Thanks.OK, found the paper magazine. The back issue seems to be available still :
https://www.hificritic.com/store/p130/HIFICRITIC_Vol_10_No_2.html
It also has some unexpected tests - for some of their tests they sent sound through FIVE cycles of WAV-FLAC-WAV conversions then listened for stereo image HEIGHT changes!
It is in HiFi Critic, April to Jun 2016 issue. It is titled "Why do WAV and FLAC files Sound Different?", written by Dr. Charles Zeilig and Jay Clawson. I have not tried to Google for it but if it's available the above details should be enough to find it. If you do, please post the link. The article refers to articles in The Absolute Sound issues 218-221, 246 and 248 but links to buy these did not work for me. The hificritic.com URL quoted in the article to download larger versions of it no longer works.
Anyway, here is my summary of their summary at the end of the article :
Load fresh metadata into WAV files before saving them if they have been converted more than once e.g. from WAV files (HF : this is more for studios but using options to seek metadata from the Internet when ripping might achieve this for the home user).
Do not use even lossless compression unless you really need to save that disk space.
If you must use the FLAC format (they strongly prefer WAV format) use either NO compression or the lowest level of compression aka Level 0 (HF : this is what I use but I have not done any comparisons). Never use (default) Level 5 compression.
Make metadata-associated cover art no larger than 800 x 800 pixels and use maximum JPG compression (! - HF) for best sound quality.
HTH,
HughF
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Hugo M Scaler by Chord Electronics - The Official Thread
- Thread starter ChordElectronics
- Start date
mlxx
100+ Head-Fier
- Joined
- Apr 7, 2012
- Posts
- 484
- Likes
- 153
If you do, please post the link.
https://www.hificritic.com/uploads/..._differences_between_wav_and_flac_formats.pdf
https://www.hificritic.com/uploads/..._differences_between_wav_and_flac_formats.pdf
Interesting comments in this video about sources/noise...
HughF
New Head-Fier
Many thanks - now I have 15 pages of homework to read after digesting the original 5-page article! Safely downloaded now.
Cheers,
HughF
JaZZ
Headphoneus Supremus
Thanks for the interesting links! I was reading the articles rather hastily, so might have missed what I miss: how exactly the auditions were done and if the reconversion to Wave would be able to fix the issue under certain preconditions.
Since my brief attempts never revealed a sonic difference between Wave and Flac versions of the same recording, my interest was peeked, so I took the logical step: mixing Wave and Flac files together with inverted phases – using WaveLab and Flac 1.3.1 encoder set to highest compression (8) in foobar2000. Test subject was the first movement of Shostakovich's 10th Symphony in 24/96, a DVD-A rip (from DGG), one of my best sounding recordings.
To be more precisely: The so-called inverted «Flac file» mixed to the Wave original is in fact a Wave file reconverted from the Flac file.
What I initially found was a distinct «noise floor» of exactly –73.33 dB and one single peak of –52 dB.
However – mea culpa! – I managed to convert the original 24/96 Wave file to a 16/96 Flac file due to the corresponding preset in foobar2000. So please forget my previous findings, and sorry for the false allarm!
The new graph shows a regular –144 dB ground floor – but again also a single peak, this time of –120.412 dB.
Now this can barely be the cause of the pretended sonic differences, but it is nevertheless a bit disturbing.
However, I still can't hear a difference between Wave and Flac. Actually in my main (computer-based) system (with M Scaler and DAVE) I exclusively listen to Wave files. Flac is reserved for my DAPs.
Since my brief attempts never revealed a sonic difference between Wave and Flac versions of the same recording, my interest was peeked, so I took the logical step: mixing Wave and Flac files together with inverted phases – using WaveLab and Flac 1.3.1 encoder set to highest compression (8) in foobar2000. Test subject was the first movement of Shostakovich's 10th Symphony in 24/96, a DVD-A rip (from DGG), one of my best sounding recordings.
To be more precisely: The so-called inverted «Flac file» mixed to the Wave original is in fact a Wave file reconverted from the Flac file.
What I initially found was a distinct «noise floor» of exactly –73.33 dB and one single peak of –52 dB.
However – mea culpa! – I managed to convert the original 24/96 Wave file to a 16/96 Flac file due to the corresponding preset in foobar2000. So please forget my previous findings, and sorry for the false allarm!
The new graph shows a regular –144 dB ground floor – but again also a single peak, this time of –120.412 dB.
Now this can barely be the cause of the pretended sonic differences, but it is nevertheless a bit disturbing.
However, I still can't hear a difference between Wave and Flac. Actually in my main (computer-based) system (with M Scaler and DAVE) I exclusively listen to Wave files. Flac is reserved for my DAPs.
Last edited:
rkt31
Headphoneus Supremus
- Joined
- Oct 14, 2014
- Posts
- 2,108
- Likes
- 776
I accuaccident pressed one of last three buttons of HMS and all three lit up. I know there is no use of these and these are for future use but pressing and lighting up these affect the sound quality ?
miketlse
Headphoneus Supremus
- Joined
- May 8, 2016
- Posts
- 5,977
- Likes
- 3,980
Did you hear the sound quality change?
Triode User
Member of the Trade: WAVE High Fidelity
I accuaccident pressed one of last three buttons of HMS and all three lit up. I know there is no use of these and these are for future use but pressing and lighting up these affect the sound quality ?
They are a non cancelable self destruct sequence. Put the MScaler in a bucket of water and hide under a table until the self destruct sequence has been completed. This is a random period selected by the firmware from a complex calculation involving the date of Rob Watts birthday and the number of minutes that the MScaler has been used since 1st April.
miketlse
Headphoneus Supremus
- Joined
- May 8, 2016
- Posts
- 5,977
- Likes
- 3,980
Careful - it is almost guaranteed that some audiophile will read this, and think you were being serious.They are a non cancelable self destruct sequence. Put the MScaler in a bucket of water and hide under a table until the self destruct sequence has been completed. This is a random period selected by the firmware from a complex calculation involving the date of Rob Watts birthday and the number of minutes that the MScaler has been used since 1st April.
dmance
Member of the Trade: AudioWise Inc
Two weeks to the Munich high end show...Chord rumored to be showing several new products. I'm sure one is going to make use of those spare glowing buttons
miketlse
Headphoneus Supremus
- Joined
- May 8, 2016
- Posts
- 5,977
- Likes
- 3,980
It is slightly stronger than a rumour.Two weeks to the Munich high end show...Chord rumored to be showing several new products. I'm sure one is going to make use of those spare glowing buttons
What Hi Fi have posted "Last year, Chord used High End to introduce both its Hugo TT 2 headphone amp/DAC and Choral Etude power amp, and it looks as though multiple new products will be in the spotlight this year. We've been sent an invite for the launch of "a number of new products". Will we see a new amplifier? Or a second streaming device? More DACs? All will be revealed on Thursday 9th May."
rkt31
Headphoneus Supremus
- Joined
- Oct 14, 2014
- Posts
- 2,108
- Likes
- 776
Mojo 2, davina, digital power amp and ?
JaZZ
Headphoneus Supremus
I would say 2Go and maybe Mojo₂ and maybe more...
lcasadonte
100+ Head-Fier
I've seen a flurry of activity around the poly software that leads me to believe they are getting close anyway.
Perfect! The 2go comes out at a low financial trough for I.
Users who are viewing this thread
Total: 28 (members: 11, guests: 17)