Although creating single files for whole albums is good, it does not allow me to play the individual tracks should I want to. As i only have a handful of albums where I find a gap in between tracks irritating, can I simply open these mp3s in audacity and then re-encode them as LAME mp3 (using Export as mp3 function) for when gapless becomes available for LAME mp3s? I know this question is more about audacity than the fuze per se, but you guys seem to know what you’re talking about!
also, as far as I am aware, the apple Ipods are able to perform a satisfactory gapless payback, is this only on m4a/aacs (or whatever they’re called)? If not, does anyone know how they do so, or is it just a mac mystery?
cheers
Andy
EDIT: apologies if this has already been asked and answered!
can I simply open these mp3s in audacity and then re-encode them as LAME mp3 (using Export as mp3 function) for when gapless becomes available for LAME mp3s?
What lame adds is information on the exact play time, which comes from the play time of the source file. IF you have an existing mp3 file then it is already too long and the exact play time isn’t stored anywhere. You could potentially in your editor manually cut off the extra before coding in lame.
If possible, it would be better to start with the source from which your mp3 was originally generated (CD, wav file, whatever).
also, as far as I am aware, the apple Ipods are able to perform a satisfactory gapless payback, is this only on m4a/aacs (or whatever they’re called)? If not, does anyone know how they do so, or is it just a mac mystery?
AFAIK, Itunes takes the exact playtime (whether from ripping a CD itself, or it can pick it up from a lame file) and stores it in the Ipod’s database at sync time. This does no good for players other than Ipod.
Edit: As has been explained a few times in this thread, the other necessary factor is the player being able to set up the next track before the current one is finished. This is a fairly recent thing for Ipod, don’t know offhand their first model to have gapless.
Edit: As has been explained a few times in this thread, the other necessary factor is the player being able to set up the next track before the current one is finished. This is a fairly recent thing for Ipod, don’t know offhand their first model to have gapless.
And the time it takes to setup the next track is by far the biggest annoyance factor!
I would just like to throw one thing out there regarding the whole tech discussion. The guys that have posted in here thus far are spot-on about the LAME tag and what it does, and why MP3 can’t be gapless as-is.
However, it is possible to guess where gaps are at the beginning or end of an MP3 stream and cut that out to create, if not a perfect reproduction of the original streams (which is pretty unlikely), gapless playback that does a great job at fooling my ears.
Lest you think I’m BSing, I wrote a Python script (which realistically only runs on Linux, but other than that I’m happy to share, if anyone’s interested) to do exactly this years ago, to take MP3s I bought from Emusic and create gapless CD images out of them. It basically decodes each MP3 in turn, checking the last fraction of a second for sound levels below a certain threshhold and chopping them off, and outputting each track in turn to a single WAV. Swap WAV output for sound output, and you’ve got yourself a gapless player.
I believe the xmms-crossfade plugin for some Linux music players does something similar to accomplish its gapless playback.
So, basically, given enough processing power on the Fuze and all the other issues that have been brought up here, it should be possible to make MP3s play back gapless even without the LAME tag.
ANother possibility for AJK’s question of de-gapping existing mp3 files once the Fuze is ready… If that lame-tag could be edited in manually in a tag-editor, you could determine the proper play length in audacity, then add the tag yourself and save the degradation that comes from re-encoding.
Conversionbox: yeah, perhaps not. I don’t really know. However, in my case, I simply chose an arbitrary threshhold and did integer compares for individual samples as far back as I could, rather than doing any actual level calculations.
It might even be “cheating” but I doubt anyone’d notice; it’s definitely better than nothing at all.
If you just run that willy-nilly, it will also eliminate gaps where there are supposed to be gaps. On the often cited example Abbey Road album, only a few of the tracks are supposed to be gapless.
I’m so glad SanDisk are looking into implementing true gapless playback. I’ve been waiting for such a player for years. I have all my music encoded in Flac, and some transcoded to Ogg Vorbis for my current (non-gapless) portable device that has a small storage space.
Will SanDisk be providing *true* gapless playback of both Ogg Vorbis and Flac files on this player?
Finally, I hope it is TRUE gapless, i.e. no gaps but also, no missed beats either. Another manufacturer of a non-portable device had thought they’d provided gapless Ogg Vorbis playback in addition to Flac when someone pointed out that the player missed a beat or two out during a track transition. Hopefully SanDisk will give us perfect true gapless playback first time.
How will the availability of gapless playback be announced?
Finally, in the meantime, we can play Ogg Vorbis and Flac files truely gaplessly on our PC usings free software called AquaLung. It’s quite new and it’s the ONLY software I have found so far that does *true* gapless playback.
I can’t wait for the gapless SanDisk firmware to come out - then I’ll buy myself a Fuze player.
I’m so glad SanDisk are looking into implementing true gapless playback. I’ve been waiting for such a player for years. I have all my music encoded in Flac, and some transcoded to Ogg Vorbis for my current (non-gapless) portable device that has a small storage space.
Will SanDisk be providing *true* gapless playback of both Ogg Vorbis and Flac files on this player?
Finally, I hope it is TRUE gapless, i.e. no gaps but also, no missed beats either. Another manufacturer of a non-portable device had thought they’d provided gapless Ogg Vorbis playback in addition to Flac when someone pointed out that the player missed a beat or two out during a track transition. Hopefully SanDisk will give us perfect true gapless playback first time.
How will the availability of gapless playback be announced?
Finally, in the meantime, we can play Ogg Vorbis and Flac files truely gaplessly on our PC usings free software called AquaLung. It’s quite new and it’s the ONLY software I have found so far that does *true* gapless playback.
I can’t wait for the gapless SanDisk firmware to come out - then I’ll buy myself a Fuze player.
My 2c
Thanks! :)
I dont know how “Truly Gapless” it will be, but I hope it is as you suggest. I have recorded some of my more essential gappless albums (Dark Side of the Moon, Led Zepelin IV, etc) as one track with no gaps to play on the fuze for now. The release of this feature will be part of a Firmware. So it will be listed as part of a chnagelog on here.
Have to admit I have not needed gapless playback for awhile, jut recently ran out of stuff to listen to and loaded in some of my old global underground mix cds and OH HORROR no gapless playback, argh, serious serious buzzkill.
Sure would be nice! Hint hint!!!
I think itunes uses lame in its mp3 encoding, so I can always convert stuff there and then bring it over here.
AFAIK Itunes puts the playtime info (for gapless) in the ipod database, NOT in an imbedded tag as Lame does. Last I tried (not sure if it’s the latest version), itunes “gapless” did not play gapless on other players know to be compatible with the Lame scheme.
If you read the thread someone said they can implement gapless but only for lame encoded mp3s. I was just saying I can convert everything to lame if that’s the issue.
Finally, in the meantime, we can play Ogg Vorbis and Flac files truely gaplessly on our PC usings free software called AquaLung. It’s quite new and it’s the ONLY software I have found so far that does *true* gapless playback.
@zonkerbl wrote:
If you read the thread someone said they can implement gapless but only for lame encoded mp3s. I was just saying I can convert everything to lame if that’s the issue.
Even if Itunes encodes with lame (doesn’t that I’ve heard), it doesn’t use a setting that puts the needed tag in. What itunes DOES do is read that tag and use it for the ipod database if you already have a gapless lame mp3 file.
2) You can’t just convert some other “gappy” file to lame and have the gap magically disappear. Lame is going to set that tag with the length of the original file, which has a gap.
I meant AquaLung was the only one that I tried that worked for me.
Xmms, Amarok and Audacious did not do proper gapless playback on my (Ubuntu) computer. Xmms did remove the gap mind you, but also missed a few beats of the song which ruined the track transition.
I don’t have a Windows PC so couldn’t try Foobar2000 or Winamp etc Sorry, I should have been more clear about that.
Finally, regarding the audio file format, rather than just mp3, I was hoping SanDisk would be implementing gapless playback for Ogg Vorbis (and even Flac would be nice but tFlac encoded files are rather large for a portable player.)
Anyway, hopefully SanDisk will post an update/announcement at some point so we can try it. Or do we just poll the release notes for the next version of the firmware?
Xmms, Amarok and Audacious did not do proper gapless playback on my (Ubuntu) computer. Xmms did remove the gap mind you, but also missed a few beats of the song which ruined the track transition.
Borderline OT, but I’m guessing you were using the xmms-crossfade plugin in its default crossfade mode. If you configure song transitions in the plugin to “None/Gapless” (IIRC), then it should use a strategy similar to the one I outlined here, stripping silent samples off the end of each track.
I guess it’s supposed to be able to tell when one song follows another, but in practice I had issues, so I just configured all song transitions to be None/Gapless.
This is rather important, by the way, to understand even if one isn’t using xmms-crossfade. A lot of people are complaining about “missed beats” with gapless strategies; if you’re missing an entire beat, you’re not using a limited silence-strip strategy. Your player is almost certainly crossfading, which is going to make mincemeat out of transitions. A limited silence-strip strategy, on the other hand, is going to do things as correctly as possible without the original track length info, and I’d be shocked if someone could really tell the difference.
This strategy is only needed for MP3 (and of course not needed if the LAME tag is present) due to the frame size. Ogg Vorbis and I believe FLAC can be of arbitrary length and so only need for the player to buffer the beginning of the next song before the current one ends.
Xmms, Amarok and Audacious did not do proper gapless playback on my (Ubuntu) computer. Xmms did remove the gap mind you, but also missed a few beats of the song which ruined the track transition.
Borderline OT, but I’m guessing you were using the xmms-crossfade plugin in its default crossfade mode. If you configure song transitions in the plugin to “None/Gapless” (IIRC), then it should use a strategy similar to the one I outlined here, stripping silent samples off the end of each track.
I guess it’s supposed to be able to tell when one song follows another, but in practice I had issues, so I just configured all song transitions to be None/Gapless.
This is rather important, by the way, to understand even if one isn’t using xmms-crossfade. A lot of people are complaining about “missed beats” with gapless strategies; if you’re missing an entire beat, you’re not using a limited silence-strip strategy. Your player is almost certainly crossfading, which is going to make mincemeat out of transitions. A limited silence-strip strategy, on the other hand, is going to do things as correctly as possible without the original track length info, and I’d be shocked if someone could really tell the difference.
This strategy is only needed for MP3 (and of course not needed if the LAME tag is present) due to the frame size. Ogg Vorbis and I believe FLAC can be of arbitrary length and so only need for the player to buffer the beginning of the next song before the current one ends.
Thanks for the info. I was using the cross-fade plugin and I am pretty sure I did configure it to use None/Gapless. The gap remover seemed to remove too much, so the track/song just stopped short and the next track began and it sounded wrong. Definitely some ‘notes’ missing though. I’m not sure where the note were missing from though (i.e. from the end of the first track? Or the start of the second track? Or both?) but I could tell there was a problem. I think the human ear is VERY good at detecting timing differences, at least, minor lip synch issues in video playback are very annoying to us.
It’s encouraging to hear that with Ogg Vorbis, one needs only to start loading up the next song before the current one finishes in order to get gapless playback (and I’m pretty sure it’s true since you can simply concatenate Ogg Voribis files together using cat!) However, I am not sure what other issues are involved in achieving this. I say this because Slimdevices had perfect FLAC gapless playback and then they added Ogg Vorbis playback to their devices, but their first attempt somehow chopped out too much ‘gap’ and there were beats missing. (A DJ had complained his tracks weren’t segueing(?) into each other correctly - the bad track transitions made them hard to dance too! And were very noticeable) Anyway, they fixed the issue at the second attempt, but one wonders how it could have gone wrong the first time if all you need to do is preload the next track in memory. Must be trickier than it sounds at first glance???
So anyway, how are SanDisk doing? Do they have ANY format playing gaplessly yet?
