If you had to choose...

[sandyk]

Deleted.

[One and a half]

Back on topic : DSD, pref .dsf for the metadata.

[Paul R]

Deleted

[Paul R]

My master library is currently in AIFF format, since ot s perfectly portable between Apple, Microsoft, and Linux. I don't buy into the anti-Apple crowds' reasons to use FLAC.

 

But f I had to choose today, I would convert /RIP everything to DSD256 in a dsf container file.

 

Paul

[firedog]

uncompressed flac

[Teresa]

Back on topic : DSD, pref .dsf for the metadata.

 

1+

My choice also.

[Maldur]

There is no contest - AIFF for PCM recordings (176,4kHz, not 192 kHz!) and rips, over an half year now I convert most of my doings to DSD for archival purposes.

[Boris75]

...only one format to rip all your cd library, which would it be?

 

ALAC because I want to use iTunes.

[jhwalker]

I'm a bit concerned about all the "DSD / DSF" responses. Doesn't seem to make much sense to me to rip / archive PCM material in a DSD container - that's a lossy process, isn't it (i.e., if you rip PCM content directly to DSD, inevitably, original format / content is irretrievably lost). I would think we would rip Redbook (at least) to 16/44 PCM (e.g., FLAC / ALAC, WAV / AIFF) to get a bit perfect archival copy, then if you want to convert to DSD later (either real-time or offline) . . .

[wgscott]

There is no loss of information, just the possibility of additional low level wideband noise/EMI due to all the additional processing involved due to imperfect computers. If this is all so cut and dried, why are so many members paying a great deal more for aftermarket Linear PSUs, both internal and external ? It would appear that you don't believe there is such a thing as Software induced "Jitter" either.

 

This is your crazy opinion that has now become a legendary internet example of how nutty Computer Audiophiles can be. Please stop claiming it is substantiated fact, over and over again.

[mwheelerk]

I'm a bit concerned about all the "DSD / DSF" responses. Doesn't seem to make much sense to me to rip / archive PCM material in a DSD container - that's a lossy process, isn't it (i.e., if you rip PCM content directly to DSD, inevitably, original format / content is irretrievably lost). I would think we would rip Redbook (at least) to 16/44 PCM (e.g., FLAC / ALAC, WAV / AIFF) to get a bit perfect archival copy, then if you want to convert to DSD later (either real-time or offline) . . .

I would agree that it would seem to make more sense to have your archival copy in its native format. I may be missing something in understanding why this is done. If SACDs are being ripped that would be one thing but I fail to understand the benefit of doing this to redbook.

[mwheelerk]

There is no loss of information, just the possibility of additional low level wideband noise/EMI due to all the additional processing involved due to imperfect computers. If this is all so cut and dried, why are so many members paying a great deal more for aftermarket Linear PSUs, both internal and external ? It would appear that you don't believe there is such a thing as Software induced "Jitter" either.

I'm sorry, I fail to connect a dotted line from PSUs to something conceived as software induced jitter. What software? When is this induced?

[Jud]

I'm sorry, I fail to connect a dotted line from PSUs to something conceived as software induced jitter. What software? When is this induced?

 

There actually is a connection between power supply behavior and jitter induced at the point of conversion. There is also a real phenomenon, written about in journals going back to the dawn of digital audio, regarding "software induced jitter." However, as far as I know (and I confess to being far from an expert here), jitter induced in connection with power supply behavior and "software induced jitter" aren't one and the same. To explain both here would just derail the thread further. If you don't find satisfactory explanations from a Google search, you're welcome to start another thread where I would be happy to relate what I've gathered (and hoping that someone who actually knows something would then jump in with corrections/further explanation).

[Paul R]

I'm a bit concerned about all the "DSD / DSF" responses. Doesn't seem to make much sense to me to rip / archive PCM material in a DSD container - that's a lossy process, isn't it (i.e., if you rip PCM content directly to DSD, inevitably, original format / content is irretrievably lost). I would think we would rip Redbook (at least) to 16/44 PCM (e.g., FLAC / ALAC, WAV / AIFF) to get a bit perfect archival copy, then if you want to convert to DSD later (either real-time or offline) . . .

 

I do not think that PCM to DSD is lossy in any way, or rather, any more lossy than any PCM file already is, but I do see your point. On the other hand, when I want to play something that is non-DSD, it is easily transcoded to any PCM format, and likely will be played from a portable device. Not sure it makes all that much difference.

 

-Paul

[Jud]

I'm a bit concerned about all the "DSD / DSF" responses. Doesn't seem to make much sense to me to rip / archive PCM material in a DSD container - that's a lossy process, isn't it (i.e., if you rip PCM content directly to DSD, inevitably, original format / content is irretrievably lost). I would think we would rip Redbook (at least) to 16/44 PCM (e.g., FLAC / ALAC, WAV / AIFF) to get a bit perfect archival copy, then if you want to convert to DSD later (either real-time or offline) . . .

 

I was thinking folks were mostly (though not entirely - I see Paul's comment upthread a little way) talking about DSD files in DSF and some other format for PCM files, but perhaps I wasn't paying close enough attention.

[christopher3393]

FLAC. I see nothing but pcs in my future.

[Paul R]

I was thinking folks were mostly (though not entirely - I see Paul's comment upthread a little way) talking about DSD files in DSF and some other format for PCM files, but perhaps I wasn't paying close enough attention.

 

I was suggesting that if I were to re-RIP or convert my master library today, my choice would be to convert everything into DSD256 in a dsf file container. A dsf file so that the metatdata would come along.

 

Something like Audiogate can do that with amazing results.

 

-Paul

[mav52]

3 TB's of FLAC

[Skeptic]

I use Apple lossless because I have some Apple hardware (iOS/AirPlay devices) and they don't support FLAC.

If they did, I would probably use FLAC instead.

 

But there's no real difference between FLAC/ALAC anyway. They're both open source formats, both compress to roughly the same size and are both very efficient to encode/decode.

Which format you choose really only depends on what is most compatible with your devices.

 

If you're only playing local files on an internal drive, I suppose you could make the argument that "disk space is cheap" and use an uncompressed format, but network bandwidth is often limited, if you are streaming wirelessly, or outside of your home network.

 

 

And compressed files have integrity checks on the data they contain.

I don't know about other players, but by default JRiver won't play a damaged FLAC or ALAC file.

With uncompressed formats, there are no integrity checks so you don't know whether the files are damaged or not.

 

With compressed formats, a single flipped bit can produce disastrous results, which is why these checks are in place.

This is easier to demonstrate with images than audio:

 

Bitrot and atomic COWs: Inside

Let's look at a graphic demonstration. Here's a picture of my son Finn that I like to call "Genesis of a Supervillain." I like this picture a lot, and I'd hate to lose it, which is why I store it on a next-gen filesystem with redundancy. But what if I didn't do that?

 

 

As a test, I set up a virtual machine with six drives. One has the operating system on it, two are configured as a simple btrfs-raid1 mirror, and the remaining three are set up as a conventional raid5. I saved Finn's picture on both the btrfs-raid1 mirror and the conventional raid5 array, and then I took the whole system offline and flipped a single bit—yes, just a single bit from 0 to 1—in the JPG file saved on each array. Here's the result:

 

 

So while I suppose you could argue that the compressed formats are less robust, every time you play them the file integrity is verified.

With uncompressed formats, your files are susceptible to bit rot.

 

 

There is no loss of information, just the possibility of additional low level wideband noise/EMI due to all the additional processing involved due to imperfect computers. If this is all so cut and dried, why are so many members paying a great deal more for aftermarket Linear PSUs, both internal and external ? It would appear that you don't believe there is such a thing as Software induced "Jitter" either.

Use a converter which performs verification on the converted files before deleting the originals.

 

If the checksum is the same, the audio stored in the file cannot be different.

No, it cannot contain jitter either - it would fail the checksum.

 

This is why you want to use compressed formats rather than uncompressed formats for storage.

 

 

I do not think that PCM to DSD is lossy in any way, or rather, any more lossy than any PCM file already is, but I do see your point. On the other hand, when I want to play something that is non-DSD, it is easily transcoded to any PCM format, and likely will be played from a portable device. Not sure it makes all that much difference.

PCM to DSD is absolutely a lossy conversion.

There are a lot of compromises made when converting to a 1-bit format.

 

Whether your DAC performs better when you send it a DSD signal rather than a PCM one is another matter though, and is something which should happen on playback.

 

You always want to store things in their original format.

The only real exception is if you're processing tracks to extract HDCD data, or deal with pre-emphasis. Then you should convert from 16-bit to 24-bit.

 

I was thinking folks were mostly (though not entirely - I see Paul's comment upthread a little way) talking about DSD files in DSF and some other format for PCM files, but perhaps I wasn't paying close enough attention.

If you are ripping SACDs, I would rip directly to an ISO file and play that.

There are some problems which can arise when you split them into individual tracks.

[wwaldmanfan]

I use several different software programs to rip, transcode, edit, tag, and organize my music library, namely XLD, MediaHuman Audio Converter, Audacity, Audiofile Engineering Sample Manager, and iTunes. AIFF is the only uncompressed universal file format supported by all of them.

[Paul R]

PCM to DSD is absolutely a lossy conversion.

There are a lot of compromises made when converting to a 1-bit format.

 

I see people saying this, but I never see any details on why they think this is true. No data is lost in a PCM to DSD transcode so far as I can tell. The DSD format is easily able to accommodate all the data in the PCM file.

 

Looking at some algorithms to do the PCM -> DSD conversion, I don't see any obvious place where any kind of "lossey" conversion takes place. Perhaps Damien or Jussi will step in here and correct me if I am wrong. But until someone can show me, I will contend that PCM -> DSD is lossless.

 

Converting from DSD -> PCM requires decimation, and that is indeed not lossless. However, PCM -> DSD is horse of an entirely different color.

 

 

-Paul

[Jud]

I see people saying this, but I never see any details on why they think this is true. No data is lost in a PCM to DSD transcode so far as I can tell. The DSD format is easily able to accommodate all the data in the PCM file.

 

Looking at some algorithms to do the PCM -> DSD conversion, I don't see any obvious place where any kind of "lossey" conversion takes place. Perhaps Damien or Jussi will step in here and correct me if I am wrong. But until someone can show me, I will contend that PCM -> DSD is lossless.

 

Converting from DSD -> PCM requires decimation, and that is indeed not lossless. However, PCM -> DSD is horse of an entirely different color.

 

 

-Paul

 

Paul, just converting sample rates in PCM requires filtering and is thus imperfect, so I can't think why conversion from PCM to DSD should be immune.

[Paul R]

Paul, just converting sample rates in PCM requires filtering and is thus imperfect, so I can't think why conversion from PCM to DSD should be immune.

 

Admittedly, dithering might be a culprit, but on the other hand, upsampling does not seem to loose any data. If anything, it adds additional data.

 

I just converted a small 16/44.1 file to 16/48.0, and was able to find and identify identical samples in the resampled file for each sample in the in the original file file. (AIFF).

 

I would agree the conversion might not be perfect, but I do not think it is "lossey."

 

Yours,

-Paul

[Cycleman]

However I cannot hear a difference between ALAC and AIFF and if I was starting out I would probably recommend ALAC.

 

I have several albums, particularly some well-recorded ones where I like/love the music, in both ALAC and AIFF formats -- just to test, to see if I can tell the difference.

 

And I, too, cannot hear a difference -- and this over time, over the past year or so.

 

So, sure ALAC for Mac computers, especially if you have enough power and memory, so that your music player unpacks the compressed file before playing it.

[Skeptic]

Admittedly, dithering might be a culprit, but on the other hand, upsampling does not seem to loose any data. If anything, it adds additional data.

The data is interpolated. What might be considered a good interpolation algorithm today could be surpassed by something much better a few years from now.

 

This is why you should store the data in its original format, and convert on playback.

Ideally you wouldn't be converting files to extract HDCD data or deal with pre-emphasis either, but I don't think any players do that in realtime, and going from 16-bit PCM to 24-bit PCM is not a lossy process as such.

 

Just imagine if you had converted your entire library to 24/96 a few years back, or 1xDSD, and now wanted to listen to them as 2xDSD.

Would you go through the whole ripping and conversion process again?

 

What if the popular format in a year or two is DSD-Wide or high sample-rate PCM instead?

I would agree the conversion might not be perfect, but I do not think it is "lossey."

It is not "lossy" in the sense that you are not discarding lots of data as you would when converting from an uncompressed or losslessly compressed file to a lossy format like MP3.

But if it is not a perfect conversion (it isn't) then you have lost something from the original data.

 

For one thing, you're adding a ton of ultrasonic noise to the files that did not exist in the source, which now has to be filtered out on playback.