FLAC vs. 320 Mp3

[bigshot]

How many blind comparisons did you do?

 

[hakunamakaka]

Stax Lambda’s had astonishingly low distortion nearly 40 years ago and as far as I’m aware none of the HPs you mentioned get close in that regard. The $6k Susuvara’s for example have a measured distortion of up to about 0.5%, while the $9 Apple dongle has a distortion measurement of DAC and amp combined of about 0.001%, orders of magnitude less distortion than nearly all HPs, even TOTL ones!

No, quality bass has always been important, although more so in the digital age, which allowed higher lvls and more accurate bass than vinyl. Much/Most electronic music does not have quality bass. It typically has a very loud bass but not an especially high quality bass.

G

Stax lambda can’t compete against electrostats today. Piercing treble, no bass and weak soundstage. It’s detailed sound, but not worth the shot even for 400$ as used to me.

I’m with headphones since teens and never came across on any vintage headphone which had good bass reproduction for modern music. Actually they are bested in every category against todays flagships.

There was no money to be made with headphones back than. Poor logistics, no availability for materials or precise manufacturing.

The holy grail of headphones - HD600 was released for a whopping 450$, which if we consider inflation would match the price for the cost of Utopia today. It wouldn’t be a fair fight if we compared them against each other.

I see lots of potential for further growth in that area as portable gear is getting more popular

 

[gregorio]

Stax lambda can’t compete against electrostats today.

Yes they can. …

Actually they are bested in every category against todays flagships.

I’ve already stated this assertion is false and provided the distortion figures of a flagship of today (0.5%), which is way higher than the distortion of the nearly 40 year old Lambdas (around 0.07% if I remember correctly).

Just repeating the same falsehood, contrary to the objective facts, with no supporting reliable evidence will NOT get you very far in this subforum. I would have expected you to know that by now, considering the amount of time you’ve been posting to this subforum!

For the record, I’m not saying there has been no improvements in transducers over the last 40 years or so, just that they have been relatively minor, more along the lines of “tweaks”. They have not “improved a lot” and nearly all of them still produce way more distortion than even today’s cheap amps and DACs.

G

 

[gregorio]

to me 320 FLAC sounded slightly less compressed, slightly more detailed and slightly brighter than 320 MP3 however 320 MP3 sounded slightly warmer and slightly richer than 320 FLAC to my ears while listening …

Do you just mean FLAC, rather than 320 FLAC? I’ve not heard of 320 FLAC.

What you describe is more indicative of different masters than of the difference between FLAC and 320 MP3. Did you create the MP3 from the FLAC yourself or did you download the different versions?

Disclaimer: This is based on my Audio setup and my results will be different to your results due to the variations of Audio setups.

Why would the results be different? Wouldn’t it be slightly more detailed, brighter or warmer/richer on other setups if that difference existed in the files being reproduced?

I have taken onboard that expectation bias can occur so I asked my Brother who does audio and video editing work for a second opinion and he was in the room with me during the initial blind and subsequent sighted comparison session as well and his results matches my results.

But you seem not to have “taken onboard that expectation bias can occur”. Expectation bias occurs in all human beings, so your brother matching your results doesn’t indicate anything more than that you are both human beings.

If you really want to take onboard that expectation bias occurs, then please try an ABX test which eliminates that likelihood. There’s free software available, an ABX plugin for Foobar2000 for example.

G

 

[hakunamakaka]

Yes they can. …

I’ve already stated this assertion is false and provided the distortion figures of a flagship of today (0.5%), which is way higher than the distortion of the nearly 40 year old Lambdas (around 0.07% if I remember correctly).

Just repeating the same falsehood, contrary to the objective facts, with no supporting reliable evidence will NOT get you very far in this subforum. I would have expected you to know that by now, considering the amount of time you’ve been posting to this subforum!

For the record, I’m not saying there has been no improvements in transducers over the last 40 years or so, just that they have been relatively minor, more along the lines of “tweaks”. They have not “improved a lot” and nearly all of them still produce way more distortion than even today’s cheap amps and DACs.

G

ok gatekeeper

 

[gregorio]

ok gatekeeper

Why should I and others have to keep acting the “gatekeeper” with you?

Again, you been posting to this subforum for ages, how is it possible you haven’t learned in all this time that you’ll be challenged if you post falsehoods?

G

 

[dondadpie]

I have tested this quite a bit on different systems, and really can't tell the difference. Would be interested to see the difference on a big club system, if there is one

 

[gregorio]

Would be interested to see the difference on a big club system, if there is one

Big club systems are usually pretty poor as far as fidelity is concerned. For starters they’re usually mainly mono, sacrifice accuracy for volume and have very uneven freq response, typically heavy on bass and mids, weak everywhere else.

G

 

[Singleton]

Do you just mean FLAC, rather than 320 FLAC? I’ve not heard of 320 FLAC.

What you describe is more indicative of different masters than of the difference between FLAC and 320 MP3. Did you create the MP3 from the FLAC yourself or did you download the different versions?


Why would the results be different? Wouldn’t it be slightly more detailed, brighter or warmer/richer on other setups if that difference existed in the files being reproduced?


But you seem not to have “taken onboard that expectation bias can occur”. Expectation bias occurs in all human beings, so your brother matching your results doesn’t indicate anything more than that you are both human beings.

If you really want to take onboard that expectation bias occurs, then please try an ABX test which eliminates that likelihood. There’s free software available, an ABX plugin for Foobar2000 for example.

G

I've downloaded the different versions from the same online digital music distribution website and the reason as to why the results could be different is due to the high variability of the following variables: the Computer Motherboard used, the Computer Motherboard BIOS version, the Computer Motherboard Firmware version and the Computer Operating System version used as any updates and changes to these variables have the potential to affect the performance of the Audio Driver installed and the sound quality that is being outputted.

I'll take onboard your suggestion and thanks for critiquing and providing your input as this helps me improve my audio experiments and its results. I agree that we are all human and its super difficult to remove all biases including confirmation bias. Will reconduct my experiment.

If you have anymore suggestions on how to reduce confirmation bias please reply to this post. I'm openminded and will consider any suggestions on how to improve the ways I conduct audio experiments.

 

[sander99]

I've downloaded the different versions from the same online digital music distribution website

They could still come from different masters. The only way to be sure that you compare the same master in different formats is by converting yourself.

 

[gregorio]

I've downloaded the different versions from the same online digital music distribution website and the reason as to why the results could be different is due to the high variability of the following variables: the Computer Motherboard used, the Computer Motherboard BIOS version, the Computer Motherboard Firmware version and the Computer Operating System version used as any updates and changes to these variables have the potential to affect the performance of the Audio Driver installed and the sound quality that is being outputted.

As sander99 says, you must create the MP3 yourself from the FLAC. How, when, from which master and if any other processing was applied before encoding are all common variables that could easily affect your results, especially given that your description of the differences match these variables but don’t really match the artefacts of MP3. Next on the list after this probability would be a test methodology fault. A very, very distant last place on the list would be the motherboard, bios, OS variables you listed.

MP3 was released in 1992 and was designed to work on cheap consumer devices of the day. For context, the neural engine in a current iPhone can apparently perform 1,500 gflops (billion operations per sec) while the world’s most powerful super computer in 1992 cost about $30m and could perform 22 gflops. Something would have to be very seriously wrong with your computer for any of the variables you listed to have any affect on the decoding of MP3s.

I agree that we are all human and its super difficult to remove all biases including confirmation bias. Will reconduct my experiment.

Yes, it can be extremely difficult, not only to remove confirmation bias but to remove any clues the brain can pick up from the test procedure. However …

If you have anymore suggestions on how to reduce confirmation bias please reply to this post.

Not beyond the suggestion I mentioned previously, an ABX test. The free software not only eliminates confirmation bias but also many of the difficulties and clues with blind and double blind testing.

G

 

[RMesser9]

Lossless and 320kbps MP3 sound identical to me. I even can't tell much of a difference between 256 and 320. I still use 320 though

 

[bigshot]

Nothing wrong with that. The difference between 256 and 320 is much smaller than the difference between 320 and lossless. You don’t have to squeeze every last drop out, just be efficient.

 

[Blackwoof]

Just did a A/B with LAME never expected It to hold up so well at 160kbps VBR, If not better than Vorbis/AAC in few areas like ambient & Vomir. With LAME 3.99.5o at Q2 + 48KHz SOX resample it transparent at 192kbps on pre echo rich music.

 

[EMINENT]

I tried one of these tests online the other day and got 2 out of 6 right. Between uncompressed wav, to 320 mp3 and 128 aac.

This was not conclusive to me as I believe I would have to test the same file that I am used to hearing compressed down to the differing amounts to form a conclusion. I don't believe my transducer is limiting. Maybe this has the opposite effect and the sh!ttier the can the worse it sounds?