Variations in mastering? CD ripping a flawed process?

[joeyc410]

Hi all,

I have posted about this before and I apologize if you recall reading about this particular issue a number of other times. I am just trying to understand why a specific issue occurs with ripped tracks from a CD. Much of the ideas I have shared below have come from people in this support forum and on other audio sites/forums I have posted in and read.

Note: please excuse the weird numbers and ampersands in this post; every time I try to edit words that have apostrophes, I hit save and my post gets messed up... I think it is still clear what I am trying to say.

If the read offset of a drive is inaudible, and certain players use fade-in/synch delay to hide/soften hard track beginnings, is it really true that almost every gapless album has annoying pops/clicks at particular track starting points? I would think that the industry standard would be to find an exact point in a track to prevent a pop sound when shuffling/selecting tracks from gapless albums. Such a task can be done with Audacity, but in my experience, it is quite tedious, which is why I do not bother with it (editing 100s of tracks, zooming in hundreds of times to the waveform of a track to find a better starting point than what is originally chosen by an engineer). Plus, editing tracks with Audacity renders them no longer accurate in the AccurateRip database. Maybe the difficulty of this is why there are so many annoying track boundaries on pressed discs.

I find that when I listen to the same tracks in gapless albums on streaming platforms/YouTube&*8217;s official audio uploads (random people&*8217;s uploads are usually not reliable because anyone can edit anything to their liking), the pop/click sounds are not an issue. I should assume this is because these players have ways to hide those sounds? Or these steaming platforms/websites are provided with different masters than what I have on CD?

Also, I was told by someone credible on hydrogenaudio that most CDs do not begin until around 10,000 samples have passed. That sounds only somewhat believable but who am I to say if it is or isn&*8217;t? I am not an audio expert, just someone trying to figure out a frustrating problem. I would love that to be true as that would truly solidify that hearing 24 samples of offset is not at all possible.

Thank you for any further clarifications!

[Spoon]

Re: Variations in mastering? CD ripping a flawed process?

> is it really true that almost every gapless album has annoying pops/clicks at particular track starting points?

Yes

>I would think that the industry standard would be to find an exact point in a track to prevent a pop sound when shuffling/selecting tracks from gapless albums

Most players will automatically fade in, so it is not an issue.

You cannot hear 24 samples, as there are 44,100 samples per second, so 24 is 0.00054 of a second

[joeyc410]

Re: Variations in mastering? CD ripping a flawed process?

> is it really true that almost every gapless album has annoying pops/clicks at particular track starting points?

Yes

>I would think that the industry standard would be to find an exact point in a track to prevent a pop sound when shuffling/selecting tracks from gapless albums

Most players will automatically fade in, so it is not an issue.

You cannot hear 24 samples, as there are 44,100 samples per second, so 24 is 0.00054 of a second

Thanks for the response. If I were to purchase a track from the Apple iTunes Store, listen to it and not hear a click at its beginning or end, would you say that would be the result of the source material being different (meaning the digital download is from a different source than what is used for the pressed CD)?

I have experimented with different pressings of different CDs and had different results (clicks on one pressing, none or at least different track positioning on another), so that leads me to believe that the process of pressing the digital audio to the physical CD introduces these complications?

[Spoon]

Re: Variations in mastering? CD ripping a flawed process?

If it was a gapless album then you would expect there to be no fade in, in the audio.

[joeyc410]

Re: Variations in mastering? CD ripping a flawed process?

If it was a gapless album then you would expect there to be no fade in, in the audio.

I think that explains why you do not often see gapless albums available for digital download.

[schmidj]

Re: Variations in mastering? CD ripping a flawed process?

The issues would appear to be in the digital players. I've never experimented with recording the output when you push "play". If you don't want some form of clicks, the player would need to fade in the audio of a ripped CD track when you play it. While audio engineers often do their digital edits at "zero crossings" of the waveform, which reduces, but by no means eliminates Fourier clicks, the closest point you can define for a track starting on a CD is one frame, 1/75th of a second, so the recording engineer really has no ability to control the track beginning to sample accuracy.

If your CD is individual tracks, you will have learned that most, if not all CD players mute the audio when paused or stopped, and fade it in, quite slowly, when you begin playing it. Therefore the beginning of a track, when your CD is of discrete tracks, had better be either silence or audio that will not appear upcut as the CD player begins playing it (such as applause or a fade-in). When I have mastered my location CDs, I have always left about .375 second of silence or such before the beginning of the track. On the occasions where I have accidentally left less, I've gotten complaints that some CD players upcut the track. Ripping any such track will not give you a Fourier click, but of course it is not gapless.

If you, after ripping, trim the silence off the beginning of the track (and if necessary off the end of the preceding track), to make the segue gapless, and don't crossfade the tracks, you may well get a Fourier click if you play the two tracks consecutively. You can avoid this possible click by in fact doing a brief crossfade between the tracks, instead of a hard edit. In fact most professional audio editing programs allow a default crossfade between clips edited together.

Now we have the concert recording. If played from the CD end to end, no clicks in the middle, the CD was recorded straight through, through the "pause" time and through the time before and after the track start "marker". If we play, on a CD player, a track from the middle, it will not have a click, because the track has been faded in by the player when I hit "play". But, whether I get a click when I hit "play" playing digital files depends on the design of the player. If the player is designed to begin playing with a fade, then you won't hear a click.

But now the concert recording played as digital files end to end. We have an issue. The CD was one giant "file", a continuous track. But the rip (or download) of it is individual files, one per "track". And when we put them together we have a dilemma. If we but them all together with no fades, they may or may not play through clean, depending on the design of the codec and the implementation of the player. We can give the appearance of a clean, clickless, gapless play by crossfading them, but when we do that, we are actually shortening the playback of the overall track by the length of the crossfade. When I edit audio, I'm careful to put that crossfade where it is inaudible, during applause or a momentary pause in the performance, or even between notes, and I listen to the result and move or change the edit until it in fact is inaudible. But your digital player doesn't have that option. It just either buts or crossfades the tracks. If the crossfade is in the middle of a note, you may well hear a glitch. If they but the tracks with no crossfade and you don't play them consecutively or the codec/player can't internally make a clean splice, you may hear a Fourier click a pop because the codec can't begin or end encoding cleanly, or a dropout if it is designed to cover those issues by fading in and out.

Interestingly, my audio editing software has a better solution that doesn't exist (AFAIK) in the consumer world. I can rip the CD "as one" from end to end as recorded, and then (and sometimes automatically) place "markers" into the file creating "regions" which are the same as the old tracks on the CD. Now if I want to play the whole concert, I just play the whole file. If I want one track, I play the region, If I want two or three consecutive tracks, I just play the consecutive regions. No pops, clicks, dropouts between track/regions. Unfortunately, the coding of the region markers is, AFAIK, not well standardized, it is proprietary to the editing software, although the different software companies seem to have some commonality, many of them do recognize the markers of competing software. But I've never seen consumer player software that recognizes these markers or offers any way to play regions from a CD length file. If that came about, it would provide a solution to your issue.

[joeyc410]

Re: Variations in mastering? CD ripping a flawed process?

The issues would appear to be in the digital players. I've never experimented with recording the output when you push "play". If you don't want some form of clicks, the player would need to fade in the audio of a ripped CD track when you play it. While audio engineers often do their digital edits at "zero crossings" of the waveform, which reduces, but by no means eliminates Fourier clicks, the closest point you can define for a track starting on a CD is one frame, 1/75th of a second, so the recording engineer really has no ability to control the track beginning to sample accuracy.

If your CD is individual tracks, you will have learned that most, if not all CD players mute the audio when paused or stopped, and fade it in, quite slowly, when you begin playing it. Therefore the beginning of a track, when your CD is of discrete tracks, had better be either silence or audio that will not appear upcut as the CD player begins playing it (such as applause or a fade-in). When I have mastered my location CDs, I have always left about .375 second of silence or such before the beginning of the track. On the occasions where I have accidentally left less, I've gotten complaints that some CD players upcut the track. Ripping any such track will not give you a Fourier click, but of course it is not gapless.

If you, after ripping, trim the silence off the beginning of the track (and if necessary off the end of the preceding track), to make the segue gapless, and don't crossfade the tracks, you may well get a Fourier click if you play the two tracks consecutively. You can avoid this possible click by in fact doing a brief crossfade between the tracks, instead of a hard edit. In fact most professional audio editing programs allow a default crossfade between clips edited together.

Now we have the concert recording. If played from the CD end to end, no clicks in the middle, the CD was recorded straight through, through the "pause" time and through the time before and after the track start "marker". If we play, on a CD player, a track from the middle, it will not have a click, because the track has been faded in by the player when I hit "play". But, whether I get a click when I hit "play" playing digital files depends on the design of the player. If the player is designed to begin playing with a fade, then you won't hear a click.

But now the concert recording played as digital files end to end. We have an issue. The CD was one giant "file", a continuous track. But the rip (or download) of it is individual files, one per "track". And when we put them together we have a dilemma. If we but them all together with no fades, they may or may not play through clean, depending on the design of the codec and the implementation of the player. We can give the appearance of a clean, clickless, gapless play by crossfading them, but when we do that, we are actually shortening the playback of the overall track by the length of the crossfade. When I edit audio, I'm careful to put that crossfade where it is inaudible, during applause or a momentary pause in the performance, or even between notes, and I listen to the result and move or change the edit until it in fact is inaudible. But your digital player doesn't have that option. It just either buts or crossfades the tracks. If the crossfade is in the middle of a note, you may well hear a glitch. If they but the tracks with no crossfade and you don't play them consecutively or the codec/player can't internally make a clean splice, you may hear a Fourier click a pop because the codec can't begin or end encoding cleanly, or a dropout if it is designed to cover those issues by fading in and out.

Interestingly, my audio editing software has a better solution that doesn't exist (AFAIK) in the consumer world. I can rip the CD "as one" from end to end as recorded, and then (and sometimes automatically) place "markers" into the file creating "regions" which are the same as the old tracks on the CD. Now if I want to play the whole concert, I just play the whole file. If I want one track, I play the region, If I want two or three consecutive tracks, I just play the consecutive regions. No pops, clicks, dropouts between track/regions. Unfortunately, the coding of the region markers is, AFAIK, not well standardized, it is proprietary to the editing software, although the different software companies seem to have some commonality, many of them do recognize the markers of competing software. But I've never seen consumer player software that recognizes these markers or offers any way to play regions from a CD length file. If that came about, it would provide a solution to your issue.

Very interesting to read. A key piece of information that I've learned has to do with the one frame limitation you mentioned. I wonder who or what caused that to be the case.

[Maxpower]

Re: Variations in mastering? CD ripping a flawed process?

Just to add on to this: more commonly, the clicks are a result of the way the ripped track is padded (beginning or end). Some older or less compliant renders incorrectly read this as audio data and you get a second or so of noise. To the listener, it is not always obvious if problems lies with the end of the previous track or the beginning of the next track. Both can be a problem. Experimenting with your ripping software can yield good results.

One of the easiest examples to replicate this padding issue is when ripping SACDs to flac / wav. Most of the open source rippers have now added an ability to customise the padding parameters. In my experience, fading does not help or only masks the problem.

[mville]

Re: Variations in mastering? CD ripping a flawed process?

Just to add on to this: more commonly, the clicks are a result of the way the ripped track is padded (beginning or end)...

Padding isn't an issue, but software that misinterprets a file format is an issue.


schmidj is correct, the problem lies with certain player software. If you playback digital audio, crudely, at a non-zero crossing, you will most likely hear a click.


Any digital audio playback software worth it's salt, will attempt to circumvent this non-zero crossing playback issue, so that the click is not perceived by the user.

[S80_UK]

Re: Variations in mastering? CD ripping a flawed process?

Very interesting to read. A key piece of information that I've learned has to do with the one frame limitation you mentioned. I wonder who or what caused that to be the case.

The framing is a standard function of the "Red Book" audio CD format. Audio tracks occupy an integral number of frames, their start and end being defined in the Table of Contents (often just TOC) at the start of the CD. Unless the audio is somehow recorded such that there is a zero crossing on frame boundaries (almost impossible in the real world if it is a gapless recording) then starting playback of a gapless recording from any frame other than the first may result in a spontaneous signal transition from the previous no-signal level to a different arbitrary level - hence the click or pop.

Some further reading here... https://www.travsonic.com/red-book-cd-format/

[Maxpower]

Re: Variations in mastering? CD ripping a flawed process?

Padding isn't an issue, but software that misinterprets a file format is an issue.


schmidj is correct, the problem lies with certain player software. If you playback digital audio, crudely, at a non-zero crossing, you will most likely hear a click.


Any digital audio playback software worth it's salt, will attempt to circumvent this non-zero crossing playback issue, so that the click is not perceived by the user.

While I do not agree with your statement that padding is not an issue . . . maybe it would be helpful to broaden the discussion and think about definitions here. 'Clicks' might be described as very short, low energy noise. The other common artifact that many people report is a more substantial noise (higher energy, longer duration).

I suspect these artifacts get reported and commented on interchangeably but they are indeed very different things with (probably) very different root causes. An interesting topic.