Topic: "ASIO"?

[Greebo]

So I had this friend, and I sent him an ogg, and it played really badly in Winamp. So he got a bit upset and said "oh I better switch to foobar".

I said why not try XMPlay.. and he said he looked into it but it didn't support "ASIO" and therefore sound quality was not as good.

I had no idea what this meant and did a bit of Google research.. relevant to mp3 playback it  apparently bypasses Windows resampling of sound and allows bit perfect blah I got bored about this stage.

If I was to, say, accuse this friend of being a completely retarded audiophile-poseur, would I be out of line? Or is he onto some HOLY GRAIL of mp3 playback that eludes the masses!

[Roj]

Do a search in the on ASIO - you'll uncover a bunch of info.  I personally view it as largely overrated for listening purposes if you have a quality sound card with quality drivers.

[Tsorovan]

http://www.kikeg.arrakis.es/measurements/

But let's dispense with all the tech talk for a minute and go straight to the point: OGG? It's a pretty shitty format if you want to attain transparency, so that dude's words ring a bit hollow in this example.

[Roj]

http://www.kikeg.arrakis.es/measurements/

But let's dispense with all the tech talk for a minute and go straight to the point: OGG? It's a pretty shitty format if you want to attain transparency.

I have read that the latest rev of the OGG encoder is supposed to offer a meaningful improvement in audio quality over earlier versions.  I haven't tested this.

Also, I have one caveat on the information at that link.  It's all pre-SP1 and significant changes were made to Window's kmixer by M-Audio in SP1.

[Tsorovan]

Well, if you actually looked at the info provided in that link you'd see that the differences were more or less negligible (for that card).

[Roj]

Well, if you actually looked at the info provided in that link you'd see that the differences were more or less negligible (for that card).

You are correct - I didn't look that far.  I looked at the header of the graph.

Mea culpa, Robert Culpa.

[Dotpitch]

OGG? It's a pretty shitty format if you want to attain transparency, so that dude's words ring a bit hollow in this example.

i think it's better than mp3 at the same bitrate

i don't think ASIO will make the difference between "bad sound" and "really nice sound", maybe just the difference between "acceptable" and "better" for people with good ears . holy grail? well, xmplay plays fine without ASIO, imho, i wouldn't benefit from it. if you're friend persists in using ASIO, just let him recode the ogg or transcode it to whatever he likes and than play it in WinAmp

[Tsorovan]

Only at very low bitrates (say, 32-128 kbps).

[Pike84]

Well, as far as Rjamorim's listening test goes, Vorbis (ogg) is actually better at mid-level bitrates. As far as HydrogenAudio's general opinion goes, both formats can attain transparency at sufficient bitrates. Also remember that variable bitrate (VBR) is always better than constant bitrate (CBR), and it just happens to be so that Vorbis doesn't even have a CBR mode.. so, in real world, there's a better chance that a Vorbis file is of higher quality than an mp3 file .

[Tsorovan]

You call that mid-level bitrates? I call it low bitrates. Your statement that VBR is always better* than CBR is quite fallacious as 320 kbps CBR will always be best*, and 256 CBR will also always be better* than, say 160 kbps ABR, to take a few examples.

*: Filesizes not taken into account, which they shouldn't when talking about audio quality/transparency at above low-level bitrates.

Anyway, maybe we should return to discussing ASIO...

[Roj]

Anyway, maybe we should return to discussing ASIO...

Nothing really to discuss now.  The findings you posted clearly indicate that the HA / Foobar set overstate its use for listening, especially when one has a quality audio card with competent drivers.

No surprize here on any of the above points.

[Pike84]

You call that mid-level bitrates? I call it low bitrates.

So be it. But 128k is often referred to as mid level, wheras low level would be something like 32-96k.

Your statement that VBR is always better* than CBR is quite fallacious as 320 kbps CBR will always be best

*sigh*
I know very well of that exception. Of course, since it's the highest an mp3 can go, but as a general rule, VBR does beat CBR. Also, I'm not sure if that 320 CBR exception applies to other formats, such as ogg vorbis.

and 256 CBR will also always be better* than, say 160 kbps ABR

ABR is a whole other form of encoding, and has nothing to do with my statement above.

[Tsorovan]

So be it. But 128k is often referred to as mid level, wheras low level would be something like 32-96k.

No one uses that. If people call 128 kbps mid-level they're wrong and stupid and should be drawn and quartered (:

*sigh*
I know very well of that exception. Of course, since it's the highest an mp3 can go, but as a general rule, VBR does beat CBR. Also, I'm not sure if that 320 CBR exception applies to other formats, such as ogg vorbis.

Nah, the thing is that blanket statements like that don't work. VBR at an average bitrate of 160 will be better than CBR 160, yes. THAT's what you should say, not that that has any bearing on anything being said here really.

ABR is a whole other form of encoding, and has nothing to do with my statement above.

I actually debated with myself whether to use that as an example or another VBR preset... I didn't really mean it as ABR, I meant it more like VBR with an average bitrate of... But that's my fault for not typing it. Let's say 256 CBR over --alt-preset standard, then? --alt-preset standard rarely goes over 256 CBR.

Anyway... ASIO...

[Pike84]

*sigh*
I know very well of that exception. Of course, since it's the highest an mp3 can go, but as a general rule, VBR does beat CBR. Also, I'm not sure if that 320 CBR exception applies to other formats, such as ogg vorbis.

Nah, the thing is that blanket statements like that don't work. VBR at an average bitrate of 160 will be better than CBR 160, yes. THAT's what you should say, not that that has any bearing on anything being said here really.

ABR is a whole other form of encoding, and has nothing to do with my statement above.

I actually debated with myself whether to use that as an example or another VBR preset... I didn't really mean it as ABR, I meant it more like VBR with an average bitrate of... But that's my fault for not typing it. Let's say 256 CBR over --alt-preset standard, then? --alt-preset standard rarely goes over 256 CBR.

Are you sure you still don't understand, that I meant it in general?? In general VBR beats CBR, and that's a fact. If you want to critisize my way of putting it in the first place (like using the word "always"), that's one thing. But I really don't think I left anything unclear here... And ABR is still a whole other thing and shouldn't be mixed here - at least not in this context.

[Tsorovan]

I acknowledged/explained that with ABR. It's a closed chapter. I still don't like the VBR/CBR thing just because you added an "in general". It's incorrect/not specified enough/a blanket statement. Can we leave this now? Hehe.

[sequestrum]

Yes please, let's leave this subject aswell as the real subject in question; ASIO.
If Greebo wants more info, he can check the previous ASIO discussion out, and see what a mess that turned out to be.

[Pike84]

I still don't like the VBR/CBR thing just because you added an "in general". It's incorrect/not specified enough/a blanket statement.

Sure, literally read, it was. But that's obviously not how I meant it.

And no, you're not getting the last word .

[Greebo]

What have I created!

[Jace]

Beats Jerry Springer though

[Dotpitch]

lol

I'd say ASIO output support would be nice, but it's not critical.

I guess it could happen in future, but there are still no plans for ASIO support at present

feel free to ask more questions (just realise what consequences it could have )

[Rah'Dick]

I wonder why people seem to "hear" differences with ASIO when they're listening to MP3s! I mean, this file format is known for (and works by) cutting away all subtle frequencies, so what's the point in trying to hear something "less resampled"?
People just don't seem to get it: MP3 is not a high quality audio codec!

By the way: Ever used headphones with a frequency range of max. 30kHz (or higher)? Yes? Still keen on ASIO?

[Roj]

I wonder why people seem to "hear" differences with ASIO when they're listening to MP3s! I mean, this file format is known for (and works by) cutting away all subtle frequencies, so what's the point in trying to hear something "less resampled"?
People just don't seem to get it: MP3 is not a high quality audio codec!

It may not be a high quality codec but to say that it removes all the subtleties of a piece of music is grossly incorrect.  There is still a wealth of detail inherent in a well encoded mp3 and that's why decent audio cards, drivers and speakers are important to those of us who enjoy that.

That being said, I'm already on record as stating that ASIO is vastly overrated under the abovementioned conditions (translation: *I* don't hear a difference with competent audio gear and question the validity of the claims of those who do).

[Rah'Dick]

With "high quality" I mean something else than a lossy codec that cuts high frequencies... Load a ripped WAV file into Audition and have a look at the frequency bands. Then encode it as MP3 and look again - everything beyond 20 kHz is gone... (And some other frequencies, too.) Unfortunately, these missing bands are sometimes significant for correct sound. I already encountered songs that didn't sound like CD even when I encoded them as 320 kbit MP3. (Rare case, I have to admit.)

[Pike84]

I assume you've ABX'd it too . If that's true, you have some pair of golden ears, I'd say.

[Roj]

With "high quality" I mean something else than a lossy codec that cuts high frequencies... Load a ripped WAV file into Audition and have a look at the frequency bands. Then encode it as MP3 and look again - everything beyond 20 kHz is gone... (And some other frequencies, too.) Unfortunately, these missing bands are sometimes significant for correct sound. I already encountered songs that didn't sound like CD even when I encoded them as 320 kbit MP3. (Rare case, I have to admit.)

If you can hear *anything* above 18Khz and are over the age of 30, I'd say that when you talk in person, the conversation goes something like this:

"woof!"

'nuff said.