| ID | Category [?] | Severity [?] | Reproducibility | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 07305 | Sound | Minor | Always | Apr 27, 2019, 13:06 | Jun 26, 2021, 10:00 |
| Tester | robcfg | View Status | Public | Platform | MAME (Official Binary) |
| Assigned To | AJR | Resolution | Fixed | OS | MacOS X |
| Status [?] | Resolved | Driver | |||
| Version | 0.209 | Fixed in Version | 0.233 | Build | 64-bit |
| Fixed in Git Commit | 9c002bd | Github Pull Request # | #8206 | ||
| Summary |
 07305: dragon32, dragon64, dragon64h, dragon200, dragon200e: DAC sound not working |
||||
| Description | I did a small assembler program to play 3000hz sampled sound through the Dragon's DAC and found that it works as intended on a real Dragon and on XRoar, but on MAME it produces no sound at all. | ||||
| Steps To Reproduce |
I've attached my program along a sampled sound, the original asm source code and a video of XRoar playing the sample for comparison. Just start a Dragon machine with the vmailsnd.cas file and type: CLOADM CLOADM EXEC 20000 The result should be as on the video, but actually MAME will remain silent. |
||||
| Additional Information | |||||
| Github Commit | |||||
| Flags | |||||
| Regression Version | |||||
| Affected Sets / Systems | dragon32, dragon64, dragon64h, dragon200, dragon200e | ||||
|
Attached Files
|
 Mametesters Dragon DAC Issue.zip (1,106,771 bytes) Apr 27, 2019, 13:06 Uploaded by robcfg | ||||
 No.16595
Osso Moderator
Jun 28, 2019, 13:02
|
As soon as the EXEC 20000 command is entered, MAME trips on an illegal opcode. |
|---|---|
 No.16596
Robbbert Moderator
Jun 29, 2019, 08:54
|
I got a period of silence followed by a screenful of garbage. Debug shows it in a loop for a while, followed by jumping into the weeds (zero-page ram). |
 No.16630
robcfg Tester
Jul 9, 2019, 19:49
|
On the cas file, there is first the sample data and then the code. So you should CLOADM once to load the data, CLOADM again to load the code and finally EXEC 20000 to run it. Please tell me if there's anything I can help with. |
|
No.16631
Robbbert Moderator
Jul 10, 2019, 01:44
edited on: Jul 10, 2019, 01:45
|
Corrected the instructions in "Steps to Reproduce". Now it spends a couple of seconds writing to FF20 then finishing. I can confirm it makes no sound. |
|
No.16632
robcfg Tester
Jul 10, 2019, 18:09
|
The sound circuit is the same for the Dragon 32 and the Dragon 64, so you can use the restored Dragon 64 schematics from Richard Harding's site: http://www.dragondata.co.uk/tech/circuit-diag/D64/FS-White-Prints/Dragon-64K_CPU_PAL_CD41505-2_of_3-B.zip |
 No.19006
tlindner Moderator
Jun 21, 2021, 02:51
|
Github pull request #8206 fixes this. |
|
No.19007
tlindner Moderator
Jun 21, 2021, 03:22
|
So… the idea of using the Data Direction Register to twiddle the DAC is an interesting one. I just tried it on my CoCo 3. I hear the bell sounds perfectly. But at the end of the sound I get static on the real CoCo 3. I loaded the audio in Audacity and hear and see near silence at the end. In your Xroar video I hear rapid beeps. I don't know what is exactly going on with the static, or Xroar's beeps. But I will keep looking into it. |
|
No.19008
tlindner Moderator
Jun 21, 2021, 16:53
|
Arjhacker has a better fix with https://github.com/mamedev/mame/commit/9c002bd2642a73e240d37f6c8528a22b880af590 |
|
No.19041
robcfg Tester
Jun 26, 2021, 10:00
edited on: Jun 26, 2021, 17:16
|
Hi, This is great news, thanks a lot! |