There are two tracks that I've always wondered about how they were made.
I would say not even the SID chip could reproduce that very special sound.
I am pretty sure there are no samples involved. Is all this stuff made with hardware-envelope-tricks?
Could you do this tracks with ARKOS-Tracker?
Does somebody know more about it?
https://zxart.ee/eng/authors/m/miguel/sweet-moments-of-lust/
https://zxart.ee/eng/authors/n/never/tortures-of-detonated-flesh/
Listen to this tunes over big speakers and the world after is not the same.
WoW!!! That's really awesome. Quite a difference to the usual soundtrakker droning :) :) :)
I'm reciting a friend of mine who does a lot of conversion to AY:
Short answers: Yes, no, no.
Long answers: "Yes, this is the ZX using their advanced hard-sound. No, it cannot be converted 100% to the CPC because our AY is inferior and only has limited hard-sound. I could get close, but it won't be as good! ^_^ And no, I do not know more about it!"
I might add that the links you gave us clearly states:
AY frequency:1.75 MHz (Pentagon)
INT frequency:48.828125 Hz (Pentagon)
The Pentagon is a Russian clone of the ZX Spectrum with quite a few enhancements.
This hard-sounds can sound like a brute resonantfilter, that is really amazing for a chip that was made for three beeps and one noise.
Okay, the CPC could do this, but maybe in some cases not as good, because the frequency of the chip differs, but could the YM do this?
Again I'm reciting my friend:
" The ZX & ST chips are pretty similar. ST has DMA, too, so SID-voice is a lot easier and nicer! But saving as .YM will produce the same sound, so yes!"
So, the YM chip on the ST can totally do this:
Here's an example of TAO who specialises in SID-voices on the Atari ST:
https://www.youtube.com/watch?v=Kz8KaDHClTA
If you mean a file in .YM format: Yes, you can do that.
Some other examples you might like (again the Pentagon Spectrum clone):
https://zxart.ee/eng/authors/m/miguel/das-verbaten-la-fuego-ver-chuta/
https://zxart.ee/eng/authors/n/nik-o/deadline-funk/
A note from the Wiki AY-3-8910 page:
"It was also manufactured under license by Yamaha (with a selectable clock divider pin and a double-resolution and double-rate volume envelope table) as the YM2149F; the Atari ST uses this version. It produces very similar results to the Texas Instruments SN76489 and was on the market for a similar period."
As far as I know, the ST was not featured with DMA, only the STE. But maybe more CPU-Power helps to generate an additional "digital" square-wave in order to frequencemodulate a Hardware-Squarewave for some pseudo-PWM-effects. The "Sid-Voice" is not the same like the hard-sound. But I am sure, I teach you nothing new, haha.
Are there any examples of CPC-Tunes which offers the SID-Voice too?
I actually doubt that this cannot be done on the CPC. I had a listen and most of the Wow was when 2 channels were playing in (almost) unison while utilizing a hardware-envelope to modulate both channels. The specs that MaV mentioned are actually the same for both, the pentagon mentioned *and* any ZX Spectrum+ or newer 128k model, basically anything having a "standard" AY equipped. The main difference here boils down to a higher clock rate, resulting in e.g. a better resolution on the hardenv generators pitch setting. Anyways, this sounds like nothing that the CPC could not do, apart from maybe a slightly less quality in matching the hardenv generators pitch with the actual note that's being played in parallel.
Speaking of SID sounds on the CPC, I have made plenty of those, for example:
https://soundcloud.com/betasoftcologne/klimper-bsc-slow
https://soundcloud.com/betasoftcologne/run
https://soundcloud.com/betasoftcologne/cpcsid
and some experiments similar to or touching the SID effect:
https://soundcloud.com/betasoftcologne/sets/amstrad-cpc-music-and-experiments
I feel almost tempted to create a small demo in Soundtrakker, of which I am pretty sure to be able to create something very similar.
Quote from: MaV on 15:58, 09 July 22No, it cannot be converted 100% to the CPC because our AY is inferior and only has limited hard-sound. I could get close, but it won't be as good!
That doesn't actually sound like that friend knows what he's talking about. The AY in the CPC is not inferior, it's *the same* chip, just with a lower clock rate, 1MHz instead of ~1.7MHz in the Spectrum. Saying it has limited hard-sound is plain wrong.
Quote from: Axel on 11:01, 10 July 22As far as I know, the ST was not featured with DMA, only the STE. But maybe more CPU-Power helps to generate an additional "digital" square-wave in order to frequencemodulate a Hardware-Squarewave for some pseudo-PWM-effects. The "Sid-Voice" is not the same like the hard-sound. But I am sure, I teach you nothing new, haha.
Are there any examples of CPC-Tunes which offers the SID-Voice too?
Our very own BSC has done a lot in that direction! :)
I think he mentioned DMA because the STe is somewhat the standard for music productions, much like the CPC6128 is for demos in the CPC scene.
Quote from: BSC on 11:49, 11 July 22That doesn't actually sound like that friend knows what he's talking about. The AY in the CPC is not inferior, it's *the same* chip, just with a lower clock rate, 1MHz instead of ~1.7MHz in the Spectrum. Saying it has limited hard-sound is plain wrong.
I think the lower clock rate is exactly what he is referring to (with his own eccentric wording.)
You can do more on the ST because you can throw more processor grunt at it as well as having better timers IIRC.
It doesn't quite come down to just that though, aside from running at different clock speeds the YM chip used in the ST does have some additional advantages such as internally using a larger range of volume levels, leading to smoother transitions etc.
I'm a long way from being an audio expert though, just stuff I've picked up along the way so may not be 100%
@ BSC:
"RUN!" sounds huge! WOW! :-* :-* :-* :-* :-* :-* :-* :-* :-* :-* :-*
I have to show this demo some 64er-Freaks the next days. :P :P :P :P :P :P :P :P :P
Your "custom engine" sounds superior to the "old way" with two in a certain way layered squarewaves.
Is a stock-CPC able to do this? You tell us, this is a "soft-synth". That sounds hungry for CPU.
Glad you liked it! :)
This has been recorded from a stock (emulated) CPC or my 6128 and yes,
it's very CPU hungry because it synthesizes those voices in real-time, leaving
not much CPU power for much else. Color cycling is possible, as you can see in RUN!
Some hardware scrolling should be possible - maybe even using some split-scree technique,
but I (and others) haven't done that yet. Maybe in the future.
There's also a little music disk available btw:
https://www.pouet.net/prod.php?which=91266
Okay, wow. This engine seems to be a spectacular method for an "unusual-sounding" titelmusic for a game. Maybe one day we will see a new breathtaking action-game with good scrolling and a muscular titelmusic made with this incredible engine?!
I'll check your musicdisk out. You should put it captured as a video on Youtube. This makes it easier to show other people what a CPC is able to do.
Hi
@BSC ,
I'm making a video of the Klimper disc, do you have approximate timings for the three tracks?
Cheers,
Peter
Hey, is it possible to realise 4-Channel-Sample-Mix-Music on the Amstrad like on the Atari ST (Turrican 2, Shadow of the beast) with the same soundquality?
Here one example:
Music starts at 0:31. The quality is immense!
Please note that this is NOT the Atari STE with enhanced soundfeatures.
This is pure YM-Power.
https://www.youtube.com/watch?v=-4LrKqxSfsg
Okay, available RAM would make a difference for the amount of samples and size, but technically?
I suspect the issue will mostly boil down to sample rate. Aside from having a lot more memory to store samples, the ST has a lot more CPU grunt to push values into its YM registers, so can manage a much higher sample playback rate even when it's having to mix samples in advance etc. With very little on screen, the whole hardware can be dedicated to pumping up audio quality.
You might be able to do an approximation with a Plus machine, which can feed the AY slightly faster than the CPU (and benefit from releasing the CPU to prepare the next samples while it plays back via DMA) but even then memory and sample rate are probably going to be limitations.
Have a look at Digitracker: https://www.cpcwiki.eu/index.php/Digitracker
Pretty amazing what is possible on the CPC.
Okay, Digitracker is really impressive, but compared to ST-Beast it sounds noisy and scratchy. How is it possible to make an ST sound so good? Is this only a question of a high enough sample-rate or can the ST produce 8-Bit-Samples instead of 4-Bit via the YM-Chip with some clever tricks?
Another question:
Listen to this both tunes, made by Ben Daglish.
This tracks offers something that I would describe as a resonant-filter-sweep. I was really impressed back then.
Is this a sample or made with hard-sound?
https://www.youtube.com/watch?v=rMnLJkSi3nE
https://www.youtube.com/watch?v=UkY12Juu7XM
Quote from: Axel on 13:46, 16 July 22Maybe one day we will see a new breathtaking action-game with good scrolling and a muscular titelmusic made with this incredible engine?!
I hope so, but the engine is a bit on the memory-hungry side, so it better be a 128k game :)
Quote from: ComSoft6128 on 15:28, 16 July 22I'm making a video of the Klimper disc, do you have approximate timings for the three tracks?
Sorry, not really, did you find it out for yourself?
Quote from: Axel on 23:51, 19 July 22Hey, is it possible to realise 4-Channel-Sample-Mix-Music on the Amstrad like on the Atari ST (Turrican 2, Shadow of the beast) with the same soundquality?
As you have already noted by listening to a Digitrakker tune, this is about the best sample based quality you can get when using the AY as a DAC. There's the digiblaster add-on, but I count that as cheating, so - the Atari has more CPU power to software-mix and that's why it sounds better afaik.
Quote from: Axel on 15:46, 23 July 22Another question:
Listen to this both tunes, made by Ben Daglish.
This tracks offers something that I would describe as a resonant-filter-sweep. I was really impressed back then.
Is this a sample or made with hard-sound?
Both tunes are using the same envelope generator trick by sliding from smaller to bigger period values over a relatively high-pitched tone.
Fun fact: I never heard the music of Techno Cop before now :)
But there is something wrong with the music of Techno Cop in this video, the channel with the accompanying arp-chords is too quiet, which causes a less balm for the ears.
Give it another try on a good emulator, it is a very good track!
And, hey, I like your conversion of "Space Debris" soooo much! :-*
I discovered it many years ago in the AY project library. :o
Nur mal so am Rande. :)
Quote from: andycadley on 13:31, 20 July 22You might be able to do an approximation with a Plus machine, which can feed the AY slightly faster than the CPU (and benefit from releasing the CPU to prepare the next samples while it plays back via DMA) but even then memory and sample rate are probably going to be limitations.
I thought a little bit better than an approximation on the Plus?
Quote from: BSC on 22:04, 23 July 22Quote from: ComSoft6128 on 15:28, 16 July 22I'm making a video of the Klimper disc, do you have approximate timings for the three tracks?
Sorry, not really, did you find it out for yourself?
I can use the timing from SoundCloud (I missed that earlier in the thread) for one and the other two will get my usual two minutes for tracks of indeterminate length.
Quote from: Axelay on 01:02, 24 July 22Quote from: andycadley on 13:31, 20 July 22You might be able to do an approximation with a Plus machine, which can feed the AY slightly faster than the CPU (and benefit from releasing the CPU to prepare the next samples while it plays back via DMA) but even then memory and sample rate are probably going to be limitations.
I thought a little bit better than an approximation on the Plus?
It's a great version, although I'd still say the samples in the ST one sound a little cleaner to me. But then they're also different tunes.
The impressive thing with the GX demo over the ST intro though is that it's doing that whilst also running something that looks more like the actual game, rather than just a static screen full of text.
Hi
@BSC , the text & video below have not yet been published on YT - if there are any corrections to be made please let me know.
2022
Klimper disc - BSC Slow by 'BSC'
From SoundCloud:
"Another tune I wrote on the CPC back in the 90s using Soundtrakker. I think it was inspired by a C64 tune by Padua from that time. The track was recorded from my SID emulation engine, consisting of 3 virtual SID channels: one saw-tooth wave and two square-wave with varying duty cycle."
Links:
https://www.cpc-power.com/index.php?page=detail&num=18667
https://cpcrulez.fr/demostestDO_klimper.htm
https://soundcloud.com/betasoftcologne/klimper-bsc-slow
https://www.pouet.net/prod.php?which=91266
Not emulated - original hardware and software.
Please note that the aspect ratio for this YouTube video is 16:9 but the CPC monitor
has an aspect ratio of 4:3 so you may wish to adjust your viewing device accordingly.
The GX-Version of Beast sounds surprisingly good! Respect!
Quote from: ComSoft6128 on 10:20, 24 July 22Hi @BSC , the text & video below have not yet been published on YT - if there are any corrections to be made please let me know.
Sounds good, go ahead! And thanks for recording & uploading :)
Cheers,
I'll schedule it for Tuesday with the other two on Wednesday & Thursday :)
Quote from: Axel on 15:46, 23 July 22How is it possible to make an ST sound so good? Is this only a question of a high enough sample-rate or can the ST produce 8-Bit-Samples instead of 4-Bit via the YM-Chip with some clever tricks?
I am no expert in this, but I have read a few articles in the past and made my own thoughts about this. So I might be completely mistaken, but I think there are 3 factors, why ST samples sound better:
1) The ST has much more RAM. This means, it can easily hold better samples in RAM. 64KB is just enough for 8s of 8Khz sample sound or 4s of 16KHz.
2) The ST has much more CPU power. Moving data to the soundchip at such a high frequency is a lot of CPU load. And for modules you also have to mix several channels. On the ST the CPU can easily do that for 8-15bit samples without much overhead. The Z80 has much more work to do - and I guess that's the reason why Digitracker is using 7bit samples instead.
3) Both soundchips do not offer a real linear DAC, that is required for 8bit samples. A 8Bit DAC produces 256 different levels of voltage on its output. This simply cannot be accomplished directly on the GI/YM soundchip. There we have 3 channels with 16 logarithmic volume levels per channel. What is now done on the ST is, that you combine all variations of 16 volume levels on 3 channels and measure the voltage level on the sound output. This gives you a list of 16^3 (4096) voltage levels. In this list you try to find 256 voltage levels that are closest to the voltage levels of a real 8bit linear DAC. On the ST this seems to work pretty well (although still not good if you compare it e.g. to the STe that has a real 8bit DAC). On the CPC it's unfortunately slightly different: The designers of the CPC wanted to have something special, so they offered "stereo" output. This means, that from the 3 output signals of the GI, 2 are combined for left and 2 for right. One channel is supplied to both sides at half the volume, so it is kind of the center. While this is a quite funny effect although mostly useless, it has a huge disadvantage for samples: you can no longer combine 3 channels, but only 2 channels. And one of these channels is at half the volume level. This leads to the situation that you have only 16^2 (256) different volumes available and within these, there are again many, many volume levels that are too close together. And even worse, there are several zones in the 0-255 level range, where you simply don't find a matching voltage level for a linear output level.
So at best, you will be able to produce a sound that is closer to a 5bit or maybe 6bit DAC. But far from the quality of the (almost) 8bit sample sound on the ST.
Thank you very much. :-*
But I must say, stereo-sound has always been a great feature for me. Listen to CPC-music via a good ghettoblaster or even a stereo-system is a great pleasure!
Play "Tempest" with big speakers...WOW! Arcade at home. Like an earthquake!
Same for the music. It is really a shame if you listen to CPC-Music over the internal mega-shitty mini-speaker.