VOGONS


First post, by arncht

User metadata
Rank Oldbie
Rank
Oldbie

a measured a couple of old isa sound cards in dos with rmaa 5.5.
i used an analog/digital allen & heath mixer for recording, and every output level calibrated to the same input level.

https://drive.google.com/drive/folders/1_aWLW … C-N?usp=sharing

My little retro computer world
Overdoze of the demoscene

Reply 1 of 9, by darry

User metadata
Rank l33t++
Rank
l33t++
arncht wrote on 2021-06-14, 06:30:

a measured a couple of old isa sound cards in dos with rmaa 5.5.
i used an analog/digital allen & heath mixer for recording, and every output level calibrated to the same input level.

https://drive.google.com/drive/folders/1_aWLW … C-N?usp=sharing

Thank you! It must have taken quite a while to test and compile.

On first glance, it is particularly intriguing to me that the GUS 3.74's amp out actually seems to have significantly better performance through its amp out than its line-out. That and its frequency roll-off .

EDIT: Shouldn't the dynamic range of the 8-bit cards have a hard limit of 48dB ?

Reply 2 of 9, by dionb

User metadata
Rank l33t++
Rank
l33t++

Wow, nice work!

Could you elaborate on exact setup you used here. Would it be feasible for others to get close enough to reproduce it?

I have a pile of sound cards I'd like to put through their paces (if I ever have time...)

Reply 3 of 9, by arncht

User metadata
Rank Oldbie
Rank
Oldbie
darry wrote on 2021-06-14, 08:59:
Thank you! It must have taken quite a while to test and compile. […]
Show full quote
arncht wrote on 2021-06-14, 06:30:

a measured a couple of old isa sound cards in dos with rmaa 5.5.
i used an analog/digital allen & heath mixer for recording, and every output level calibrated to the same input level.

https://drive.google.com/drive/folders/1_aWLW … C-N?usp=sharing

Thank you! It must have taken quite a while to test and compile.

On first glance, it is particularly intriguing to me that the GUS 3.74's amp out actually seems to have significantly better performance through its amp out than its line-out. That and its frequency roll-off .

EDIT: Shouldn't the dynamic range of the 8-bit cards have a hard limit of 48dB ?

you are welcome.

i used a mixer and there is the possibility to calibrate the input level, i set the amplified and the line out to the same level - so probably the background noise is bigger, than how worse is the amp on the soundcard.
i expected worse thd, but i could not see too much disadvantage. by my ears, i always heard the amplified output slightly better.

it could be a biggest problem with a nonstandard hifi input/output, than with a studio equipment.

rmaa55 does not really support 8 bit well, especially 44khz mono, 22khz stereo, etc. so finally after many tests on the recording side, i kept the default 16bit 44khz setting. it is ok for the comparison.

Last edited by arncht on 2021-06-14, 15:30. Edited 2 times in total.

My little retro computer world
Overdoze of the demoscene

Reply 4 of 9, by arncht

User metadata
Rank Oldbie
Rank
Oldbie
dionb wrote on 2021-06-14, 12:43:

Wow, nice work!

Could you elaborate on exact setup you used here. Would it be feasible for others to get close enough to reproduce it?

I have a pile of sound cards I'd like to put through their paces (if I ever have time...)

i used zedi8 mixer, and the config was:
p200, dtk 430hx board, matrox millennium and a recapped tagan 330w psu (cleaner than the generic psu from the 90s).

i compared later with a dx4 (with fsp at psu), it looked quite close... but be honest on the pre 1993 cards the internal noise so big, it always modulates the frequency, you cannot really make a proper test.
the doom runs well on these cards 😀

My little retro computer world
Overdoze of the demoscene

Reply 5 of 9, by dionb

User metadata
Rank l33t++
Rank
l33t++

So you're playing and recording with the same DUT? I'd personally be tempted to try to use a known-good card (AWE64Gold, SBLive or similar) to test against, to be sure whether it's recording or playback that gives bad results.

Reply 7 of 9, by arncht

User metadata
Rank Oldbie
Rank
Oldbie
dionb wrote on 2021-06-14, 22:47:

So you're playing and recording with the same DUT? I'd personally be tempted to try to use a known-good card (AWE64Gold, SBLive or similar) to test against, to be sure whether it's recording or playback that gives bad results.

How do you mean? I played in dos with the isa card, and recorded with my mixer on my modern pc.
The loopback is not really good, it could compensate the weaknesses.

My little retro computer world
Overdoze of the demoscene

Reply 8 of 9, by dionb

User metadata
Rank l33t++
Rank
l33t++
arncht wrote on 2021-06-16, 04:45:

[...]

How do you mean? I played in dos with the isa card, and recorded with my mixer on my modern pc.
The loopback is not really good, it could compensate the weaknesses.

Then I misunderstood what you posted earlier. Could you make a little diagram of the setup? Or at least specs of the modern PC (which sound card?)

Reply 9 of 9, by arncht

User metadata
Rank Oldbie
Rank
Oldbie
dionb wrote on 2021-06-16, 08:53:
arncht wrote on 2021-06-16, 04:45:

[...]

How do you mean? I played in dos with the isa card, and recorded with my mixer on my modern pc.
The loopback is not really good, it could compensate the weaknesses.

Then I misunderstood what you posted earlier. Could you make a little diagram of the setup? Or at least specs of the modern PC (which sound card?)

i think the chance, that you can build exactly the same setup, it is very low... 😀 the mixer, as i wrote, is a zedi8, it has a usb part.

My little retro computer world
Overdoze of the demoscene