Musical potential of the NEXT.

This section is for discussing everything about Next hardware and latest updates.
Alcoholics Anonymous
Posts: 505
Joined: Mon May 29, 2017 7:00 pm

Re: Musical potential of the NEXT.

Post by Alcoholics Anonymous » Tue May 01, 2018 5:55 am

The FM cores are quite large and that's why they were ruled out in favour of the SID which is also too large now. I don't recall the original FM chip under consideration but it was the smallest FM chip. If you can find a small FM implementation then maybe it's a possibility but I don't think these exist.

User avatar
mitja_i
Posts: 223
Joined: Tue May 30, 2017 6:27 am
Location: Ljubljana, Slovenia
Contact:

Re: Musical potential of the NEXT.

Post by mitja_i » Thu May 10, 2018 6:31 am

https://www.facebook.com/groups/specnex ... 22R2%22%7D
Victor Trucco
Please stop spreading rumors. I, Victor Trucco, removed the SID to make some room for other features and to make a "better" use of the space in the FPGA. Nothing to do with Paul, IP, or whatever, just my decision. Actually the SID still there at the sources, as the FM module (YM2413), I'm just building the core without these modules included.

Victor Trucco
To my surprise I did some tests including the SID in the last compilations and still fits, but it stays in the limit of fpga. Then it might be that including it means having no room for future development and not even for bitstream patches.

Victor Trucco
Anyway, the build with SID included can be used as a secondary core, since the multiple cores boot was tested and it worked great.

Victor Trucco
Long story short: wait for the "launch" release (the one with the cased Next), I have not decided anything yet at this moment.
Backer 423 @ 37% funded 29 days left
---
ZX Spectrum Next glossary

PiyoTaro
Posts: 176
Joined: Thu Jun 01, 2017 11:13 am

Re: Musical potential of the NEXT.

Post by PiyoTaro » Wed May 16, 2018 3:00 pm

Victor Trucco's remarks were followed by a query tree after the remarks quoted here.

From this remark, I misunderstood that the area of ​​9152 cells of "FPGA SLX 9" is used only half, that is, when upgrading to SLX 16 with 14579 cells, the mounting area is only increased from 4500 to 7000 cells.
Certainly, as a feature of ZXSpectrumNext 's hardware, it will operate as "Anti - brick system" even if it fails to rewrite the FPGA.
Since there is the word "launch release", I hope to combine these to become the hardware that consumed the SLX 16's area.

Postscript 1.
FB's thread began as a place to enjoy the rumor that "SID was canceled was the circumstances of the trademark". Also since last week, Facebook group has been unable to display from my web browser bookmark. When I browse the group, I will log in with some account.

Postscript 2.
Since "Multicore" was SegaMasterSystem or C64 emulator, I learned that "SID" and "YM 2413" not related to ZXSperctrum were nominated for ZXSpectrumNext sound chip.

Postscript 3.
If there is something like YM2413 installed in ZXSpectrumNext, I hope that users can rewrite preset 15 instruments. (I think that it does not differ from OPL 2 such as YM 3812 on specifications)

"OPLL" has a lineup of YM2420, YM2423 and YMF281 which changed the preset instruments.
In Atari ST's "FM Melody Maker" linked from Wikipedia's "YM2423" item, I think the tone of banjo and bell is clearly different from YM2413. https://en.wikipedia.org/wiki/Yamaha_YM2413
The reason that there are variations in OPLL seems to have been adopted for "pachislot" in Japan after it was worn out by "MSX" personal computer.
**modified about the number of cells of FPGA
  • ***Postscript 2019/3/24:
    The FM Sound core "VM2413" implemented in "1 chip MSX" has "Extended mode", and the tone parameters of 15 instruments and 3 drums can be registered in separate registers.
    Contents: "A YM2413 clone module written in VHDL." on GitHub. Author digital-sound-antiques.
    https://github.com/digital-sound-antiques/vm2413
Last edited by PiyoTaro on Sun Mar 24, 2019 2:48 pm, edited 1 time in total.

PiyoTaro
Posts: 176
Joined: Thu Jun 01, 2017 11:13 am

Re: Musical potential of the NEXT.

Post by PiyoTaro » Sun Jun 10, 2018 5:42 pm

Previously, I posted that I found a post by a person developing YM2203 or YM2151 in the forum of ZX-UNO.

Last week I found a post that he helped develop "Turbosound FM" using YM2203 for the ZXSpectrum core for FPGA computer "MiST".
Also, in his post there is a description "While still working on a new generic core FM for all Yamaha fit the ZX-A", and it was also difficult to implement YM2203 in "ZX-UNO"(FPGA SLX 9) It was.

PiyoTaro
Posts: 176
Joined: Thu Jun 01, 2017 11:13 am

Re: Musical potential of the NEXT.

Post by PiyoTaro » Tue Jun 19, 2018 12:40 pm

There are people still posting questions on Facebook about which models are implemented in "FMSound". :|
(link Wikipedia: Yamaha YM3526 OPL(*YM3812 OPL2,*YM2413 OPLL))
17 June,2:13
Antonio Ruggiero:
Victor Trucco just for sake of curiosity : how difficult could be implement in Next board fpga an additional sound chip like ym3526?

Victor Trucco:
Physical ic? The boards were already produced and waiting for the plastics. FPGA ic? No space left. Almost all FM fpga modules are huge and some even bigger than SLX16 itself. A year or so ago, I saw one that uses 55K LEs. More than three times the space of the Next, for the ic ALONE, without any other modules.

Antonio Ruggiero:
in my ignorance I fancied just a fpga implemtation. Obrigado .
In the facebook posting, the function of "changing the output waveform of AY" was requested.

I imagine Taito's game board with a tone control IC TA7630 for volume control outside the synthesizer IC MSM5232.
Although this is not "FM Sound", I would like to consider waveform memory tone generator like "Konami SCC" and "Namco 163" instead of additional AY. (link: "Namco 163 audio" on NesDev)
**As I follow Wikipedia, the original sound chip called "waveform memory sound" is said to be Ensoniq 5503 adopted by "Apple 2GS". The inventor is also a person who invented "SID".
11 June,20:35
Alan Ishøj Bostrup-Sørensen:
Many people seent to want sid on spectrum next, even sid was not been part of kickstart and in fact spectrum newer got that chip anyway. I'm also releazed ay can also do some awesome thing when you now have many more channels instead and in fact music do not got interrupted by sfx.
Howover how about possible to modify a channel to play a (clean) triangle sound in a same why it's did on a Nintendo NES? It's was very smart to just replace a square channel into a triangle by a system call. I'm wonder how ay could have sound with such if a feature......
Last edited by PiyoTaro on Wed Jun 20, 2018 7:31 am, edited 2 times in total.

PiyoTaro
Posts: 176
Joined: Thu Jun 01, 2017 11:13 am

Re: Musical potential of the NEXT.

Post by PiyoTaro » Tue Jun 19, 2018 12:43 pm

There are people still posting questions on Facebook about which models are implemented in "FMSound".:|
17 June,2:13
Antonio Ruggiero:
Victor Trucco just for sake of curiosity : how difficult could be implement in Next board fpga an additional sound chip like ym3526?

Victor Trucco:
Physical ic? The boards were already produced and waiting for the plastics. FPGA ic? No space left. Almost all FM fpga modules are huge and some even bigger than SLX16 itself. A year or so ago, I saw one that uses 55K LEs. More than three times the space of the Next, for the ic ALONE, without any other modules.

Antonio Ruggiero:
in my ignorance I fancied just a fpga implemtation. Obrigado .
There are people still posting questions on Facebook about which models are implemented in "FMSound".

In the days before I posted this thread on this forum, I found out that there was the following exchange at the external forum (perhaps the question column of "kickstarter").
May 2017
Q:How much LEs takes SID implementation in Next
A:SID implementation 900 LEs. AY, 600 each.
Unfortunately, "big" FMs, only at external bus,
I tested one with more than 50000LEs, but cant remember which FM now.
I thought whether it is doubtful whether 50,000 cells are necessary to reproduce one YM 2203. It seems that Yamaha's FM tone generator was not planning to adopt it at the beginning.
And it seems that it was trying to adopt SID6581 which is an analog synthesizer as "FM sound" :o . It seems that he was considering adopting "YM2413 (used as a sound chip of the Sega MasterSystem)" which is historically unrelated to ZXSpectrum.

By the way, is it possible to mount "FM tone generator emulator" of "TurboSound FM" equivalent function on the sub board's Raspberry pi zero? 8-)
Can Raspberry pi of sub board use sound output of SpectrumNext body?
Last edited by PiyoTaro on Wed Jun 20, 2018 7:19 am, edited 2 times in total.

User avatar
sol_hsa
Posts: 90
Joined: Fri Jun 02, 2017 10:10 am

Re: Musical potential of the NEXT.

Post by sol_hsa » Tue Jun 19, 2018 12:49 pm

Thanks for echoing stuff from the facebook as some of us never go there..

I'm pretty surprised that SID is smaller than some FM:s, considering that FM is "basically" just digital math whereas SID requires analog (and thus "analog simulation") filters. I guess there's way more going on in those FM synths than I've considered.

The raspi should easily be able to software-simulate basically anything thrown at it, including the next itself, but I'm yet to see any explanation on how it will be used in the next, what the interfaces are, what kind of access does it have to alter the next's behavior, etc.

User avatar
SevenFFF
Posts: 218
Joined: Mon Jun 05, 2017 5:30 pm
Location: USA

Re: Musical potential of the NEXT.

Post by SevenFFF » Tue Jun 19, 2018 2:25 pm

The details are still being worked out and explored, and the focus is currently on getting a bug-free and stable core with the maximum possible feature set before the cases are ready. TK-Pie is a proof of concept, but it’s not intended to be the end result.

If you have any technical suggestions about what you’d like to see the Pi do or what the interface should be like, please give them. I’ll make sure they get passed on to Victor.
Robin Verhagen-Guest
SevenFFF / Threetwosevensixseven / colonel32
NXtel Spectron 2084blog

User avatar
SevenFFF
Posts: 218
Joined: Mon Jun 05, 2017 5:30 pm
Location: USA

Re: Musical potential of the NEXT.

Post by SevenFFF » Tue Jun 19, 2018 2:41 pm

One suggestion was to mimic the PiTubeDirect protocol on the BBC Micro/Master/Electron, which would give access to half a dozen copros, including a native ARM running at GHz speed, for free.

But the Tube bus runs at 2MHz and the interface is only four memory mapped bytes in each direction, so I think we can do better than that.

The next runs at 28ish MHz internally, and simulates dual port RAM, so perhaps we can do better than that.

It would be good to explore if entire 8k MMU banks could be mapped in both directions. Or the Pi could write copper programs and have it’s own DMA with flat access to the full 25-bit memory map. Or a flat access DMA blitter.
Robin Verhagen-Guest
SevenFFF / Threetwosevensixseven / colonel32
NXtel Spectron 2084blog

User avatar
SevenFFF
Posts: 218
Joined: Mon Jun 05, 2017 5:30 pm
Location: USA

Re: Musical potential of the NEXT.

Post by SevenFFF » Tue Jun 19, 2018 2:49 pm

Back to the musical potential, one popular suggestion is to make a second “official” music core, with the SID, quad DACS, the smallest FM implementation, and standard 128K spectrum features, but dropping the most space-consuming Next-only features.

Making it an official core would add legitimacy and avoid fragmentation, and encourage people to develop for it.

We have most of the framework in place necessary to run multiple cores at startup now, so this could be a good compromise.
Robin Verhagen-Guest
SevenFFF / Threetwosevensixseven / colonel32
NXtel Spectron 2084blog

Post Reply