Poll
Question:
MP3
Option 1: ~100 kbps or less
votes: 10
Option 2: ~130 kbps
votes: 32
Option 3: ~145 kbps
votes: 12
Option 4: ~160 kbps
votes: 28
Option 5: ~180 kbps
votes: 45
Option 6: ~200 kbps
votes: 66
Option 7: ~225 kbps
votes: 38
Option 8: ~260 kbps
votes: 60
Option 9: 320 kbps
votes: 62
Option 10: I don't encode to MP3
votes: 127
It was a long time since poll like this. It will be interesting to see what bitrates people use in 2011.
Please, move this topic to the poll section.
Ogg Vorbis, q5.0 (LancerMod SSE3).
I would rather target a quality than a specific bitrate.
hm.. I answered ~180 for MP3 because that the average of a few albums I just selected.
But I agree with Nick, since I just use -V2.
The biggest change since the last poll I participated in is that storage is so cheap now that I really don't need to care about bitrate.
The complicated thing is that if options would be target bitrate then I should mention ALL of settings like VBR/ABR/CBR and even 2-pass ABR/VBR. Plus it would make the poll unreadable with all settings.
If I go only for -Vx then it's limitation for not LAME users http://www.hydrogenaudio.org/forums/index....st&p=640471 (http://www.hydrogenaudio.org/forums/index.php?s=&showtopic=72265&view=findpost&p=640471)
So the decision was to go as generic as possible. No settings.
BTW some quality labels are really funny like "deaf, radio, insane.... braindead?". How about "superman, batman, wolverine, ironman and wonder woman"
Where's guru and the 2011 ripping/encoding general poll?
ogg vorbis @ q5 (~160kbps). i've never bothered with listening tests (too lazy) so i had a search around and this seemed to be a popular choice of codec/setting and i just went with it.
I was using Vorbis at q 4 previously but found several tracks in my own library with obvious artifacts (mainly but not only metal tracks), so I did some ABX tests and concluded that q 5 is safe.
MP3 requires too much bitrate to be transparent, so I'm not using it anymore.
I was encoding @mpc-q5 till 2006,switched to oggV till 2009 @q5,and 2010 i just go lossless couse hd are so cheap
I use LAME V6 for the portable as well as AAC at about 128.
On the desktop I use LAME V0 and have a FLAC/TAK collection
AAC (Nero or QuickTime) at VBR around 130 kbps.
Quality is really strong to my taste and for portable use. There are some minor annoyance from time to time, but I easily forgive them. I have everything in lossless formats at home.
(thanks for this poll too)
I rip only to FLAC so bitrate is pointless to argue about there.
When I'm transferring to a DAP, I transcode to ogg vorbis at -q2 (~96Kbps)
If I'm transferring to a DAP that doesn't support ogg vorbis (of which I own none), then I transcode to mp3 using lame -V2 (~196Kbps)..
To me vorbis loses most annoying artifacts at -q2 (~96Kbps), and becomes transparent at -q5 (~160Kbps).
To me mp3 loses most annoying artifacts at -V2 (~196Kbps), and becomes transparent at -V0 (~256Kbps).
aoTuV Beta6.02 @ q5 still sound transparent in my own opinion for the given bit rate and file size.
I use NeroEnc -q 0.45 when I need lossy files, which has a bit of a range depending on the audio.
Vorbis @ q6 here.
Non-losslessly speaking, still using Musepack --standard here, SV7 due to slow adoption of SV8 decoders (W*cough*nAmp).
Why Musepack: still appears have the fastest decode speed of lossy formats according to my tests using foo_bench.dll; psy-model/VBR appears to perform well (mpcbits.exe showed bitrates at --standard between 3kbps to over 1000kbps); APE tags; and the comfort of using a format less known for artist exploitation (translation = screw the leechers ).
In terms of lossless, I use FLAC so bitrates are, as said "pointless"
In terms of lossy, I still use the ancient MP3. And even though most people (here) are using VBR, I'm still using CBR, 320kbps though small quantities of my library are encoded in LAME VBR V0 and V2
I have start to encode in Nero AAC q 0.45 (~150 kbps) after I started to hear differences between Lossless (ALAC) and AAC 128
One good song where I really can hear difference is Red Hot Chili Peppers - Californication, in the AAC 128 encode there is sounds that is missing, but in q 0.45 it is (for me) fully transparent.
foo_abx 1.3.4 report
foobar2000 v1.1.2
2011/03/05 20:37:23
File A: C:\Red Hot Chili Peppers\Californication\06 Californication.m4a
File B: C:\Red Hot Chili Peppers\06 Californication.m4a
20:37:23 : Test started.
20:37:56 : 01/01 50.0%
20:38:31 : 02/02 25.0%
20:39:21 : 03/03 12.5%
20:40:02 : 04/04 6.3%
20:40:40 : 04/05 18.8%
20:41:22 : 05/06 10.9%
20:42:06 : 06/07 6.3%
20:43:09 : 07/08 3.5%
20:43:35 : 08/09 2.0%
20:44:17 : 09/10 1.1%
20:44:45 : 10/11 0.6%
20:45:41 : 11/12 0.3%
20:46:04 : 12/13 0.2%
20:46:25 : 13/14 0.1%
20:46:54 : 14/15 0.0%
20:47:09 : 15/16 0.0%
20:47:39 : 16/17 0.0%
20:47:57 : 17/18 0.0%
20:48:27 : 18/19 0.0%
20:48:45 : 19/20 0.0%
20:49:18 : 20/21 0.0%
20:49:32 : 21/22 0.0%
20:49:54 : 22/23 0.0%
20:50:12 : 23/24 0.0%
20:50:35 : 24/25 0.0%
20:51:07 : 25/26 0.0%
20:51:16 : Test finished.
----------
Total: 25/26 (0.0%)
Lame V3 (Voted ~180 kbps).
I used V5 for the longest. But even phones today come with 16GB standard. It's not such a big deal anymore. My library grew a couple of gigs. *shrug*
I use FLAC or WavPack for ripping, but if I need lossy, I'll use LAME V2. If I need really lossy, I'll use Nero AAC -q 0.3.
Since large storage is cheap I just rip to Wavpack with switch -hx.
For my Rockboxed portable I use lame 3.98.4 -V4
256 kbps AAC at home, 128 kbps AAC for my iPhone and iPad, 64 kbps mono MP3 for the car so it fits on one CD-R.
256 kbps AAC at home, 128 kbps AAC for my iPhone and iPad, 64 kbps mono MP3 for the car so it fits on one CD-R.
May I ask: so that what fits on one CD?
64 kbps mono MP3
okay...
For lossy neroaacenc -q0.55.. Occassionally mp3 -v0. Voted ~260/~220
My lossy compression of choice is Ogg (AuToV whichever's-latest) at q6, or ~192kbps (so i selected ~200, as the results usually shoot high of 200). These days I've gotten my EAC set to rip straight to FLAC (compression level -6, which is the max anyone should bother using IMHO), then I make my lossy copy with oggdropxd (still q6).
Musepack SV8 Standard Bitrate when used in conjunction with EAC ver 1.0 beta 2
then copy files across to various iPod 5th Gen players with Rockbox ver 3.8.1 which
has support for Musepack SV8
also very happy with VLC Media Player on various Windows laptops churning out SV8
encoded files via various headphones, speakers etc around work and home :-)
I would rather target a quality than a specific bitrate.
Agree with him. I use mostly -v2.
I switched computing platforms from Windows to Mac OS X earlier this year so my lossy encoding preferences changed. Before I would rely on Nero AAC at -q0.45 and ALAC for my archives. The later hasn't changed but I now use QuickTime AAC at 160kbps true VBR as it is built right into the OS and XLD can harness that rather easily. I encode at 128kbps VBR_constrained with iTunes AAC for uploading to my Amazon cloud account.
Why Musepack: still appears have the fastest decode speed of lossy formats according to my tests using foo_bench.dll; psy-model/VBR appears to perform well (mpcbits.exe showed bitrates at --standard between 3kbps to over 1000kbps); APE tags
I was requested to upload MPCbits.exe for another HA user, and having found the ancient binary I decided to re-examine my claim. It turns out my above statement is not entirely correct. Unable to edit my previous post, I wanted to clarify that huge bitrates reported by MPCbits.exe are usually one or two frames, and usually at the beginning of the audio file. This appears to be a necessary design of the gapless function (forgot to add that in my list of features ) and usually appears when the audio is at the very start of the track.
An example of a song that has 1000+ bits frame at the beginning is off Metallica's Ride The Lightning, track 06 (note: my pressing is very early, so I'll mention the RG reported -4.01dB for the track on my CD). I specifically used Musepack version 1.1.6 with no commandline arguments (-q5).
Looking at the results of this and a previous poll for a public AAC test (http://www.hydrogenaudio.org/forums/index.php?showtopic=77809), it seems that people use >130-kbps bitrates much more often than say <100-kbps or so. Which leads me to conclude that - if there will ever be another high-bitrate test - people would be more interested in testing 130 kbps than lower bitrates. Correct?
Chris
Looking at the results of this and a previous poll for a public AAC test (http://www.hydrogenaudio.org/forums/index.php?showtopic=77809), it seems that people use >130-kbps bitrates much more often than say <100-kbps or so. Which leads me to conclude that - if there will ever be another high-bitrate test - people would be more interested in testing 130 kbps than lower bitrates. Correct?
While i can not pick -v4 from the original, i encode to v2, because i like to play music with friends, and i rather be safe than sorry, so yes, for me 130 is more interesting than 100.
Looking at the results of this and a previous poll for a public AAC test (http://www.hydrogenaudio.org/forums/index.php?showtopic=77809), it seems that people use >130-kbps bitrates much more often than say <100-kbps or so. Which leads me to conclude that - if there will ever be another high-bitrate test - people would be more interested in testing 130 kbps than lower bitrates. Correct?
Chris
I'm also interested in 128 kbps public test.
Although it makes sense to conduct it first at 96kbps.
As for me when I can't hear the artifact at high bitrates I will go to lower bitrate to figure out the nature of it and then get back to spot it at previous high bitrate.
Now the listeners who have participated in previous 64 kbps test are trained to perform new one at 96 kbps. Jumping to 128 kbps will be too abrupt.
And then we will see.
Just my thoughts.
While i can not pick -v4
That means that you would be able to provide useful results if there would be -V 5 (130 kbps) test and even lower bitrate for AAC, Vorbis, Opus.
Just a thought on the choices.
Though the results don't show a lot down there I would have expected people choosing <130 kb/s would have more reason to make a conscious choice and more significant difference in quality and space driving that choice compared to choosing between, say, 200 and 225 kb/s. But "100 and less" is the only choice in that range.
If I'm encoding for the portable, generally vorbis in the q0 through q3 range depending on how much I like the piece.
For just talk, especially audio books that often run 6 hours, I encode at 10-20 kb/s in speex.
Music on the PC, mostly vorbis or mpc in the 190 range.
Mp3 is mostly things I get already encoded in the 180-256 range from emusic, some music podcasts, etc.
I just started using MPC extreme (~200 kbps) on my Rockboxed Sansa Clip+ (it's skipped by the annoying OF indexer as a bonus). I was previously using Vorbis q6 on it and VBR MP3 around 220kbps (I guess it was V2) on my Samsung mobile phone. I also keep everything in FLAC on a PC.
~260, although I only use lossy on my underused DAP. . .
On my PC, where storage is unbelievably cheap, I refuse to settle for anything other than lossless, just for the sake of having guaranteed "perfect" archival quality. I know it's overkill, but when a terabyte of storage can be had for $60, I think overkill is justified. Furthermore, having lossless files on my PC simplifies transcoding if I change my mind about what I want on my portable. Within the lossless domain, I use FLAC because it's the best supported.
On portable equipment, where storage is an order of magnitude more scarce, I use Vorbis -q2. When encoded from the FLACs on my PC, I find it near transparent. I can ABX some stuff but the arifacts are extremely small nitpicks and not at all annoying and there's plenty more that I can't ABX. Vorbis -q2 is roughly 96 kbps and lets me fit tons of stuff on my cheap yet very good 4GB Sansa Fuze and Clip. I don't care about "archival quality" or "safety factors" here because I have the FLACs on my PC.
I know that my encoding preferences sound like two extremes, but in my experience once you get to "near transparent", increasing the bitrate of lossy compression further gets you to "completely transparent" in a very slow, asymptotic manner, if at all. Even though I can't ABX most stuff at -q2 in Vorbis, there's one killer sample that I've successfully ABX'd at -q7 (nominal bitrate 224 kbps). Therefore, I tend to believe in lossless, which is guaranteed transparent, when space is cheap and relatively low "near transparent" bitrates when space is at a premium.
I'm using MP3 for my "high-end" equipament, headphones (in case, SONY MDR-V900HD) of my pc, and aac to my mobile phone with cheap phones, due to compatibility with "tagged" covers isn't MP3, replaced with aac (Nero).
LAME 3.98 -v5 or QuickTime AAC q65 (which usually ends up at about ~120 kbps for my collection). I also have the original FLACs of my music on my main computer, but for some reason I prefer listening to the lossy rips.
I try to walk the line between file size and quality, and those bitrates are perfect for that. I still have a rockboxed Sansa e260, but I also have a new 8 GB iPod touch. Neither have all that much space, and when I'm listening to my portable player on the bus/subway, I wouldn't be able to hear low bitrate artifacts anyway. (But I'm not a very discerning listener compared to many people here -- LAME -v4 sounds transparent to me.) I think higher bitrates would also consume too much battery for my liking.
I use mp3 for universal compatibility, and AAC on my iPod. I would have completely swiched to AAC by now, but it's not as efficient to decode, i.e. it uses more battery on rockbox devices.
I prowled through these forums about codec information, and these two links: http://www.hydrogenaudio.org/forums/index....60117&st=50 (http://www.hydrogenaudio.org/forums/index.php?showtopic=60117&st=50) and http://www.hydrogenaudio.org/forums/index....showtopic=66949 (http://www.hydrogenaudio.org/forums/index.php?showtopic=66949) told me that low bitrate QuickTime AAC was actually incredibly good.
I feel like lame mp3 or aac have reached their peak in quality at the bitrates I encode at, for at least a year now, and that makes me relieved. I've been replacing my older mp3 collection recently (like 128kbps mp3s from ten years ago; they sound too strange and lo-fi for me now). I'm confident I'll be able to preserve my current collection without having to re-rip everything in the future.
FLAC w/cue sheets for archival/storage purposes.
ALAC for most music on my iPod Touch 64 Gb. (Probably not entirely necessary, but I don't care.) AAC VBR @256 Kbps for some.
HE-AAC @ 64 Kbps (usually stereo) for spoken word recordings.
(I can ABX a LAME-encoded MP3 @ 128 Kbps 100% of the time, but that's trivial. The differences I've heard in such tests were nearly negligible and I probably would not have noticed them if I wasn't "testing" something. I've never tested anything higher than that...it doesn't interest me too terribly.)
I don't encode to a bitrate, it's whatever -V 2 gives me.
Like others I also don't encode to a bitrate.
For regular albums I have that are alright, but which I also don't listen to that often I use:
LAME -V 2
For albums I love and listen to allot, i.e.; several times each year I use:
LAME -V 0
I also have many albums archived on DVDs using FLAC.
I use LAME -V0, which I think puts me in the ~260kbps MP3 slot
I use lossless (TAK) and convert to MP3 VBR V2 for portable/car. Excluding digital purchases which are usually 320 kbps MP3.
I have been using LAME 3.98.4 at -V0, which is around 256 kbps.
FLAC copies on Genuine Verbatim DVD's, although the last time I promised not to do DVD backup anymore as it is a tedious and consuming process.
I use LAME 3.98 with V0. I can't tell the difference between V2-V0 and lossless even on pretty revealing headphones, yet my hearing can go up to 19kHz(and I'm 20 years old lol).
Guess that goes to show that LAME does an excellent job in my opinion.
Home computer - FLAC
Portable Player - MP3 @ -V0 (245kbps)
External HD for Work - Vorbis @ q 4.0 (128kbps)
Perhaps its overkill, but I paid for the space, so why not use it? The speakers I have with the work system really aren't that super, so vorbis works great for the stuff I want to take to work.
I didn't realize lossy formats were still this popular.
Lossy formats are much more popular than non lossy ones.
LAME v0, mostly for for PC and ipod
FLAC , mostly for home theater and ipod with Rockbox
I ripped very carefully to FLAC with dBpoweramp and then converted to mp3 with LAME. My priorities were to never rip or transcode again, and treat disc space as cheap or unlimited. I never conducted a robust comparison between LAME v0 and FLAC on my ipod, PC or home theater, but with crude comparisons I could not recognize or recall any differences between these formats on any playback device. I will not rule out issues with my old Irish ears. I find myself listening to my LAME library much more than FLAC because I spend much more time in front of my computer instead of my home theater.
FLAC all the way... On my home digital receiver.
I don't listen to music on phones or other similar devices.
At PC with studio headphones: FLAC or MP3 320 kb/s... bottommost MP3 V0 ... I do not like lowpass on my files.
For my car hifi MP3 V2, that's enough. Signal-to-noise ratio is not that good in my car.
I use LAME -V0, which I think puts me in the ~260kbps MP3 slot
I've started to use -V0 as well, though it seems to usually stay between 180-210 kbps sometimes, and usually doesn't get above 220 kbps... hopefully, I'm not doing anything wrong. The kbps was even lower before I switched to "slow" ripping v0 (I use dbpoweramp).
I used to rip everything to 256 kbps CBR.
It depends on the genre you mostly listen to. Some music requires lower bitrate to reach the same fidelity.
FLAC for my archive, mp3 -V4 for mobile use. Average bitrate for 11426 (almost completely self-ripped) mp3s: 153kbit/s.
Don't see a reason to switch to a post-mp3-codec because mp3 can sound as good as a cd (at least with -V2 and above) and no other codec is as widely supported and accepted.
@krafty: Copies on dvd? Are you serious? An external hdd lasts longer, is cheaper (per GB) and is way more comfortable (and smaller and faster and ...).
This has kept me happy.
Music: Archive to Flac, and MP3 (LAME) -V2 -q0 --lowpass 19.7 for on the go.
Audiobooks: AAC (Nero) -q 0.22 and MP3 (LAME) -V9 -q0 --lowpass 17 -b56
All my CD collection is duplicated in FLAC and all music I download I try to acquire in FLAC from downloads. After I have them like that I transcode to Ogg Vorbis q5
foobar2000:
qtaacenc --cvbr 64 --he - %d
Celt 0.11.2 with fb2k
celtenc --bitrate 64 %s %d
extension file: oga
and flac, Tak 2.2.0 use for archival
AAC:
Qaac @ (~100kbps) for my phone
Lossless:
Zip/WavPack (zip.wv) for PC playback and archiving with AR/log/art/etc...
I know this thread is old, but here's what I use now:
Rip CDs to FLAC with foobar2000 for archive, these are put on my external 1TB HDD.
Encode to V2 or V0 depending on music complexity for portable use, latest LAME version is always used .
I know this thread is old
It probably should be closed so we can gather new statistics in a new poll.
OK, make considerations, something that will be useful to see and feel free to open the poll.
I use 320 mp3. mp3 is compatible with everything and storage is so cheap and vast that I don't bother with any vbr. Plus, I get some comforting sense of having the best possible mp3 that soothes my OCD. I archive everything in FLAC with logs and cue sheets.
I use 320 mp3. mp3 is compatible with everything and storage is so cheap and vast that I don't bother with any vbr.
I only listen to MP3 on portable players, all with 32GB or less of storage. While that storage is relatively cheap, it is limited, so I use VBR at about 180kbps, which nearly doubles the number of files I can get on the devices compared to 320 kbps. And considering that I'm usually listening to those players while in a noisy gym, the sound quality is way more than adequate.
I use 320 mp3. mp3 is compatible with everything and storage is so cheap and vast that I don't bother with any vbr.
I only listen to MP3 on portable players, all with 32GB or less of storage. While that storage is relatively cheap, it is limited, so I use VBR at about 180kbps, which nearly doubles the number of files I can get on the devices compared to 320 kbps. And considering that I'm usually listening to those players while in a noisy gym, the sound quality is way more than adequate.
That's cool. I don't use any portable music devices. At the gym they have TVs on the treadmills. I do all my listening either at home on my computer, where I shuffle my music, or in my car, where I just listen to CDs old school style.
I don't use any portable music devices. At the gym they have TVs on the treadmills. I do all my listening either at home on my computer, where I shuffle my music, or in my car, where I just listen to CDs old school style.
Then why don't you simply listen to the FLAC files that you have? Why a need for MP3s at all?
I don't use any portable music devices. At the gym they have TVs on the treadmills. I do all my listening either at home on my computer, where I shuffle my music, or in my car, where I just listen to CDs old school style.
Then why don't you simply listen to the FLAC files that you have? Why a need for MP3s at all?
FLAC files are very large. Thus, instead of keeping them on my computer I archive them. Plus, I do a lot of manipulation to the mp3 files that I don't want to do to the flac files, things like renaming, retagging, replay gain, etc. The FLAC files are directly from rips and archived exactly as they are. Even with a 2TB drive flac fills it up pretty quickly whereas I can have weeks and weeks of music to shuffle with mp3 and still tons of room to add more. So, if I ever need to burn a disc, I have the flac album. If I ever need to encode to some other lossy format, I have the flac, etc.
You just got done saying "storage is so cheap". Which it is - you can store about 6000 CDs on a $100 2TB hard disk.
You allow your ripping software to tag those files, don't you? I retag my FLAC files after ripping, add ReplayGain tags, move the files, sometimes rename them. All of that "manipulation" is to the metadata only, and leaves the audio portion of the files completely untouched. I can still burn an exact duplicate of the CD using the CUE sheet that I generate during the ripping process, and use the files as source for generating MP3s.
Plus, I do a lot of manipulation to the mp3 files that I don't want to do to the flac files, things like renaming, retagging, replay gain, etc.
You probably should have said they are less than half the size and left it at that. Now we all get to watch you be disabused.
<sarcasm>
Joy!
</sarcasm>
Eh, the way I do it makes sense for me and makes me happy. Cheap is a relative term I guess.
You just got done saying "storage is so cheap". Which it is - you can store about 6000 CDs on a $100 2TB hard disk.
You allow your ripping software to tag those files, don't you? I retag my FLAC files after ripping, add ReplayGain tags, move the files, sometimes rename them. All of that "manipulation" is to the metadata only, and leaves the audio portion of the files completely untouched. I can still burn an exact duplicate of the CD using the CUE sheet that I generate during the ripping process, and use the files as source for generating MP3s.
If you rename them then you can't, unless you are willing to manually edit the cue sheet.
Also, I should have mentioned that I have some stuff where I have the mp3s but not a lossless rip. Long story short, it soothes my autism to do it the way I do it because I want them all to be uniform. We all have our ways I guess.
And yes, I use EAC to tag the files, but my desires for my mp3 tags are different. I use as little information as possible in those tags: artist, title, album, track # and year. That's it. For all 70 bajillion of them. I don't consider the tags important in an archived lossless format so I don't mess with them.
I have actually experienced file corruption when overwriting during tagging, so to some extent I understand the “don't touch my rip archive!” attitude. (Using generous padding upon first encoding should reduce the writing, though ... for those who wish to comfort their semiparanoia.)
But then I don't understand the purpose of a FLAC archive never to be played, and some MP3 everyday drive (unless for portable use, where space is still an issue) – when just by upsizing you can have another backup of your FLACs. My second set is in my office, on external USB drives. Files are created at home, with file ownership being Porcus@home, so my work computer denies my every attempt at changing or deleting ... I have to live with those junk tags. And unlike an untouched backup, this one is test run 250 days a year.
As a guard against “what the heck did I just write to file?”, I did once fiddle around with an OpenSolaris server setup with the awesome ZFS file system for snapshotting. But nowadays Windows users can set up the Volume Shadow Copy as well.
But then I don't understand the purpose of a FLAC archive never to be played, and some MP3 everyday drive (unless for portable use, where space is still an issue) – when just by upsizing you can have another backup of your FLACs.
It's because I'm extremely anal about the collection I listen to. I want them all to be uniform and not every one (but most) come from EAC Rips. Every single album in the collection I listen to is 320 and tagged the same way, etc. Plus, I have other things on the drives I use such as movies, etc.
It's just what makes me happy, it doesn't have to make sense to everyone. It's fun to hear what other people do though. I'm sure that when we have 50 tb hard drives or whatever it won't be an issue. At that point FLAC won't even be necessary, except for tagging I guess, but in my mind no tags are better than bad ones.
Not really sure how to answer.
I mostly use ogg -q5 but I also use ogg -q1 for my portable.
For mp3 I just use -preset standard which I think is V2 ???
But I rarely do mp3 these days.
And yes, I use EAC to tag the files, but my desires for my mp3 tags are different. I use as little information as possible in those tags: artist, title, album, track # and year. That's it. For all 70 bajillion of them. I don't consider the tags important in an archived lossless format so I don't mess with them.
I do the same thing - lots of tags for flac (and my ogg vorbis has same as flac) but when I transcode to mp3 I just the basic 6 - artist album title track year genre - plus a comment (comment has version of lame used to encode)
I'm not a big fan of ID3 - maybe I'd be more tag happy in mp3 if I used ape or something, I don't know.
for stereo movie soundtracks (sharing edits (for preview purposes) over intra or internet);
afconvert -v -f "m4af" -s 3 "$tmp.aif"
(around 100 kbps average i think)
for production backups; flac or alac or i just keep them aiff.