Integer is King? - the challenge

[thorkz]

Yep. And it is not true that you cant hear it. You just dont pay attention. I'm listening on old HIFI speakers that I found on the street.

Thanks to Christian, for me it is now fact that fixed point processing produces much less ugly artifacts, very useable high-end, steady sound, deep and solid bass. I will not hesitate to use this plugin even though its "just a simple" EQ.

While X and Y for me are both what I'm used from floating point. They both loose bass. Y is a little worse in the high freq distortion department. It gets "foggy" if you tweak to extreme. The sound gets sourrounded by some "clowd" that gets really hot on my ears.
X is somthing in between ( not integer though), it also gets clowdy, but has a little better high end.

To me Z is of much higher quality.

I want more plugins in integer.

pleeeaase?

thorkz

[whyterabbyt]

Refer to and use Yamaha FM synths if you want to know what does what. There are many patches, including preset ones, taking advantage of specific aliasing in their synths.

Taking advantage of aliasing in a patch is not the same as the FM implementation containing 'smart alising'. Please provide references for your claim.

For the record, John Chowning specifically requested operator key-scaling in Yamaha's implementation of FM, so that patches could compensate for aliasing. That, however, is not 'smart aliasing' in the chip.

If you have no documented proof, then we'll just assume that I have a DX7 to hand, and that as a result I have proof that your claim is false, okay?

[whyterabbyt]

I've never heard of a bit identical emulation of any FM chip. If you have any reference to anywhere that demonstrates it, please show it.

Oh, the irony.

Tell you what. I have one here. If you compared it, then you'd know it was perfect. Voila, proof the way you offer it yourself.

[Shy]

whyterabbyt: Proof for what? That Yamaha's FM synths have vastly different aliasing from one another and that in each, the company has taken advantage of that specific aliasing's character in various patches? Not to mention countless artists doing the same, much better?

FM synthesis without aliasing is incredibly sterile. Good luck doing anything useful with that. If there ever was a synthesis method with which a company could be creative and turn the current technology's implementations' limitations into advantages, it's frequency modulation.

Today you have lots of FM synths that have an uninteresting sound thanks to much different aliasing, or lack of, and people's twisted opinion is that "FM synthesis is cold because it's digital" finally has some ground, because most if not all of what they know is poor software implementations by people who haven't realized the forumula to interesting and creative use of FM, and that's aliasing. Yamaha knows it well, and even today the expansion FM cards they sell have intentional aliasing. Do you think they couldn't remove it if they wanted with the processing power they have available now? Of course they could. And of course they wouldn't.

[living sounds]

I remember what a letdown the audio part of old computer game emulators were whenever I used one. But maybe they got better.

[vonRed]

Well, I can say that the OPL3's emulation is not *perfect*. Afaik, MAME's code doesn't emulate the pseudo-exponential curve of the enveloppes (it uses real exponential curves instead!),

Right, that one's not there yet. But others, e.g. OPL2 are, and OPL3 is getting the same treatment.

As for aliasing, emulators might use different sampling frequencies, which means that sounds making intential use of aliasing (hihats and other cymbals) won't sound right

The usual solution for that problem is running the chips at their native samplerates and then do a samplerate conversion. Depending on the quality of that, it can sound identical.

but aside from that, there's really nothing special about aliasing in yamaha FM synths. They simply didn't have the number crunching power to do anything about alias, and if they had it, they would have spent it on other stuff (filters, more oscillators/channels, sample playback...).

Exactly.

[whyterabbyt]

whyterabbyt: Proof for what? That Yamaha's FM synths have vastly different aliasing from one another and that in each, the company has taken advantage of that specific aliasing's character in various patches? Not to mention countless artists doing the same, much better?

moving the goalposts are you?

As for the less simple ones, those with very characteristic aliasing (which is not by chance and not random, but -designed- aliasing by Yamaha), nothing comes anywhere close, and it's no wonder why some Yamaha chips and various FM synths sound very characteristic and have favored sound. They have smart aliasing in various parts of the FM, and that's what really makes most of their sound.

Proof of this 'smart aliasing' designed by Yamaha, please, which you have also referred to elsewhere as 'intelligent aliasing'. As I have already said, that is quite a different thing from the patches utilising aliasing you have referred to subsequently as 'proof'.

[Shy]

whyterabbyt, I really have no idea what more you want me to say, sorry. I guess you think Yamaha's FM synths sound like they do by chance.

[whyterabbyt]

whyterabbyt, I really have no idea what more you want me to say, sorry.

I'd like some reference to this 'smart'or 'intelligent' aliasing that you claim Yamaha built into their FM.

I guess you think Yamaha's FM synths sound like they do by chance.

what does my opinion have to do with proving the existence of 'smart aliasing'

[Shy]

What is there to prove? Do you think Yamaha's aliasing in their FM synths is stupid and that's why their synths sound good? Do you have any idea about the effort it takes to make aliasing actually compliment generated sound rather than ruin it? It doesn't take a genius to realize their aliasing is very smart, or at least far smarter than uncontrolled aliasing. There is nothing to "prove".

[AdmiralQuality]

[whyterabbyt]

What is there to prove?

The existence of 'smart' aliasing by providing some reference to it. Can you actually do that, at all, rather than whiffle on about how f**king miraculous it is?

It doesn't take a genius to realize their aliasing is very smart, or at least far smarter than uncontrolled aliasing.

which is why Chowning asked them to allow key-scaling of operators.

There is nothing to "prove".

Then it shouldnt be hard to provide a reference, then. Start with Chowning's original papers and work forwards, that'll help.

[Shy]

I understand that all you're concerned with is papers and you like to ignore the fact that even today, their FM synth expansion cards don't have eliminated aliasing even though they could very easily do that, just like they do with their other stuff. If you want to "prove" something that's subjective in nature, good luck with that, you won't. But if you don't want to be sensible and realize that their aliasing has its purpose, which is to make a specific, distinctive sound, for a purpose of making things sound good, there's nothing I can say, obviously. I can however say what I said above, they could eliminate their aliasing from their current FM cards easily, as they do with their other synthesizers, and they don't. Draw conclusions, use your hearing, and there you have your "proof".

[whyterabbyt]

I understand that all you're concerned with is papers

Nope. Just one single reference would do.

and you like to ignore the fact that even today, their FM synth expansion cards don't have eliminated aliasing even though they could very easily do that, just like they do with their other stuff.

I havent ignored anything, its f**king irrelevant.

If you want to "prove" something that's subjective in nature, good luck with that, you won't.

You're saying that what you have described as 'smart'or 'intelligent' aliasing is actually subjective. So not designed, then.

But if you don't want to be sensible and realize that their aliasing has its purpose, which is to make a specific, distinctive sound, for a purpose of making things sound good, there's nothing I can say, obviously.

Aliasing is to make things sound good? Erm, is that the subjective bit, then?

I can however say what I said above, they could eliminate their aliasing from their current FM cards easily, as they do with their other synthesizers, and they don't. Draw conclusions

Okay. Conclusion : They dont care because they can sell them without having to invest any more money in old technology.

, use your hearing, and there you have your "proof".

Im glad you finally put that word in inverted commas.

[Shy]

Now you're starting to talk about the way I write? Shows how serious you are.
So, I guess you think synthesizers are designed to sound objectively good? I wonder how someone could define what is objectively good sound. You think people can't design things for subjective purposes? That's interesting.