan 8 2014, 22:24
Post #12





Group: Members
Posts: 1842
Joined: 30-November 06
Member No.: 38207



QUOTE (drfisheye @ Jan 4 2014, 21:24) *
If a single algorithm to make music sound better existed, then it would already be mastered in the recording. Since recording engineers are not stupid.


The loudness war is the answer to the OP's prayer then laugh.gif


--------------------
One day in the Year of the Fox came a time remembered well
Go to the top of the page
+Quote Post
AliceWonder
post Feb 27 2014, 22:21
Post #13





Group: Members
Posts: 119
Joined: 13-July 12
From: California
Member No.: 101393



I seem to remember there being a filter that could add a vinyl "sound" to digital audio, but I don't recall where or by who.
Go to the top of the page
+Quote Post
d125q
post Feb 27 2014, 22:40
Post #14





Group: Members
Posts: 57
Joined: 4-May 13
Member No.: 107966



iZotope Vinyl. iZotope products are, as far as I'm aware, considered state of the art.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 29th August 2014 - 14:44