Support for exclusive mode is needed in both applications and drivers.
The Windows sound system model changed after Windows XP. More or less, starting with Windows 7 and above, exclusive mode is supported by Windows under Direct Sound and WASAPI driver models, with WASAPI being the latest model.
The idea with shared mode is that multiple applications can access the sound card simultaneously, and Windows will do SRC automatically as needed to make it possible. Under that model, its not possible to let simultaneous applications keep changing real sound card settings in potentially conflicting ways, so any changes set in applications are virtual, not real. The real sound card settings can only be set from control panel -> advanced by a human administrator.
EDIT: For the sake of completeness, in exclusive mode, applications can control the real sound cards settings, and also take exclusive control of the sound card, thus preventing simultaneous access by other applications. Whether or not these things are allowed depends on the checkbox settings for exclusive mode under control panel -> advanced.
Last edited:
This is where things deviate from what they are supposed to do. The exclusive mode still makes no difference, apparently Audacity does not demand 24 bit samples so if I set the shared mode to 16 bits I get 16 bits no matter what. I can't control how the developers want their application to behave you just have to suss it out.
"Can" being the operative word.
Last edited:
It sounds like Audacity may be operating in shared mode instead of exclusive mode. If attempts to connect via exclusive mode fail, it may default to shared mode. Just sayin'.
RNM,
I would suggest patience. Actually without the usual noise there are very useful tidbits being bandied about, not just on implementation but actually buried in there are some really important basic bits.
You might even see these bits implemented in gear such as AP in a few years.
ES
I would suggest patience. Actually without the usual noise there are very useful tidbits being bandied about, not just on implementation but actually buried in there are some really important basic bits.
You might even see these bits implemented in gear such as AP in a few years.
ES
Don't quite get your point, I have no control over how the drivers and software were written. Things don't behave according to the Microsoft developers guidelines what else is new. If something does not work it does not work, the user can't fix it no matter what the failure mode. AFAIK Audacity could care less about the resolution and takes what it gets in any case, the developers have already stated that 24 bits can't matter to anyone.
WASAPI vs. Apple?
I know most are doing this stuff on window's 'puters
but how is apple's implementation different?
Why are they (apple mac computers) noted by many
people as having much better sound AND quality
audio than the win machines?
closed system vs. Microsoft and everybody making the same
or similar boxes Vs Apple doing it exclusively?
If I knew, I wouldn't ask.
Holy Smokes it feels like and ongoing small earthquake here in
North Texas and I don't know why.
I know most are doing this stuff on window's 'puters
but how is apple's implementation different?
Why are they (apple mac computers) noted by many
people as having much better sound AND quality
audio than the win machines?
closed system vs. Microsoft and everybody making the same
or similar boxes Vs Apple doing it exclusively?
If I knew, I wouldn't ask.
Holy Smokes it feels like and ongoing small earthquake here in
North Texas and I don't know why.
My point was that if changing Windows settings for shared mode changes how many bits you see, the implication of that is that the system is probably operating in shared mode, even if one has selected to use exclusive mode in one's chosen application. That would suggest to me that there is a problem in the application, although there may be problems elsewhere too. If people like you who find problems don't contact application developers to complain, it may be reasonable for them to assume there is no problem. After all, how many people do you think actually check like you did? Probably not many. And of those, how many call to complain and ask for a fix? Maybe not many.
Sound on Mac computers at best, isn't any better than on Windows computers, at best. Apple's support for AAC over mp3 is better. Also, it may be that the poorest Windows sound quality is worse than the poorest Mac sound quality, as Apple does have control over any hardware they choose to use.
There has to be threading in Macs, and there is: https://www.google.com/search?q=obj...eid=chrome&ie=UTF-8#q=objective+C+threading&*
I was wondering what is the fundamental amplitude, can't tell without knowing the bin size. Otherwise, THD of -123dB @1KHz is nothing to call home about, at least for $2500. Mr. Marsh could get two Panasonic 722's or a couple of Shibasoku for that kind of money. Those will have all the bells and whistles of an instrument, from autoranging to GPIB.
But I understand, sound cards are so much more fashionable.
There is but a not so subtle difference between objective C and C++.
Each object must complete in specified amount of time. This is not so in Windows.
What happens to win audio if you open 100 apps. Wouldn't work to well. On a Mac it might.
Lots a tidbits on the net about this.
This couldn't be more of topic. For further discussion we start another thread.
Actually tying a modest sound card to the monitor output of the analyzer is simple and gives lots of info. Getting a sense of calibration is not simple but taking advantage of the analog processing and gain from the analyzer enables looking much deeper.
You don't need the bin size, there is a red line of text at the bottom of the graph, indicating "RMS = -7.2dBV"
So the magnitude of the fundamental is around -7.2dBV. I included the word "around", to count for the noise and the harmonics, that are mostly negligible here.
BTW, the unit was a 4-year old used one when I got it, so I paid much less than the $2500 you mentioned, primary use as a DAC to listen to music, but since it can also do measurement so why not 🙂
Last edited:
Got Portaudio working with ASIO support now. Moving up the curve.
Scott I don't anything about Python let alone PYaudio.
You'll have to help me that part.
Scott I don't anything about Python let alone PYaudio.
You'll have to help me that part.
I thought this was pretty good:
Note: Because of their underlying kernel support, operation objects can often create threads more quickly. Rather than creating threads from scratch every time, they use pools of threads already residing in the kernel to save on allocation time.
From the Apple Developer, Guides and Sample Code, Threading Programming Guide, Thread Management,
All those threads in a pool moving like the grass in a tidal pool, waving, looking for
something to do, some process to be called, or I/O to perform. Waiting, Wanting.
The grass becomes tentacles reaching into everything that is linked, Wi-Fi-ed,
networked for only "it's" benefit. Well call ORWELL "lite" because
that it what it was compared to the new reality.
Wow, we're all going to be screwed when AI finally takes root.
We'll all end up being F***ed when it becomes self aware...
and we were worried about N. Korea. We ain't seen nothing yet.
We'll be made to serve the machine, the de-facto Omnipotent god.
It's probably already there in the Utah desert, they just aren't
telling anyone.
Cheers,
Note: Because of their underlying kernel support, operation objects can often create threads more quickly. Rather than creating threads from scratch every time, they use pools of threads already residing in the kernel to save on allocation time.
From the Apple Developer, Guides and Sample Code, Threading Programming Guide, Thread Management,
All those threads in a pool moving like the grass in a tidal pool, waving, looking for
something to do, some process to be called, or I/O to perform. Waiting, Wanting.
The grass becomes tentacles reaching into everything that is linked, Wi-Fi-ed,
networked for only "it's" benefit. Well call ORWELL "lite" because
that it what it was compared to the new reality.
Wow, we're all going to be screwed when AI finally takes root.
We'll all end up being F***ed when it becomes self aware...
and we were worried about N. Korea. We ain't seen nothing yet.
We'll be made to serve the machine, the de-facto Omnipotent god.
It's probably already there in the Utah desert, they just aren't
telling anyone.
Cheers,
Last edited:
To implement WASAPI & Exclusive mode (to get 100.0% bit perfect) on Windows is some additional challenge/effort 😀 and is not done as a snap 😱
Also the application needs a multi threading lift and proper done to use SMP (Symmetric Multi Processing). I did all of them (ASIO, WASAPI Exclusive & SMP) and may some stopped regarding the ROI question. While it gets a xxxx hours question.
Keep in mind on Windows 384kHz is the MS limitation where on ASIO things are only limited to the DWord, while DSD native deals with such xx MHz sample rates.
On MAC, the HW developer has to send the driver to Apple for validation and gets added. Now on Windows 10, the HW developer has to send it for validation too to MS.
Cheers and happy implementing
Hp