Visual studio 2012 c++ and wasapi minimalist player

[nige2000]

i5-2500T, 8Gb, umpteen USB sticks with different shades of Server2012 & optimisation to suit different players.

RME HDSPe soundcard locked to EmmLabs DCC-2SE.

Switch mode PS - umpteen Maplin PSs waiting for me to come out trance.

What have I missed?

 

was more thinking about what im missing,

did you get yourself a couple of maplins?

they might not look the business in anyones equipment signature but they sound real good at the price

 

ive got a couple of lifepo4 12 v batteries on route and one of them 16-24v pico's that pauls modding as an experiment

i think the batteries will make a difference but not sure on the pico but its not expensive at $30 odd worth a try

would be interested in a better dc to dc converter than the pico if such things exist.

 

bty i havent listened since this morning but i got the impression 2.48 ntdqdqnt r8 might have slightly more vocal detail and or just detail than 2.49 no buff which is still my fav

[lekt]

2.49 no buff definitely adds real vocal emotion. it is no false effect due to brightness. but 2.49 no buff lacks vocal fullness that 2.49 rax has.

 

2.49 sse2 intel is very interesting. it is not delicate as eax, rax, no buff. its sound is "rustic", but so much clarity and emotion, and certainly, healthy, and so warm. there everything is strong, power (sound is a bit more bright). its vocal have more emotion than other versions. i like it. try to listen it more, jesuscheung.

[jrling]

Nige - you not only have to get up early to keep up with SBGK but go to bed late too by the look of the timing of some of his postings!! What energy and enthusiasm.

 

Sorry if this is covered elsewhere but does anyone have a WaveIO USB/SPDIF converter and Thesycon v 1.67 driver and got 24/96 WAV files working with MQn?

 

I am amazed at the detail of SQ from 16/44.1 but would love to try 24/96 on MQn.

 

Jonathan

[realmadrid]

think the 1.2 version sounds ok, but slightly sibilant as Nige said not in the same league as 2.48/9.

 

If so, give me the 1.48/ version then...

I can´t play the 2 versions for now

 

Tanks

 

Juan

 

Juan

[sbgk]

If so, give me the 1.48/ version then...

I can´t play the 2 versions for now

 

Tanks

 

Juan

 

Juan

 

Juan Juan, you'll have to wait until the weekend. People love free stuff, give me mo yo free stuff.

[sbgk]

Uploaded mqnplay.exe 2.49 tlb cache r8 & mqnplay.exe 2.49 mfc stat tlb cache, primes the tlb and only puts data in l2 cache in first loop. MFC uses MFC static library setting.

 

Predictably the MFC version is bass heavy, but otherwise sounds terrible.

 

The mqnplay.exe 2.49 tlb cache r8 is the best version I've heard.

[goon-heaven]

why would you do that ?

 

The reason it is quicker is that the data is in cache after the first play, so if you want to impress someone then load the albums into cache first, the only way at the moment is to stop playing after the data is loaded.

That is what I do now.. just rather inelegant, and I dont know how bad the mess left for the wet team.

 

When you say starting MQncontrol directly are you typing it into a dos prompt, not sure what you mean.

Yes. Type in mqncontrol --file "Files.bat"

 

There is a utility called MQnplay.bat which allow different versions of MQn to be played, a good trick is to stick the same version on twice and see if you can spot that it is the same version, very difficult because the mind wanders etc, but over time it is possible to identify different versions.

Where do I find that?

[goon-heaven]

was more thinking about what im missing,

did you get yourself a couple of maplins?

Yes, as per your recommendations. Nick very kindly arranged a batch of 4 to be sent out to darkest Brittany. 1 for CPU, 1 for PICO, 1 for the HDSPe, and 1 for PCIe network card.

Then I'll try some batteries and mole-grips.

 

ive got a couple of lifepo4 12 v batteries on route and one of them 16-24v pico's that pauls modding as an experiment. i think the batteries will make a difference but not sure on the pico but its not expensive at $30 odd worth a try would be interested in a better dc to dc converter than the pico if such things exist.

I'm very interesting in learning your impressions. You're way in front on this. Everything here just ends up in a bit heap waiting for my time.

 

What drives do you have your OS and music on?

[goon-heaven]

Uploaded mqnplay.exe 2.49 tlb cache r8 & mqnplay.exe 2.49 mfc stat tlb cache, primes the tlb and only puts data in l2 cache in first loop. MFC uses MFC static library setting.

 

Predictably the MFC version is bass heavy, but otherwise sounds terrible.

 

The mqnplay.exe 2.49 tlb cache r8 is the best version I've heard.

Wow, thanks. You are releasing far quicker I than can listen..

[nige2000]

Yes, as per your recommendations. Nick very kindly arranged a batch of 4 to be sent out to darkest Brittany. 1 for CPU, 1 for PICO, 1 for the HDSPe, and 1 for PCIe network card.

Then I'll try some batteries and mole-grips.

 

 

I'm very interesting in learning your impressions. You're way in front on this. Everything here just ends up in a bit heap waiting for my time.

 

What drives do you have your OS and music on?

 

the mole grips are famous

sometimes i think im a sort of comic relief

 

I should have got some shares in maplin, bought a few in there myself

 

was the question where do i have my music or where should it be?

as thats two different things

should be on usb3 flash powered seperately or sata ssd powered seperately

id like to try nas storage via meicord but i have neither

[sbgk]

That is what I do now.. just rather inelegant, and I dont know how bad the mess left for the wet team.

 

 

Yes. Type in mqncontrol --file "Files.bat"

 

 

Where do I find that?

 

Uploaded it as mqntest.bat in the test folder

 

it is made up of the following blocks of code, just change the

 

copy "mqnplay.exe 2.49 tlb cache r8" mqnplay.exe line to the version/s you want to test, it then loops round. You need to select a track and start/stop it using normal mqn.bat so there is something in files.txt, 1 track at a time works best.

 

taskkill /fi "PID gt 0" /IM mqnplay.exe /F >nul

taskkill /fi "PID gt 0" /IM mqncontrol.exe /F >nul

 

 

del mqnplay.exe

 

 

copy "mqnplay.exe 2.49 tlb cache r8" mqnplay.exe

 

 

start MQncontrol --file ".\Files.txt"

 

 

powershell -command "Start-Sleep -s 80"

[sbgk]

the mole grips are famous

sometimes i think im a sort of comic relief

 

I should have got some shares in maplin, bought a few in there myself

 

was the question where do i have my music or where should it be?

as thats two different things

should be on usb3 flash powered seperately or sata ssd powered seperately

id like to try nas storage via meicord but i have neither

 

I have a 5m meicord from my squeezebox days, make me an offer.

[sbgk]

Wow, thanks. You are releasing far quicker I than can listen..

 

best thing is to listen to the latest version, wish I'd done that - would have saved a lot of time.

[sbgk]

Uploaded mqnplay.exe 2.49 tlb cache r8 & mqnplay.exe 2.49 mfc stat tlb cache, primes the tlb and only puts data in l2 cache in first loop. MFC uses MFC static library setting.

 

Predictably the MFC version is bass heavy, but otherwise sounds terrible.

 

The mqnplay.exe 2.49 tlb cache r8 is the best version I've heard.

 

this version just needs sse2. Am now down to hearing differences in the sound caused by how many parameters are passed to the memcpy function, struth. more = worse.

[nige2000]

I have a 5m meicord from my squeezebox days, make me an offer.

 

your the reason i dont have one!!!

 

think i have a new fav too

[goon-heaven]

the mole grips are famous

sometimes i think im a sort of comic relief

 

I should have got some shares in maplin, bought a few in there myself

 

was the question where do i have my music or where should it be?

as thats two different things

should be on usb3 flash powered seperately or sata ssd powered seperately

id like to try nas storage via meicord but i have neither

I'll have to try separating OS and music, but didnt like what I heard on SATA. Perhaps I'll retry with a linear.

 

I have NAS and meicord and giso on the heap ever since I heard the magic of mqn and unplugged the ethernet from the Audio PC.

I plan to chop the meicord in 2 and add XLR plug and 25 pin D-sub on cut ends. I want to tryout giso on the soundcard-DAC AES/EBU link. Another project waiting patiently for my time.

[goon-heaven]

Uploaded it as mqntest.bat in the test folder

Thanks - that appears straightfwd, if a little weird to keep killing mqn off - i'm surprised the device buffers dont get their knickers in a twist.

[nige2000]

your the reason i dont have one!!!

 

think i have a new fav too

 

spoke too soon

bass better in 2.49 no buff compared to the 2.49 tlb cache r8

this is an issue for bass reliant tracks, rock and metal have no emotion

[goon-heaven]

this version just needs sse2. Am now down to hearing differences in the sound caused by how many parameters are passed to the memcpy function, struth. more = worse.

Thats the trouble with these bureaucratic functions - trying to appear busy shuffling bytes from somewhere to else.

 

Divide and conquer in assembler?

[sbgk]

spoke too soon

bass better in 2.49 no buff compared to the 2.49 tlb cache r8

this is an issue for bass reliant tracks, rock and metal have no emotion

 

uploaded 2.51 sse4 intel, think I prefer it to the others, maybe prefer it a bit warmer sounding, but it's ok to listen to. The problem with 2.49 tlb cache r8 was I was using a sub instruction which has always caused problems so had to do a negative add instead, in the end I went back to 1 loop as it's a lot simpler.

 

there's definitely a cycle of thinking a change is the best thing ever then after a while realising it could be improved further.

[sbgk]

Thanks - that appears straightfwd, if a little weird to keep killing mqn off - i'm surprised the device buffers dont get their knickers in a twist.

 

It's the only way to change versions, trust the device drivers have exception handling, never had a problem so far.

[jesuscheung]

Uploaded mqnplay.exe 2.49 tlb cache r8 & mqnplay.exe 2.49 mfc stat tlb cache, primes the tlb and only puts data in l2 cache in first loop. MFC uses MFC static library setting.

 

Predictably the MFC version is bass heavy, but otherwise sounds terrible....

 

2.49 mfc stat tlb cache

it just has a bass tune. it is far from being having a full and complete bass like 1.2.

[jesuscheung]

uploaded 2.51 sse4 intel, think I prefer it to the others, maybe prefer it a bit warmer sounding, but it's ok to listen to. The problem with 2.49 tlb cache r8 was I was using a sub instruction which has always caused problems so had to do a negative add instead, in the end I went back to 1 loop as it's a lot simpler.

 

there's definitely a cycle of thinking a change is the best thing ever then after a while realising it could be improved further.

 

2.51 has improvements over 2.49. focus got better due to better volume.

 

vocal emotion

2.49 buf>2.51. with 2.49 buf, at the point of emotion, it is stronger and last longer. 2.51 runs out of emotion too quickly.

 

bass

it is actually pretty great. just bothers me i instinctively prefer 1.2 for bass. pretty sure it is the same problem like emotion. when someone hits the drum, the strongest point of impact last longer with 1.2. even though 1.2 has less weight less volume, but the impact lasts longer, that should be why i prefer 1.2. with 2.51, the impact has good weight and volume, but the impact runs out too quickly.

 

2.51 just like every other new version has improved clarity.

 

no major difference in stage or volume or most things with 2.51 and 2.51 eax. think 2.51 is more free.

[lekt]

2.49 tlb cache r8 & 2.49 mfc stat tlb cache not good as other 2.49 sse4, except 2.49 tlb cache r8 vocal emotion better than others.

 

2.51 sse4 intel eax vs 2.49 no buff:

-soundstage: v1.2 >~ 2.51 sse4 intel eax > 2.49 no buff. can accept.

-drum/bass volume/weight lesser, i think in norm. bass vibration better (micro vibration), more soft but enough power => better

-mid/treble volume larger, more brightness => more energy?, but i think inconsistent with mellow bass (like as 2.22_atom)

-piano: micro-vibration better, pretty

difficult to conclude which one better.

 

v1.2 vs 2.49/2.51:

-brightness lesser, i think below norm in MQn case

-soundstage better (wider but vocal focus still true!). streo effect very good.

-bass have multi-layer (very good) => deeper. vibration OK, diffcult to compare because not in same brightness. i feel micro-vibration lesser a bit.

-vocal < no buff

 

for me no buff better than all other MQn.

[sbgk]

2.49 tlb cache r8 & 2.49 mfc stat tlb cache not good as other 2.49 sse4, except 2.49 tlb cache r8 vocal emotion better than others.

 

2.51 sse4 intel eax vs 2.49 no buff:

-soundstage: v1.2 >~ 2.51 sse4 intel eax > 2.49 no buff. can accept.

-drum/bass volume/weight lesser, i think in norm. bass vibration better (micro vibration), more soft but enough power => better

-mid/treble volume larger, more brightness => more energy?, but i think inconsistent with mellow bass (like as 2.22_atom)

-piano: micro-vibration better, pretty

difficult to conclude which one better.

 

v1.2 vs 2.49/2.51:

-brightness lesser, i think below norm in MQn case

-soundstage better (wider but vocal focus still true!). streo effect very good.

-bass have multi-layer (very good) => deeper. vibration OK, diffcult to compare because not in same brightness. i feel micro-vibration lesser a bit.

-vocal < no buff

 

for me no buff better than all other MQn.

 

Uploaded 2.51 sse4 intel rax no zi, zi debug info setting was adding a touch of sibilance