Ogg Vorbis, MP3 and DAP battery life?

Topic: Ogg Vorbis, MP3 and DAP battery life? (Read 6179 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.

Ogg Vorbis, MP3 and DAP battery life?

2005-03-09 13:40:25

Hi all,

With all the talk about how Vorbis really drains a portable player's battery faster than MP3 (not sure about other formats), I would like to gather some user experiences and opinions on the topic.

I have an iRiver H120 with very good earphones, so having high quality files is important.
I can ABX q7 Vorbis from the source, but I am not too bothered at this level (just means that I do not want or like files below 160 Kbps/q5).
Battery life is about as important actually.

So if you were to get, say, 12 hours of battery life on ~192Kbps VBR MP3s, and about 7 hours with q7 Vorbis files (average bitrate around 200 Kbps), where is the sweet spot according to you?
At those bitrates I assume the Vorbis files will sound better, but (no matter how subjective) not 70% better.

So what do people choose betwen Vorbis and lame MP3, what quality/presets, etc, on their portable players?
Right now I'm hesitating between 192-256 Kbps lame MP3 (using presets, but I forgot which), and q6/q7 Vorbis. I'll probably make my own decision, just want to know what the opinions are.

Ogg Vorbis, MP3 and DAP battery life?

Reply #1 – 2005-03-09 14:25:12

Quote

Hi all,

With all the talk about how Vorbis really drains a portable player's battery faster than MP3 (not sure about other formats), I would like to gather some user experiences and opinions on the topic.

I have an iRiver H120 with very good earphones, so having high quality files is important.
I can ABX q7 Vorbis from the source, but I am not too bothered at this level (just means that I do not want or like files below 160 Kbps/q5).
Battery life is about as important actually.

So if you were to get, say, 12 hours of battery life on ~192Kbps VBR MP3s, and about 7 hours with q7 Vorbis files (average bitrate around 200 Kbps), where is the sweet spot according to you?
At those bitrates I assume the Vorbis files will sound better, but (no matter how subjective) not 70% better.

So what do people choose betwen Vorbis and lame MP3, what quality/presets, etc, on their portable players?
Right now I'm hesitating between 192-256 Kbps lame MP3 (using presets, but I forgot which), and q6/q7 Vorbis. I'll probably make my own decision, just want to know what the opinions are.
[a href="index.php?act=findpost&pid=280589"][{POST_SNAPBACK}][/a]

I have an HP120 also, with E2C Shure headphones. I'm currently in the process of ripping all (most of) my music to FLAC and then I'll do a batch run overnight to convert to...
Well, at the moment it's gonna be mp3 with LAME (of course) and I was gonna choose the latest alpha with -V 4 -vbr-new (preset medium fast). However, as I haven't done much testing with this quality setting I cannot say how I'm going to find it.

I have tried ogg vorbis at -q4 on my DAP and am happy with the quality. Indeed, I had intended to actually use ogg vorbis for all my music on my DAP... but then found out that I get less battery life. Compound this with the next LAME 3.97 alphas and the (hopefully) impending LAME 4.0, then mp3 for me is a good choice. It seems to be still a popular format.

Although...

When rockbox release their HP120 & 140 firmware, and if it improves battery life for ogg vorbis files (which they seem to imply will happen - for all audio formats), then I'll take another look at ogg vorbis on my DAP.

It's exciting times for both formats, as the recent work and improvements to some of the ogg vorbis encoder versions and the development on LAME, look highly promising.

I guess what I would say: is have all your music stored as lossless and that way you can easily transcode to the lossy format of your choice very easily.

I've not really answered your question have I... the 'sweet spot'?? Judging by my battery life experience with ogg vorbis I'd say that q7 is too high. Might be worth trying 3.99 with the latest aoTuV beta 3 compile?? Otherwise, stick with LAME at say
-V4.

Ogg Vorbis, MP3 and DAP battery life?

Reply #2 – 2005-03-09 15:10:07

Quote

I can ABX q7 Vorbis from the source, but I am not too bothered at this level (just means that I do not want or like files below 160 Kbps/q5).[a href="index.php?act=findpost&pid=280589"][{POST_SNAPBACK}][/a]

Are you sure you have a current version of the vorbis encoder? Samples that can be ABXed at this rate should be quite rare. If you've found some problem samples at that setting, you may want to upload some snippets, as they could be very helpful for improving the encoder -- you shouldn't have to use a setting that high to achieve transparency on most music.

I use -q 6 on my portable (Rio Karma) and am satisifed with the quality and battery life (and from the numbers you mention, it sounds like the HP120 has pretty similar battery life.) It may be overkill on a some music. There are a few things that have noticeable preecho on sharp transients at -q 6, so for those cases I use impulse_trigger_profile=3 switch (look for the special ITP compile in the recommended encoders section). The default value for that variable is 2.7 at -q6 and 3.0 at -q7. I used to use the gt3b2 version of oggenc, which had a similar tweak.

Ogg Vorbis, MP3 and DAP battery life?

Reply #3 – 2005-03-09 16:19:26

>I have an HP120 also, with E2C Shure headphones.

Nearly got those ended up buying Etymotic ER6's. Slightly more expensive, but (IMHO) slightly higher quality, and above all, easier to place in ears/take out.

>When rockbox release their HP120 & 140 firmware, and if it improves battery life >for ogg vorbis files (which they seem to imply will happen - for all audio >formats), then I'll take another look at ogg vorbis on my DAP.

I am not certian how Rockbox can improve that... but I'll keep my fingers crossed anyway.

>I guess what I would say: is have all your music stored as lossless and that way
>you can easily transcode to the lossy format of your choice very easily.

Yes, that's the plan. I've already done step #1 though, which is why I asked

Anyway I'll check LAME, do some comparisons. MP3's battery usage seems so low compared to Vorbis that I won't mind increasing the bitrate a little if necessary. Though the last encodes I did with LAME presets really impressed me.

>Are you sure you have a current version of the vorbis encoder?

It's one I compiled under Gentoo/Linux, IIRC it is version 1.1.0 Aotuv.
Maybe not the best encoder (I don't know, didn't check for that yet), but certainly respectable?

>Samples that can be ABXed at this rate should be quite rare.

I tried an album I know really well (Annett Louisan's Boheme) and did the ABX with Foobar2000's plugin, on 3 songs. In all of them I could tell which was the Vorbis within 30 seconds or so. Maybe I'm just too picky, because it does sound fine. Just that the source sounds better -> not transparent

I do not think they are problem samples per se, it's just the voice sounds... less pure? Difficult to explain.

To be honest I would probably be satisfied with a well-encoded q6 Vorbis file.
But given the above mentioned battery issue, think I'll re-encode my FLACs and CDs to LAME MP3 using presets.

Thxx

Ogg Vorbis, MP3 and DAP battery life?

Reply #4 – 2005-03-09 19:35:53

Quote

I do not think they are problem samples per se, it's just the voice sounds... less pure? Difficult to explain.

If they are not transparent at -q 7, they're problem samples. I would be interested in examining them. Would it be possible to upload a clip or two (less than 30 seconds) to the upload forum in a lossless format? They might be useful for improving vorbis quality.

Ogg Vorbis, MP3 and DAP battery life?

Reply #5 – 2005-03-09 19:35:56

i have an ihp140 i use koss sporta pros which sound good.

I use mareo to encode to flac -6 and lame 3.96.1 alt preset standard at the same time, i sometimes experiment with mpc ogg etc. (no mpc support yet..)

i find it hard to tell the difference from lame aps and ogg q6 and mpc q5. (not abxing) but if i do find something that sounds odd ill up ogg to 6.5 or lame to extreme. i dont bother going higher because
A. you most likely have problem track
B. higher quality eats battery (especially ogg)
C. do you really need transparent on a dap when you probably get outside noise from the hussle and bustle of everyday life. (if you do?, just use flac on a karma to be sure) its not like your gona abx whilst walking down the high street.

i havnt measured ogg battery use in a scientfic way but the fact that i can actually notice a shorter life without measureing sugests it must be fairly big (maybe 20% worse) so i stick with mp3 for now, but will almost certainly move to mpc if it is ever possible.

if you really can abx q7 ogg try even higher, you prolly found problem songs.

Ogg Vorbis, MP3 and DAP battery life?

Reply #6 – 2005-03-09 22:21:24

My choice fell on vorbis q6 on a H140. Even though 40 gb might sound much for around 160 kbps music, the hd space ran out faster than I thought I haven't yet found a lame mp3 switch that produce 160 kbps files sounding as good as vorbis aotuv b3.
A bit shorter battery life is not a big issue for me. The time is already good enough for my use.

Ogg Vorbis, MP3 and DAP battery life?

Reply #7 – 2005-03-10 03:50:25

Vorbis -q 5 user on my iHP-120. I tested my battery life and got around 12 hours for it so I felt pretty content I wasn't wasting away all my battery too quickly.

Ogg Vorbis, MP3 and DAP battery life?

Reply #8 – 2005-03-10 06:30:55

I'm doing some serious testing this week on Ogg vs. MP3 battery life, but my Samsung YP-MT6x gets a ridiculously long battery life with either, so it will take at least a week.

I'm using low bitrate audiobooks, then next week I'll test with music @ q6/7. Identical source files, of course. Maybe even a nice little graph when it's finished

Ogg Vorbis, MP3 and DAP battery life?

Reply #9 – 2005-03-10 10:16:25

@phong:
>If they are not transparent at -q 7, they're problem samples. I would be
>interested in examining them. Would it be possible to upload a clip or two (less
>than 30 seconds) to the upload forum in a lossless format? They might be useful
>for improving vorbis quality.

I'll be happy to do that. Soo you want me to upload the same 30 seconds, one in FLAC from the source, one in FLAC from q7 Vorbis?
(it might take a lot of listening to notice - good luck!)

@dobz
>C. do you really need transparent on a dap when you probably get outside noise >from the hussle and bustle of everyday life. (if you do?, just use flac on a karma >to be sure) its not like your gona abx whilst walking down the high street.

My earphones block about 36 dB, so outside noise was only a serious issue when Concorde flew over my town but I see your point, I won't ABX while sitting on a train, let alone *need* transparency. I do think 160 Kbps MP3 is good enough, but then if 192Kbps sounds better... you get the idea.

> havnt measured ogg battery use in a scientfic way but the fact that i can
>actually notice a shorter life without measureing sugests it must be fairly big
>(maybe 20% worse) so i stick with mp3 for now,

My experience would say the difference is 30% or a little over that, but YMMV (see DreamTactix291's response).

@DreamTactix291
>Vorbis -q 5 user on my iHP-120. I tested my battery life and got around 12 hours
>for it so I felt pretty content I wasn't wasting away all my battery too quickly.

12 hours with Vorbis q5? I am surprised, actually. I would have expected sthg like 10 hours max. Hey, that's good, since this is roughly what I get out of ~192Kbps VBR MP3s. I'll do some serious comparison.
If battery life hovers around the 12-hour mark, I'll be quite happy to pick the highest quality i can get for those 12 hours (Vorbis or lame MP3).

(why o why is it not simpler, like back when cheap DACs were crappy - we've all become very demanding, haven't we?)

Ogg Vorbis, MP3 and DAP battery life?

Reply #10 – 2005-03-10 14:06:01

Quote

I'll be happy to do that. Soo you want me to upload the same 30 seconds, one in FLAC from the source, one in FLAC from q7 Vorbis?
(it might take a lot of listening to notice - good luck!)

Just the source - I can encode the vorbis version myself. Also, a description of the nature of the problem and where in the sample it occurs.

Ogg Vorbis, MP3 and DAP battery life?

Reply #11 – 2005-03-10 15:15:20

Quote

Quote
I'll be happy to do that. Soo you want me to upload the same 30 seconds, one in FLAC from the source, one in FLAC from q7 Vorbis?
(it might take a lot of listening to notice - good luck!)

Just the source - I can encode the vorbis version myself. Also, a description of the nature of the problem and where in the sample it occurs.
[a href="index.php?act=findpost&pid=280972"][{POST_SNAPBACK}][/a]

OK I'll try and do that tonight from home (can't edit audio at work, and can't be bothered to install audacity here...).

Ogg Vorbis, MP3 and DAP battery life?

Reply #12 – 2005-03-10 17:33:04

I'm the user of iAudio M3L. The producer advertise this (L) version to have 35 hours playback, exacly:

Quote

For iAUDIO M3L can play continuously up to 35 hours and battery can be recharged fully within 6 hours.

This dap „supports MP3, OGG, FLAC, WMA, ASF & WAV.“
I use Koss PortaPro headphones, so quality is important. To save HD space is important too.
So decision lies between Lame and Vorbis. I tried to find transparent (for me ) settings with many encoder versions and switches. Fortunately my ears are deaf enough, so finally AoTuV b3 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).

Now I was wondering about real battery life with these two settings so I made a test:

Code: [Select]

Vorbis AoTuV b3*  -Q4  -   Avg bitrate 116 kbps**   -   25 hours
Lame Lame 3.96.1  -V4  -   Avg bitrate 153 kbps**   -   28 hours

*I'v tested Vorbis twice: with original AoTuV b3 encoder & Abcher b10 (based on AoTuV b3) - with exactly the same results.
** different bitrate, but similar transparency

Of course this is an ideal measurement without switching the DAP on/off, changing volume etc., but IMHO the conclusion (at least for me) is:
- minus 10% battery life is con for Vorbis
- plus 30% size (and similar quality) is bigger con for Lame.
So I will stick with Vorbis for now.

Frankly speaking, there is still "the gapless problem": M3L doesn't play gapless, but gaps are shorter with Mp3 !!! than with Vorbis. But this could be solved with next firmware, I hope.

Edit: typo

Ogg Vorbis, MP3 and DAP battery life?

Reply #13 – 2005-03-10 17:47:29

Well, one little off-topic question for iRiver (1xx, 3xx series) users: are here any differences in gap lengths between lame and vorbis too (as iAudio M3, according to my previous post here)?

Ogg Vorbis, MP3 and DAP battery life?

Reply #14 – 2005-03-10 22:19:53

Quote

Well, one little off-topic question for iRiver (1xx, 3xx series) users: are here any differences in gap lengths between lame and vorbis too (as iAudio M3, according to my previous post here)?
[a href="index.php?act=findpost&pid=281040"][{POST_SNAPBACK}][/a]

Not significant to my ears. Might be a trifle shorter with MP3's but it could also just be a placebo effect.

@phong: will upload a sample in a minute in the relevant forum section.

Ogg Vorbis, MP3 and DAP battery life?

Reply #15 – 2005-03-11 00:40:04

I've also uploaded two samples which I find interesting for testing (at the same thread as mickywicky)

Ogg Vorbis, MP3 and DAP battery life?

Reply #16 – 2005-03-11 00:46:21

Quote

I'm doing some serious testing this week on Ogg vs. MP3 battery life, but my Samsung YP-MT6x gets a ridiculously long battery life with either, so it will take at least a week.

I'm using low bitrate audiobooks, then next week I'll test with music @ q6/7. Identical source files, of course. Maybe even a nice little graph when it's finished
[a href="index.php?act=findpost&pid=280883"][{POST_SNAPBACK}][/a]

Let me know how it goes with the Samsung. I just got that as a birthday gift and am playing around with it (mostly in the car). Just to play around, i encoded all the Rush albums and some boots to ogg q2.5 and have been fairly happy for car use. Sure, I hear artifacts, but I can deal.

Ogg Vorbis, MP3 and DAP battery life?

Reply #17 – 2005-03-11 03:36:31

Quote

Well, one little off-topic question for iRiver (1xx, 3xx series) users: are here any differences in gap lengths between lame and vorbis too (as iAudio M3, according to my previous post here)?
[a href="index.php?act=findpost&pid=281040"][{POST_SNAPBACK}][/a]

Actually depending on whether or not the next song has been loaded into RAM the gap is pretty inconsistent. Anywhere from 500ms (this is the case if you're listening to stuff in album order) to 2s (this is the case if you change songs during the middle of a song). Doesn't seem to be much of a difference in gap length between mp3 and Vorbis.

Ogg Vorbis, MP3 and DAP battery life?

Reply #18 – 2005-03-11 08:49:06

Quote

Quote
Well, one little off-topic question for iRiver (1xx, 3xx series) users: are here any differences in gap lengths between lame and vorbis too (as iAudio M3, according to my previous post here)?
[a href="index.php?act=findpost&pid=281040"][{POST_SNAPBACK}][/a]
Actually depending on whether or not the next song has been loaded into RAM the gap is pretty inconsistent. Anywhere from 500ms (this is the case if you're listening to stuff in album order) to 2s (this is the case if you change songs during the middle of a song). Doesn't seem to be much of a difference in gap length between mp3 and Vorbis.
[a href="index.php?act=findpost&pid=281150"][{POST_SNAPBACK}][/a]

Yeah. The gap might be shorter with MP3 for me because I never re-encoded those from 128 or 160 Kbps, whereas the Vorbis files I have are much larger. So it makes perfect sense now

Ogg Vorbis, MP3 and DAP battery life?

Reply #19 – 2005-03-13 07:08:34

Quote

Actually depending on whether or not the next song has been loaded into RAM the gap is pretty inconsistent. Anywhere from 500ms (this is the case if you're listening to stuff in album order) to 2s (this is the case if you change songs during the middle of a song). [a href="index.php?act=findpost&pid=281150"][{POST_SNAPBACK}][/a]

That 's a good reason to go with a flash memory player. I get a 0.075 sec gap between tracks with MP3s split with musiCutter, 0.119 sec with tracks encoded encoded separately on my iFP-799.

Ogg Vorbis, MP3 and DAP battery life?

Reply #20 – 2005-03-13 08:12:11

Quote

Well, one little off-topic question for iRiver (1xx, 3xx series) users: are here any differences in gap lengths between lame and vorbis too (as iAudio M3, according to my previous post here)?
[a href="index.php?act=findpost&pid=281040"][{POST_SNAPBACK}][/a]

iFP series cuts off the ends of Ogg Vorbis tracks, if that's what u want to know... then the gap appears, which is a bit more than MP3 gaps IME.

EDIT: clarity

Notice