Sequencer Resolution "ppq" Question (in both song & auto accomp style recording)

Posted by: Scottyee

Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/12/99 12:52 PM

Hi all,

Can anybody here tell me what the sequencer's "timing resolution" is, in parts per quarter note (ppq), on the Solton X1 arranger keyboard as well as other arranger keyboard brands such as Yamaha, Korg, Roland, and GEM?

An important feature of a sequencer/arranger keyboard is what the maximum sequencer "timing resolution" is for both "song recording" and "style pattern" recording. Most older hardware based sequencers have a maximum resolution of 96 ppq. Some of the newer hardware sequencers like the Yamaha QY70 support 480 ppq. Computer software sequencers such as Cakewalk, Logic, and Cubase support “much higher” timing resolutions like 960 ppq and even 3,840 ppq.

The higher the timing resolution (ppq), the more accurately the recording reproduces the original “live” performance. The lower the ppq resolution, the more the performance gets altered and quantized (robotic). The difference between 480 ppq and 96 ppq may not seem substantial, but it actually can effect the outcome of the arranger style produced (a style with a live sounding spontaneous energy vs. a less inspired canned sounding one).

I have heard from reading various postings here that the Solton X1 has better “live” sounding styles than the Technics KN5000/KN6000. I'm wondering if this has anything to do with the sequencer "timing resolution" that the X1 may support. The Technics KN5000 has a maximum sequencer timing resolution of 96 ppq. Even if I import sequences that I originally recorded live on my midi software sequencer at 3,840 ppq, the final result when imported to my KN5000 is only 96 ppq. This changes the performance dramatically (quantizes it down from 3,840 ppq to 96 ppq) which alters the result producing a less spontaneous “live” sound than the original. This difference may be subtle but it does makes the difference between a fantastic “exciting” live sounding style and one that is just "very good".

Because the sequencer's "timing resolution" (ppq) is so important in achieving the best song/style quality, I am very anxious to find out what the sequencer "timing resolution" is on other brands/model arranger keyboards (Solton X1, Yamaha PSR9000/8000, Korg, GEM, Roland, etc).

Thanks in advance for all feedback.

Happy arranger keyboarding to all,

- Scott

[This message has been edited by Scottyee (edited 12-13-1999).]
Posted by: freddynl

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/12/99 03:44 PM

Interesting question;
My "old" Roland G 800 has 120CPT resolution for the styles, so guess this is the same as ppq (I have a dutch Manual)
You might make a list of all answers!
Posted by: DonaldS

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/16/99 05:43 AM

I have a ppq concern also but for a different reason. The Roland XP-80/60 has been known to have some timing problems. I suspect the sluggish timing is due to the 96 ppq that the performance gets quantized to. This quantizing to 96 ppq allows for fewer ticks for midi events to get disperced to, causing more then the normal amount of events to be called on the existing ticks putting more pressure on the processor at the occurence of these ticks. I have also noticed that the users who don't encounter timing problems have mentioned that they only use external software sequencers.

The Roland XP-80/60's sequencer records at 96 ppq, but plays back at 480 ppq. Why the difference? I don't know. My unofficial observations make it seem that the number of ticks between record and playback on the Roland XP-80/60 is the cause of timing problems with the synth.

Sorry, but I don't know the tick resolutions of the other synths you mentioned. Perhaps their websites would have the specs you are looking for.
Posted by: Alex K

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/16/99 09:08 AM


I have programmed a (hardware-based) sequencer in my college years. From that experience, I can tell you, that for sequencer (as for most devices) it is much harder to record than to play back. Without going into too much detail, the sequencer does not know when the next midi event is going to come in, so whenever it does, the processor has to interrupt whatever it is doing and record the midi event accordingly. Unless you have a very fast processor, capable of handling a lot of interrupts, the recording resolution will be limited, especially keeping in mind that that same processor has to do many other things, such as driving the user interface and generating sounds. (There are other ways to record, such as polling the MIDI input, but these too have to be occuring at a finite frequency).

On the other hand, playing back a sequence is much simpler - the computer knows when the next note will need to be played, and can schedule its activities accordingly. Hence, the recording resolution may well be set well below that of playback. In the XP-80 case, one tick of the recording resolution corresponds to 5 ticks of playback, hence using the internal sequencer to record causes quantization error.

I think that you are right in attrubuting the poor sequence quality to the quatization error - the sequencer does not record exactly what you are playing in the first place, but rather something more robotic sounding.


Having said that,
Posted by: Fran Carango

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/16/99 12:29 PM

Guys ,96 resolution is not all that bad ! Some of the first hardware sequencers were only 24 resolution...Maybe we are getting spoiled with the numbers game.. Some problems are the stream of unnecessary advents we record in our sequences,[filter them out]they overwork the processor needlessly...Fran
Posted by: Scottyee

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/16/99 01:20 PM


I agree with you that 96 ppq is not bad (at least for a few years ago) but I certainly do think we as consumers have the right to demand more as the technology improves to provide hardware/software sequencers with higher timing resolutions because it really does affect how "accurately" your performance is recorded and then "reproduced". My little Yamaha QY70 hardware sequencer/arranger/sound module supports 480 ppq so obviously the hardware sequencer technology is able to support this level of timing resolution now. I would expect that the current high-end arranger keyboards should also support this higher level of timing resolution.

One of the 'personal signatures' of a great musician is how they choose to play behind/ahead or on top of the beat. The higher the timing resolution, the more midi will be able to capture these nuances.

I can tell quite a difference when a song was recorded originally (live pro keyboard playing) on a software sequencer that supports 3,840 ppq and then converted to another sequencer format that only supports 96 ppq. The converted (quantized) result seems to lack the spontaneous pizazz of the original performance.

As far a timing goes because of "midi clog" (too much extraneous controller data being sent at once, not 'note data' as I was refering to above), your suggestion to "filter these out" is excellent.

Happy arranger keyboarding,

- Scott
Posted by: Fran Carango

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/16/99 05:22 PM

Scott, I think there should be a clarification. I will use a Roland G1000 as my example. It has a resolution of 120[that is 120 steps per quarter note]. Do not confuse this with CPT [clock pulse time]. This is the smallest step unit used by the G1000. How small?. Up to 3,825 steps,the default is 480, . How small do we need to edit the timing you suggest? As you can see ,the high end units are just fine...Fran
Posted by: Scottyee

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/16/99 06:01 PM

Hi Fran,

I'm a bit confused about your latest explanation regarding sequencer timing.

You say that your Roland G1000 has a sequencer resolution of 120 steps per quarter note.

My Technics KN5000 User's manual (page 192) states that the sequencer resolution is 96 pulses per quarter note.

My Yamaha QY70 Sequencer User's manual (page 229) states that its' sequencer resolution is 480 clocks per quarter note.

Aren't 'steps', 'pulses' and 'clocks' refering to the SAME type of measurement?

If so, then the sequencer resolution of the Yamaha QY70 is a much higher (finer resolution) than the G1000 or KN5000 and would more accurately record and playback a midi recorded "live" performance.

I do not understand what you are referring to re: CPT (clock pulse time) and the other figures you mention: 3,825 steps - default 480.

I didn't think that 'clocks' when referred to in the context of sequencer timing resolution referred to 'clock pulse time', but that it instead referred to the # of divisions per quarter note.

If indeed the Yamaha QY70's sequencer resolution is the same as the Roland G1000 and Technics KN5000, can you please explain.


- Scott
Posted by: Fran Carango

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/16/99 06:53 PM

Scott, The resolution, G1000 [120 parts per quarter note ] are fixed. The CPT [clock pulse time] are selectible settings and are EDIT functions. The companies that refer to the high numbers, are actually refering to this edit function[CPT]. For example,if you deside to edit a sequence and you select a bar[1-9999] than a beat[1-],than the CPT:This is where you specify the CPT position of the begining and end...Fran
Posted by: Scottyee

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/17/99 12:18 AM


I finally spoke with a Yamaha technical support specialist and he assured me that the 480 ppq (parts per quarter note) specification on the Yamaha QY70 Sequencer/arranger was calculated using the same type of measurement as was used to calulate the 120 ppq on the Roland G1000 and 96 ppq on the Technics KN5000.

He assured me that 480 parts per quarter note is the latest advancement for a hardware sequencer and that other keyboard hardware sequencers will soon follow suit with this much improved higher "timing resolution".

I am now curious to find out if the Solton X1, Technics KN6000, or Yamaha PSR9000 implements this new improved higher 480 ppq "timing resolution" feature.

- Scott
Posted by: Les Swartz

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/17/99 01:14 AM

Whew!!!! Am I glad I am an amateur. I go by the sound of instruments and not ppg. I do not know if ppg on arrangers make for better sounds as I don't know everything a ppg incorporates in their sound or end result. I do know that the X1 has "nuances" in some of their instrumentation that others don't have. Or maybe it is in their adjustments "attack" "sustain", etc maybe it is in my ears only. But the violin can be, with proper touch and after touch, very sad. And the blowed sax can come very very close to a Kenny G sax. And blowsaxes is the perfect "Auld Lang Syne" sax if I ever heard one.

I don't know a lot of technical stuff and probably will never use the X1 to its full potential. But I will use it to "my" fullness which no other keyboard I've seen allows me to do.

Hopefully you find your answers, but will they really indicate quality of instrumentation beyond the ears? Does 128ppq indicate the sound will be better than 96ppq.

As for me. I am not sure, nor would I buy on the ppg. I bought on a number of X1 users raving about the X1. I bought. I won - in my ears at least.

Really is it this old amateur or do the solo sounds of the X1 have a nuance in their sounds that others don't, and does the touch and after touch enhance the nuances beyond other arrangers. To me they do. And the end result of the best keyboard is that which, on an individual level, each of us enjoy.
Posted by: Scottyee

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/17/99 08:02 AM

Hi all,

The 'bottom line' is how the keyboard sounds and styles sounds to you personally. Some people may not even notice, hear (or even care) about the difference in timimg resolution between 480ppq and 96 ppq. I originally asked this question because I had noticed that both songs and styles originally recorded on higher resolution hardware sequencers (480 ppq) and then converted to my Technics KN5000 sequencer (96 ppq) sounded slightly altered and lacked the pizzaz and spark of the original recording . These are subtle differences in timing that some people wouldn't notice but I did because I was the one who layed down the original keyboard tracks in the first place.

Irregardless, I do think that advancements made in timing resolution is a good thing and that we must be aware that different arranger keyboard sequencers have dirferent specs in regards to "timing resolution". It's very important to realize when you import (convert) styles and songs from other sources that the successful result of your conversion is partly dependent on your sequencer's 'timing resolution' to capture all the nuances of the source song or style.

Nuf said?

Happy keyboarding to all,

- Scott
Posted by: Alex K

Re: Sequencer Resolution "ppq" Question (in both song & auto accomp style recording) - 12/17/99 09:35 AM

Les, I don't disagree with you that it is the end result that is important. And indeed, the sound quality is not related to the ppq resolution.

However, as Scott has indicated above, low resolution of the sequencer could be a serious limitation, depending on what you are doing with the instrument. Obviously, you can't fault one for trying to make an informed decision before making a commitment to a new instrument (and to spending significant sums of money).

Putting that aside, there is a curiosity factor, of trying to understand what makes the X1 styles as good as they are, and whether there is any inherent limitation in other keyboards from playing styles as well as well as X1.

With that, I have another question to the X1 users and dealers: As I understand, the main advantage of the X1 styles is their use of the grooves. Aside from the grooves in the instrument's ROM, are there any external grooves which are available on the market? How easy is it to include those in the styles? I presume that there is a flash RAM memory area which stores grooves, which could in theory use third party grooves. Is that true?


The other re