Quality settings for lossless format?!

[sprnva]

Not true, daddy-o. VLC for example can play audio that is contained within zip files.

Foobar too. But it does this by unzipping the file to a temporary folder and playing the files from there. Not the same as playing a FLAC file where there's no temporary file created.

Anyway, I know all about containers and formats and whatnot. I'm just trying to keep the explanations as simple as possible without going into too much geekspeak.

[fluffy_little_something]

So, what file format is the "unzipped" FLAC file?

[Daags]

Not true, daddy-o. VLC for example can play audio that is contained within zip files.

Foobar too. But it does this by unzipping the file to a temporary folder and playing the files from there.

No shit. Nevertheless, you phrased it in user terms, and contrary to that 'you' dont need to unzip it to play it at all - it's an abstract process.

[aciddose]

But from I have read here one can make a lossless FLAC file of a crappy sounding audio file

In fact it should be possible to create a .wav containing flac data. I'm not sure if they got a registration for flac, but there are ways around this.

It just gets ridiculously complicated to deal with which is why they used another container format instead. They have their own format, there is also support for OGG containers. There isn't much reason to believe that it wouldn't be possible to store flac compressed data in many other containers.

[Numanoid]

It's a joke, the indication is you apparently came to believe the FLAC file sounds better based in the concept 'lossless'. Like someone that believes in superiority of 'vinyl'. But .wav is not lossy compression, (as a file) it's undergone no compression.

In the end it is about saving storage space

MP3 was big when computers shipped with HDD's like 40-80 GB in size 10-15 years ago

When HDD became affordable up to 1TB and beyond it is much more affordable to make lossless compression, which takes more space than mp3 but less than WAV

And now when HDD can be affordable for 10 TB+ users just gonna be working with WAV and skipping compression, lossless or not, because there is space.

What is happening now is that one want to go from 16 bits to 24 bits to 48 bits WAV files, everybody just loves eating HDD space

[Numanoid]

So, what file format is the "unzipped" FLAC file?

You can experiment with the FLAC frontend prog

Lets you compress and decompress what you have compressed

http://flacfrontend.sourceforge.net/

[Numanoid]

But from I have read here one can make a lossless FLAC file of a crappy sounding audio file

Of course, FLAC does nothing with that

Shit audio quality in means you get compressed lossless shit audio quality out as a FLAC file

It always amazes me when somebody rips their vinyl collection to 320 kbps mp3 or WAV, and think that will matter if they use a rather shitty old vinyl player and a primitive amp to record the audio from.

Oh yeah having big files then is super important, want to hear crackles in pristine audio

[aciddose]

So, what file format is the "unzipped" FLAC file?

Question doesn't make sense.

Data doesn't come in a "file format", they're two distinct concepts. A file format is a particular way of arranging some data.

The data inside flac data is in... flac format.

https://xiph.org/flac/format.html

[sprnva]

So, what file format is the "unzipped" FLAC file?

Uncompressed it's raw audio just like what is coming out of your DAW. The zip analogy was my attempt to explain audio compression vs data compression. With audio compression there's no need to unpack the audio file into an uncompressed form to play it back.

FLAC contains raw audio data compressed, same as MP3. When you play back an MP3 it doesn't have to unpack the raw audio to a file, it can decompress it on-the-fly. Same goes for FLAC or any other audio (or video) compression format.

[fluffy_little_something]

Ok, got that
What format is the live DAW sound in? Since it is digital, it must have some sort of format, right, unlike analog signals.

[whyterabbyt]

What format is the live DAW sound in?

Primarily depends on the DAW, and the user's preferences of that DAW. In ProTools for example, though, the file format for each project is set when you create the project; it's either WAV or TIFF, in 16,24 or 32 bit, either interleaved or non-interleaved.

[sprnva]

I was going to say PCM, but I think float is the right answer for the audio you hear during playback in the DAW.

[whyterabbyt]

I was going to say PCM, but I think float is the right answer for the audio you hear during playback in the DAW.

Cant get float into a DAC, so audio you hear will ultimately be in the format you set for that ie 24-bit integer.
OTOH, AFAIR, the VST format expects 32-bit floats so the 'working' format internally in a VST DAW is likely to be that, (except where internal processing is 64-bit float. or something else again, like in an old TDM protools system ).

Really, that's the data format, though, and, as with PT, any audio clips in the project may be coming off disk, in which case that'll have a file format - unclear whether fluffy has conflated file format with data format, I suspect so.

[Numanoid]

To put on my HDD I just compressed a CD I bought the other day:

608 MB in WAV -> 306 MB FLAC

So obviously there is lots of space to save doing this

And it took less than half a minute transfering the WAV's to FLAC

[aciddose]

Ok, got that
What format is the live DAW sound in? Since it is digital, it must have some sort of format, right, unlike analog signals.

The "live sound" is transmitted by the interface. If the interface is VST, it is most likely "raw" blocks of single channel audio with a single 32-bit floating point value per sample.

An interface is defined by code... sort of like how the interface to your PC is a keyboard and mouse. In order to use the PC (or, the plug-in) the user (or the DAW) must communicate in a specific way.

In VST, this looks like this: (in C)

process_function(float *inputs[channels], float *outputs[channels], int samples)

The datatype is "float" as I explained. The "*" means the datatype is pointed to, in other words each element 0 ... channels contains an address where some float data can be found.

When you type "name[length]", this defines an array. Arrays are defined both by the language and the compiler, although in this case we can safely assume each element appears directly after the last in a solid block.

float buffer[samples];

Once again, just like the array of addresses, each address points to the beginning of an array of samples which we call a "buffer". This is just like the one-after-the-other list of addresses. Each sample is stored sequentially in memory.

So the format is simply a block of memory filled with raw numbers.