VOGONS


First post, by dionb

User metadata
Rank l33t++
Rank
l33t++

Ensoniq cards are famous for their good quality. Unfortunately the best of the lot (S-2000 and particularly Elite) are also getting famous for astronomical prices, so the cheaper alternatives (OEM S-2000-ish cards, Opus and VIVO90) might be more attractive. But then things get complicated. It looks like very few Ensoniq cards share the same sample ROMs. I spotted an OEM S-2000-like card for an attractive price on eBay, then started digging and managed to get quite confused.

There was an earlier thread on this topic Re: Comparison of Soundscape revisions with 1 MB and 2 MB - but all its relevant links are dead an it doesn't go into the difference between the 1MB/2MB ROMs. There are also a few Youtube films with an "Ensoniq DB" in them, without specifying if it's the Rev A (2MB) or Rev B (1MB)

So, this is what I found so far:

Original Soundscape S-2000:
P/N: 4001022851 (rev B)
ROM: 1350901601
Size: 2MB
Subjective quote: "Great"
Sample: (Doom E1M1) http://www.tenorman.info/assets/pc/s2000/doom_e1m1.mp3

Soundscape Elite:
P/N: 4001027201 (Rev B)
ROM: 1351000601 2M
Size: 2MB
Subjective quote: "Different, not necessarily better. Percussion more prominent"
Sample: (Doom E1M1) https://www.vogonswiki.com/index.php/Ensoniq (nasty embedded flash player)

OEM Soundscape #1:
P/N: 4021025099
ROM 1351000501 1M
Size: 1MB

OEM Soundscape #2:
P/N: 4001025001 (Rev B)
ROM: 1351000701 1MG (not 100% sure, potatocam pic)
Size: 1MB (assumedly)

Soundscape Odie:
P/N: 4001032101 (Rev A)
ROM: 135100801 WP
Size: 1MB

Soundscape Opus:
P/N: 4001030001 (Rev A)
ROM: 1351000801 WP
Size: 1MB
Sample (Doom E1M1) http://tenorman.info/assets/pc/ssopus/doom_e1 … 1_ssopus_gm.mp3

]Soundscape VIVO90:
P/N: 4001024701 (Rev B)
ROM: 1351000801 WP
Size: 1MB

And then you've also got the Wavetable boards:

Ensoniq Soundscape DB Rev A:
P/N: 4001024401 (Rev A)
ROM: 1351000301 2M
Size: 2MB

Ensoniq Soundscape DB rev B:
P/N: 4001024401 (Rev B)
ROM: 1351000701 1M2
Size: 1MB

So, that gives us these ROMs:

ROM: 1350901601 (original 2MB)
ROM: 1351000601 2M (Elite 2MB)
ROM: 1351000301 2M (DB rev A 2MB)
- 3 different 2MB ROMs
ROM 1351000501 1M (OEM 1MB)
ROM: 1351000701 1M2 (OEM and DB rev B 1MB)
ROM: 1351000801 WP (all the later PnP Odie, Opus and Vivo90 cards)
- 3 different 1MB ROMs

Now, a lot has been written about the 2MB sample ROMs, and it's generally very positive. My guess is that the 2MB ROM on the DB is basically the instruments on the Elite, with pre-computed echo&reverb added.

It's also clear the 1351000801 on the later cards is less highly regarded, and certainly, if I compare the sample from S-2000 vs Opus the difference is striking and not in a good way. But what about the other 1MB ROMs? I'm particularly interested in the 1351000701 1M2 from the DB rev B and some OEM boards.

Reply 2 of 42, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++

I should have cards with most of the different ROMs.

The S-2000 cards are easy to tell which have the 1MB ROM as the ROM on those is the much smaller chip compared to the full size DIP package chip from what I have been able to gather... Or are the some with the smaller chip that are 2MB?

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 3 of 42, by dionb

User metadata
Rank l33t++
Rank
l33t++
CHiLL72 wrote on 2020-10-11, 16:08:

As I have a rev B DB, I can make a recording or two. Same Doom track?

Would be great 😀

cyclone3d wrote on 2020-10-11, 16:17:

I should have cards with most of the different ROMs.

The S-2000 cards are easy to tell which have the 1MB ROM as the ROM on those is the much smaller chip compared to the full size DIP package chip from what I have been able to gather... Or are the some with the smaller chip that are 2MB?

Never seen one, but given the diversity I wouldn't be surprised if later ones might - then again, they might just have downsized to 1MB - in fact, that's exactly what they did on the OEM versions.

Reply 5 of 42, by dionb

User metadata
Rank l33t++
Rank
l33t++
konc wrote on 2020-10-11, 20:36:

I believe the 2MB ROM chip has "2M" written on it and the 1MB "1M" or "1M2", I've never seen a small 44-pin chip being 2MB though.

Er, did you read the list of different chips I compiled in the topicstart?

Take a look at the Elite, or for that matter on the Soundscape DB rev A, both 'small' 44-pin chips with 2MB:
Esselitenoesp.jpg

Reply 6 of 42, by swaaye

User metadata
Rank l33t++
Rank
l33t++

I have recordings of the Opus (1MB), Vivo (1MB), S-2000 (2MB) and Elite (2MB).

http://www.mediafire.com/folder/uc37ntgrkgxhr … Card_Recordings

The takeaway was 2MB is better. Believe it or not. 🤣 Not to say that the 1MB sets aren't enjoyable.

Reply 7 of 42, by konc

User metadata
Rank l33t
Rank
l33t
dionb wrote on 2020-10-11, 21:17:
Er, did you read the list of different chips I compiled in the topicstart? […]
Show full quote
konc wrote on 2020-10-11, 20:36:

I believe the 2MB ROM chip has "2M" written on it and the 1MB "1M" or "1M2", I've never seen a small 44-pin chip being 2MB though.

Er, did you read the list of different chips I compiled in the topicstart?

Take a look at the Elite, or for that matter on the Soundscape DB rev A, both 'small' 44-pin chips with 2MB:
Esselitenoesp.jpg

Yes, I was mostly replying to cyclone3d that even though I also haven't seen (hadn't until your photo!) a 2MB small chip it apparently exists, so maybe a better way to distinguish between the two is the 1M/1M2 /2M marks rather than the chip size.

Reply 8 of 42, by dionb

User metadata
Rank l33t++
Rank
l33t++
swaaye wrote on 2020-10-12, 06:13:

I have recordings of the Opus (1MB), Vivo (1MB), S-2000 (2MB) and Elite (2MB).

http://www.mediafire.com/folder/uc37ntgrkgxhr … Card_Recordings

The takeaway was 2MB is better. Believe it or not. 🤣 Not to say that the 1MB sets aren't enjoyable.

Do you hear any difference between the Opus and the Vivo? They sound pretty similar to me - and with the same sample ROM I'd expect that.

Reply 9 of 42, by swaaye

User metadata
Rank l33t++
Rank
l33t++
dionb wrote on 2020-10-12, 09:03:

Do you hear any difference between the Opus and the Vivo? They sound pretty similar to me - and with the same sample ROM I'd expect that.

Yeah the Opus and Vivo sound the same to me.

Reply 11 of 42, by dionb

User metadata
Rank l33t++
Rank
l33t++
imi wrote on 2020-10-12, 19:09:

heh, I was about to buy that card but then found this thread, thanks for clarification, I guess this is not the one I'm looking for then ^^

Thing is, we don't know how it sounds as it is a different 1MB ROM to the Opus and Vivo90, rather it has the same ROM as the Soundscape DB rev B. So, it probably sounds different, possibly better - but unlikely to be at S-2000 or Elite level. Want to take one for the team and buy it to find out? 😉

Reply 12 of 42, by cyclone3d

User metadata
Rank l33t++
Rank
l33t++
dionb wrote on 2020-10-12, 19:38:
imi wrote on 2020-10-12, 19:09:

heh, I was about to buy that card but then found this thread, thanks for clarification, I guess this is not the one I'm looking for then ^^

Thing is, we don't know how it sounds as it is a different 1MB ROM to the Opus and Vivo90, rather it has the same ROM as the Soundscape DB rev B. So, it probably sounds different, possibly better - but unlikely to be at S-2000 or Elite level. Want to take one for the team and buy it to find out? 😉

What card are we talking about exactly that is different than the Opus and Vivo?

I probably have one.

Last edited by cyclone3d on 2020-10-12, 22:38. Edited 1 time in total.

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 13 of 42, by mkarcher

User metadata
Rank l33t
Rank
l33t

In case you are collecting part numbers, maybe this one (currently at my hand) might be interesting, too:

Spea Media FX (S2000 OEM edition, sold e.g. at Vobis in Germany)
P/N 4001022801 Rev D
Assy Number 40900228 01 Rev E
ROM: 1350901601
Size: 2MB

The main difference between the Media FX and the original SoundScape 2000 is a different set of supported IRQs (The four IRQ pins of the sequoia chip are routed to different pads on the ISA card edge connector than on the S2000). If you use an old SSINIT version designed for the SoundScape S2000 with the Media FX, the IRQ5 and IRQ7 options are swapped (also the high IRQ on the Media FX is 15, while the S2000 has 10). A friend hat the Media FX back in the days, and downloaded a driver update from a BBS (which was meant for the S2000 only), and kept getting weird IRQ problems.

BTW, when we are discussing SoundScapes: I'm still looking for the Version 1.0 firmware and the associated Windows driver that allowed to wave playback two voices through the Sequoia chip in addition to the WSS voice. The windows drivers offered "Wave Out A (330)", "Wave Out B (330)" and "WSS (534)". The driver packages I found on the internet (including vogons driver library) only have the 1.2 or newer firmware that provides only one wave channel on the Sequoia chip, because the second DMA channel (card-side DMA, not host-side DMA) is used for FM emulation in all operation modes. On the old firmware FM emulation was a different operation mode of the card that disabled MIDI support, and in MIDI mode, most likely both card-side DMA channels could be associated to host-side DMA channels and used for independent wave playback.

Just from experience: Don't even think about trying out FM or SB emulation on the version 1.0 firmware (except for giggles), it's about the worst i've ever encountered (with SB 1.0 support only). SB DSP emulation on the 1.2 is still very mediocre, but at least it works properly in a couple of games. The whole idea to emulate FM using wave table samples is just so broken that I don't need to discuss the uselessness of it.

Reply 14 of 42, by dionb

User metadata
Rank l33t++
Rank
l33t++
mkarcher wrote on 2020-10-12, 22:02:
In case you are collecting part numbers, maybe this one (currently at my hand) might be interesting, too: […]
Show full quote

In case you are collecting part numbers, maybe this one (currently at my hand) might be interesting, too:

Spea Media FX (S2000 OEM edition, sold e.g. at Vobis in Germany)
P/N 4001022801 Rev D
Assy Number 40900228 01 Rev E
ROM: 1350901601
Size: 2MB

Tnx - looks very close to the regular S2000.

[...]
Just from experience: Don't even think about trying out FM or SB emulation on the version 1.0 firmware (except for giggles), it's about the worst i've ever encountered (with SB 1.0 support only). SB DSP emulation on the 1.2 is still very mediocre, but at least it works properly in a couple of games. The whole idea to emulate FM using wave table samples is just so broken that I don't need to discuss the uselessness of it.

That's one of the things I'm looking forward to hearing. I had a very good laugh with my Prometheus Aria doing the same 😉

Reply 15 of 42, by CHiLL72

User metadata
Rank Member
Rank
Member

So I made recordings of both my Ensoniq DB revisions A and B. During recording I noticed that the signal level on the right channel of the rev A board was rather low, much lower than the left channel. I suppose there is some kind of malfunction or bad part in the analog section of the board.
However, I normalized the recording to restore the balance between the left and right channel.
The recording of the rev B board was not normalized or filtered in any way. I think you'll all agree that the quality of the rev A board with the 2MB ROM is a lot better.

Filename
Doom E1M1 - Ensoniq DB Rev A.mp3
File size
3.62 MiB
Downloads
76 downloads
File comment
Recording of Ensoniq DB rev A
File license
Fair use/fair dealing exception
Filename
Doom E1M1 - Ensoniq DB Rev B.mp3
File size
3.62 MiB
Downloads
70 downloads
File comment
Recording of Ensoniq DB rev B
File license
Fair use/fair dealing exception

Waveblaster MIDI boards: https://waveblaster.nl - online now!

Reply 16 of 42, by dionb

User metadata
Rank l33t++
Rank
l33t++

Thanks! And very interesting. Comparing the DB rev B to both Opus recordings (swaaye's and tenorman's that I linked to) the two Opus sound identical - as expected - but your DB rev B definitely isn't. The samples are so similar it could be the same, but at an analog level something very different is going on, with yours being more bass-heavy with less clarity in the top. Could this be the sound card? (SB16 filtering?)

In any event unless my (admittedly untrained) ears deceive me, we can consider 1351000701 1M2 (on your DB rev B and some OEM S-2000) and 1351000801 WP (all the later PnP Odie, Opus and Vivo90 cards) to be for all intents and purposes the same, and not exactly great.

That leaves the older 1351000501 1M from other OEM cards. Somehow I'm not very hopeful...

Reply 17 of 42, by CHiLL72

User metadata
Rank Member
Rank
Member
dionb wrote on 2020-10-13, 21:09:

...but at an analog level something very different is going on, with yours being more bass-heavy with less clarity in the top. Could this be the sound card? (SB16 filtering?)

FYI: the Ensoniq DB was on a Serdaco MIDI Interface Board 'CHiLL' (that has a very clear output), recording was done using a Creative X-Fi USB HD. As I stated, there were some issues with the output of this specific DB, where neither channel was particularly loud and the right channel being much weaker than the left. I tried to correct that by normalizing the resulting recording in Audacity. That may also be a cause of any apparent difference in the sound.

Waveblaster MIDI boards: https://waveblaster.nl - online now!

Reply 18 of 42, by florian3

User metadata
Rank Newbie
Rank
Newbie

I recently acquired an early version of the Spea V7 media fx (FCC ID FODSWFX 1000). It has a socketed 2MB ROM (1350901601).
I also have a OEM "Odie" card with 1MB ROM (135100801 WP). I dumped the 1MB ROM and programmed an EPROM for the media fx card. Only did some minimal testing by playing Doom but the music sounded ok, similar to the Odie card. This is obviously a downgrade, but I'm thinking of replacing the ROM on the Odie card and trying to upgrade it to the 2MB ROM. Has anyone tried this before?
I'm also curious to try the ROM of the Elite, but I don't have that ROM.

Reply 19 of 42, by mkarcher

User metadata
Rank l33t
Rank
l33t

I'm surprised that the media fx sounds OK with the 1MB ROM. The ROM only contains the raw sample data. The ROM layout (which instrument is stored where) is part of the firmware that gets loaded to RAM using SSINIT. Did you notice that SSINIT uploaded a different SNDSCAPE.COx file after you swapped the ROM?

Using the ROM of the Elite might sound too dry on the other cards, because the Elite has an chorus/reverb effects processor, while the standard soundscape ROMs have some chorus/reverb precomputed on patches that are typically used with chorus/reverb.