Tried it with Blood 1.11, and it's working here in 0.74-3 and SVN. You didn't mention it, but can we assume you have the necessary GUS MIDI files installed (i.e. C:\ULTRASND\MIDI\ containing a bunch of .PAT files and an ULTRAMID.INI file)? If not, there's a guide of sorts for installing the files: Install Gravis Ultrasound Properly, All files, No Errors (With Pictures!)

Found the culprit: WOODFLUT.PAT

I did that ULTRASND installation so long ago, I'm not 100% sure but I think this is from the update 1.60 to 1.61.

1.61 version is 25kb (ULTRASND\MIDI)
92034e230bc31b60395bf201b954cc9c61a286b96bce9b9f80e1fb689732256f WOODFLUT.PAT < PROBLEMATIC

1.60? version is 4kb ( ULTRASND\MIDI\BACKUP)
24aa668c5a6809a8e4869706bc3aeca1c3b80011fc11988ec27f3f069fe3ec0a WOODFLUT.PAT

After replacing it with the one from the BACKUP (1.60?) folder it works

It is present in the PPL150.ZIP and PPL160.ZIP, the assumption is the old comes from the GUS411.ZIP
92034e230bc31b60395bf201b954cc9c61a286b96bce9b9f80e1fb689732256f WOODFLUT.PAT

The 4kb version comes from the GUS 4.10 installation, the 25kb from PPL150/160
After replacing it with the one from the BACKUP (4.10) folder it works 😁

The patch looks not correct, I'm no expert but
4.10 WOODFLUT.PAT

1GF1PATCH110
2ID#000002
3Copyright 1992,1993 EYE&I Productions and Advanced Gravis
4Wooden Flute
5NoName

PPL1.50/1.60 WOODFLUT.PAT

1GF1PATCH110
2ID#000002
31994 Jesus Villena                                         
4WOODFLUT.SDX   
5panflu

Ah, well, I only use the official 4.11 patch set; but it's good to be aware that PPL could be problematic in some cases. Maybe the PPL version of the "bad" patch could be fixed with a tweak of ULTRAMID.INI, though I don't really understand how the numbers in there work.

ripsaw8080 wrote on 2020-10-10, 07:31:

Ah, well, I only use the official 4.11 patch set; but it's good to be aware that PPL could be problematic in some cases. Maybe the PPL version of the "bad" patch could be fixed with a tweak of ULTRAMID.INI, though I don't really understand how the numbers in there work.

Is posted a 'need help' flag in this thread: Re: Install Gravis Ultrasound Properly, All files, No Errors (With Pictures!)

ripsaw8080 wrote on 2020-10-10, 07:31:

Ah, well, I only use the official 4.11 patch set; but it's good to be aware that PPL could be problematic in some cases. Maybe the PPL version of the "bad" patch could be fixed with a tweak of ULTRAMID.INI, though I don't really understand how the numbers in there work.

AFAIK the numbers are for redirecting patches to other instruments for each possible memory configuration of the card, to ensure the widest possible range of unique patches is used.
Games often use their own internal ultramid.ini to best suit their tracks.

The issue is probably caused by the patch file taking up too much space in the GUS ram (when combined with all the other patches that Blood is loading).

jmarsh wrote on 2020-10-10, 14:11:

AFAIK the numbers are for redirecting patches to other instruments for each possible memory configuration of the card, to ensure […]
Show full quote

ripsaw8080 wrote on 2020-10-10, 07:31:

Ah, well, I only use the official 4.11 patch set; but it's good to be aware that PPL could be problematic in some cases. Maybe the PPL version of the "bad" patch could be fixed with a tweak of ULTRAMID.INI, though I don't really understand how the numbers in there work.

AFAIK the numbers are for redirecting patches to other instruments for each possible memory configuration of the card, to ensure the widest possible range of unique patches is used.
Games often use their own internal ultramid.ini to best suit their tracks.

The issue is probably caused by the patch file taking up too much space in the GUS ram (when combined with all the other patches that Blood is loading).

Thanks for taking the time,

After replacing the patch with a 30k patch it does not fail.
I even copied in the original header (I don't know if GUS detects identical files), and it still worked.

I'm no expert but I think we can rule that one out.

After checking other patch files for similarities I found this PAT to be different,
after changing it to be like the other PAT files, it started working.

Keep in mind that I have no clue what I did 😁

1original
2< 0000f0 61 6e 66 6c 75 00 00 [fc 61] 00 00 b0 4c 00 00 22
3patched
4> 0000f0 61 6e 66 6c 75 00 00 [62 00] 00 00 b0 4c 00 00 22

That's the sample length (in bytes).

jmarsh wrote on 2020-10-10, 16:47:

That's the sample length (in bytes).

Hèh, and I reduced it to 62 bytes 🤣
It seems your assessment was spot on.

After some trial and error, I got it to spit out a proper error message.
Can we somehow increase that memory ?

Akuma wrote on 2020-10-11, 08:31:

After some trial and error, I got it to spit out a proper error message.
Can we somehow increase that memory ?

No, the GUS (classic) had a maximum of 1MB RAM and that's what DOSBox emulates.
Any patch replacement sets should use individual patches that are the same size as the originals, because the original sizes are what game devs used when choosing the patches used in their games.

That's a shame, but it is what it is.
So the PPL patch sets were intended for GUS cards with more memory ? (like the PnP)

Akuma wrote on 2020-10-12, 14:16:

That's a shame, but it is what it is.
So the PPL patch sets were intended for GUS cards with more memory ? (like the PnP)

PPL patch sets use .pat files that can use up to 1MB max of GUS memory , no matter which GUS you use, PnP or otherwise. The PnP uses a single file bank in Interwave native mode where it can use more than 1 MB. It uses .pat files in original GUS compatibility mode and is limited to 1MB in that scenario.

Game devs tested and optimized their games for the official patch set so that everything would fit in memory . They likely never considered testing with unofficial third party patchsets like PPL . Consequently, issues like this are bound to pop up when using third party patches .

Thanks darry, that was real helpful 😁