In reality there is no such thing as 'un-quantized' because what 'unquantized mode' means is the maximum 'timing resolution' (abeit quantized) that particular keyboard's sequencer can support. On the Yamaha PSR2000, unquantized mode supports 1,920 ppq (parts/quarter note), and on the Technics KN6500, unquantized mode only supports 96 ppq. Because of this, the PSR2000 will 'more accurately' capture & record the timing of the original live performance than the KN6500.

I agree that hardware sequencers have a ways to go before they approach software sequencers, but still, compared to the competition, the Yamaha PSR2000 sequencer (both its song & style sequencers) timing resolution (1,920 ppq) is FAR SUPERIOR to any other hardware sequencer available today. All other arranger keyboard hardware sequencers only support 96ppq or 192 ppq at best. Even the Yamaha QY100 supports only 480 ppq, so seeing Yamaha incorporate 1,920 ppq in the PSR2000 is at least an encouraging sign that the manufacters are finally paying attention, because the higher the timing resolution, the more accurate, dynamic and realistic the arranger STYLES PATTERNS will sound.

Esh, you might be able to verify your 9000pro's max timing resolution (unquantized mode) by going into the step record window and see if there is a PPQ number displayed. On my PSR2000 it displays: 1920PPQ. I believe it was a technical rep at Yamaha that told me that the PSR9000/9000pro supported 192 ppq. Perhaps someone here could check & verify this.

I agree that the PSR9000/9000pro has almost double the polyphony (126) support over the PSR2000 , yet I personally haven't experienced any polyphony problems playing on my PSR2000. Afterall, I got only 10 fingers.

Scott

[This message has been edited by Scottyee (edited 06-07-2002).]

Scott:

Hmmmm... as suggested, when I call up the step record mode on my Yamaha 9000 Pro it says 384ppq (this is also clearly visible in the illustrations in 9000's manual under step recording). It gets even weirder if you go to chord step recording: 2ppq ! I'm not at all sure if the 384ppq rate would also apply to real-time recording. It's possible that this could be a limitation of the current OS rather than the processor also. This is interesting but inconclusive. There is no other reference to the 9000's ppq rate anywhere else I know of. I'd like to know more.

There's no doubt that the PSR2000 is using breaking technology, such as the XF Scorch music display I envy. But I wouldn't under-rate the importance of polyphony, especially with Yamaha keyboards. Many sounds are layered by default (such as organ patches) and then you can add layers yourself (piano/strings for example). Add sustain pedal and accompaniment or song file playback with your performance and you've got far more than ten-fingers worth of polyphony happening. That's why I added the PF expansion card (64) and the DX expansion card (16) to my 9000 Pro (126 stock: 206 note polyphony total) to accomodate everything I can do at the same time with ten fingers, two feet and one brain.

maybe off topic, but about the ppq issue:
I thought the stability of the clock was far more important than the ppq spec.

In other words a 96 ppq sequencer with high quality rock solid crystal clock, would be superior to a 960 ppq spec sequencer with a cheap varying resonator put in to save money.

Scott's was not a valid comparison because he quantised an existing recording. The test would be to play into a 1920ppq sequencer, and then play again into a 120 ppq sequencer of equal clock stability. If the clock was good enough I don't think you would hear the difference.

PC sequencers are another thing because there you divide down from a pc clock that might not be of best stability in the first place, so have more chances of error magnification maybe, and need high ppq never to notice? Some people still say their Atari's were the most rock solid sequencer of all...

If this is not the case, why have we not been talking for years about our sequencers not being capable of reproducing the exact nuances of our performances?

I have never read any perceptual studies that the human ear can devolve down to microseconds, which is the implication if this all were to be true?

Maybe this is more of a paper spec than real world issue?

I think BOTH are important.



Alex, A while back, a couple of friends (musicians) and I performed a similar test to the one you suggested. We recorded to both a software sequencer (Logic Audio-Mac), both (at 96ppq) and (at 960+ ppq) and compared the recordings. We noticed a substantial difference. The 960 ppq sequence captured the original playing (timing) spontaneous nuances (spark) whereas the 96 ppq version sounded quantized (flat), and somehow lacked the pizazz of the original performance.



I definitely agree with you re: the Atari. I owned the Atari 1040ST, and the timing was rock solid. Ahhh .. . Those were the days . . . too bad Atari ignored the musician market and focussed on games instead.



Well, I can't speak for others here, but I've been raising this issue for some time now (do a search on past postings). It just seemed other people here weren't that interested. In the meantime, a number of musicians in the pro music community have been complaining about just this issue for quiate a while, and now the manufacters have finally begun to listen. I just hope Technics has listened too, and will implement a MUCH HIGHER timing resolution in their KN7000.



Alex: I'm not relying on scientific studies. I use MY EARS and can hear the difference. My musician friends concur (after testing/hearing for themselves).



Well, if this is true, then WHY are both software and hardware manufacters now steadily increasing the timing resolutions on their sequencers then? Just for laughs?

I guess this all boils down to your ears. If it doesn't matter to you, then fine. Hey, some people are happy as a clam with MP3 sound quality. But as a musician, I want a midi sequencer which can MOST accurately record & reproduce the original keyboard performance.

Scott

Hi Scott,
no argument that ears are the final arbiter, but without double blind testing, ears are notoriously unreliable.

the logic test was again not a valid comparison because performed on a pc sequencer, as I have tried to explain I don't believe that this is directly comparable in such a simplistic sense to the keyboard case without taking all other factors into account. I'm not sure that a pc sequencer at 96 ppq would be much use at all?

the really good question would be what was the clock and ppq on the old atari to be as good subjectively as it was?

you make the common mistake of putting words and implications into my mouth which I did not state, I raised a series of questions, and did not make any of the absolute assertions that you reply to. I respect your opinions but you do yourself no favours by replying to assertions I never made...

I never stated that one value of ppq was sufficient, or did not need to be improved upon, or that future improvements were not neccessary, or that manufacturers should not increase resolution. In fact I gave no hint of what my stance on any of these subjects are.

The main reason resolutions are rising is because micros get faster for the same money and it is easily possible for the manufacturers with each new generation of chips at any given price point.

I merely made the perfectly valid point that the forums have not been full of complaints that keyboard sequencers before the days of 1920 ppq somehow had noticeable problems that meant a great many people felt they did not mirror the nuances of their original performances. If you made some posts about this subject in the past and no-one else was interested, you have provided the proof of exactly what I originally said.

Also that originally you drew a conclusion from an experiment that was totally flawed by quantising an existing recording. This still remains the case.

with respect,

Alec,

I apologize if I came off rather strong, as I meant no disrespect. I can see both yours and UD's points now, that quantizing an existing recording (from 1,920 to 96) is not an accurate comparison, but as I mentioned in my last reply, I had also, on another occasion, made 2 separate recordings of the same song (same sequencer) at 2 different resolutions and both myself (and friends) heard subtle yet significant differences, so I still stand by 'initial' findings that the higher the resolution, the better.

Alec, I realize you're an avid Technics KN Keyboard fan (as I have been too), and just wonder if your views may be influenced by the fact that Technics arranger keyboard sequencers have still yet to support a higher midi note timing resolution than 96 ppq.

I of course realize that sequencer midi clock accuracy is equally important, but with 'equal' clock accuracy taken into account, I still believe that the higher the note (timing) resolution, the more accurate the recorded result.

I look forward to discussing (debating?) this issue further in a shared effort to find the best way to encourage the manufacters to further improve the life like 'realism' of arranger keyboard styles and midi sequencer recordings as well. This way, we ALL win !

Scott

Best regards,

Scott

Hi Scott,
no disagreements at all here. I am not defending anything, never even mentioned Technics equipment, just pointing out that you cannot define different equipment with different qualities and operating systems by a simple figure of ppq without investigating all the underlying differences that also contribute to a subjective result.

Maybe an analogy would be the difference between a 128 poly and 64 poly instrument. Everybody assumes that the 128 must be better from a paper spec, to then find that the voices are made up of twice as many partials, and the dropout algorithms are worse, so in real life use may cop out subjectively more noticibly earlier.

As for me being biased for Technics, I could equally accuse you of puffing this psr2000 ppq thing, when since the psr2000 was launched, and before you highlighted the issue I can't remember anyone spontaneously and without forewarning remarking Wow! the sequencer on this machine sounds much superior to anything that has gone before...

I doubt that we will see much hard research on what ppq is neccessary because technology has overtaken the subject and the high ppqs will become a norm in keyboards just because cheap micros make it easy and possible. If so that is great!
However every feature is a cost/quality compromise, and I doubt if many people will notice the difference spending money on the sequencer ppq, whereas I'm pretty sure that everyone would notice the difference if the money was spent on wave rom, for instance.

But I still feel the pc case is quite different to the keyboard case, since have seen much comment on pc sequencer 'solidity' over the years but very few complaints about keyboard sequencers. Thus I feel these 2 cases have different sets of criteria, that a simple ppq cannot explain.

all the best

You guys are really splitting hairs. I have NEVER had a sequencer that could not reproduce my exact nuances in an UNquantized state. Forget the tests - if you try something that sounds wrong to you, THEN look for an alternative. Don't look for problems before they exist. It's just like the MP3 vs. CD comparisons. IN certain situations, they are virtually the same ... especially when played through a comercial sound system. I'm sure a real audiophile could tell the differance in a high end studio setup, but in a normal "PA" - forGETTAbout it !

I was gonna say that Dave.
DonM