MP3, AAC, FLAC, etc.

[DavidW]

The CD is cheap enough on Amazon.

No it's not!  Anything greater than $0 for me to pay for you to support your assertion is too much.  Especially since you're knocking the accepted results of double blind tests as conducted by Hydrogen Audio and other sources.

[Gurn Blanston]

No it's not!  Anything greater than $0 for me to pay for you to support your assertion is too much.  Especially since you're knocking the accepted results of double blind tests as conducted by Hydrogen Audio and other sources.

Yeah, but Dude; Sammy Davis! It's win::win for you!   

[bigshot]

Sammy Davis Jr in his prime too!

The posts at Hydrogen I've seen say that both AAC and LAME both achieve transparancy, but for LAME it's at 320 while AAC is transparent somewhere between 192 and 256. I use AAC 256 VBR and it's completely transparent.

[DavidRoss]

Sammy Davis Jr in his prime too!

The posts at Hydrogen I've seen say that both AAC and LAME both achieve transparancy, but for LAME it's at 320 while AAC is transparent somewhere between 192 and 256. I use AAC 256 VBR and it's completely transparent.

To you with your ears, equipment, and music choices. I found congestion and blurring in complex orchestral passages, especially with voice. Did you try any Mozart operas?

[mahler10th]

 
May I ask what you mean by 'transparency' in digital files? 

[Gurn Blanston]

May I ask what you mean by 'transparency' in digital files?

John,
Transparency is an imaginary state of hearing wherein one can forget that one is listening to a (pick one: mp3, flac, aac etc) and simply dwell on the music. It actually only exists in the conduit between the ears and the imagination, thus the threshold differs from person to person.

[DavidRoss]

May I ask what you mean by 'transparency' in digital files? 

Transparency in audio means the same thing as transparency in vision. Just as a truly transparent pane of glass would transmit 100% of the light passing through it, with no distortion of colors, no scratches, no blurring, no wavy lines, so "transparent" audio transmits 100% of the recorded sound, without tonal distortion, artifacts, smearing, or warbling.

Equipment is critical, but so is the listener, not only his physical equipment--"ears"--but his software--the capacity and training of his mind to hear details. The millions of kids who are perfectly satisfied with dynamically flattened 128kbps sound files played back on iPods with stock earbuds suggests that they have not learned to hear. To continue the visual analogy, they are like casual observers who look at a painting and see, "It's a tree"--unlike the trained and skilled painter who sees the mixing of pigments, the mastery of light and shade, the balance of composition, the allusion to historical tree paintings, the shift of hue in the shadows and reflected light, the direction, depth, and length of brush strokes, and so on.

Is that clear (transparent)?

[mahler10th]

Thank you Gurn and David.  I 'can see clearly' now.  I switched to encoding in Lame 256 variable from Franhaufer 320 constant (or however you spell it), and I can hear things much better and less one dimensional with Lame, it seems to suit my stuff.  I cannot tell the difference between CD and MP3 Lame 256 variable on my stuff.  I suppose I can say that Lame, by far, brings more transparency (or complete transparency) for my setup, which is not nearly as effective as it could be (5.1 playing out of a DVD Player with 24 bit upsampling output I think, very poor for a Classical Music buff! )  But I can hear the difference in the codecs I mentioned.
Thanks for explaining it Gurn and David, I know what you mean now by 'transparency', and it's something I am very familiar with without ever knowing how to describe it.   

[DavidW]

The posts at Hydrogen I've seen say that both AAC and LAME both achieve transparancy, but for LAME it's at 320 while AAC is transparent somewhere between 192 and 256. I use AAC 256 VBR and it's completely transparent.

Individual posts are just people like you and me just on a different forum.  Doesn't count.

I was wrong to say that there were dbts over there for 320 lame vs 256 aac.  Never happened.  They had a showdown at 128 over there in which aac was on average very slightly better.  The general opinion is that both codecs reach transparency at the bitrates that we're talking about, but I'm open to supported assertions otherwise. 

If you can consistently hear the difference then post some abx logs to convince me.  If it was just that one cd in that one passage, maybe you should back off the whole 256 AAC kicks 320 LAME butt opinion that you've got going.

[DavidW]

Thank you Gurn and David.  I 'can see clearly' now.  I switched to encoding in Lame 256 variable from Franhaufer 320 constant (or however you spell it), and I can hear things much better and less one dimensional with Lame, it seems to suit my stuff.  I cannot tell the difference between CD and MP3 Lame 256 variable on my stuff.  I suppose I can say that Lame, by far, brings more transparency (or complete transparency) for my setup, which is not nearly as effective as it could be (5.1 playing out of a DVD Player with 24 bit upsampling output I think, very poor for a Classical Music buff! )  But I can hear the difference in the codecs I mentioned.
Thanks for explaining it Gurn and David, I know what you mean now by 'transparency', and it's something I am very familiar with without ever knowing how to describe it.   

Now you're one step closer to sounding like one of those rootin' tootin' fancy audiophiles.

[Gurn Blanston]

Thank you Gurn and David.  I 'can see clearly' now.  I switched to encoding in Lame 256 variable from Franhaufer 320 constant (or however you spell it), and I can hear things much better and less one dimensional with Lame, it seems to suit my stuff.  I cannot tell the difference between CD and MP3 Lame 256 variable on my stuff.  I suppose I can say that Lame, by far, brings more transparency (or complete transparency) for my setup, which is not nearly as effective as it could be (5.1 playing out of a DVD Player with 24 bit upsampling output I think, very poor for a Classical Music buff! )  But I can hear the difference in the codecs I mentioned.
Thanks for explaining it Gurn and David, I know what you mean now by 'transparency', and it's something I am very familiar with without ever knowing how to describe it.   

Always my pleasure to be of assistance, mi amigo.   

[DavidRoss]

Thank you Gurn and David.  I 'can see clearly' now.  I switched to encoding in Lame 256 variable from Franhaufer 320 constant (or however you spell it), and I can hear things much better and less one dimensional with Lame, it seems to suit my stuff.  I cannot tell the difference between CD and MP3 Lame 256 variable on my stuff.  I suppose I can say that Lame, by far, brings more transparency (or complete transparency) for my setup, which is not nearly as effective as it could be (5.1 playing out of a DVD Player with 24 bit upsampling output I think, very poor for a Classical Music buff! )  But I can hear the difference in the codecs I mentioned.
Thanks for explaining it Gurn and David, I know what you mean now by 'transparency', and it's something I am very familiar with without ever knowing how to describe it.   

Any time, John. It's a pleasure to share a perspective on something with someone who's interested in understanding others' points of view. Look for pm in the near future.

[mc ukrneal]

Individual posts are just people like you and me just on a different forum.  Doesn't count.

I was wrong to say that there were dbts over there for 320 lame vs 256 aac.  Never happened.  They had a showdown at 128 over there in which aac was on average very slightly better.  The general opinion is that both codecs reach transparency at the bitrates that we're talking about, but I'm open to supported assertions otherwise. 

If you can consistently hear the difference then post some abx logs to convince me.  If it was just that one cd in that one passage, maybe you should back off the whole 256 AAC kicks 320 LAME butt opinion that you've got going.

This matches what I have read is the general consensus - that AAC was better at lower bitrates and while theoretically better at higher bitrates, the difference was too small  to notice (if it existed in the first place)and that they were more or less on par. One can avoid this whole mess by going lossless of course.

[bigshot]

To you with your ears, equipment, and music choices. I found congestion and blurring in complex orchestral passages, especially with voice. Did you try any Mozart operas?

I can't remember which operas I used, but I think I did. It's funny. The things that sounded complex to my ear encoded perfectly. It was just certain sounds, like the massed string tone on that Sammy Davis Jr record that caused problems. Crackly 78s were problematic at lower bit rates, but they cleaned up by AAC 192.

[bigshot]

I cannot tell the difference between CD and MP3 Lame 256 variable on my stuff.

There really is no reason not to use VBR. It allows the bitrate to go as high as 320 if needed, and it drops the bitrate in simple sections to save filesize. CBR doesn't give better sound quality, only bigger files.

[bigshot]

Individual posts are just people like you and me just on a different forum.  Doesn't count.

I was wrong to say that there were dbts over there for 320 lame vs 256 aac.  Never happened.  They had a showdown at 128 over there in which aac was on average very slightly better.

Generally AAC is comparable to MP3 LAME at one notch higher in bitrate... ie: AAC 128 = MP3 LAME 192.

As for counting or not. If you want a scientist in a lab coat to make a determination for you, you'll have to hire one. I was doing line level matched A/B switchable comparisons for myself to determine the proper bitrate setting to rip for my media server. I determined that the lowest bitrate for perfectly artifact free sound was AAC 256 VBR, and that's what I've used for thousands of rips with perfect results. Feel free to use 320 LAME. It sounds about the same. It just has a little bigger filesize.

[bigshot]

This matches what I have read is the general consensus - that AAC was better at lower bitrates and while theoretically better at higher bitrates, the difference was too small  to notice

That is completely true... EXCEPT for that doggone Sammy Davis Jr CD. For some reason it had a sound that stubbornly refused to encode without digital gurgling, except at a high bitrate (320 LAME / 256 AAC). Everything else was fine at 192 AAC, 320 MP3 and 256 LAME.

[DavidRoss]

Generally AAC is comparable to MP3 LAME at one notch higher in bitrate... ie: AAC 128 = MP3 LAME 192.

Not even Apple made this claim in their promotion of AAC as a superior codec. Their claim was that AAC is better than standard Frauenhofer mp3 (not LAME) at equivalent bitrates and comparable to standard Frauenhofer mp3 (not LAME) at "one notch higher bitrate."

YOU may find this is the case for YOUR listening on YOUR equipment with YOUR music, YOUR ears, and YOUR mind. That does not make it objectively true.

There really is no reason not to use VBR. It allows the bitrate to go as high as 320 if needed, and it drops the bitrate in simple sections to save filesize. CBR doesn't give better sound quality, only bigger files.

Whether CBR gives better SQ depends upon the algorithm used by the VBR codec and whether the information that it chooses to drop really is audible or not. Have you ever encountered a discrepancy between theory and practice?

[mahler10th]

Whether CBR gives better SQ depends upon the algorithm used by the VBR codec and whether the information that it chooses to drop really is audible or not.

These are the settings I use, does the trick.

[petrarch]

These are the settings I use, does the trick.

Those look like a good set of settings. However, it all depends on what equipment you plan to use to listen to those files. If I listen to music with ipod-like earphones, that level of quality is perfectly acceptable, and it is what I use when encoding files for use on the iphone. However, it is insufficient if I listen to music on my hi-fi system. So, when I encode stuff I don't have on CD myself, I go for lossless, so that the level of quality is not compromised when I listen to those tracks on my system, and which can always be reencoded in mp3 to create another set of smaller files suitable for listening on the iphone.