Help the clueless - mp3 quality.

[Thisp]

This reeks of placeabo. Please do provide some ABX results if you're to claim you can tell something, as many people flinch telling 160k songs from the FLACs when they're blind to which is which.

Use --alt-preset standard with LAME 3.90.3 and let the codec worry about the bitrate. It's smarter than you, so it should be choosing the bitrate.

 

[Mr.Radar]

Quote:

Originally Posted by hackeron FIrst of all, increased bitrate with the same format, ofcourse. -- Testing with different formats is not a fair test.

I did test increased bitrate with the same format (LAME -APS vs. LAME -API), but threw in Ogg Vorbis just to mix things up a bit. I'm going to run the test again with more bitrates for MP3 to make it more "fair" by your standards.

Quote:

Also, thats where I come in -- I dont use that poor excuse for a GUI, so results are comming.

I wish that I could use Linux but I can never get X working right. Windows works, and I'm sticking with it. BTW, did you read part "b" of why I used Foobar2000?

 

[Mr.Radar]

Here are some more results for MP3:
LAME (Alt Preset Medium, 152 kbps): 1901 ms
LAME (no command line options, 128kbps): 1791 ms

EDIT: More Ogg Vorbis
Q3 (105 kbps): 2844 ms
Q4 (118 kbps): 2650 ms
Q6 (177 kbps): 3078 ms

 

[hackeron]

OGG_ENCODER = oggenc 1.0.1
MP3_ENCODER = lame 3.96 mmx
DECODE1 = mplayer 1.0pre5-3.3.4
DECODE2 = ogg123 1.0.1 , mpg123 0.2.10

CODEC BIT ENCODE DECODE1 DECODE2 OTHER
ogg 190 11m55s 35.61s 36.60s -managed -b 192 (full bitrate management)
ogg 161 3m36s 36.21s 40.43s -b 192 (VBR)
ogg 135 3m57s 35.42s 38.68s -q 5
mp3 192 3m29s 49.49s 59.42s -b 192 -h
mp3 160 3m23s 47.86s 56.19s --abr 160 -h
mp3 149 3m35s 48.30s 56.61s --vbr-new -V3

mp3 is significantly slower to decode!

Let me clear this up:
mp3 takes 59.42s 56.19s 56.61s to decode using a native mpg decoder
ogg takes 36.60s 40.43s 38.68s to decode using a natice ogg decoder

The source is a Chopin piano classics CD. About 1 hour in length.

 

[hackeron]

Quote:

Originally Posted by commando If you read my post a little more carefully you'll see that i've described preloading of songs on a hard disk into memory on the player. I think this is a common problem with you posts - lack of reading/understanding what people have said, and the constant editing of your posts shows that you post first and think second. I don't mean that to be offensive, just a suggestion about how to get along here better. Yes, hdd players know the order of the playlist, even on random, so it can preload the next few songs into ram. They do spin the drive down while it's not needed. They can have 64MB of ram or so, enough for a lot of music at 128kbps. I'd be very suprised if any portable player currently on the market doesn't do this caching. If the user manually skips songs the cache will probably be emptied and whatever the user has selected is loaded. That's why you get better battery life if you use playlists rather than jumping randomly around your music. Flash is a type of RAM, you don't need to cache it.

Flamebeit aside, I did read your message and think about it logically.

Hard drive takes 90% of the battery. If using the ipod constantly to move around through songs manually, it lasts about 7 hours. If letting it play non stop, you get what, another extra 3 hours?

Thats blatant proof the hard drive doesnt spin down. Also, please point me to a specs sheet of a player that has 64mb ram please.

Anyway, if the hard drive would indeed spin down for 30 minutes, thats a major gain in battery life. So then the battery should be 7 hours if constantly used, and up to 60 hours if player left to play music. -- but then again you didnt respond to the lag issue of waiting for hard drive to spin back up.

I clearly explained this in my post so it is clearly you who is misreading my post. Thats your problem, you try to somehow convince yourself by convincing others of complete junk. No offence.

 

[commando]

64 was a guess from the article I read a while back - seems the iPod has 32MB (link).

I think the assumption that hdd takes 90% of the power of a player need to be verified before any of your other assumptions are worth considering. I don't see any "blatent proof" there that's not based on your assumptions.

I didn't think your lag comment needed me to comment on it - yes, if you need to spin up a hard disk it could take a while. How long does it take to spin up a hard disk with a 1 inch diameter? I have no idea. I don't have a hdd based music player to test it with either.

It'd be interesting if someone could do a test - have the player play a playlist, wait say 10 minutes for it to get part way thru it's cache, then change to a song not in the playlist. How long does it take to start playing the song? Is it possible also that once you start scrolling thru your playlist that it automatically spins up the disk? That way the disk could be idle most of the time, but when the user starts to use their player it could respond quickly.

 

[DerrickW]

Quote:

Originally Posted by hackeron MP3_ENCODER = lame 3.96 mmx ... mp3 is significantly slower to decode!

Wanna run this with LAME 3.90.3 --alt-preset standard? Most of us stay away from the newer versions on purpose.

 

[hackeron]

Quote:

Originally Posted by commando 64 was a guess from the article I read a while back - seems the iPod has 32MB (link). I think the assumption that hdd takes 90% of the power of a player need to be verified before any of your other assumptions are worth considering. I don't see any "blatent proof" there that's not based on your assumptions. I didn't think your lag comment needed me to comment on it - yes, if you need to spin up a hard disk it could take a while. How long does it take to spin up a hard disk with a 1 inch diameter? I have no idea. I don't have a hdd based music player to test it with either. It'd be interesting if someone could do a test - have the player play a playlist, wait say 10 minutes for it to get part way thru it's cache, then change to a song not in the playlist. How long does it take to start playing the song? Is it possible also that once you start scrolling thru your playlist that it automatically spins up the disk? That way the disk could be idle most of the time, but when the user starts to use their player it could respond quickly.

Hmm, 32Mb sounds about right for 20Gb. About 20Mb of it is used to cache the filelist, 2mb used initially, rest is used to cache some time forwards and backwards in raw stream as an anti shock measure and for quick responding fast forward.

Sounds just right

EDIT: ah, here: "Built-in DRAM provides over 20-minutes of skip protection." -- sounds about as much as better CD players on the market

-- skip protection none the less.

 

[hackeron]

Quote:

Originally Posted by DerrickW Wanna run this with LAME 3.90.3 --alt-preset standard? Most of us stay away from the newer versions on purpose.

Well, might as well try with an outdated version of oggenc while you're at it

 

[commando]

Quote:

Originally Posted by hackeron Hmm, 32Mb sounds about right for 20Gb. About 20Mb of it is used to cache the filelist, 2mb used initially, rest is used to cache some time forwards and backwards in raw stream as an anti shock measure and for quick responding fast forward. Sounds just right EDIT: ah, here: "Built-in DRAM provides over 20-minutes of skip protection." -- sounds about as much as better CD players on the market -- skip protection none the less.

MP3 at 128kbps is about 1MB/minute, correct? Which would mean 20MB is used for caching the music, is that what you meant rather than 20MB to cache the file list? A 20MB list of files would be massive!

I still wonder if they really spin the hdd for those 20 minutes that it's not really needed. I suspect it'd be more power efficient to read the files they need and spin the disk down, rather than leave it running.

 

[hackeron]

Ok, lets see:

* Flash player: 20 hours from 520mAh 13mW (others more efficient)
* HD player: 10 hours from 1200 mAh no mW specs, but quieter
* Reading from ram takes significantly less power than reading from flash memory.

Put the 3 together and figure it out commando

EDIT:

OK, I was pretty off on the hard drive specs, but have a look:
Start: 4.7 watts
Seeking: 2.6 watts
Reading: 2.3 watts
Writing:2.3 watts
Idle: 0.9 watts
Standby: 0.25 watts
Sleep: 0.1 watts

 

[DerrickW]

Quote:

Originally Posted by hackeron Well, might as well try with an outdated version of oggenc while you're at it

Hackeron, please read the thread before you post!

Quote:

Originally Posted by Imyourzero 3.90.3 may be older, but the reason it is so widely suggested is because it has been very thoroughly tested and was found to have the best sound & least amount of issues. Newer versions may work as well, but without them being tested as much as 3.90.3 I'd be hesitant to use them. I'd hate to rip my entire collection with a newer version only to find out that 3.90.3 would have given me slightly better sound quality or smaller file size, ya know?

 

[hackeron]

Quote:

Originally Posted by DerrickW Hackeron, please read the thread before you post!

I did read it.

So what? -- older versions of OGG take off more inaudiable highs so quality is also slightly better overall.

 

[commando]

I still think you're making too many assumptions based on the data - I can't see how you can assume that the hard disks stay spun up instead of being powered down. I'm not saying which is correct, just that we don't have the information here to make that assumption/logical leap.

Of course flash products are going to use less power, they don't have to power motors to throw metal disks around. I basically consider flash memory to be ram, i'm not aware of any power issue reading from it rather than standard dynamic issues, but I know very little about power consumption of different types of memory!

If anyone knows the answer for sure i'd quite like to hear it.

 

[hackeron]

Quote:

Originally Posted by commando I still think you're making too many assumptions based on the data - I can't see how you can assume that the hard disks stay spun up instead of being powered down. I'm not saying which is correct, just that we don't have the information here to make that assumption/logical leap. Of course flash products are going to use less power, they don't have to power motors to throw metal disks around. I basically consider flash memory to be ram, i'm not aware of any power issue reading from it rather than standard dynamic issues, but I know very little about power consumption of different types of memory! If anyone knows the answer for sure i'd quite like to hear it.

There is a reason why palms and other handhelds dont use flash memory, and the reason is a huge difference in power consumption.