Jump to content
IGNORED

Visual studio 2012 c++ and wasapi minimalist player


sbgk

Recommended Posts

no, put the pcm in memory and then just need a chip to pass the data between memory and device

 

I think this is in essence what Chord do with the long buffer settings on their highend DACs. And it seems it should be what happens in integrated streamers like the Naim.

 

There are some very good digital replay systems out there, but translating that to what is a general purpose computing platform like the windows pc brings you up against all the other stuff Microsoft believe needs to be happening all the time.

 

And to go back to your earlier question, sbgk, (did you try tot align ? it's a bit more easy flowing than 2.59 8 ) yes I did try tot align. It was my previous favourite, for several hours until I got into 2.59 8 8. I will go back and try tot align again just for good measure, but the 2.59 8 8 version seems like a big step forward. It plays tunes, it brings out immense levels of detail and it has a solid, robust quality perhaps missing in some of the earlier iterations. If I had to stick with 2.59 8 8 on an un-optimised W8 pc from here on, I'd not complain. I'd be too busy playing tunes.

 

A big thanks.

 

Mark

Link to comment

Like others here, I have following the development over time. I just have to chime in my thanks - I've had some of the best sounds ever here today. Really incredible work with 2.59.

 

Here's the thing. As this just gets better and better, we are all going to start wanting more features..... eg for me, flac playback would be great (tried all the suggestions back up this thread and couldn't make it work). These things will seem like demands - and in reality will be demands on your time. I really think you should consider putting a paypal button on your site. That way you will get a monetary reward for your time, and we won't feel so bad about asking for hires or flac or whatever.

 

For your own sake, and for ours!!!

 

Once again, fantastic work, many thanks!

 

Fran

Link to comment
..... eg for me, flac playback would be great (tried all the suggestions back up this thread and couldn't make it work).

 

Did you try my Autohotkey script posted earlier in this thread? It does not use Sox.exe but Flac.exe. It also enables you to start MQn from within Foobar or Jriver : just select the track(s) and press the hotkey. Only stopping MQn still requires typing X in the MQn window. If you have problems with it : send me a PM

Check my profile for my audiosystem.

Link to comment
Did you try my Autohotkey script posted earlier in this thread? It does not use Sox.exe but Flac.exe. It also enables you to start MQn from within Foobar or Jriver : just select the track(s) and press the hotkey. Only stopping MQn still requires typing X in the MQn window. If you have problems with it : send me a PM

 

Hi, haven't tried it yet, but seems promising. plan on mqncontrol to stop everything after playback ends. have no plans for flac etc conversion inside mqncontrol.

 

to stop before then you can create a batch file to stop control & play

 

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

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

There is no harm in doubt and skepticism, for it is through these that new discoveries are made. Richard P Feynman

 

http://mqnplayer.blogspot.co.uk/

Link to comment
BTW, I just thought I'd add in here that fideliser doesn't play nice with 2.59 - just in case anyone is using it and having issues with 2.59 not running.

 

 

Fran

 

any details ? I know the system timer setting to 0.5ms affects the sound of MQn

There is no harm in doubt and skepticism, for it is through these that new discoveries are made. Richard P Feynman

 

http://mqnplayer.blogspot.co.uk/

Link to comment

Sorry - no details. I had been using the first 2.59 file you loaded on drive this morning - and that worked fine. I read about the new versions (8 4 and 8 8) and thought I'd try 8 8. Loaded it up, wouldn't play - command window flashed for a sec and no play. Then I did a restart and it did work - but stopped once fideliser kicked in (I had fideliser on a scheduled start 30 secs after boot up). It was only then I realised what was going on. So just to be clear - this wasn't a SQ issue, it was a no-play issue.

 

Fran

Link to comment

sbgk,

 

Thanks a bunch for an incredible music player. I have a much larger soundstage, and just an amazing amount of micro detail. It is just amazing that software is able to do this. Some of your versions give a real sense to room cues, while others as has been mentioned you can hear the skin of the drums. Very nice.

Link to comment
I think this is in essence what Chord do with the long buffer settings on their highend DACs. And it seems it should be what happens in integrated streamers like the Naim.

 

There are some very good digital replay systems out there, but translating that to what is a general purpose computing platform like the windows pc brings you up against all the other stuff Microsoft believe needs to be happening all the time.

 

And to go back to your earlier question, sbgk, (did you try tot align ? it's a bit more easy flowing than 2.59 8 ) yes I did try tot align. It was my previous favourite, for several hours until I got into 2.59 8 8. I will go back and try tot align again just for good measure, but the 2.59 8 8 version seems like a big step forward. It plays tunes, it brings out immense levels of detail and it has a solid, robust quality perhaps missing in some of the earlier iterations. If I had to stick with 2.59 8 8 on an un-optimised W8 pc from here on, I'd not complain. I'd be too busy playing tunes.

 

A big thanks.

 

Mark

 

Hi - I have a Chord DAC 64 MkII with the 4 second buffer ability which I always use and it sounds great.

 

But I think SBGK was going a lot further than that. The Chord is receiving S/PDIF encoded data in a stream originating in a PC (bad) and in my case via a USB S/PDIF converter (more bad stuff) and then buffering and reclocking 4 seconds worth of the stream (more potential for bad stuff).

 

What I have always thought would be miles better is doing away with the PC except for using it to select the Playlist and just loading a whole playlist into memory embedded in the DAC over say Ethernet. Simple file transfer over Ethernet of WAV files in other words. The whole WAV file(s) is then 'played' as SBGK suggests from the onboard memory via a controller chip in the DAC and directly connected to the DAC chip.

 

I have read the likes of John Swenson think it's doable and actually not that technically challenging or requiring expensive parts.

 

Happy to be told this is all rubbish! But if not, it would be great to move the idea forward in a DIY way?

Link to comment
this donate button is harder than you think - paypal it's for non profits and google had a support payment option but that's finishing, might do a charity just giving option instead.

 

really nothing easy is there,

could you setup a new email account and then a paypal account we can just send it to?

 

question

i know nothing about the coding your doing for this but when you improve 16 44 in say 2.59 are you improving all the other 24 bit formats too or just the 16 bit

Link to comment
really nothing easy is there,

could you setup a new email account and then a paypal account we can just send it to?

 

question

i know nothing about the coding your doing for this but when you improve 16 44 in say 2.59 are you improving all the other 24 bit formats too or just the 16 bit

 

Yes, they are all improved as they use the same optimised memcpy, the MMCSS pro audio time setting should ideally be changed for each format, which may be why you hear less improvement. 16/44.1 transfers 2048 samples in 8192 bytes and the 24 bit 1024 samples in 8192 bytes, it might be that it would be better to have 24 bit transfer 2048 samples at a time, but haven't investigated it, but larger chunks of data have their own problems.

There is no harm in doubt and skepticism, for it is through these that new discoveries are made. Richard P Feynman

 

http://mqnplayer.blogspot.co.uk/

Link to comment
Yes, they are all improved as they use the same optimised memcpy, the MMCSS pro audio time setting should ideally be changed for each format, which may be why you hear less improvement. 16/44.1 transfers 2048 samples in 8192 bytes and the 24 bit 1024 samples in 8192 bytes, it might be that it would be better to have 24 bit transfer 2048 samples at a time, but haven't investigated it, but larger chunks of data have their own problems.

 

with other players hires was a good bit better than 16bit,

with MQn 16bit sounds much much better than hires did on the other players

hires on MQn isnt as proportionately better than 16bit on other players

i suppose i was expecting a larger gap in SQ

 

im surprised we havent come across more 24 in 32 dacs at this stage

Link to comment
with other players hires was a good bit better than 16bit,

with MQn 16bit sounds much much better than hires did on the other players

hires on MQn isnt as proportionately better than 16bit on other players

i suppose i was expecting a larger gap in SQ

 

im surprised we havent come across more 24 in 32 dacs at this stage

 

well, like I said the mmcss setting may have a role to play, smaller things have bigger impact, I thought you weren't using your best dac with hirez, maybe it's not optimal, also the 24/96 24 bit version used an unoptimised loop and memcpy. When I've tried hirez I've been impressed with it.

There is no harm in doubt and skepticism, for it is through these that new discoveries are made. Richard P Feynman

 

http://mqnplayer.blogspot.co.uk/

Link to comment
I know that, it can. I played it so many times by JRiver and Foobar.

 

right now for MQn, playing 24/96 requires a correct version of mqncontrol. this depends on your DAC. my DAC uses 24bits in 24 container. in my case, i use

 

mqncontrol.exe 2496 24 bit v5

mqncontrol.exe 2496 1644 24 bit v6

 

you can run wasapi_test to check the type of container needed to play 24/96.

 

by the way, this is my first time playing 24/96. sounds very good.

 

sbgk, just double checking, even with 24/96, the buffer size is still 1024? coz i am trying to test the clock rate in registry.

 

1024*3/96000 = 320000. no rounding!! exactly!

this is the perfect clockrate mathematically.

Link to comment
right now for MQn, playing 24/96 requires a correct version of mqncontrol. this depends on your DAC. my DAC uses 24bits in 24 container. in my case, i use

 

mqncontrol.exe 2496 24 bit v5

mqncontrol.exe 2496 1644 24 bit v6

 

you can run wasapi_test to check the type of container needed to play 24/96.

 

by the way, this is my first time playing 24/96. sounds very good.

 

sbgk, just double checking, even with 24/96, the buffer size is still 1024? coz i am trying to test the clock rate in registry.

 

1024*3/96000 = 320000. no rounding!! exactly!

this is the perfect clockrate mathematically.

I am loosing it...sorry. Do you pass 2496 1644 24 bit v6 as parameter in .bat file or you pointing to the version that I should find somewhere. I am a bit of confused here.

Link to comment

the regular MQn play(e.g 2.59) and control files can play hires with dacs that accept 24bit in a 32 container

thing is these dacs seem to be uncommon, i have one Sbgk has one and i think Pearse has one

most of the dacs seem to play 24 in 24, my main dac does all but 24 in 24

but im experimenting with a arcam rpac which is 24 in 32 capable

 

Sbgk were you saying that doing a 24 in 24 might loose a bit of SQ as well as been a pita?

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...