The iBasso DX50 Thread - Latest firmware: 1.9.5 - June 30, 2016
Dec 14, 2013 at 11:50 AM Post #8,626 of 18,652
  Honestly, i'd like the crossfade if it was correctly made. Losing the first few beats off African Drop from a Knife Party remix, or the first few seconds out of a NIN song, really bothers me. The crossfade just doesn't sound right to me. Actually, i'm missing most of the beginnings of all songs. I guess it could be the 5 seconds setting, but turned off still does some sketchy fades. It could be a nice compromise to add both Gapless and Crossfade, but i don't think Crossfading would be "bit-perfect", would it?

Regarding the battery drain, i've been having some weird issues too, specially yesterday, when i left my DX50 resting, but turned on for around 30 minutes or so, was almost full, but when i came back to pick it up, was almost to the 2 last bars, on the last few minutes of the second to last bar. I came back to the 1.2.6 because i was doing some testing between the firmwares and i might do some try-outs to see if it is the firmware, because if i remember correctly, on 1.2.2 didn't had this issue...
 
Btw, thank you musicheaven and Sorensiim, i was looking for the 1.1.6, and actually had no idea there was a 1.1.5, mine came with 1.1.0 and went right to 1.1.6, this could help further the testing i've been doing with the different sound signatures between the firmwares.

 
Please by all means share you discovery, this is one aspect we have not touched upon due to the very personal nature of sound signature and preferences but a neutral and fair comparison between them is truly welcome.
 
Dec 14, 2013 at 12:28 PM Post #8,627 of 18,652
  Thank you!
 
I absolutely detest the new broken gapless, completely ruins live albums for me 
mad.gif

By the way Soren, the first page has never looked so good, you did a fantastic job gathering all of the info that was there in one sizeable see it all page, a true pleasure reading.
 
I have noticed a little problem but I really don't know if this can be fixed because it would require you know that one is browsing through a mobile device. The version table is clipped on the right so it makes it almost impossible to read anything past the first two columns. Not that anyone would be able to download anything on their mobile phone but if it had descriptive information, one would not be able to see it. Is there such feature from the editor that allows one to resize the table or add a different one depending if one is browsing from a portable device?
 
Dec 14, 2013 at 12:54 PM Post #8,628 of 18,652
   
 
I agree that the cross-fade can be funky on certain studio albums. However funky is fine as a choice but it is no substitute for gapless (which was not perfect but actually pretty good on 1.2.5). The issue is not just live albums but also all the classic concept albums from the Who, Genesis, Pink Floyd etc. For example, when the Lamb Lies Down On Broadway transitions into Fly On A Windshield, that is a continuous piece of music divided into 'individual tracks' for cueing convenience. Cross-fade on this is a disaster.
 
Cross-fade is cross-fade; it is not gapless and no pretense should be made that it is.

 
I have disagree with the gapless 'not being perfect' in 1.2.5, as I have been experiencing nothing short of perfection with it.  It is everything I want gapless to be, it needed not a single tweak.  And yet they tweaked it for 1.2.6.  lol...
 
Dec 14, 2013 at 1:15 PM Post #8,629 of 18,652
Guys, all this talk about the gapless issue in 1.2.6 being a "crossfade" feature is driving me nuts.  I'm sure iBasso had no intention of adding crossfade to the 1.2.6 firmware.  They just screwed up the gapless coding in the firmware and you're hearing the result. 
 
For those who don't understand what gapless means, it means that it should play the the original source CD exactly as it was encoded.  If there were fades and pauses of silence between tracks, then it should playback with those fades and pauses intact, exactly as intended.  However, if the source CD had no pause between tracks and the tracks merge into each other seamlessly (like a live concert recording or a prog rock "concept" album a la Pink Floyd), then it should be played back without adding even the briefest gap of silence between tracks. 
 
In other words, there should never be any reason to shut the gapless feature off on a DAP.  In fact, it's not even a "feature."  It's essential if you want perfect playback of the source material.  I'm perplexed by the fact that these DAPs even have an on/off switch for gapless.  There is no conceivable reason to ever shut it off. 

 
Dec 14, 2013 at 1:23 PM Post #8,630 of 18,652
  Guys, all this talk about the gapless issue in 1.2.6 being a "crossfade" feature is driving me nuts.  I'm sure iBasso had no intention of adding crossfade to the 1.2.6 firmware.  They just screwed up the gapless coding in the firmware and you're hearing the result. 
 
For those who don't understand what gapless means, it means that it should play the the original source CD exactly as it was encoded.  If there were fades and pauses of silence between tracks, then it should playback with those fades and pauses intact, exactly as intended.  However, if the source CD had no pause between tracks and the tracks merge into each other seamlessly (like a live concert recording or a prog rock "concept" album a la Pink Floyd), then it should be played back without adding even the briefest gap of silence between tracks. 
 
In other words, there should never be any reason to shut the gapless feature off on a DAP.  In fact, it's not even a "feature."  It's essential if you want perfect playback of the source material.  I'm perplexed by the fact that these DAPs even have an on/off switch for gapless.  There is no conceivable reason to ever shut it off. 
 

Hear hear!! If an album is gapless it should play gapless, if it's not it shouldn't. Simple.
 
Dec 14, 2013 at 1:39 PM Post #8,631 of 18,652
Damn we are getting alot of replies in this thread:D
 
Regarding the screen protector, I think it's pretty good, except you have to clean the screen first before installing it and not pull it back afterwards. So take care when installing it and it should be fine.
 
Actually I think the protector itself is quite high quality or iBasso has copied the way they make the piece from a highend manufacturer. I think it was Gorilla Glass or somebody else topnotch manufacturer who had that kinda flaps (for removing the covers for the protector itself) in their protectors. Or I'm totally imagining things and everybody has the similar flaps.
 
I'm still running with my original patch #7 FW. I don't bother to update it as I have no issues with it.
 
Dec 14, 2013 at 2:11 PM Post #8,634 of 18,652
  Guys, all this talk about the gapless issue in 1.2.6 being a "crossfade" feature is driving me nuts.  I'm sure iBasso had no intention of adding crossfade to the 1.2.6 firmware.  They just screwed up the gapless coding in the firmware and you're hearing the result. 
 
For those who don't understand what gapless means, it means that it should play the the original source CD exactly as it was encoded.  If there were fades and pauses of silence between tracks, then it should playback with those fades and pauses intact, exactly as intended.  However, if the source CD had no pause between tracks and the tracks merge into each other seamlessly (like a live concert recording or a prog rock "concept" album a la Pink Floyd), then it should be played back without adding even the briefest gap of silence between tracks. 
 
In other words, there should never be any reason to shut the gapless feature off on a DAP.  In fact, it's not even a "feature."  It's essential if you want perfect playback of the source material.  I'm perplexed by the fact that these DAPs even have an on/off switch for gapless.  There is no conceivable reason to ever shut it off. 
 

 
Actually, the first time i read about the gapless feature using time separation (5 sec or 10 sec), i actually thought it would turn on some sort of crossfade, or some break buffer between songs, and off was "No crossfade". Your definition of gapless is what i was expecting at first, but so far i haven't experienced it with all my records. My "gapless testing disc" would be The Slip, from NIN (specially the 999.999 to 1.000.000 song). Among the firmwares i've used, the only one that came close to gapless, was the 1.2.5. The rest, either had 1 or 2 extra seconds between tracks, or it had some sort of fade in, small, but i sensed it.
 
Dec 14, 2013 at 2:31 PM Post #8,635 of 18,652
Actually, the first time i read about the gapless feature using time separation (5 sec or 10 sec), i actually thought it would turn on some sort of crossfade, or some break buffer between songs, and off was "No crossfade". Your definition of gapless is what i was expecting at first, but so far i haven't experienced it with all my records. My "gapless testing disc" would be The Slip, from NIN (specially the 999.999 to 1.000.000 song). Among the firmwares i've used, the only one that came close to gapless, was the 1.2.5. The rest, either had 1 or 2 extra seconds between tracks, or it had some sort of fade in, small, but i sensed it.


That was my point how can you do gapless if there is a setting in there that injects something 5 of 10 seconds before the song ends, it just does not make sense. Like I said if you want to do it, then just do and only provide on/off if you want to let the songs end and starts naturally without removing any silence between tracks, meaning turn gapless off. On Apple products you don't even have the option, it just does it period.

As far as 1.2.5 I checked the song timeline and it seemed to me it was stopping say 5 seconds before it was scheduled to finish and the other one kicked in so that does not sound gapless to me. It was seamless but with a loss of few seconds at the end of a song.

Anyway I think we spent more than enough time on this feature and should instead enjoy the player and then rejoice when they finally fix it, again with whatever there setup means. I would urge them to use Apple type gapless as these guys have done it for years now and it works flawlessly.
 
Dec 14, 2013 at 2:55 PM Post #8,637 of 18,652
That was my point how can you do gapless if there is a setting in there that injects something 5 of 10 seconds before the song ends, it just does not make sense. Like I said if you want to do it, then just do and only provide on/off if you want to let the songs end and starts naturally without removing any silence between tracks, meaning turn gapless off. On Apple products you don't even have the option, it just does it period.

As far as 1.2.5 I checked the song timeline and it seemed to me it was stopping say 5 seconds before it was scheduled to finish and the other one kicked in so that does not sound gapless to me. It was seamless but with a loss of few seconds at the end of a song.

Anyway I think we spent more than enough time on this feature and should instead enjoy the player and then rejoice when they finally fix it, again with whatever there setup means. I would urge them to use Apple type gapless as these guys have done it for years now and it works flawlessly.

 
There is no loss of any seconds at the end of any songs in 1.2.5 with gapless on the 5 second option.  Every time, with every song and album I have that take advantage of having no gaps between tracks on the original cd, the dx50 would flawlessly end one track and roll into the next without even a hint at a pause.  Allows for my live albums to play through uninterrupted.
 
I still don't get what the 5 or 10 second options even actually mean.  If I set it to 10 seconds, I don't lose 10 seconds of any song, or even half a second, it still rolls perfectly into the next song unless you select the default option of 'close' or 'off', which would introduce a gap.  It is silly that the default is off.  Android phones the default setting is on, and there's not even a way to turn it off.  But it works perfectly on my HTC one with flac files, much as it also works perfectly on the DX50 on either the 5 or 10 second setting with flac files.  Unless you're using firmware 1.2.6, then, while it still does play gapless, it has a volume fade at the end of and start of the next song.  I wouldn't really define it as a crossfade though, because the two songs aren't playing on top of each other, or fading on top of each other either. 
 
Dec 14, 2013 at 3:18 PM Post #8,638 of 18,652
Dec 14, 2013 at 3:28 PM Post #8,639 of 18,652
Mind the Gap.

This is going to be a bit long so bear with me. It also suggests a reason why for the two different buffer sizes so again bear with me.

What is "gapless playback" really? To answer that question I need to describe why there are gaps in recordings. There are two reasons for this.

The first is because vinyl LP records have gaps between tracks, typically about 2 seconds long. When the Red Book audio spec was being drafted it was decided to retain these gaps to keep the same listening experience across the different mediums. The way this was implemented was dirt-simple: every track should have 2 seconds of silence at the end. So says the spec. There were lots of arguments about this, and about live recordings. Suffice it to say that it is not a strongly enforced requirement of the spec. Many CD-ripping tools have options to eliminate this silence. It's easy to do: scan backwards from the end of the file and then cut off the end when sound is detected.

The second has to do with how music is stored digitally. Audio recordings on analog mediums are continuous streams of one sort or another. Digital recordings are broken up into chunks called frames. Different encoding formats call for different sized frames but within a given file the frame size is constant. Assume for the explanation that a given file has frames that are 1 second long. What happens with a track that is 59.2 seconds long? You get 0.8 seconds of zeros to fill that final block. This can't be made to go away at encoding time; the frames are set by the encoding spec. Making it go away at playback time is tricky: you have to scan backwards from the end of the file and do it while the file is playing which means twice as much processing and I/O load during playback. It's easier to do with flash media than with rotating disks because flash has no seek times while rotating media has seek times measured in milliseconds.

Now, if my new reasoning is correct then what I wrote previously about gapless is incorrect. It isn't a buffer. It's how many seconds the player will scan backwards from the end of the currently playing file looking for the end of the audio in the track. In other words, it can find up to five or ten seconds worth of total silence at the end of a track while it is playing. If I'm figuring things right, DX50's player overlaps the start of the next track on top of this silence. If I'm figuring things right, the bug is that the backwards scan is failing to correctly detect the trailing silence and instead is overlapping the next track for the full five or ten second scan duration.
 
Dec 14, 2013 at 4:06 PM Post #8,640 of 18,652
There is no loss of any seconds at the end of any songs in 1.2.5 with gapless on the 5 second option.  Every time, with every song and album I have that take advantage of having no gaps between tracks on the original cd, the dx50 would flawlessly end one track and roll into the next without even a hint at a pause.  Allows for my live albums to play through uninterrupted.

I still don't get what the 5 or 10 second options even actually mean.  If I set it to 10 seconds, I don't lose 10 seconds of any song, or even half a second, it still rolls perfectly into the next song unless you select the default option of 'close' or 'off', which would introduce a gap.  It is silly that the default is off.  Android phones the default setting is on, and there's not even a way to turn it off.  But it works perfectly on my HTC one with flac files, much as it also works perfectly on the DX50 on either the 5 or 10 second setting with flac files.  Unless you're using firmware 1.2.6, then, while it still does play gapless, it has a volume fade at the end of and start of the next song.  I wouldn't really define it as a crossfade though, because the two songs aren't playing on top of each other, or fading on top of each other either. 


Ok, fair, I moved to version 1.2.5 on the DX50 to try to see if the gapless was ok, enabled gapless by setting it up to 5 seconds and no loss of music at the end of the first song but the next song started after a one second delay so not quite gapless. I changed the setting to 10 seconds and no difference. Gapless setting to close yielded the worse result with a two second lost at the end of the first song.

I think the crossfade is the most seamless setting so far on 1.2.6. The only thing I wish is for them to re-introduce true gapless with no delays whatsoever and make the cross fade optional. Now that would be a fantastic setup, everyone would be happy.

I am going to move back to 1.2.6 and check the gapless results with the same above setup.
 

Users who are viewing this thread

Back
Top