Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Ogg Vorbis or mp3 for portable use? (Read 18119 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Ogg Vorbis or mp3 for portable use?

I just bought a Samsung YP-MT6X and would like to convert my FLAC files for use with the portable. Keep in mind I will be using $20 KSC-75 clip-ons, not $1000 iems. According to Wikipedia:

"Many users feel that Vorbis reaches transparency (sound quality that is indistinguishable from the original source recording) at a quality setting of -q5, approximately 160 kbit/s. For comparison, it is commonly felt that MP3 reaches transparency at around 192 kbit/s, resulting in larger file sizes for the same sound quality."

Does that mean I should use Ogg q5 for the best compromise between space and quality?

Thanks,

-Jonathan.

EDIT: Sorry if my question parrots other threads, I didn't see any that explicitly answered my question.

Ogg Vorbis or mp3 for portable use?

Reply #1 –
Have you tested, as in ABX, OGG vs MP3 to hear for yourself?  One test is worth a thousand opnions.   
Nov schmoz kapop.

Ogg Vorbis or mp3 for portable use?

Reply #2 –
I use ogg q2 on my MT6 and have very few SQ issues.  Tthe only problem is a slight cut at the end of .ogg tracks.
"You can fight without ever winning, but never win without a fight."  Neil Peart  'Resist'

Ogg Vorbis or mp3 for portable use?

Reply #3 –
On a portable (and this depends ultimately on yourself and your hearing) you can achieve perceptual transparency with way lower bitrates.

I use MP3 V4 for my portable. I get 140 kbit/s in average. I listen thorough Sennheiser's PX100.
It is transparent to me.

And, IIRC, if you go with Vorbis AoTuV, you can go lower.

Read some of the listening test threads
I'm the one in the picture, sitting on a giant cabbage in Mexico, circa 1978.
Reseñas de Rock en Español: www.estadogeneral.com

Ogg Vorbis or mp3 for portable use?

Reply #4 –
Quote
Have you tested, as in ABX, OGG vs MP3 to hear for yourself?   One test is worth a thousand opnions.    
[a href="index.php?act=findpost&pid=327844"][{POST_SNAPBACK}][/a]


No because the mp3 player is currently...in transit.

EDIT: Foobar only lets me select commandline ogg vorbis q4-q6, what happened to all the the other options?

Ogg Vorbis or mp3 for portable use?

Reply #5 –
Quote
No because the mp3 player is currently...in transit.

EDIT: Foobar only lets me select commandline ogg vorbis q4-q6, what happened to all the the other options?

Those are the presets that foobar has pre-defined. You can make your own. Go to Diskwriter prefrences and click the "add new" button under the preset selection box. Either use the built in ogg vorbis, or choose a the command line. I'd use the command line so that you can point it to an autuv version. Or select one of the ogg vorbis comand line settings and duplicate it and then edit it to a different q number.

I think what encoder and quality you use will depend on how you use your portable. If you don't like to change out the tracks all the time, you should go for lower bitrate so you can fit more. If you will be transcoding from flac every time you change music, that might make you less inclined to do it all the time. On the other hand, mass storage flash means loading new songs is really quick. Some method of automatic random fill might be neat.

As for mp3 vs ogg, I would do a little test to see if there is any effect on battery life. If ogg doesn't change power req'd, you could go 10 kbps lower than the same mp3.

Ogg Vorbis or mp3 for portable use?

Reply #6 –
Aiming for full transparency is not the best idea for flash based portables where space comes at a premium, most people encode at the lowest bitrate that still sounds fine to them (personally i encode Vorbis at -q2, that's 96kbps average), to get the best quality/filesize tradeoff.

If you do some listening tests, you'll probably find you can go to lower bitrates with OGG than MP3, enabling you to cram more music on your player. The downside is you may have less battery life, i don't think anyone has tested this for the Samsung Yepp.

Another plus OGG may have over MP3 on portables is a single tagging format, so you don't run into incompatibilities/issues over id3v1/id3v2, also with MP3 there are a lot of encoders out there, and it's easy to produce files with sub-par quality if you don't know your way around them. Vorbis also has different encoders, but quality wise you're less likely to go wrong.

Here is one person's listening test at 80kbps, comparing, amongst others, OGG and MP3.
Veni Vidi Vorbis.

Ogg Vorbis or mp3 for portable use?

Reply #7 –
I've been thinking about this myself. My concern isn't quality, but battery life. I figure with all the ambient noise, mp3 and ogg at the same bitrate sound "good enough." I'd like to switch to ogg, but I haven't tested battery life of my player between mp3 and ogg at the same bitrate. Based on ancient manuscripts found in Cyprus, using mp3 would yield better battery life.

Maybe I should just test it myself... but that's soo much work!

Ogg Vorbis or mp3 for portable use?

Reply #8 –
I did some ABX and battery tests recently, for exactly that purpose: fit as much music as I could on my portable music player (SPV C550 smartphone with tcpmp) and maximize battery life. Sorry I do not have the formal results here, but I can present you with my conclusions:

I used foobar2000 ABX (with a minimum of 8 results everytime) for evaluating quality. On the phone I used tcpmp to run a song in a loop for 1 hour, and compared the battery levels before and after.

* mp3 vs original: -V6 with lame 3.97 was the lowest I could go before the sound quality degraded too much.

* mp3 vs ogg: I left lame at -V6 and started with ogg (lancer) at -q4. Ogg won. I kept going down all the way to -q0 and was amazed that ogg still beat mp3 at that level!

* aac vs original: I used the new winamp plugin at 48kbps, and although I could clearly hear the difference, this sounded just about acceptable to me; not sure I would use it though.

* aac vs ogg: aac as above, ogg -q0 won clearly. Pushed it down to -q-1 and to my amazement, ogg still won. Went to -q-2 and quite a few annoying artifacts started to appear; not quite sure which ones were less annoying.

* ogg vs original: since I had establish -q0 as more than acceptable, I started at this level. Went down to -q-1, and was unable to tell the difference between -q0 and -q-1!!! -q-2 could easily be ABXed though, and sounded quite nasty.

Quality conclusion:

ogg at -q-1 is the optimal quality vs size for me. Second choice would probably be aacplus v2 at 48kbps, and third choice lame mp3 at -V6 (but the bitrate increase is very large).

Battery tests:

mp3 -V6 for 1 hour = 3-4% battery life
ogg -q-1 for 1 hour = 4% battery life
aac = 12% battery life

I was very surprised by the high battery consumption of aac since I was under the impression that it was designed for low battery consumption. It was just the opposite with ogg: all the tests I had seen claimed that ogg use significantely more power than mp3, but the extra battery usage I observed was minimal.

Power comsumption conclusion:

Both mp3 and ogg are equivalent and very good. Stay clear from AAC.

Ogg Vorbis or mp3 for portable use?

Reply #9 –
battery lasts more with mp3 than with ogg in most portable players. search the forum and you'll find more information.

Ogg Vorbis or mp3 for portable use?

Reply #10 –
Maybe with Vorbis being at quite a lower bitrate than the mp3 tested, it used less battery.

Vorbis on my iAudio U2 uses about 30% more battery than mp3 - but that was with -q4 Vorbis vs 128 CBR mp3.  I think I need to try with a lower Vorbis

Ogg Vorbis or mp3 for portable use?

Reply #11 –
Quote
Maybe with Vorbis being at quite a lower bitrate than the mp3 tested, it used less battery.

Vorbis on my iAudio U2 uses about 30% more battery than mp3 - but that was with -q4 Vorbis vs 128 CBR mp3.  I think I need to try with a lower Vorbis
[a href="index.php?act=findpost&pid=330153"][{POST_SNAPBACK}][/a]

Yep, I think that's the key. 
With Vorbis it seems one can achieve transparency at a lower bitrate, and thus a smaller file size, than with mp3.  But it also seems, certainly with my experience with Vorbis on my iRiver H120, that it uses up more battery life than with mp3s at similar bitrates.  But since Vorbis files at the same quality will be smaller in size & bitrate, then the battery life might well be the same as the original mp3 battery life.  So, if the above is true, then Vorbis would be better as you'll get more audio files onto the player.

Having said all that, I'm still using LAME '-V4 --vbr-new'.  But I've run out of disk space to do a mass transcode of all my FLACs into Vorbis, so until my new disks arrive, I shall be keeping my mp3s (which I'm very happy with, due to the great strides the LAME devs have made with LAME recently [3.97b1]).

Ogg Vorbis or mp3 for portable use?

Reply #12 –
Quote
Maybe with Vorbis being at quite a lower bitrate than the mp3 tested, it used less battery.


Quote
since Vorbis files at the same quality will be smaller in size & bitrate, then the battery life might well be the same as the original mp3 battery life.  So, if the above is true, then Vorbis would be better as you'll get more audio files onto the player.


That is what I suspect, which if true would definitely make vorbis the format of choice for portable player, as long as people realize they do not need the same bitrate as for mp3.

I remember a few months ago doing a test to see what the effect of bitrate would have on battery life. I encoded some mp3 files at constant bitrate (64,128,192,256,320). What I expected and which testing proved, was that higher bitrates would lead to shorter battery life (plus stuttering of playback problems in some cases). Since then I have always transcoded my mp3 with a maximum bitrate of 192 for portable use.

I think it would be time to repeat the test for ogg quality settings and post the results here. It would be interesting to have mp3 in the test as well for reference. Which settings for lame would you like me to use? Bear in mind that this is for portable use, and far from archival quality.

Ogg Vorbis or mp3 for portable use?

Reply #13 –
I made a test on my iAudio M3L to compare battery life when playing Vorbis and MP3
Files were encoded with these settings:
Vorbis: Lancer (based on AoTuV beta4) Q4
MP3: Lame 3.96.1  -V4

These settings are "transparent enough for portable use" to my ears (Exactly AoTuV b4 encoder at Q4 is transparent for me with most samples. I can't ABX it even on loudspeakers (handmade, quality one). For Lame 3.96.1 --V 4 is almost transparent (slightly worse then Vorbis)

I made the test with headphones connected (Koss PortaPro - Impedance 60 Ω, Sensitivity 101 dB / 1 mW) and with volume level 25

Results:
Vorbis -Q4 (Avg bitrate 116 kbps) - 25 hours
Lame  -V4 (Avg bitrate 153 kbps) - 28 hours
(Advertised battery life on M3L is 35 hours  )

Of course this is an ideal measurement without switching the DAP on/off, changing volume level, without any EQ or screen light on.
Is there a difference between yes and no?

Ogg Vorbis or mp3 for portable use?

Reply #14 –
Quote
Have you tested, as in ABX, OGG vs MP3 to hear for yourself?   One test is worth a thousand opnions. [a href="index.php?act=findpost&pid=327844"][{POST_SNAPBACK}][/a]
Not if those opinions are themselves based on listening tests 
God kills a kitten every time you encode with CBR 320

Ogg Vorbis or mp3 for portable use?

Reply #15 –
Quote
Vorbis -Q4 (Avg bitrate 116 kbps) - 25 hours
Lame   -V4 (Avg bitrate 153 kbps) - 28 hours


That is the kind of difference I would expect. I consider this to be minimal (ie: less than 11% less battery life with ogg).

I am currently running a test a ogg -q5. Once done, I'll pit it against ogg -q-1, and then mp3. It will be a few days before I can post any significant results.

Ogg Vorbis or mp3 for portable use?

Reply #16 –
Interesting results.

I have an explanation other than reduced bitrates for ogg not using all that much battery compared to what's expected. There's also a non-xiph vorbis decoder on the market, focussing specifically on use with DSP's. It's mentioned in this thread.

I suppose for such a thing to be viable as a commercial product it has to outperform Tremor in the resources required to decode, which is usually linked to battery usage.

I did a little benchmark, i downloaded vov.exe (requires cygwin) from vinjey.com, and used OggDecT from rarewares.org.

I took a 1:07:35 podcast, which, at 128kbps nominally made for a file of 66.5MB, and started benchmarking. Here are the results:

Code: [Select]
D:\Download>timethis "start /realtime /b /wait oggdect.exe -o file.ogg > NUL:"

TimeThis :  Command Line :  start /realtime /b /wait oggdect.exe -o file.ogg > NUL:
TimeThis :    Start Time :  Wed Sep 28 18:41:37 2005

TimeThis :  Command Line :  start /realtime /b /wait oggdect.exe -o file.ogg > NUL:
TimeThis :    Start Time :  Wed Sep 28 18:41:37 2005
TimeThis :      End Time :  Wed Sep 28 18:42:49 2005
TimeThis :  Elapsed Time :  00:01:12.156

and
Code: [Select]
D:\Download>timethis "start /realtime /b /wait vov.exe file.ogg NUL:"

TimeThis :  Command Line :  start /realtime /b /wait vov.exe file.ogg NUL:
TimeThis :    Start Time :  Wed Sep 28 18:43:21 2005

VOV:: Got 0x201 while trying to get packet = 415631
VOV:: Time taken to decode is 96

TimeThis :  Command Line :  start /realtime /b /wait vov.exe file.ogg NUL:
TimeThis :    Start Time :  Wed Sep 28 18:43:21 2005
TimeThis :      End Time :  Wed Sep 28 18:44:57 2005
TimeThis :  Elapsed Time :  00:01:36.046


I ran each of these twice, but because of the realtime switch my system freezes and background apps cannot interfere, run times were identical for OggDecT, and the second vov run was 18ms slower. I consider these results highly accurate. There was no noticeable harddisk activity during the benchmarks. My system is an Athlon XP 2500+ with 1GB of PC2700 memory, slightly overclocked.

As you can see, tremor is 33% faster than vov, however this doesn't say much because vov isn't targeted for x86 systems, and i don't know the compiler and tunings that were used for it. It could still be god's gift for vorbis on DAPs
Veni Vidi Vorbis.

Ogg Vorbis or mp3 for portable use?

Reply #17 –
Quote
Battery tests:

mp3 -V6 for 1 hour = 3-4% battery life
ogg -q-1 for 1 hour = 4% battery life
aac = 12% battery life

I was very surprised by the high battery consumption of aac since I was under the impression that it was designed for low battery consumption. It was just the opposite with ogg: all the tests I had seen claimed that ogg use significantely more power than mp3, but the extra battery usage I observed was minimal.

Power comsumption conclusion:

Both mp3 and ogg are equivalent and very good. Stay clear from AAC.
[a href="index.php?act=findpost&pid=330121"][{POST_SNAPBACK}][/a]

I'm sorry for having to remark this, but this test is totally meaningless. Battery level indicators are nonlinear and extremely imprecise (edit: in fact, the expected uncertainty is larger than the measurements themselves). It would have been better if you would have exhausted a complete battery with mp3, another one with ogg, and a third one with aac, and compare the times. And even then, you'd have a significant uncertainty because of uncontrollable factors like the temperature. The ideal way to go would be to measure the consumed energy directly with specialised electronics.

Ogg Vorbis or mp3 for portable use?

Reply #18 –
Quote
I suppose for such a thing to be viable as a commercial product it has to outperform Tremor in the resources required to decode, which is usually linked to battery usage.


Don't forget also that current Vorbis encoder is using a lot of extra code that's not needed. I.E there is no reason to include floor 0 anymore at least for the time being ;-D.

Quote
I did a little benchmark, i downloaded vov.exe (requires cygwin) from vinjey.com, and used OggDecT from rarewares.org.


Vinjey systems seems like an interesting company with focus on embedded processors I could see them creating a faster decoding library.  They have also written a simple peeler too, which I mentioned on wikipedia, despite the fact that they mention a working example doesn't exist.  . Integerized MDCT routines have been a big plus though. I have cygwin on my HD I might have to play around with it after, although I don't have the time right now. 
budding I.T professional

Ogg Vorbis or mp3 for portable use?

Reply #19 –
Yeah, i was introduced to Vinjey through the peeler thread, that's why i linked to it in my post.

There's really not much to vov, it.s vov.exe in.ogg out.wav, no options, almost no verbosity. Their site mentioned it's mainly there to test the output for compliance. Haven't done that yet, i don't think i've the tools or experience to do that properly. I would expect it's results to be audibly indistinguishable from, if not identical to Tremor's.

Does anyone know if using cygwin instead of a standalone app has a significant performance impact?
Veni Vidi Vorbis.

Ogg Vorbis or mp3 for portable use?

Reply #20 –
Quote
Battery level indicators are nonlinear and extremely imprecise (edit: in fact, the expected uncertainty is larger than the measurements themselves). It would have been better if you would have exhausted a complete battery with mp3, another one with ogg, and a third one with aac, and compare the times. And even then, you'd have a significant uncertainty because of uncontrollable factors like the temperature. The ideal way to go would be to measure the consumed energy directly with specialised electronics.
[a href="index.php?act=findpost&pid=330245"][{POST_SNAPBACK}][/a]


I wish I had the equipment to do it. I am well aware the measurements are too short and can be in variable conditions.

However... I am now repeating them in as much as possible controlled conditions, for a longer period of time. I know it is not very accurate, but I consider it good enough to get a good idea of the relative power consumption of different codecs/bitrates.

Ogg Vorbis or mp3 for portable use?

Reply #21 –
I'm thinking about doing a test myself. My player has a built in voltage meter for the battery. I could just plot voltage vs playing time. Disable the backlight... disable all EQs and FX...

I guess ideally, I would have a camera pointed at the screen doing time lapse. then I could also lower the volume... use a better set of samples...

That's not gonna happen though, plus the info screen with the battery info goes away after a couple of seconds.

Ogg Vorbis or mp3 for portable use?

Reply #22 –
Quote
Battery tests:

mp3 -V6 for 1 hour = 3-4% battery life
ogg -q-1 for 1 hour = 4% battery life
aac = 12% battery life
[...]
Power comsumption conclusion:

Both mp3 and ogg are equivalent and very good. Stay clear from AAC.
[{POST_SNAPBACK}][/a]



Quote
Results:
Vorbis -Q4 (Avg bitrate 116 kbps) - 25 hours
Lame   -V4 (Avg bitrate 153 kbps) - 28 hours
[a href="index.php?act=findpost&pid=330184"][{POST_SNAPBACK}][/a]


Interesting. It could be compared to the [a href="http://www.hydrogenaudio.org/forums/index.php?showtopic=36391&view=findpost&p=321716]weird calculations made by oki[/url], claiming that "ARM7 needs about 10MHz to decode MP3 and about 180Mhz for OGG (for this purpose Rio Karma has 2 x ARM7 processors running at 90MHz) this is 18 times more power for ogg than for MP3!!! and which conclude with:

Quote
Concluding,
1st: Wogg ~= 18 x Waac;
2nd: Waac ~= Wmp3 in the newest implementations.
Wavpack Hybrid -c4hx6

Ogg Vorbis or mp3 for portable use?

Reply #24 –
When i benchmark vorbis, musepack and mp3 on my strongArm pocketpc the difference isn't nearly that big. Could provide some numbers if there's a demand for it.
Veni Vidi Vorbis.