Which file format do you use for digital music storage?

Many thanks, Sevoflurane. I'll try that.

 

I use FLAC, but for the super great ones, AIFF. I believe it was Barry Diament who said he heard AIFF better, and it does indeed sound a bit smoother.
For DSD discs, I use DFF.

Btw, my FLAC's of The Blue Nile's Hats UK cd (NOT the US A&M) sound infinitely better streaming from a dedicated external hard drive to the Oppo 105 than the disc itself (except for some slight midrange issues).
HUGE, wide, detailed and deep. I couldn't believe the difference.

 

Are you using XLD? The best ripper for Mac I've found. And it's free.

 

The original CD release of The Blue Nile "Hats" has pre-emphasis. The later release does not. If you've got the original release then you are going to hear an obvious difference between the CD and the ripped files. Unless you've applied de-emphasis digitally to the ripped files. The Oppo 105 should detect if the CD has pre-emphasis and apply the de-emphasis filter when you play that CD. But if you rip the CD to FLAC then that de-emphasis will not automatically happen.

 

Mixed.

CDs are ripped to ALAC in this Mac household.

SACDs are ripped to DSF (after a steeeeep learning curve).

Downloaded FLAC purchases stay that way.

JRiver plays 'em all.

 

That's interesting. But pre-emph or not (and I've heard plenty of cds and rips with both) the DAC in the Oppo 105, running from a dedicated hard drive, has enormous sound. If you have one, try it. Pre-emph may be an issue, but the huge soundstage is NOT.

 

I resemble that post!

Only difference is that I use 320MP3 instead of FLAC. As I've posted before, I cannot tell the difference between a 320MP3 rip and the CD it was ripped from.

Jeff

 

Find out if you have the CD version with pre-emphasis. If your CD has pre-emphasis then comparisons between the CD playing in a CD player and the ripped files is going to be complicated. Lots of issues to deal with that can affect how you perceive soundstage.

One issue is with the digital filter used in the CD player to do de-emphasis. De-emphasis can be done with an analog circuit. But in modern CD players it is going to be a digital filter to do de-emphasis. Different ways of doing digital filters can make a difference for soundstage. Whether the filters are linear phase, minimum phase, or whatever will make a difference. I've done comparisons of different digital filters to do de-emphasis and have heard a difference in soundstage width and depth with different de-emphasis filters. It makes a difference. Don't assume that all CD players do de-emphasis with equal sound quality.

I don't have an Oppo. Can't do a listening comparison test of its de-emphasis.

For software de-emphasis of ripped files I've found SoX to do the best de-emphasis to my ears with my gear and my DACs. With different gear and a different DAC I could find a different filter to work better. It's not a simple comparison where one solution will end up best for everyone and everything.

If you're hearing an obvious difference between the CD played in a CD player and playing the rip through the same DAC then I suspect that there are some pre-emphasis and de-emphasis issues going on that are causing that.

 

Good. I have no use for MP3.

 

Flac -v from the command line or install Foobar2000 and drag and drop the files in there or setup the library function and right click and pick verify.

 

Many of my digital files are on CD .

When I rip, it's MP3. This was originally due to compatibility, as I was doing sound system testing at the time for Mercedes and BMW etc etc, so I had to have the most compatible format. Later this persisted because I got into doing iPod/iPhone adaptor design, so had to test with MP3 as well as iTunes DRM and ALAC.

I still rip within iTunes because it is SO easy, error correction on, 224k VBR to highest quality. This gives really musical sounding recordings, which I could stick on my little Clip Zips for listening while I run (they have more capacity now but back then space was a significant issue). [EDIT: I swear this used to say LAME encoder, but just noticed now it does not]

Somewhat-in-depth testing made me feel 192k sounded pretty good but not quite lossless quality. Nevertheless the sound could be ENJOYABLE which is the main point. I stepped up to 224k VBR for more quality but not-huge files...I know some rip to 320k MP3 but at that point you're halfway to lossless so I just didn't do that.

I had originally also ripped a lot of stuff to ALAC, but since I still have the CDs that seemed silly, as I'm not set up for streaming off a network drive. Instead of that, more likely I'll upload to iTunes music cloud match which I'm forgetting at the moment what they call that...except I think that transcodes to 256k AAC? In which case I should just rip from lossless to that...but too much re-ripping...actually I guess it's just for certain items they are missing like Maiden Japan or Whitesnake's Live in '84 - Back To The Bone and so on, so it's not that big of a project.

*I'm thinking to make a CD with pairs of the same track alternating between the full WAV and MP# ripped as above expanded back to WAV. Then give to a friend for a totally blind test, to see what he comments. Then maybe ages later if I've forgotten the track order I can listen as well (gotta put the decoding in a list of course). Huh, if I had more energy I could mail them out to people, or post up the tracks someplace...

 

+1, love XLD, I send the developer a bit of money periodically.

 

Yeah, I doubt I could under truly blind conditions. Maybe on a stupendous system with stupendous material...but not on my Sansa Clip Zip with earbuds while running around the Rose Bowl, listening to old bootlegs
...and for some recordings, LESS fidelity could be a mercy...

 

=AFLAC?

 

I have never done more than an internet quiz not tell the difference between 256 kbps

 

Mainly high bit rate mp3. But I have poor high frequency hearing, so I can't hear a difference between mp3s and lossless formats. I also can't tell pre-emphasis in many cases.

 

Beat me to it. Early version of Hats had pre-emphasis, there is also a UK version without it. Can't remember which way round but one has a blue label and one pink.

 

I have never done an in-home comparison of mp3 vs lossless, I have taken online "can you tell the difference" tests. I thought I could hear some occasional variation at 256 kbps, I could not differentiate 320 kbps from lossless...
I don't know how accurate an internet sound test is.

 

This should have been a multiple choice poll. Wav --> Flac for the archive, Wav --> Mp3 for portable music listening. I don't use any other format/encoding scheme.

 

I keep my needledrop "master format" in WAV 192/32float, compressed in the lossless VW format (wavpack.exe) with a compression rate of ~58%. This is the highest quality audio format (even better than DSD 5.6 MHz 1bit, according to the research papers of some audio formats developers). I don't use FLAC because actually it can't compress the PCM 32-bit floating point, only 16-bit or 24-bit.

The best format I use for desktop audition is WAV 48/32float (the optimal balance between size/quality). The size of this format is only 25% of the needledrop "master format".

For smartphone audition I resample to standard CD WAV 44/16 (11% size of the needledrop "master format") or even mp3 (3% size of the needledrop "master format"). mp3/lame/48KHz/320kbps/VBR/V0 with proper settings: filters disabled, developer settings enabled, ATH 5, minimal psychoacoustic. In the ABX tests you can't hear any differences between WAV 44/16 and this lame MP3 format.

 

AIFF for both CD rips and needle-drops since I still use iTunes in combination with Roon on a Mac within a whole-house Apple ecosystem.

 

Whatever iTunes creates when I load .wav files into the iTunes music library.

 

I use XLD. The problem is iTunes doesn’t play flac nor does my iPhone and iPod.

I’ve done tests and noticed the Oppo handles FLAC files better than ALAC. But for me to maintain a FLAC set and a ALAC set would be too much work.

 

Ain't no big deal. Remember that 32-bit is only 24-bit fixed point with an eight-bit mantissa. You do not lose any sound quality between 32-bit > 24-bit.

Maybe you can't, but many people, like myself, can.

 

Ok, let's test it! I am very curious.

Here is the ABX package and samples. You need java runtime installed.

abchr-java.zip - MegaUpload