Did you miss the part where the CPU does all the mixing itself?
The definition of 7.1 is not mixed. It is 7 discrete sound channels that are not mixed. Each channel goes to a separate speaker.
Even when downmixing all the channels to two stereo channels? Did you actually read the text? You don't have any idea of what you are talking about if you think the DSPs of these chips aren't capable of internally mixing a huge amount of mixer inputs. DirectX and a lot of programs that use it make perfect use of hardware mixing and hardware resampling. The CPU doesn't have to deal with it, but since both these things are so computationally cheap, a lot of programs do it in the CPU anyway.
If you want mixing done you must use the CPU to do it. This sound chip is 10 dumb DACs. They just sit there refusing to do anything until the CPU has done all the work of mixing a 64 channel mod into either A) Stereo format or B) 7.1 format.
Again, no idea about what you are talking about. See above.
How do you think a 12 channel mod would magically play on this sound chip?
The sound chip does not support 12 channels.
The sound chip does not support mods.
The sound chip does not support 8-bit samples. (most mods have 1 or more 8bit samples in them)
If the chip doesn't "support" mods, neither does Paula. Yes, the mod format is closely modeled after the features of the Paula, but the chip doesn't support it on chip level more than any other sound chip.
And 12 channels, are you kidding me? These sound chips frequently support 128 channels mixed by the DSP (i.e. no CPU needed) with >24 bit internal mixing depth. Mod players ran almost CPU-less already in the days of the AWE32 and GUS, the former which was able to play them back with its wavetable synthesizer, and the latter which was able to play it back on its multi-channel variable sampling rate architecture.
The sound chip does not support variable playback frequencies so by definition it cannot ever play a mod.
Maybe its firmware doesn't (and really, why would it? We don't live in the 80s where band-limited resampling could actually have been a performance hit) but neither of us know exactly what the internal DSP is capable of.
Exactly. Now u r getting it.
7 individual unmixed sound channels.
7 channels is like 1980s TFMX or Okalyzer. Your sound chip can match a 7Mhz Motorola 68000. I am impressed.
Wait a second, TFMX is software mixed. In that case, the whole variable sample rate rant you've got going doesn't really matter in any way, and neither does the "free mixing" rant. If we are going to compare apples to apples, PCs are capable of software mixing 7 channels, and 7 to a couple of powers of ten...
Since nobody writes 7 channel mods for decades the sound chip can't play them. Its up to the CPU to render your 24 channel mod down into 7.1 format and to scale all the sample rates up or down as required. All the mixing work is done by the CPU.
Not necessarily. See above.
No. They are a lot less competent since I never even got started on its complete lack features.
Oh, please get started. It's specs are a lot unlike Paula, but are they really lacking in features? Please do a side-by-side comparison.
I am sure u could somehow buy such a soundcard as u r dreaming of.
Nobody would be using it and thus nobody would support it with software so it would not be particularly relevant.
The other 99.999% of ppl with intel compatible computers would still be using lame nonmxing cpu-driven soundchip. Somebody cut&pasted a feature list into the thread a while back. Perhaps you missed it?
I saw the feature list, but you obviously didn't do more than skim over it. It's not a dream sound card I'm talking about. You'll have a hard time finding a PC sound chip these days that doesn't support DirectX hardware accelerated mixing and resampling, which is all done by the DSP.
Here is a quote:
Notice how it refuses to accept 99% of all existing sample rates?
Wanna play a mod? It tells you to go screw yourself until you have the CPU do all the mixing itself and convert all samples into one of those 4 formats and convert all 64 tracks down into 7.1 format (or stereo format).
This is an architectural difference between the Paula and modern sound chips. You are really stupid to think a modern sound chip would have anything to gain from minutely variable sample rates. But rest assured that the CPU doesn't have to deal with either resampling nor mixing.
You have obviously never made a mod before or you would understand that you can't make a mod with only 4 sample rates. Especially not those 4 particular sample rates.
Oh, I've made quite a few mods, both on the Amiga and the PC. It's landing me live gigs every couple of months or so. I perfectly understand the differences between the variable sample rate Paula and modern fixed-rate PC soundcards. It's a difference, alright, but in no way a limitation of the latter. Variable sample rates is an artefact from a time where a dedicated DSP to handle high-quality resampling was infeasible and resampling on the CPU actually had a noticeable performance impact.
The YM2149 is capable of playing 300 khz tones, something which the Paula (and most modern PC sound chips) is utterly incapable of. Does that make the YM2149 superior?
Good lucking trying to make music where every note sounds exactly the same.
Oh, we're talking about actual use now? Are you sure you want the discussion to head that way?
