Fake Flac? I.e fake lossless.

[Retronic]

I can re-rip BRI 16 if needs be.

[Uvuzau]

Greta & Retronic , thank you a lot for your kind offers, if you catch some time i would really like to listen BRI 16 in better quality or better rip ..

[Andrew]

Hi,

BRI 16 is a fairly quiet record compared to others.  Here's a picture of the waveform to show how quiet it is on the record.
(the darker brown lines at the top and bottom of each are 0db, which is the loudest)

[Porn Library]

Going back to the "flac is not 100% lossless"....Porn Library....you killed me. I simply DON'T WANT to believe it

It all depends on how the flac is made and with what....that's what I was trying to get at when I said it's not 100% lossless. Like it's no guarantee it's gonna come out lossless depending on software, etc. What was in my head and I was trying to get out in writing didn't come across right LOL

Basically, all I was trying to say in a nutshell without the tech explanations is...flac is a pain in the f**king ass, i'm not a fan of it, never will be and I will ALWAYS use the original uncompressed lossless digital format called WAV to avoid any potential screw ups  Plus it's easy for a middle aged audiophile grump like me to burn CD's with the original WAV 

[Greta]

Yes PL, I got all that.
I just didn't want to believe you when you said it's basically like a 320 mp3!
I agree with uncompressed wav, but with that kind of media a new problem take place: space. Thousands of records would fill expensive hhd very quickly, which, I guess, is the same problem for many people, one step below, with mp3 vs flac.

[Greta]

Now, supposing it's a matter of compression software, is there any suggestion on which one it's more reliable?
I'm on Mac, and I mostly use XLD, but also fre:ac and xACT. XLD and fre:ac use the FLAC 1.3.3, which is the latest version of the codec (dated august 2019), xACT uses the previous version 1.3.2.
I'm guessing if it's only a matter of codec or also a matter of software itself and how the software uses the codec.
Then, is the matter the same with other lossless codecs like ALAC or APE ?

[Andrew]

I stored all my original files in APE format.  I picked APE because it had the highest compression.
I'm just now converting them to FLAC.  They are all lossless formats.

Lossless formats only store what you give them.  If you give them an MP3, then they will sound like an MP3.
Similarly, you can create a bad WAV file by converting an MP3 file to a WAV file.

There also appears to be some people out there who think that if they convert an MP3 to a FLAC it will somehow improve the sound.
It doesn't work that way either lol

https://www.reddit.com/r/audiophile/comments/9kjwpb/is_flac_really_lossless/

[niknak]

If I keep my recording chain lossless for instance..

Record/CD/Dat/R2R -> Deck with Analogue or Digital Output -> Computer with Software Set to capture in PCM/Wav/AIFF/DSD/Other non lossy -> Converted to Flac in *Any* software that offers that as a format...

It will be lossless. Has to be.

I say *any* software that offers Flac because they all pass the process off to Flac.exe/dll (or at least the very exact same conversion Math) in one way or another - they have too or it couldn't be a classed as Flac and could lead to compatibility issues...

But If I stick crap in I'll get crap out. lossy in - lossy frozen in time in a lossless shell out - but still lossy.

A little side note about Opus codec, Opus is a great codec, technically incredible - it's lossy but has full 20-20khz stereo bandwidth at 128kbps but the thing with it is if you convert a 128kbps opus file to wav or flac it 'can' fool checking software into thinking its true lossless. This is why spectrograms are useful tools as they tell you so much about the waveform, Opus has a tell tale 20khz Hard cut off, high kbps Mp3 shelves at 16khz and 18khz
 

Some official lossless releases are trash, somewhere around the early 2000s when digital remastering and digital brick walling became a thing - sometimes they were from lossy archive masters or the remastering process had a lossy link in the chain.. Classical CD's were pretty bad and anything that was remastered from the 50s etc..

Now as for recording volume, as long as you aren't clipping the signal, record! Once its done find the normalize function and run that over the file- what that does is find the loudest peak and set that as 0db reference (or what ever you pick as reference -0.2db is ideal) and raise the rest of the track up to match. You'll get the same result as what you would if you had recorded it louder, It won't introduce crap because we are in the digital realm here - it wont add anything that isn't already there. But use Normalize, not set gain or anything like that Normalise retains the integrity of the Volume throughout where as othere methods can cut off anything over 0db killing transients and effectively brick walling the file..   

If your recording chain is 32bit (unlikely right now but maybe one day soon) you can even record totally clipped and normalise back down to 0db and not lose any transients - it's amazing to see!

Little observation - There seems to be a mistrust of noise reduction - it's a shame because this especially has been solved from a technical standpoint for years now. Every movie/Tv show and every individual sound in those, every song in recent years, every voice recording have all been treated and you'll have not known. Its really that good now. Its just the operator using the tools that can go overboard or not understand the operation correctly that can mess it up.

Sorry for the ramble - too much coffee today

[Andrew]

Thank you very much for the information niknak!

[kimhill]

FLAC is lossless.

This can be checked by starting with an uncompressed file, converting to FLAC, and then converting back to uncompressed. The before & after uncompressed data can then be compared in software.

This is not about waveforms, it's about 1's and 0's — and computers are quite adept at checking those. If anyone can demonstrate that there's actually a problem, there will be a storm on the net like few have seen.

No storm? No problem.

[Snowdog]

I'm on a Mac & I rip all my CDs using dBpoweramp to AIFF format. iTunes won't play FLAC files so I use Foobar & VLC player for that.

I'm unclear about the WAV argument, though. You can take any lossy file & convert it to WAV, FLAC or AIFF & it will still be a lossy source so the fact that it's a WAV file is no guarantee of it being lossless any more than any other format, surely? Unless I've misunderstood.

[niknak]

That's correct, it would be impossible to turn a lossy file into a true lossless file because the data isn't there anymore

[Snowdog]

That's correct, it would be impossible to turn a lossy file into a true lossless file because the data isn't there anymore

Exactly, but the point I was making (& I probably should've quoted to make it clearer) was in reference to Porn Library's apparent assertion that a WAV was the only safe true lossless format, which isn't necessarily the case, any more than an AIFF or whatever.

Unless I misunderstood & they were saying that converting to FLAC could cause data loss & that the only true way of preventing that was to keep the file as a lossless WAV, in which case the same is true of AIFF, etc.

I feel like a right pedant now, though. 

[niknak]

That's correct, it would be impossible to turn a lossy file into a true lossless file because the data isn't there anymore

Exactly, but the point I was making (& I probably should've quoted to make it clearer) was in reference to Porn Library's apparent assertion that a WAV was the only safe true lossless format, which isn't necessarily the case, any more than an AIFF or whatever.

Unless I misunderstood & they were saying that converting to FLAC could cause data loss & that the only true way of preventing that was to keep the file as a lossless WAV, in which case the same is true of AIFF, etc.

I feel like a right pedant now, though. 

I skim read your post so I didn't get the gist, but yeah you are right, you cannot assume because something is a Wav/Flac/AIFF/etc that it is in fact lossless unless you know the recording chain or can view the spectrogram or have CRC's from a logchecker or trust the uploader etc..

A couple of years ago here there was some Italian (I forget the label) lossless releases pulled directly from a vendors site that turned out to be (pretty bad) upscales, The label themselves even posted here adamant at first that they were true lossless and had been remastered etc.. but it turned out that at some point in their history the tracks had been archived into a lossy format and those were used..

[stackjackson]

A couple of years ago here there was some Italian (I forget the label) lossless releases pulled directly from a vendors site that turned out to be (pretty bad) upscales, The label themselves even posted here adamant at first that they were true lossless and had been remastered etc.. but it turned out that at some point in their history the tracks had been archived into a lossy format and those were used..

I think that would be this old thread => https://librarymusicthemes.net/index.php?topic=1142.0