[Scrith]

The inability of us to easily ABX this stuff just continues to fuel this endless, senseless speculation. I really, really wish Foobar2000 weren't converting everything to WAV for its ABX support. Of course when I mentioned that on HydrogenAudio I got banned for a week (let's just say open-mindedness is heavily discouraged there).

Oh well...I tried it again, making a WAV out of one of my FLAC files that I know really well, and flipped back and forth so many times between them that I started losing track of which was which (so it was semi-blind I guess...save it purists, I know the only things that matters to you is fully blind).

Anyway, on my hyper-detailed system (In my experience the Qualia outperforms even the HE90 in this department), they sound EXACTLY the same to me. I won't be converting my huge library to WAV anytime soon.

 

[Publius]

Quote:

Originally Posted by Patrick82 /img/forum/go_quote.gif I have already compared 0% vs 100% CPU load in a blind test and it was easy to hear.

Your sound card is inferior. Upgrade.

 

[Jon L]

Quote:

Originally Posted by Scrith /img/forum/go_quote.gif Oh well...I tried it again, making a WAV out of one of my FLAC files that I know really well, and flipped back and forth so many times between them that I started losing track of which was which (so it was semi-blind I guess...save it purists, I know the only things that matters to you is fully blind). Anyway, on my hyper-detailed system (In my experience the Qualia outperforms even the HE90 in this department), they sound EXACTLY the same to me. I won't be converting my huge library to WAV anytime soon.

LOL. That's exactly what I've been doing for a long time. I've made Wav, Flac, and WMA lossless files of the same songs, then over months compared them at random times semi-blinding myself.

And I still haven't come to a conclusion. I have somewhat decided that the time of day(RFI/EMI?), your mood(psychoacoustics), how long the equipment/tubes have been warmed up, etc greatly contributes to the perceived differences also.

I also suspect the different decoder program used for Flac, WMA lossless, etc contribute to whatever minimal difference you hear for some unknown reason.

The software used (Foobar, Winamp, iTunes) also yields much greater sound differences than the lossless format/wav differences.

As far as CD vs. hard drive, I can understand people who prefer the CD sound, but I don't believe that's b/c CD is "better." IME, spinning the actual CD has a tendency to sound more "familiar" to your ears b/c that's what most people have grown up with. It has a bit more livelier, brighter tinge to things, a tiny bit more "shine" and contrast in the presence region, and a bit more upfront and "insistent" compared to hard drive.

I personally believe hard drive playback is closer to the actual recording if implemented properly. I will also note that very expensive, extravagant CD transports tend to minimize the CD effects I mention above.

 

[aphex944]

Guys, these theories and questions on quality differences between lossless encodings aren't going to lead anywhere. When you understand the technical aspect of computers(specifically how data is stored and accessed), this all becomes irrelevant.

Any true lossless encoder(meaning pure data compression, exact 1:1 replica of original input file once decompressed) HAS to physically sound exactly the same as any other. This is of course assuming the rest of the playback chain is the same: no dithering, plugins, eqing, etc. That's the whole point of lossless audio, there's no room for differences or changes.

There's not much room for error in a computer system, either. Audio files are pure data until they processed by the sound card. Because of this, they are governed by the same rules as any other data in the system. Playing a FLAC file will be EXACTLY the same as playing a WAV file. In fact, I'd be willing to bet that all a FLAC decoder in a player does is decode a FLAC to WAV on-the-fly.

Playing a CD shouldn't be any different, either. Assuming you have DAE enabled(and no scratches on the CD), it's going to produce the exact same data as it would if you had "ripped" the CD in something like EAC.

Speaking of differences in playback quality between differing levels in CPU usage is silly. This would be really easy to test for, but I doubt that'd convince Patrick anyway.

 

[lan]

Quote:

Originally Posted by aphex944 /img/forum/go_quote.gif Guys, these theories and questions on quality differences between lossless encodings aren't going to lead anywhere. When you understand the technical aspect of computers(specifically how data is stored and accessed), this all becomes irrelevant. Any true lossless encoder(meaning pure data compression, exact 1:1 replica of original input file once decompressed) HAS to physically sound exactly the same as any other.

Playback it is dependent on the whole system. If you take a look at the whole system approach to anything, there are differences. Being an exact replica (digital) doesn't mean it has to physically sound the same (we are in analog world).

e.g. there are analogies of systems in many places.

Take $1. You can represent it as 4 x .25 or 10 x .10 or 100 x .01. They are worth the same. Physically they are different and anything related to that is different. How much they weigh, how you carry them, how much space them take, how you use them, etc. There are many interactions. Some of them obvious, some not, some irrelevent, some annoying.

There are also many interactions within the computer system.

 

[maarek99]

Quote:

Originally Posted by lan /img/forum/go_quote.gif Pf you take a look at the whole system approach to anything, there are differences.

Like what?

Quote:

Being an exact replica (digital) doesn't mean it has to physically sound the same.

Hmm...yes it does.

 

[Lou Erickson]

I wonder if what he's hearing as brightness is the CPU changing frequencies and bleeding in to some part of the audio system. If he'd turn off Cool & Quiet so the CPU was always at full speed, perhaps that difference would go away. (And perhaps everything would be bright all the time! Ick.)

I'm more likely to believe that a very different clock frequency on the CPU is audible than I am FLAC itself makes a difference.

 

[Wodgy]

I really hate seeing threads like this here. This is material for AudioCircle or AudioAsylum. This is all placebo.

FLAC and WAV sound identical. Jitter isn't an issue! A sound card uses a fixed-frequency crystal to generate its clock. CPU load does not affect at what rate (and amount of jitter) a quartz crystal oscillates.

The only way there can be a difference is if your software is configured in a non-identical fashion (e.g. FLAC is using ReplayGain), if you're using a USB device which generates its own clock, or if increased electrical activity under CPU load is polluting the ground.

 

[Redo]

Quote:

Originally Posted by Wodgy /img/forum/go_quote.gif I really hate seeing threads like this here. This is material for AudioCircle or AudioAsylum. This is all placebo.

It's true. There needs to be about 4-5 different PC systems, all with identical DAC's/AMP's setup. And some outside party needs to be controlling the test, so the user doesn't know what file format is being played. It's the only way to be for sure.

Until then, it's easy to go around saying there's a difference in sound. You have no way to prove it one way or another, it's all hearsay. As stated earlier, the difference between FLAC and WAV is 50x less than 320kbps mp3 cbr and WAV. That statement alone should raise red flags.

 

[lan]

Quote:

Originally Posted by maarek99 /img/forum/go_quote.gif Like what? Hmm...yes it does.

Looking at the system approach of something requires you to open your mind to the whole world and it's intricacies. Do you believe a hurricane in the pacific affects a person's weather in another part of the world? How about your refrigerator affecting your TV? There are some people who don't because they see them as unrelated. There are other people (like me) who witness interactions amongst many things on a both macro and microscopic level. Some of which are predictable in our own systems. I find patrick rather extreme. I wouldn't make a public statement as flac is brighter than wav. While it maybe valid in his system, it may not be on other systems. In a computer system, there are many combinations of hardware and software. I believe it's impossible to recreate the exact experience as patrick.

maarek99. Looking at the system approach to this, the system doesn't work exactly in the same conditions. FLAC uses system resources differently as well as the media players or any other piece of software. It's not enough to say that if something is digitally/logically/theoretically the same the playback/analog/perception wise they will be also. How granular do you view things? To what scales?

 

[aphex944]

Your money analogy does not work in relation to this specific 'issue'. FLAC is decoded to the same format as WAV before being sent to the sound card. It's decompressed on-the-fly, and the data output is exactly the same when it reaches the sound card.

Any changes in the data between these steps indicates a severe problem, and the computer system probably wouldn't function. Computer systems are HIGHLY controlled and engineered environments-- random changes, happenings, or anomalies to the data are almost non-existant.

I'll 100% agree with you there are many interactions between systems on a micro and macro scale. However, in an environment where system function depends HIGHLY on controlled variables and outcomes, interactions can be significantly narrowed down and explained. In this case, it is easy to follow the path and flow of data from the source to the endpoint.

 

[003]

Quote:

Originally Posted by lan /img/forum/go_quote.gif Being an exact replica (digital) doesn't mean it has to physically sound the same (we are in analog world).

Assuming it is properly decoded on the fly without any error or replaygain or misconfigured software and hardware, yes it does mean it will sound the same. Digital is an all or nothing type of thing in this circumstance. Jitter can not even effect a digital flac file being decoded to wav. The only place where jitter can come into play is the digital output on the sound card, and the wav file passes through this as well.

Quote:

Originally Posted by lan /img/forum/go_quote.gif Take $1. You can represent it as 4 x .25 or 10 x .10 or 100 x .01. They are worth the same. Physically they are different and anything related to that is different. How much they weigh, how you carry them, how much space them take, how you use them, etc. There are many interactions. Some of them obvious, some not, some irrelevent, some annoying.

That is a bad analogy for this situation. Computers do not work like that, period. A digital replica on a hard drive will ALWAYS be EXACTLY the same as the original. Period.

Think of it this way. Lets talk about a digital data file, an executable to be more specific. You make a second copy of it and have it on your hard drive along with the original. If a SINGLE digital bit was flipped, changed, removed or in any way, shape or form modified, the copy would no longer work. This is how exact and precise digital is. It does not magically become prone to a margin of error (there IS no margin of error) when we are talking about audio.

As already stated in this thread a crappy sound card, misconfigured software or other oddities in the on-the-fly decoding process of the FLAC file could possibly result in an error which would change the sound. But a healthy, working PC with proper hardware and properly set up software WILL sound EXACTLY the same with a WAV file and a FLAC-compressed version of that WAV file.

 

[Patrick82]

I'm converting all my FLACs to WAV as I'm writing this. I also found a WAV file inside one of the folders, I have done A/B-ing before but I didn't hear a difference so I kept using FLAC. That was 5 months ago.

But after further tweaking I have heard huge differences from things I haven't heard before, especially warm-up of the computer, a cold computer is masking the differences of tweaks for the computer.

After 12 hours of warm up I couldn't hear a difference between 1 vs 2 harddrives. After 24 hours the difference was subtle. After 36 hours it was more apparent, but the signature of the "cold" system was still added on top of the music, it was masking everything. A couple weeks warm-up is needed for it to sound neutral, the last time I tried it I heard nothing in front of the music.

 

[Redo]

Quote:

Originally Posted by Patrick82 /img/forum/go_quote.gif I'm converting all my FLACs to WAV as I'm writing this. I also found a WAV file inside one of the folders, I have done A/B-ing before but I didn't hear a difference so I kept using FLAC. That was 5 months ago. But after further tweaking I have heard huge differences from things I haven't heard before, especially warm-up of the computer, a cold computer is masking the differences of tweaks for the computer. After 12 hours of warm up I couldn't hear a difference between 1 vs 2 harddrives. After 24 hours the difference was subtle. After 36 hours it was more apparent, but the signature of the "cold" system was still added on top of the music, it was masking everything. A couple weeks warm-up is needed for it to sound neutral, the last time I tried it I heard nothing in front of the music.

You have one of those really old computers made out of vacuum tubes?

 

[rhymesgalore]

Quote:

Originally Posted by Patrick82 /img/forum/go_quote.gif I'm converting all my FLACs to WAV as I'm writing this. I also found a WAV file inside one of the folders, I have done A/B-ing before but I didn't hear a difference so I kept using FLAC. That was 5 months ago. But after further tweaking I have heard huge differences from things I haven't heard before, especially warm-up of the computer, a cold computer is masking the differences of tweaks for the computer. After 12 hours of warm up I couldn't hear a difference between 1 vs 2 harddrives. After 24 hours the difference was subtle. After 36 hours it was more apparent, but the signature of the "cold" system was still added on top of the music, it was masking everything. A couple weeks warm-up is needed for it to sound neutral, the last time I tried it I heard nothing in front of the music.