Certainly in the case of the SB Live series and I believe the others as well, resampling is handled in hardware if not at an earlier point. I'm wondering if you've noticed any quality differences between cards.
All hail the Great Capacitor Brand Finder
wrote:Certainly in the case of the SB Live series and I believe the others as well, resampling is handled in hardware if not at an earlier point. I'm wondering if you've noticed any quality differences between cards.
There are huge quality differences, but I can't attribute this to re-sampling or anything specific. Might be something for you to look into if you are keen?
I don't have the cards 🙁
All hail the Great Capacitor Brand Finder
I typed up this nice reply talking about the AU8830's internal SRC converter and why it sounds so clean.
Then I bumped that damned back button on my mouse and lost the reply.
Short reply: Au8830's 26th order Kaiser windowed SINC filter is all sorts of good and fast with 120dB SNR.
I might type up the long reply again later.
I can't speak to the Aureal cards/chipsets, but the asynchronous sample rate conversion method used by the Sound Blaster Live! is the subject of several of the most epic and heated Usenet flame-fests known to man.
The Wikipedia entry for the SB Live! is equally critical, and when confronted with statements about the "intermodulation distortion" introduced in the sample-rate conversion process, or how "the SB/Live had great difficulty with resampling audio-CD source material (44.1 kHz) without introducing audible distortion," any merits of the card are diminished by the notion that it's instead some sort of gigantic steaming pile.
An IEEE article about the EMU10K1 addresses some of the thought behind its SRC implementation. Given that the SB Live! is a consumer-level card, the design decisions seem fairly reasonable and justified. Notwithstanding, and given the criticism and evidence surrounding the trade-offs of those decisions, there's the interminable question of whether or not it really matters.
Here's a straight 44.1kHz WAV rip of an audio CD track (~155MB):
http://www.symphoniae.com/soundcard/Creative/ … RK_OP36_RAW.wav
And here's that same track, as played from a S/PDIF-connected CD-ROM into a SB Live! model CT-4760, and as digitally captured (at 48kHz) from the S/PDIF output of the CT-4760 into a secondary system running Audacity, using an Edirol UA-1D (~169MB):
http://www.symphoniae.com/soundcard/Creative/ … OP36_CT4760.wav
You'll have to draw your own conclusions, but for me (and while the use of Audacity arguably renders the exercise "non-scientific" at best), visual and aural comparisons of the two samples makes the rampant criticism of the sample-rate conversion process seem wildly unconvincing. We're obviously not dealing with bit-perfection, but I think it's reasonable to suggest that the imperfections aren't likely to be noticed by the majority of consumer-level users with consumer-level sound systems.
When I listen to them side by side, the S/PDIF rip sounds like its missing a little something. But it sounds cleaner to my ears and I prefer it over the straight up rip. I know I shouldn't like it more (if I understand the argument made against the Live!), but I do. It seems to favor highs a wee bit more and to my ears that is a good thing. Or my ears are just not picking up on its lows as much as the straight rip, which leaves the S/PDIF recording a bit less muddled to my hearing.
Back in the day I hated Creative for how they killed Aureal off via lawsuits without merit. I didnt have much of a choice though and ended up replacing my Aureal SQ2500 with a SB Live! on a new build. I learned to like it for gaming, as I bought the version with the yellow digital out jack and a set of Cambridge Soundworks FPS2000 speakers to go with it. I still have them and the card (though they are tucked away right now). I even liked the setup for watching DVD's at the rig while I was working. Wasn't the best virtual surround, but it was pretty clean sounding, and was small enough to stay out of the way. I guess for serious audio work or a real HTPC, the Live! is pretty lame and for those who bought it based on adds for those reasons, I would agree they have a right to be angry with the card. But for late era Win 9X and early XP games, its not a bad card (at least with the right speakers).
I've been debating what hardware to use on a new PIII Win 9X build and I keep coming back to the Live! and FPS2000 combo as lead contender. The other two options are:
1. My Diamond MX300 (with a quality GM DB of course), paired with a decent modern 2.1 speaker set and quality headset for surround.
2. My Gullimot MaxiSound with optical TOS-LINK out to an old Pioneer 500 watt receiver driving a pair of monitors.
The MaxiSound has a real OPL3 for FM in DOS, and XG. It also has the ability to run EAX 1.0 and A3D 1.0 via software (on a PIII-S 1.4 GHz, thats not much of a drain). I just don't have the room to have the receiver, the studio monitors, the 2.1 speakers, and the headset for running multiple cards. So its needs to be paired down to one card. I already have way to many rigs running way to many screens and speaker setups as well as classic consoles. Slowly my x86 PC collection is pushing out my other computers and consoles from the 70's and 80's. I have to start trimming it down a bit.
Sorry to ask this in your thread Phil, but since Cloudschatze brought up the Live! quality I had to see what his opinion on the matter would be. I know you have played around with some of the stuff I mentioned above and others have too. I'd be happy to hear any thoughts on the matter. Rebuilding multiple times to to comparisons on all three setups is a bit of a pain and well educated opinions on the hardware could save me time and effort.
Haha, well to me the Audigy 2 ZS sounds the best, so that would be my pick for a Creative card. Positional 3D over headphones, if the game supports A3D, there is no way around a Vortex 2. That card is king.
Guys, I got another Super Socket video!
What is the best graphics card for Super Socket 7? Part 4: ATI
Thank you Phil, I always enjoy your videos! 😀
I had the same resource problem with the Radeon 8500 128MB. The 64MB version should work fine, since other people were using it on super socket 7 machines. I read somewhere on the net that the Radeon 8500 128MB & Radeon 9100 128MB versions do have this resource problems even on 440BX boards. There is only one resource memory configuration and the only solution would be a BIOS update for the motherboard or the graphics card itself. Ati was blaming the motherboard manufacturers.
wrote:Thank you Phil, I always enjoy your videos! 😀
I had the same resource problem with the Radeon 8500 128MB. The 64MB version should work fine, since other people were using it on super socket 7 machines. I read somewhere on the net that the Radeon 8500 128MB & Radeon 9100 128MB versions do have this resource problems even on 440BX boards. There is only one resource memory configuration and the only solution would be a BIOS update for the motherboard or the graphics card itself. Ati was blaming the motherboard manufacturers.
Ah, good to hear someone else having this issue! I think I saw the card working in a Pentium 4. So the 64 MB version would have been fine? Good to know, I'll keep an eye out for one.
The last SS7 video testing graphics cards
Best Super Socket 7 GPU? Part 6: Showcasing the final build
Showcasing the final SS7 machine!
Games are shown, DOS, DOS Glide, Windows, Windows Glide. Driver tweaks, MS-DOS mode, MiniGL drivers, dual sound cards and more!
Not sure what's next, likely something DOS era stuff 😊
What to do next...
Go Athlon 64 FX and "kill" Pentium 4 EE CPU from project before SS7 😉
Last video : LINK
Fast graphics for the 486 DX 33
What is a decent graphics card for the 486 DX 33?
There are lots of options, ISA or VLB, old and new. So we will benchmark a range of cards with 3DBench, Chris's 3DBench, PC Player Bench, Wolfenstein 3D, Doom and Quake.
We will also look at value, the top graphics card might be hard to get and really expensive. What card offers good value and is a decent ISA card enough for this processor? Some cards use jumpers and dip switches, what is that all about and what can I do when the sockets on the graphics card are empty?
Over the next few weeks I will do videos about the 486, DOS, MIDI and sound cards.
Enjoy this video!
Well done on the latest video, Phil. Enjoyed.
The more I learn, the more I realize how much I don't know.
OPL3 FM vs. Roland MT-32 vs. General MIDI DOS Game Comparison
Let's benchmark our systems with cache disabled
DOS PCI Graphics Card Benchmarks
wrote:Thank you Phil, I always enjoy your videos! 😀
I had the same resource problem with the Radeon 8500 128MB. The 64MB version should work fine, since other people were using it on super socket 7 machines. I read somewhere on the net that the Radeon 8500 128MB & Radeon 9100 128MB versions do have this resource problems even on 440BX boards. There is only one resource memory configuration and the only solution would be a BIOS update for the motherboard or the graphics card itself. Ati was blaming the motherboard manufacturers.
Would any other ATI DX8 card with +128 MB be affected as well? Also the Radeon 9100 is just a rebranded 8500 LE (underclocked 8500), so the 8500 LE 128 MB would be affected as well
486 DX2 66 The 3D Game Changer!
The 486 DX2 66! What an iconic processor!
Almost doubling the performance and making many early 90s 3D PC games fully playable.
In this video we are looking at some benchmarks and gaming footage comparing the performance between a DX 33 and upgraded to a DX2 66.
Enjoy this video!
Testing 14 ISA and VLB cards on a 486 DX2 66.
Cirrus Logic are excellent cards for DOS gaming.
With the S3 on the other hand there are a lot of abnormalities.
I saw this mentioned in the video comments, but I think it would be cool if at some point you could include PCI equivalents of some of your VLB cards (they'd have to be benched on a different board, unfortunately) to see what effect the PCI bus has compared to VLB, as well as to get the full range of graphics cards that were available to 486 users back in the day. 😀
The more I learn, the more I realize how much I don't know.
OPL3 FM vs. Roland MT-32 vs. General MIDI DOS Game Comparison
Let's benchmark our systems with cache disabled
DOS PCI Graphics Card Benchmarks
This is next to impossible, as the PCI board has a different chipset, so it's hard to isolate just VLB vs PCI. I do have one Acer OEM board that has both, but I don't have a single card in both types 😀
The ET4000 would be great to compare ISA vs VLB vs PCI, but prices are out of control it seems.
