Sulphur on 17/2/2017 at 04:44
Ah, WASAPI. It gives so much and takes away so much. I pipe WASAPI out to the receiver through foobar2K, and of course that means it takes exclusive control of the audio pipe, so every time I find a random youtube video without sound, I need to go check if I've paused a random song on fb2k first.
As for games, I'm sure the final sounds are designed with the Windows mixer's default sample rates in mind, heywood. For all anyone knows, a game's sound libraries could have different effects mastered at different rates and bit-depths all playing at once, or even if they've all been meticulously created to the same standard, the game still has to mix music and voice in with them on the fly, so there's a limit to the utility of a direct passthrough compared to movies and music.
zoog on 17/2/2017 at 14:33
Quote Posted by heywood
I try to avoid sample rate conversion as much as possible. My mobo is an MSI Z87-G45 Gaming X, with an onboard Realtek ALC1150 which is happy outputting anything up to 24/192. For music I use WASAPI to get "bit perfect" output bypassing the Windows mixer with no SRC or volume control. But most games mix everything at 48 KHz within the sound library (e.g. FMOD) and send it to the Windows mixer. I haven't found a way to get around that.
Software SRC is designed to get rid of (almost always) poor quality hardware SRC, so it's a good thing. Don't know about software SRC built-in in vista, but can guess it's crappy too, so one must use SSRC plugin for winamp/foobar to listen to the music, even lossy. In games/movies it's completely irrelevant, as it's sound quality is inferior comparing to, say, CD, + first of all 48→48/96/192 SRC makes as good as no artifacts to worry about.
Quote:
For music I use WASAPI to get "bit perfect" output bypassing the Windows mixer with no SRC or volume control
Anyway 44.1→192 SRC
is performed and if it's not you who set up and control it - then it's B. Gates who hardly will let you get hi-quality sound.
zoog on 17/2/2017 at 14:41
Quote Posted by Sulphur
a game's sound libraries could have different effects mastered at different rates
Never heard anything like this - all movie/pro/gaming sounds are made at 48k, even in 90s, though then they had to convert it to lower rates. Theoretically I may be wrong but in these cases all SRCs are omitted (48→48) or make no losses (48→96, 48→192).
Sulphur on 17/2/2017 at 16:52
I don't know -- I'm just thinking about theoretical possibilities. If 48 KHz is a standard for all sound mastering industry-wide, that's great.
Muzman on 20/2/2017 at 04:41
Now things are probably ok on that front. But go back only a few years and they were still all over the shop (well, my 'a few years' might be getting a little long in the tooth). You'd definitely get sfx at 22khz running with music oggs at 44.1 etc at the very least. Go back to Thief and the like and it's a case of whatever the file can bear to get the drive footprint down: you had 11khz mainly with a couple of 22khz here and there, a mixture of 16bit, 8 bit and 4 bit ADPCM all going together etc.
Maybe that's beside the point and it's what the game/direct Xs output runs the mix at. I don't know.
heywood on 20/2/2017 at 12:52
Quote Posted by zoog
Software SRC is designed to get rid of (almost always) poor quality hardware SRC, so it's a good thing. Don't know about software SRC built-in in vista, but can guess it's crappy too, so one must use SSRC plugin for winamp/foobar to listen to the music, even lossy. In games/movies it's completely irrelevant, as it's sound quality is inferior comparing to, say, CD, + first of all 48â†'48/96/192 SRC makes as good as no artifacts to worry about.
Anyway 44.1â†'192 SRC
is performed and if it's not you who set up and control it - then it's B. Gates who hardly will let you get hi-quality sound.
See this:
(
https://msdn.microsoft.com/en-us/windows/hardware/drivers/audio/kmixer-driver-sample-rate-conversion-and-mixing-policy) https://msdn.microsoft.com/en-us/windows/hardware/drivers/audio/kmixer-driver-sample-rate-conversion-and-mixing-policy
When multiple applications are sharing access to the audio device, so called shared mode, there is also a default bit depth and sample rate specified in the Advanced tab of the Properties dialog for your output device. In the Advanced tab, there are two more options: "Allow applications to take exclusive control of this device" and "Give exclusive mode applications priority". If you have these options checked (by default they are checked), then applications can request exclusive access to the audio device, in which case Windows doesn't perform any mixing or sample rate conversion.
Most decent player applications, e.g. JRiver or Foobar2000, can be set to use Windows Audio Session API (WASAPI) exclusive mode output. In this mode, the stream goes straight from the application to the audio device, bypassing KMixer and avoiding any SRC and gain/volume adjustment. The SPDIF output is at the sample rate and bit depth of the file. That is my preferred way to play.
There are lots of audiophiles who like to upsample their 16/44.1 CDs in software so they can experiment with different reconstruction filters, or convert to DSD for who knows what reason. I haven't found any particular advantage in doing so.
There is oversampling going on in the DAC regardless, but that's a whole other topic.
Chimpy Chompy on 20/2/2017 at 13:52
I have a soundblaster Z, and it maaaaybe sounded a bit better than onboard? But then sound started getting heavily distorted, I dunno how to describe it in techical terms, basically a loud rustling. I've gone back to onboard sound for now. No idea what the problem was - screwy drivers? Bad connection? Damage to the soundcard?
Nameless Voice on 21/2/2017 at 13:43
I use a "USB soundcard" that's build into my keyboard (Logitech G110). No idea if it's actually any better than the onboard or not, but the headphone socket is much more convenient for me.
Quote Posted by Renzatic
Wow. I don't think I've bought a soundcard since I grabbed an AWE64 back in the late 90's. Other than losing that supercool onboard wavetable synthesizer (which made MIDI files actually sound good), I haven't felt like I've missed out on much of anything since going with integrated.
I could see how some people could justify getting one if they produce music, or are just really, really into it. But for me? Eh.
You can get soundbanks these days which can emulate those old wavetable soundcards in software. (
http://midkar.com/soundfonts/) Timbres of Heaven, (
https://sourceforge.net/p/fluidsynth/wiki/SoundFont/) Fluid R3 and (
http://www.schristiancollins.com/generaluser.php) GeneralUser GS are the ones I have installed (from a couple of years ago when I last looked into it.)
You need a media player which supports them. I use AIMP3 (an old-Winamp-style player) these days, which has a built-in BASSMIDI plugin, and will play midi files with it.
I don't actually have that many midi files.
I did put up several version of the System Shock 1 midi music rendered out using some of these soundbanks a few years ago. Those are (
https://www.systemshock.org/index.php?topic=835.msg74469#msg74469) here if anyone cares.
baeuchlein on 23/2/2017 at 15:29
I have not bought any new sound card since about 2002, when Realtek's drivers for their onboard sound chip crashed several times a day. They needed about one year to put a usable driver on the net for download. That one just crashed about once a day. Only years later with Vista, I got drivers which were what you could call "stable" - with the exception being Linux, where drivers were stable and present much earlier.
In these days in 2002, I was pissed-off enough to go into some shop nearby and buy a Sound Blaster card which turned out to be pretty good, except for the fact that any setup with more than two speakers did not work well. The rear speakers' volume was far, far below the one of the others, and there was no reasonable way to correct this. Many years later, Creative Labs' latest driver ever for this card was corrected in that regard, but it was only useful in Windows XP, and none of the games I played with XP these times was played with "3D sound".
So I'm used to getting stereo sound only from games, and since I only have one modern machine by now, a laptop with stereo sound only, there will not be any "3D sound" in the near future, regardless of whether done with an onboard sound or a sound card.
Older machines still around in my realm will still use their ancient soundcards, since they don't have anything else, while newer machines most likely will be used with onboard sound. I have not heard a difference in sound in the last ten years between sound cards and chips, so even if there really is some difference, it does not matter to me. Might be because I mainly play a few older games and listen to audio plays I record from the radio, and not so much to music.
Thus, it will most likely be sound from onboard chips for me in the future. They serve me well enough. Your own mileage may vary, of course.
As for Realtek, the onboard sound of said laptop is by Realtek again, and just like before, the chip seems to be acceptable, while the drivers don't seem to be that good. No crashes, but a lot of functions not working very well. Almost all of the configurable things meant to improve sound quality actually decrease it, for example. What a steaming piece of dino poo.:nono:
heywood on 24/2/2017 at 00:10
All that was ever needed for 3D sound was a pair of stereo headphones and HRTF processing either in software or hardware.
