Visual studio 2012 c++ and wasapi minimalist player

[nige2000]

Im running this on haswell

Sometimes im getting quite a bit of distortion and noise on 2496

Havnt got to 2.42 yet

anyone seen this

[Robertas Morkunas]

ah, but your error message said 16 in 24 format not supported, so maybe needs actual conversion, suppose you could do it in sox using the same method as the flac conversion.

 

After 16/44.1 conversion to 24/96 everything works with mqncontrol v5, but not 24/44.1. But converting files is a hassle, though possible.

[Robertas Morkunas]

Im running this on haswell

Sometimes im getting quite a bit of distortion and noise on 2496

Havnt got to 2.42 yet

anyone seen this

 

Nige, aren't you using JK Ciunas? All bit depths and samples rates ok?

[nige2000]

Yes jk ciunas

when theres no distortion its great

Mqncontrol v6 doesnt play 1644 only 2496

[markvandepas]

Hi Mark - just checking one point.

Regedit says that DWORD is for 32 bit and QWORD is for 64 bit.

Is it DWORD to edit if one is using WS2012 64 bit?

 

Thanks

Jonathan

 

From internet search I understand that one can use both.

But using a QWORD (quadruple word length) is only needed when too communicate larger/longer value's than a DWORD value (double word length) can 'contain'.

We want too communicate the value: '1'.

This value fit's inside a DWORD as well as inside a QWORD

So, if I'm correct, both can be used too communicate value: '1'

 

 

Dword or Qword?

[jrling]

From internet search I understand that one can use both.

But using a QWORD (quadruple word length) is only needed when too communicate larger/longer value's than a DWORD value (double word length) can 'contain'.

We want too communicate the value: '1'.

This value fit's inside a DWORD as well as inside a QWORD

So, if I'm correct, both can be used too communicate value: '1'

 

 

Dword or Qword?

 

Many thanks. I now get the point and I am sure you are right.

 

I have done it [DWORD] and the USB Polling tweak and like you think I heard a difference (improvement)!

 

I have to say though that 2.41 is so good already that it is difficult to detect any improvement from these mods, but I am keeping them.

[sbgk]

uploaded 2.42 sse4 intel v2 for comparison with 2.42 sse4 intel, be interested to know what people think

 

finding 2.40/2.41 hard to listen to. have uploaded a 2.41 dqa dqa version which is a bit easier on the ear.

 

are these streaming loads and writes just adding noise ? This version doesn't use them.

 

uploaded a 2.41 dqa dqa 256 which uses all 16 128 bit registers

[sbgk]

After 16/44.1 conversion to 24/96 everything works with mqncontrol v5, but not 24/44.1. But converting files is a hassle, though possible.

 

just need to convert it to 24 bit, v5 is just a 24/96 24 bit test version at the moment

[jesuscheung]

uploaded 2.42 sse4 intel v2 for comparison with 2.42 sse4 intel, be interested to know what people think

 

2.42v2, 2.42

-better than 2.41

-losing the great vocal of 2.38 8 8!! be careful!

-stage/3D... still a lot of work to be done. use hotel california for example. everyone have heard it right? listen the first 10/20secs of guitar, 2.40 to 2.42 not bringing out the soul. instruments not connecting/not integrating properly. surely you have heard better from a CD player setup? 2.38 8 8 sounds much better.

[lekt]

finding 2.40/2.41 hard to listen to. have uploaded a 2.41 dqa dqa version which is a bit easier on the ear.

 

are these streaming loads and writes just adding noise ? This version doesn't use them.

 

uploaded a 2.41 dqa dqa 256 which uses all 16 128 bit registers

 

4 versions 2.41 and 2 versions 2.42:

2.41 movq is better than others, is most properly in sound dimension and weight, according to my own concepts. Then i feel not tired when long playing, more true sound.

other versions bass vibration is not good as 2.41 movq => vocal and piano sounds are slighter, softer, not enough weight (because in vocal and piano sounds need to have bass layer as like as in nature.

 

2.41 movq vs nopgomemcpy: 2.41 movq more clearly/clarity.

But nopgomemcpy have interested nuance: give better drum sound and other non-musical beats, knock,.. more true => drum+bass sound (combined ) is more pretty, have bouncing as bouncing ball, piano sound also have reasonable bouncing and weight.

[Robertas Morkunas]

just need to convert it to 24 bit, v5 is just a 24/96 24 bit test version at the moment

 

ok, understood. Take your time, and hope you'll develop fully working 24 bit version for 44.1/48/96/192. Until then i'll make conversions as i want to listen more to mqn

[Robertas Morkunas]

Yes jk ciunas

when theres no distortion its great

Mqncontrol v6 doesnt play 1644 only 2496

 

what type of distortion it is you hear?

i get few louder or quiter clicks/pops before the begining of song from time to time on rDac..

[jesuscheung]

4 versions 2.41 and 2 versions 2.42:

2.41 movq is better than others, is most properly in sound dimension and weight, according to my own concepts. Then i feel not tired when long playing, more true sound.

other versions bass vibration is not good as 2.41 movq => vocal and piano sounds are slighter, softer, not enough weight (because in vocal and piano sounds need to have bass layer as like as in nature.

 

2.41 movq vs nopgomemcpy: 2.41 movq more clearly/clarity.

But nopgomemcpy have interested nuance: give better drum sound and other non-musical beats, knock,.. more true => drum+bass sound (combined ) is more pretty, have bouncing as bouncing ball, piano sound also have reasonable bouncing and weight.

 

i am very impressed with your ears for analog sound! especially you only use onboard DAC if i am not mistaken? and you dont even tune your machine! sbgk said 2.41 movq sounded bad. so i didn't bother trying it. 2.41 movq is very very very bouncy, which is excellent for piano listening! this version is my new favorite! yesterday i went to buy a piano. most yamaha pianos there didn't sound good(e.g. not bouncy, terrible bass). and there is this 5 grand (US dollars) german made piano has bouncy sound. loved it.

 

other versions of 2.40 to 2.42 has 3D piano effect. this 2.41 movq is unique on its own. totally recommends it for piano. i would say 2.41 movq needs more weight, yes, like you said, reasonable enough.

[jesuscheung]

2.41 movq is seriously pretty damn good. 2.82 reproduces the soul of a 20 grand pianos very well. just lacks many details to justify a piano is 20 grand. 2.41 makes a 20 grand pianos sounds at least like a 10 grand piano, yes this is a compliment! most players on earth make 20 grand pianos sound like an electronic keyboard.

[lekt]

i am very impressed with your ears for analog sound! especially you only use onboard DAC if i am not mistaken? and you dont even tune your machine! sbgk said 2.41 movq sounded bad. so i didn't bother trying it. 2.41 movq is very very very bouncy, which is excellent for piano listening! this version is my new favorite! yesterday i went to buy a piano. most yamaha pianos there didn't sound good(e.g. not bouncy, terrible bass). and there is this 5 grand (US dollars) german made piano has bouncy sound. loved it.

 

other versions of 2.40 to 2.42 has 3D piano effect. this 2.41 movq is unique on its own. totally recommends it for piano. i would say 2.41 movq needs more weight, yes, like you said, reasonable enough.

 

have a nice day with new piano! jesuscheung.

i have only laptop onboard DAC. versions 2.x for USB dac, then on my system sound is crunch, clicks/pops (2.xx not compatible with my soundcard and HDMI). but okay, I'll try more test with piano sound. i see 2.41 movq is correctly in many criteria.

[sbgk]

2.42v2, 2.42

-better than 2.41

-losing the great vocal of 2.38 8 8!! be careful!

-stage/3D... still a lot of work to be done. use hotel california for example. everyone have heard it right? listen the first 10/20secs of guitar, 2.40 to 2.42 not bringing out the soul. instruments not connecting/not integrating properly. surely you have heard better from a CD player setup? 2.38 8 8 sounds much better.

 

I have it on 24/192 and it sounds ok using 2.40 sse4 intel which I find to be the most tuneful version

 

I don't understand how you and Lekt find the versions eg nopgomemcpy, 2.41 movq any good as they are the noisiest poorest sounding ones.

[lekt]

I have it on 24/192 and it sounds ok using 2.40 sse4 intel which I find to be the most tuneful version

 

I don't understand how you and Lekt find the versions eg nopgomemcpy, 2.41 movq any good as they are the noisiest poorest sounding ones.

 

I don't understand, you said "...have it on 24/192...", it mean DAC make resampling to 24/192 ?

my decice setting: "2 channel, 24 bit, 44100 Hz (Studio Quality)" or "2 channel, 16 bit, 44100 Hz (Studio Quality)" when i test 16/44100. not use resample for testing.

[sbgk]

I don't understand, you said "...have it on 24/192...", it mean DAC make resampling to 24/192 ?

my decice setting: "2 channel, 24 bit, 44100 Hz (Studio Quality)" or "2 channel, 16 bit, 44100 Hz (Studio Quality)" when i test 16/44100. not use resample for testing.

 

the wav file is 24/192 from HDTracks, MQn renders it as 24 in 32 / 192 and the dac plays it as 24/192

[Clive]

I've managed to confuse myself....which is not difficult. I've tried playing a 24/96 file using mqncontrol.exe 2496 1644 24 bit v6 with 2.40 sse4

 

This didn't work and it took me a while to get 16/44.1 working again (reboot, reset files, reset dac). What combinations of mqn do I need to play 16/44, 24/96 and 24/192?

 

I'm using a Metrum Octave mkII via the Young driver.

[Clive]

BTW running 2.40 sse4 I find the treble a trace hot and this is using a NOS dac. The sound overall very clear and controlled with a lot happening all around the soundstage.

[lekt]

I have it on 24/192 and it sounds ok using 2.40 sse4 intel which I find to be the most tuneful version

 

I don't understand how you and Lekt find the versions eg nopgomemcpy, 2.41 movq any good as they are the noisiest poorest sounding ones.

 

i test on headphones and on speakers (always on both equipments). by headphones can feel micro-differences, almost soundstage, total weight of sound. last versions have now become very good, then very hard to find out the difference and easy mistake to make comments. i will try 2.40 sse4 intel once again. its soundstage is acceptable or not? i think everything must be properly, correctly, in own technical norm. then sound will come true, automaticaly. from which the listener can adjust according to own preferences, but only on the amplifier. i think that.

[sbgk]

i test on headphones and on speakers (always on both equipments). by headphones can feel micro-differences, almost soundstage, total weight of sound. last versions have now become very good, then very hard to find out the difference and easy mistake to make comments. i will try 2.40 sse4 intel once again. its soundstage is acceptable or not? i think everything must be properly, correctly, in own technical norm. then sound will come true, automaticaly. from which the listener can adjust according to own preferences, but only on the amplifier. i think that.

 

Which headphones do you use, I just have a cheap pair and can't really tell if a version is any good until I hear it on the main system. It's not guaranteed that MQn will ever sound technically perfect, but it's proved that there are different ways of doing things.

[jesuscheung]

I have it on 24/192 and it sounds ok using 2.40 sse4 intel which I find to be the most tuneful version

 

I don't understand how you and Lekt find the versions eg nopgomemcpy, 2.41 movq any good as they are the noisiest poorest sounding ones.

 

... don't get discouraged by the noise. listen to the vibration, bouncing sound- it adds realism.

[nige2000]

I have it on 24/192 and it sounds ok using 2.40 sse4 intel which I find to be the most tuneful version

 

I don't understand how you and Lekt find the versions eg nopgomemcpy, 2.41 movq any good as they are the noisiest poorest sounding ones.

 

thats my findings as well,

maybe we need the magic bios settings?

[lekt]

Which headphones do you use, I just have a cheap pair and can't really tell if a version is any good until I hear it on the main system. It's not guaranteed that MQn will ever sound technically perfect, but it's proved that there are different ways of doing things.

 

i also use cheap headphones, but very cheap (Sennheiser MX400, China, 10$). it is earbuds, not in-ear headphones. earbuds better, less sound effect. for listening need a good headphones, but for testing simple headphones maybe enough, because good headphones usually can make sound better by many tricks. for me, with simple headphones easily detect the difference between players sounds.

i think you are extra high level developer (assembly programming), but i have some thinks.

MQn = minimalist shell + minimalist core, create by assembly programming. Excelent.

But shell difficult to use, or cause errors and too less functional. could you solve problem with "Long file name/patch with unicode text" and some other tasks for user interface in shell, although it is minimalist. I like JEP_Shell, it is reasonable minimalist user interface, doing very little work should not affect the core. sorry, I was probably wrong.