The one in the wikipedia ("MELT.AMD"? is that file available for download? Or is it part of a certain game?), I've listened to it just after I read the above statement by Ace yesterday ("all YMF7xx ... "). I did notice the difference. The second also gave me the impression of being somewhat slightly more distorted (like with a worn tape/cassette), so at first I thought the difference in pitch was somewhat irregular, not just a constant 1Hz difference like James said above.

What I'm asking really is, which OPL3 chip/block is really in the YMF724 and YMF744? How did you come to the conclusion that "ALL YMF 7xx cards use YMF289"?

P.S. Oh right, I also have an Aztech SG NX Pro with the YMF262 chip... so back to that MELT.AMD file, how to obtain it? (Adlib Tracker II? Looking into it...)

archsan wrote:

What I'm asking really is, which OPL3 chip/block is really in the YMF724 and YMF744? How did you come to the conclusion that "ALL YMF 7xx cards use YMF289"?

It's the YMF289. I heard the same drop in sound pitch as the ISA YMF7xx chipsets, and what's this? Look at which OPL3 block is defined to be inside the YMF724 and YMF744 in their datasheets:

YMF724: http://www.alldatasheet.com/datasheet-pdf/pdf … AHA/YMF724.html
YMF744: http://www.alldatasheet.com/datasheet-pdf/pdf … TC/YMF744B.html

Yes, it's a YMF289. I heard it, and Yamaha themselves mention there is a YMF289 core inside both chips, specifically on page 29 of the YMF724 datasheet and page 34 of the YMF744 datasheet.

In fact, even the YMF754 has a YMF289 core as mentioned on page 33 of its datasheet: http://www.alldatasheet.com/datasheet-pdf/pdf … A/YMF754-R.html

Same deal with the YMF740 on page 25 of its datasheet: http://www.alldatasheet.com/datasheet-pdf/pdf … TC/YMF740C.html

YMF262 designs use a 14.31818 MHz clock source, YMF289 and later are different (24.576 MHz and 33.8688 MHz).

archsan wrote:

I've listened to it just after I read the above statement by Ace yesterday ("all YMF7xx ... "). I did notice the difference. The second also gave me the impression of being somewhat slightly more distorted (like with a worn tape/cassette), so at first I thought the difference in pitch was somewhat irregular, not just a constant 1Hz difference like James said above.

You really should do a double blind (ABX) listening test, the subconscious have huge part in a non-blind test even with the smallest suggestion, or personal preference.
The audiophile industry makes millions selling fake audio "improvements" like magic ether stones to be glued to the audio cable or Power Cables that make the "sound better", people believe that and part with BIG sums of money.
Anything but double blind listening test is biased and should be considered as placebo where the person hears anything he wants to hear.

Well I can't hear the difference in MELT.AMD 😒

Catchy tune though 😀

PhilsComputerLab wrote:

Well I can't hear the difference in MELT.AMD

That's because you are unbiased.
archsan hears "slightly more distorted" and "pitch was somewhat irregular" only AFTER Ace's suggestion that YMF289 is inferior. 🤣

@Ace
OK, that suffices as proof, thanks. 😀

@James
Of course it was because it was presented that way, that I would notice a difference in the first place. Otherwise I wouldn't have a point of reference at all. And I only really caught the difference with the main/voice instrument to be honest (segment by segment, not the whole minute length). It's hard for me to describe it since I'm not a trained musician and I don't have the right vocabulary.

I've read enough of the audiophile review stuff but I don't think changing cables, putting expensive spikes etc would have any effect on pitch consistency. 😀

Let me put it more clearly: if I were playing games or listening normally (to whole tracks) I'm quite sure I would not notice a difference because I wouldn't constantly have another point of reference anyway.

P.S. James, I just read your post above, I don't know why you have to put it that way though. I have nothing against the YMF cards.

According to OPL4 Application manual FM block uses Clock / 684 for internal rate, which gives 33.8688MHz/684=49515,78...Hz(6,98 cents lower than YMF262's rate) Also OPL4 output sample rate is 44100Hz, so it definetly uses resampler. I think YMF289 uses the same resampler(Same input clock, same output rate) and pitch is 6.98 cents lower than supposed.

@archsan
Any instrument that smoothly changes it's pitch or bends/slides/swirls between notes is very poor choice to judge by.

khokh2001 wrote:

I think YMF289 uses the same resampler(Same input clock, same output rate) and pitch is 6.98 cents lower than supposed.

7 cents equals to 1.75 Hz, but I measured something closer to 1Hz (4 cents).

And that's just exactly where I noticed it.

Alright James, if you think it's a placebo effect on my part, that's fine. I can see this is going nowhere (I mean really, what are you trying to argue now, that my hearing/sense is invalid or what...). I even already said I compared it segment by segment, or second by second... you would notice it that way, right?

I just wanted to say that your assumption that I "took Ace's word that it was inferior" is just false. It's just a damn slight pitch difference that's all.

Ace wrote:

It's not TOO noticeable, but some games make the lower pitch more noticeable than others (it was really apparent in the floppy disk version of X-Wing to me)

It's interesting to know other games, besides X-Wing, where you've noticed worse FM sound on YMF289 and YMF7xx chips.
The good thing with YMF7xx ISA cards should be lower noise than most SB16, no hanging notes, wavetable connector (AWE64 has not it). While 7x4 are PCI with the most accurate FM on today.

As such, I suggest you get a SoundBlaster 16 with a YMF262 or CT1747, the latter of which is an ASIC in which a YMF262 is built in, not a YMF289 like Yamaha's chipsets.

Some assist that even CQM is ok compared to OPL, 🤣. Not taking into account doubtful and sometimes weird OPL clones on a lot of cards. So when people prefer CQM without SB16 noise, issues of YMF289 compared to 262 look like a joke.
If to choose between YMF262 and YMF289 SB16s - then it's reasonable to choose 262. But compared to YMF cards the situation is not so straight. I saw here people being glad to make SB16 -> YMF7xx change because of noise.

Or just play with a Roland and avoid all of this 😵

Well give me a GM/GS/XG Syndicate version then. 😀

archsan wrote:

I mean really, what are you trying to argue now, that my hearing/sense is invalid or what...

God no, this is not my intention.
I only want people to be aware of the scientific side of things, and always keep the subjective side in question.

Tertz wrote:

issues of YMF289 compared to 262 looks like a joke.

Yes... they are beyond nitpicking for 20 years now.
But that's where all the fun is. 😀

PhilsComputerLab wrote:

Or just play with a Roland and avoid all of this 😵

Roland "just works"... where's the fun in that?
I remember dripping cold sweat each time I reinstalled Windows 95, some intense times back then. 😁

archsan wrote:

Well give me a GM/GS/XG Syndicate version then. 😀

Haha, true, you win 😊

I just said that because X-Wing sounds awesome with Roland 😀

James-F wrote:

PhilsComputerLab wrote:

Or just play with a Roland and avoid all of this 😵

Roland "just works"... where's the fun in that?
I remember dripping cold sweat each time I reinstalled Windows 95, some intense times back then. 😁

Give it some time, it will become an "issue" just like the Sound Blaster 16 not being compatible with digital Stereo of the Sound Blaster Pro. And even if people can't hear it, there is just something about knowing everything is "right" if that makes sense?

I don't even give a crap about 1Hz difference. It's barely noticeable if you play one sample after the other beacause you lose your point of reference. Best way is to use a mixer. I did it with my PAS 16 and YMF-718. It's obvious to me that the PAS 16 is slightly sharper than my YMF. Does it matter? Not really. It actually produces a very cool effect when the cards are playing.

I can make my PAS 16XL to sound much lower pitched than my YMF by forcing the card to use its internal crystal oscillator instead of the ISA OSC signal. I didn't measure it at the time but it sounded at least a tone and a semitone lower than my YMF card. If you don't have a point of reference like i did by using a mixer or otherwise, you probably won't even notice it, let alone a 1Hz difference.

The idea that someone might imply the YMF289 is inferior because of this is just plain retarded.

Oh and BTW, CQM is NOT emulation. It is just a different implementation of FM synthesis that sounds sifferent to Yamaha OPL. But even Yamaha itself has different FM synths like the OPN series (of which the OPN2 was used in the MegaDrive/Genesis) or the OPP (Yamaha FB-01 and IBM Music Feature) or even the OPS (used in the legendary Yamaha DX7). I personally don't like CQM because of the way it sounds, just like i don't like the OPN2.

carlostex wrote:

The idea that someone might imply the YMF289 is inferior because of this is just plain retarded.

As the bad difference can be heard sometimes, it's evident negative factor.
Without direct comparision and with cheap table acoustics to notice the difference between CQM and OPL is also not a simple task.

Not long ago I saw YMF289 based card for price low enough to buy "just to test". The seller appeared to be asshole, so I got only a compensation instead of the card. It's possibly to use YMF, but as they should resample everything to 48 KHz, this would be lesser clean comparision.

CQM is NOT emulation. It is just a different implementation of FM synthesis that sounds sifferent to Yamaha OPL.

Call it approximation or imitation by OPL3 clone, if you don't like the term emulation. The problem is it is intended to be used for music made for normal OPL, where sounds noticably differently and worse.

even Yamaha itself has different FM synths like the OPN series

Yamaha, unlike Creative, had brains to don't make SB compatible cards on it.

Tertz wrote:

As the bad difference can be heard sometimes, it's evident negative factor.

This thread has yet to receive a solid comparison to back the trashing of the YMF289.
Where's the evidence?

Without direct comparision and with cheap table acoustics to notice the difference between CQM and OPL is also not a simple task.

Are you serious?
The two are completely different even when listened on "cheap acoustics".
Even without a direct comparison I immediately noticed CQM didn't sound like I remembered it from the 90's, especially in games I loved the most.

Here's a direct comparison:
* Note that the YMF719 (mine) has 0.04 semitone lower pitch in comparison to DOSBox.

James-F wrote:

Where's the evidence?

There is the example on pedia.

> Without direct comparision and with cheap table acoustics to notice the difference between CQM and OPL is also not a simple task.
The two are completely different even when listened on "cheap acoustics".

There was also about direct comparision with "and". I doubt you'd hear "metallic" CQM details on common 10-15 cm speakers with internal amplifier.
If you heard same music on both chips, even if 1st chip was years ago, - it's direct comparision, as you may remember how it sounded.

Here's a direct comparison

In this theme it's interesting to see YMF289 vs 262 comparision. We need YMF262 example also.
I'd prefer Dune, - music there is one of the most complex and quality FM in games (it needs to know an issue there).

James-F wrote:

Tertz wrote:

As the bad difference can be heard sometimes, it's evident negative factor.

This thread has yet to receive a solid comparison to back the trashing of the YMF289.
Where's the evidence?

He has none.

Tertz wrote:

James-F wrote:

Where's the evidence?

There is the example on pedia.

See?

Tertz wrote:

Call it approximation or imitation by OPL3 clone, if you don't like the term emulation. The problem is it is intended to be used for music made for normal OPL, where sounds noticably differently and worse.

I don't have to call it nothing whether i like it or not. Facts are facts, CQM is true hardware FM synthesis. The technical documentation should be around. And to say it sounds worse proves nothing. It may sound worse to you (sounds worse to me too) but it might sound better to someone else.

Tertz wrote:

Yamaha, unlike Creative, had brains to don't make SB compatible cards on it.

This statement makes no sense whatsoever.