I have constructed a blindest between 3 different sample rates. The original file was 24bit 96KS/s and the worst of my files is 44.1KS/s. Please tell your result from blind testing if you have done multiple independent listenings. https://www.dropbox.com/sh/j1535de2zh51ooe/AACMBuey9leK4_5JizDghiAra?dl=0 The names is of course random. The bit-rate is also somehow random since i might have added noise to some files in order to increase the bit-rate and make cheating harder(I can also use different compression settings when creating the flac file). Please do not tell which file is which if you have cheated using suitable program, send a PM if you have questions of any kind. Link to the songs Overgrown / Frozen Dandylions - Frozen Dandylions / Overgrown - Encyclopaedia Metallum: The Metal Archives i used "untitled" and edited out 75% of the song for several reasons.
Edit... Alternative download from Dropbox as a single zipped folder.
https://www.dropbox.com/s/6zhatlwmep0p5gi/Blind.zip?dl=0 (frozen_D has the wrong sampling rate).
Edit... Alternative download from Dropbox as a single zipped folder.
https://www.dropbox.com/s/6zhatlwmep0p5gi/Blind.zip?dl=0 (frozen_D has the wrong sampling rate).
Last edited:
Do you want listening impressions and comments posting here in this thread or should anyone pm you with their thoughts ?
One file stands out at a quick listen as very inferior. The other two I will give a closer listen later.
I'm going to move this to 'The Lounge' where it will get more exposure.
One file stands out at a quick listen as very inferior. The other two I will give a closer listen later.
I'm going to move this to 'The Lounge' where it will get more exposure.
if you refer to the files as low,mid,high or something it is ok to tell all your findings in this thread, but if you use the names(D,E,F) you should not give any valable information.
Unless you have a high end system it is highly unlikely you will hear any difference. A 30khz cabable tweeter is recommended.
I've given them all a good listen via headphones (Sony MDR V7 + Acer laptop) and listening now I can't repeatably identify them.
When I was quickly listening earlier I must have been jumping to a slightly different point in the tracks because one seemed to be limited in 'width' or soundstage bu listening via Foobar I'm not hearing that.
The only way I can play these via my main system is to convert all to wav (which I guess misses the point) as I haven't a separate DAC and so on. Would have to put them on USB as WAV to use the DAC in my Marantz Pearl Lite. I'll perhaps try it later though.
When I was quickly listening earlier I must have been jumping to a slightly different point in the tracks because one seemed to be limited in 'width' or soundstage bu listening via Foobar I'm not hearing that.
The only way I can play these via my main system is to convert all to wav (which I guess misses the point) as I haven't a separate DAC and so on. Would have to put them on USB as WAV to use the DAC in my Marantz Pearl Lite. I'll perhaps try it later though.
well if you are testing with headphones you can of course post your results in this thread, the same goes for conventional tweeters(tweeters limited to 20khz or worse) the results will not have any value(unless you got something statistically sure).
I am hoping your main system is suited for this test 🙂
I have made a picture of the difference between the 2 best files. No difference below 26khz which tells us that i did a good job at downsampling+upsampling using audacity. These tracks appears to have pretty powerful ultrasonics, a user on another forum has showed that these ultrasonics are in fact related to the music and not noise.
Attachments
I had another listen to these 🙂
The middle (Frozen_E) I felt was the 'tightest' in presentation, as if the HF content didn't overspill. The first was similar to the last. So the middle is the one you could say was 'odd one out'. If I was picking up on differences it was minute.
The middle (Frozen_E) I felt was the 'tightest' in presentation, as if the HF content didn't overspill. The first was similar to the last. So the middle is the one you could say was 'odd one out'. If I was picking up on differences it was minute.
I couldn't play Frozen_D as 192kHz isn't supported by my DAC. Initially Frozen_E sounded more open and sparkling hence more attractive. After several retries I found that even slightly subdued in comparison, Frozen_F provided more convincing illusion of live performance, especially in more complex parts. This feeling gradually creeps in, peeking at about 2:00.
There is a lot of people who think that there would be any audible difference between frozen_E and frozen_F under any circumstances. I recommend doing a few A/B sessions to verify your findings. Foobar has an ABX plugin but i think AB blind-testing is more suitable for this task. I might upload a 96 KS/s version of frozen_D instead.
I suspect my PC and its soundcard aren't suitable for this kind of tests. It seems some SR conversion goes in background while playing hires files. Paradoxically this downsampling actually helps notice hires content. Due to conversion imperfection I suppose some ultrasonic content is aliased down in audio band and makes subtle difference. So foobar ABX is not reliable in my case.
I use Raspberry Pi + Volumio + standalone DAC for serious listening. Unfortunately no ABX tool available there.
Bitperfect audio in windows is easy to achieve(wasapi, asio is alternatives). I blindtest in linux using deadbeef, looping and just play a random song in the playlist(i did recommend AB testing, not ABX). I didn't think it would be necessary to mention bit perfect playback in this forum, everyone should know it already. BTW: i have reuploaded the files with proper metadata and frozen_D with 96KS/S samplingrate now(first link in the thread). There should not be any aliasing if you have a proper setup with 96KS/s capable dac(with no sample rate conversion). I will also upload "frozen_delta" which is the difference between the 96KS/s version and 54KS/s version.
Last edited:
I simply cannot enjoy music with PC running in the room. Also I think the difference I noticed was not evident when rapidly switching between tracks. It was more like looking a 3D picture printed in magazine when you have to stare longer and make eyes focus behind the page and suddenly hidden image pops out. So in case of FrozenF it takes about 20-30 sec listening to begin feeling increased sense of presence (or perhaps just imagining).
Tomorrow I'll try the new FrozenD.
Tomorrow I'll try the new FrozenD.
About ultrasonic hearing
According to this paper http://www.tinnitus.vcu.edu/Pages/Ultrasonic Hearing.pdf
The ultrasonics removed by the down-sampling has very low amplitude which suggests that these should not be audible, however the threshold for audibility might be lower when the ultrasonics is mixed with normal sounds. This study (does not appeared to be verified by others yet) suggest that ultrasonics is unconsciously audible when mixed with normal sounds Inaudible High-Frequency Sounds Affect Brain Activity: Hypersonic Effect | Journal of Neurophysiology . The study has a long reference list which i might look at more when i have time.
Ultrasonic speech perception Science Magazine: Sign In
According to this paper http://www.tinnitus.vcu.edu/Pages/Ultrasonic Hearing.pdf
The ultrasonics removed by the down-sampling has very low amplitude which suggests that these should not be audible, however the threshold for audibility might be lower when the ultrasonics is mixed with normal sounds. This study (does not appeared to be verified by others yet) suggest that ultrasonics is unconsciously audible when mixed with normal sounds Inaudible High-Frequency Sounds Affect Brain Activity: Hypersonic Effect | Journal of Neurophysiology . The study has a long reference list which i might look at more when i have time.
Ultrasonic speech perception Science Magazine: Sign In
Last edited:
It's interesting that only 1 in 3 people who voted in the poll think that they hear any difference, with only 3 votes the poll is essentially worthless. This will be my last post unless someone else replies. We have 2 persons claiming audible difference between frozen_E and frozen_F but i have not received any blindtesting statistics yet.
The very obvious fact that any differences are so subtle that raving audiophiles have to disagree and argue about them means that the same differences are more or less miniscule at best, to completely inaudible at worst.
Can we get over rates and move on?
Can we get over rates and move on?
- Status
- Not open for further replies.