Visual studio 2012 c++ and wasapi minimalist player

[sbgk]

thats devotion for you

2.48 sse4 for the tinyest tinyest bit bass heavy it was

most people would have left it, here its not good enough

strive for perfection here is unbeliveable

 

was listening to Van Morrison In the days before rock n roll and couldn't hear the high notes of the instruments and the bass was very prominent, so it wasn't a subtle difference in my system.

[jesuscheung]

was listening to Van Morrison In the days before rock n roll and couldn't hear the high notes of the instruments and the bass was very prominent, so it wasn't a subtle difference in my system.

 

sbgk it is time to do machine code.

Smallest x86 ELF Hello World

someone manages to compile helloworld to only 142bytes from 6363bytes!

[jesuscheung]

agree with you, 2.48 SSE4 intel no mmcss better than others.

all 2.48 versions have good bass, but not like as 2.22_atom, only ~, 2.22_atom have best bass.

they give true sound, natural (bass, drum, beat/knock, guitar, vocal, treble,.. are very clarity)

 

3 frequency ranges (bass/mid/treble) of these 2.43 have big volume, over norm => SQ like as using normalize volume by DSP => difficult to listening (more rustic), although sounds is good, clarity, good vibration (seem more sharpness, but sharpness is normal. there's volume!). true sound on incorrect volume.

i happy with their bass, better than 2.4x - 2.47

What is the standard? difficult. jesuscheung, i tested mnq.exev1.2, very good on many things, as technical standard, excluding brightness, a bit lesser. i very prefer it for my HDMI setup. try to listening it.

 

2.48 sse4 intel sub al, 1 vs mnq.exev1.2 in bass

-listening to Metallica, coz it has lots of bass. i first listen to 1.2, i never felt more bass from Mqn. i was surrounded by real/non-artificial bass. the air has bass. and then i listen to 2.48, its bass is sharper, has more weight, but i just felt there isn't enough bass. some layers of bass are missing. e.g. the air lacks bass!

-conclusion. if 1.2 reproduces 100% bass, 2.48 only has 90% in Metallica. also, 1.2 is very easy to follow the drums, which is important for Metallica. 2.48 is difficult to focus.

 

ignoring weight, 1.2 has the best bass fullness. the air has bass. more than exenopgomemcpy and 2.22_atom.

thx lekt! i took out a non-bass non-rock cheap headphone hd555, never thought i could enjoy Metallica using it.

i could live without exenopgomemcpy and 2.22_atom using 2.43-2.48. but there isn't yet a new version that can replace 1.2.

 

 

[sbgk]

2.48 sse4 intel sub al, 1 vs mnq.exev1.2 in bass

-listening to Metallica, coz it has lots of bass. i first listen to 1.2, i never felt more bass from Mqn. i was surrounded by real/non-artificial bass. the air has bass. and then i listen to 2.48, its bass is sharper, has more weight, but i just felt there isn't enough bass. some layers of bass are missing. e.g. the air lacks bass!

-conclusion. if 1.2 reproduces 100% bass, 2.48 only has 90% in Metallica. also, 1.2 is very easy to follow the drums, which is important for Metallica. 2.48 is difficult to focus.

 

ignoring weight, 1.2 has the best bass fullness. the air has bass. more than exenopgomemcpy and 2.22_atom.

thx lekt! i took out a non-bass non-rock cheap headphone hd555, never thought i could enjoy Metallica using it.

i could live without exenopgomemcpy and 2.22_atom using 2.43-2.48. but there isn't yet a new version that can replace 1.2.

 

 

 

2.48 sse4 intel had plenty bass, did you try it ? not convinced the sub 1 version is the answer, have a few more things to try, but like the bass of 2.48 sse4 intel, maybe need to move the speakers further from the wall.

[sbgk]

sbgk it is time to do machine code.

Smallest x86 ELF Hello World

someone manages to compile helloworld to only 142bytes from 6363bytes!

 

you don't have 88kb for MQn spare ? the assembly code is 1k so that leaves 87 k of bloat. I have looked into reducing the size, but there are bigger gains in SQ to be had for less effort.

[jesuscheung]

2.48 sse4 intel had plenty bass, did you try it ? not convinced the sub 1 version is the answer, have a few more things to try, but like the bass of 2.48 sse4 intel, maybe need to move the speakers further from the wall.

 

you are right, i have discounted 2.48 sse4 intel too early because of nommcss.

 

2.48 sse4 intel > 2.48 sse4 intel sub 1.

2.48 sse4 intel sub 1 has too much weight. rather unfocus. bass not as good as 2.48 see4 intel.

 

2.48 sse4 intel has very good bass.

 

with Metallica:

-1.2's bass stays in the air. 2.48 lacks this.

-1.2's bass flows. 2.48's bass is choppy.

-1.2's bass is fuller, much fuller.

 

i never heard bass like 1.2 from any player. don't get me wrong. 2.48 is one of the best bass in new versions.

 

(still waiting for a version that blossoms in vocal. statistically, happens once every 20 versions)

[jesuscheung]

you don't have 88kb for MQn spare ? the assembly code is 1k so that leaves 87 k of bloat. I have looked into reducing the size, but there are bigger gains in SQ to be had for less effort.

 

lol... just checking if you are crazy.

[adolfo.a.aguiar]

I tried 2.48 sse2 atom and am really impressed, coming from JPlay Control and Audio PC.

If you analyse the sound, bass is very tuneful and vocals natural. But after a few seconds, I just stop with the critical listening and ENJOY the music.

One issue though: my collection is 100% flac, so I tried using MQnflac.bat and didn't succeed. I noticed that the track is loaded in the MQn.bat directory as a .wav so the conversion is OK. The file param.txt is created but the Files.txt is not. In the window, every 3 sec, "Playback commenced" and the track name are displayed but not the line "Space bar - Pause/Resume...". And unfortunately, no sound.

[lekt]

2.48 sse4 intel sub al, 1 vs mnq.exev1.2 in bass

-listening to Metallica, coz it has lots of bass. i first listen to 1.2, i never felt more bass from Mqn. i was surrounded by real/non-artificial bass. the air has bass. and then i listen to 2.48, its bass is sharper, has more weight, but i just felt there isn't enough bass. some layers of bass are missing. e.g. the air lacks bass!

-conclusion. if 1.2 reproduces 100% bass, 2.48 only has 90% in Metallica. also, 1.2 is very easy to follow the drums, which is important for Metallica. 2.48 is difficult to focus.

 

ignoring weight, 1.2 has the best bass fullness. the air has bass. more than exenopgomemcpy and 2.22_atom.

thx lekt! i took out a non-bass non-rock cheap headphone hd555, never thought i could enjoy Metallica using it.

i could live without exenopgomemcpy and 2.22_atom using 2.43-2.48. but there isn't yet a new version that can replace 1.2.

 

 

you are all right. super standard bass an drum sound!

bass is low layer of many sounds, very important, it make many other sounds become so better. thanks for testing.

[lekt]

.....

 

(still waiting for a version that blossoms in vocal. statistically, happens once every 20 versions)

 

i found out good vocal on mqn.exe 2.79 atom (and mqn.exe 2.79 32 bit atom, but less brightness), very mellow, clearly, clarity. 2 these versions have vocal focus as in one point.

there's bass is good. especialy stereo effect is very good (as v1.2), noise seems minimal!! could you test vocal emotion, i think OK.

[sbgk]

i found out good vocal on mqn.exe 2.79 atom (and mqn.exe 2.79 32 bit atom, but less brightness), very mellow, clearly, clarity. 2 these versions have vocal focus as in one point.

there's bass is good. especialy stereo effect is very good (as v1.2), noise seems minimal!! could you test vocal emotion, i think OK.

 

don't understand how you can comment on 1.2 it gives lots of static if played on my dac or is ok on the laptop and the other way round for 2.79, are you actually trying these versions ?

[sbgk]

interesting post no 5 on jplay forum from loworbit which I'll answer here

 

MQN thread - who's using it and how? | Computer Audio | Forum | JPLAY - hi-end audio player for Windows

 

He says it's about as good as an optimised jplay 2 pc set up, sound is very revealing, very dynamic, he then has a moan about it being flaky, can't play unicode filenames, is wav only, one man operation and won't play hirez on his 24 bit dac.

 

This is astounding progress for MQn, something written by a complete novice in a couple of months is as good, in terms of SQ, as a multi award winning leading audiophile player that has taken 2 people many years to develop and MQn still has a number of things to try that will improve SQ.

[nige2000]

interesting post no 5 on jplay forum from loworbit which I'll answer here

 

MQN thread - who's using it and how? | Computer Audio | Forum | JPLAY - hi-end audio player for Windows

 

He says it's about as good as an optimised jplay 2 pc set up, sound is very revealing, very dynamic, he then has a moan about it being flaky, can't play unicode filenames, is wav only, one man operation and won't play hirez on his 24 bit dac.

 

This is astounding progress for MQn, something written by a complete novice in a couple of months is as good, in terms of SQ, as a multi award winning leading audiophile player that has taken 2 people many years to develop and MQn still has a number of things to try that will improve SQ.

 

ya seen that earlier

seemed a bit biased even if you didnt want to convert your flac's(dont think it does any harm anyway)

the following over here is growing .im sure there will be plenty of good reviews,

 

no computer audio ever sounded like this

 

anyway post below loworbits is more honest

[sbgk]

uploaded 2 versions that use movaps instruction, one is an sse4 streaming load/write and the other is a simple loop so sse2 should be able to use. Details in link below.

 

the sse4 movaps version is a bit special

[goon-heaven]

interesting post no 5 on jplay forum from loworbit which I'll answer here

 

Seems like some punters are seeing positive comments but then are being dissappointed

because it doesnt flac or put the cat out.

I suggest perhaps plaster a few dire 'development cycle' health and safety warnings in your blog...

Me? I'm completely mesmerised by what I'm hearing from 16bit thro mqn, (but still working towards getting best server2012 setup for mqn).

[lekt]

uploaded 2 versions that use movaps instruction, one is an sse4 streaming load/write and the other is a simple loop so sse2 should be able to use. Details in link below.

 

the sse4 movaps version is a bit special

 

2.48 intel movaps 1 loop & 2.48 intel movaps:

- difficult to detect their difference, ~

- both is very like as mnq.exe 2.79 atom: clean (very low noise), stereo effect is full (mono is mono, stereo is stereo) => detail (more detail than all other 2.3x-2.4x. 2.3x - 2.4x have noise, but it is not noise, it is echo raised by incorrect stereo effect). instruments sounds have correct position, vocal focus is best.

- bass more flat than 2.79 atom, lesser depth => emotion is lesser.

2.79 atom better. i feel atom versions are good, even many of atom is better, emotion. i don't known why, maybe system power and lesser high technologue (virtualization, sse, AVX,...)? for music playing request very low system resoure, 1%CPU.

[jesuscheung]

2.48 intel movaps 1 loop & 2.48 intel movaps:

- difficult to detect their difference, ~

- both is very like as mnq.exe 2.79 atom: clean (very low noise), stereo effect is full (mono is mono, stereo is stereo) => detail (more detail than all other 2.3x-2.4x. 2.3x - 2.4x have noise, but it is not noise, it is echo raised by incorrect stereo effect). instruments sounds have correct position, vocal focus is best.

- bass more flat than 2.79 atom, lesser depth => emotion is lesser.

2.79 atom better. i feel atom versions are good, even many of atom is better, emotion. i don't known why, maybe system power and lesser high technologue (virtualization, sse, AVX,...)? for music playing request very low system resoure, 1%CPU.

 

 

i am starting to notice that SSE2 and atom versions tend to have better stage layers. not just 2.79, also 2.48 SSE2, atom etc.

 

for example, let say the music has instruments A, B and C. assume A>B>C.

with atom versions it is very easy to detect A is more important than B, and B more important than C.

 

with non-atom versions, C can sometimes be as important as A, which is incorrect, because A>B>C.

 

lekt, is this one reason you like atom?

 

that 2.79 atom does have some good micro detail in vocal and very good stage layers.

[sbgk]

Seems like some punters are seeing positive comments but then are being dissappointed

because it doesnt flac or put the cat out.

I suggest perhaps plaster a few dire 'development cycle' health and safety warnings in your blog...

Me? I'm completely mesmerised by what I'm hearing from 16bit thro mqn, (but still working towards getting best server2012 setup for mqn).

 

thought someone would have had a go at sorting the flac issue out, can't be that complicated, shall try tonight. The issue with flac and MQn is that you are going to have to wait for the decode time before each play which can be painful for hirez.

[sbgk]

2.48 intel movaps 1 loop & 2.48 intel movaps:

- difficult to detect their difference, ~

- both is very like as mnq.exe 2.79 atom: clean (very low noise), stereo effect is full (mono is mono, stereo is stereo) => detail (more detail than all other 2.3x-2.4x. 2.3x - 2.4x have noise, but it is not noise, it is echo raised by incorrect stereo effect). instruments sounds have correct position, vocal focus is best.

- bass more flat than 2.79 atom, lesser depth => emotion is lesser.

2.79 atom better. i feel atom versions are good, even many of atom is better, emotion. i don't known why, maybe system power and lesser high technologue (virtualization, sse, AVX,...)? for music playing request very low system resoure, 1%CPU.

 

thought it was sometimes a bit phasey, have uploaded a 2.48 ntdqdqnt intel version that has the same code as movaps, but uses movdqa.

[jesuscheung]

thought it was sometimes a bit phasey, have uploaded a 2.48 ntdqdqnt intel version that has the same code as movaps, but uses movdqa.

 

2.48 ntdqdqnt intel's stage sounds weird. also interesting. enjoying music differently.

[sbgk]

2.48 ntdqdqnt intel's stage sounds weird.

 

uploaded 2.48 ntdqdqnt r8 which is hopefully better

[jesuscheung]

uploaded 2.48 ntdqdqnt r8 which is hopefully better

 

stage

-every 2.48 has a different stage.

-2.48 ntdqdqnt has an interesting stage.

 

piano

-2.48 movaps and 2.48 ntdqdqnt are very good. movaps is single point focus. ntdqdqnt is two points focus. both have same great tune+crispiness. compare to 2.43 loop, the most balance version for piano, it only bests these two in having correct weight. 2.48 has more weight than norm.

 

bass rock

-1.2 is the man!

[sbgk]

stage

-every 2.48 has a different stage.

-2.48 ntdqdqnt has an interesting stage.

 

piano

-2.48 movaps and 2.48 ntdqdqnt are very good. movaps is single point focus. ntdqdqnt is two points focus. both have same great tune+crispiness. compare to 2.43 loop, the most balance version for piano, it only bests these two in having correct weight. 2.48 has more weight than norm.

 

bass rock

-1.2 is the man!

 

2.48 ntdqdqnt r8 is a bit different to 2.48 ntdqdqnt, did you try it ? Thought 2.48 movaps was a bit phasey, did you find this ?

[jesuscheung]

2.48 ntdqdqnt r8 is a bit different to 2.48 ntdqdqnt, did you try it ? Thought 2.48 movaps was a bit phasey, did you find this ?

 

what is phasey?

 

tried 2.48 movaps & movaps 1 loop on a lot of music. didn't find it apply well on them... just piano was particularly good.

 

instinctively prefer ntdqdqnt over ntdqdqnt r8. i just enjoy it better. they only differ in stage, right? will need critical listening to describe why. been listening 1.2 a lot. 1.2 is currently my norm for stage... hmm...

 

one thing for sure. 2.48 all of them need better layers in stage. as in, if A>B>C, A should always sound most important. C should never be more important than A.

[goon-heaven]

bass rock

-1.2 is the man!

 

Where do I find 1.2 please?

 

 

 

 

...not that I am an ancient rocker or anything