Several ideas/suggestions....

[davegee]

Gary,

Try some tests where you have the last Key Frame at the END of the Slide Duration in both KFSD and Non-KFSD.

The penultimate and Last Key Frames can have EXACTLY the same settings and will not affect the overall appearance but might throw some light on what is happening.

See my reply to Stu.

I think that there is a valid reason to ask Igor to confirm or otherwise that FSD is always used when calculating Key Frame adjustments with Scale Key Frames ticked.

If he confirms that KFSD is always used, I'm happy with that and understand the maths but I really think that the calculations should be different for Non-KFSD and KFSD Modes.

DG

[goddi]

Gary,

Try some tests where you have the last Key Frame at the END of the Slide Duration in both KFSD and Non-KFSD.

The penultimate and Last Key Frames can have EXACTLY the same settings and will not affect the overall appearance but might throw some light on what is happening.

See my reply to Stu.

I think that there is a valid reason to ask Igor to confirm or otherwise that FSD is always used when calculating Key Frame adjustments with Scale Key Frames ticked.

If he confirms that KFSD is always used, I'm happy with that and understand the maths but I really think that the calculations should be different for Non-KFSD and KFSD Modes.

DG

=======================

Dave,

I gave your suggestion a try. What I see is the problem with using KFSD with Scale is that the Keyframe that was placed at the precise end of the Slide Duration (beginning of the ending transition) gets moved to the end of the ending transition zone. And, if the last keyframe is placed just to the right of the beginning of the transition zone(just into the ending transition zone), and KFSD and Scale is applied with an increase in duration, this keyframe moves to the left of the beginning of the ending transition zone.

My logic expects only the Keyframes that are in between the transition zones would be moved relatively (or scaled) when applying KFSD/Scaled. I follow most of this discussion, but I must admit that I do get a bit foggy with some of it.

Gary

[davegee]

Stu / Gary,

The plot thickens!!

While trying to make an example for Gary I found that this happened - bottom half is before Adjustment - Top half is after Adjustment.

Three out of four correct is not Bad??

DG

[nobeefstu]

Dave,

I think that there is a valid reason to ask Igor to confirm or otherwise that FSD is always used when calculating Key Frame adjustments with Scale Key Frames ticked.

If he confirms that KFSD is always used, I'm happy with that and understand the maths but I really think that the calculations should be different for Non-KFSD and KFSD Modes.

Im thinking under the premise ... there can be only one Full Duration value (bottom staus bar) of a Slide whether in mode KFSD or Non-KSFD.

Testing with older v6.04 (before KFSD was introduced) ... it too also uses the Full Duration value (named just Duration then) and reacts the same as our current Non-KSFD.

[davegee]

OK.

Now what about post # 40?

DG

[nobeefstu]

Dave,

Reference to post 40 :

I could not duplicate your result. When I doubled the 8.0s Slide Duration value (of the bottom picture) using the same key point values my results showed 6,10,14,18,22 respectively with no overlaps.

*I did not test/use the same amount and type of objects as yours.

[davegee]

I will check again in morning but got that result at least twice.

DG

[davegee]

Gary/ Stu,

The following test is repeatable on both my Vista and W7 PCs and results in the anomaly seen in Post#40.

Open a NEW PROJECT.

Set KFSD on and Slide Duration and Transition Time to 8 seconds and 2 seconds respectively.

Add 3 Blank Slides.

Open Slide 2 in O&A.

Set KFs at 2 second intervals i.e 6, 8 10, 12 and 14 seconds.

Go to Customize Slide.

Tick Scale KFs and adjust Slide Duration to 16 seconds.

The result is shown below.

Either my understanding of Scale Key Frames is flawed or there is a Bug / Logic Problem there somewhere??

DG

[goddi]

Gary/ Stu,

...The result is shown below.

Either my understanding of Scale Key Frames is flawed or there is a Bug / Logic Problem there somewhere??...

DG

=======================

Dave,

I gave it a try and, yes, I got the same results.

Gary

[davegee]

Gary,

Can you also confirm in my Bug Report?

DG

[goddi]

Gary,

Can you also confirm in my Bug Report?

DG

====================

Dave,

Done. And thanks for getting into this question.

I hope my other 'ideas and suggestions' will get some discussion going, too. :D/>

Gary

[fh1805]

I have, I think, successfully split out the dialogue that was relevant to Robert Albright's post in this topic and have created a new topic "Problem with spacebar in O&A window" in the general part of the forum. I have also deleted the three posts (mine and Dave's) in which we briefly discussed the need for the action that I have taken. I also PM'd Robert and let him know what I was intending to do.

Hopefully all the i's have been dotted and all the t's crossed!

If anyone feels that there are still loose ends that need tidying up, please let me know.

regards,

Peter

P.S. I got around to this sooner than expected because my dinner party guests called off because of the weather conditions. We are getting heavy snowfalls (by UK standards) and my guests are well into their 70s and didn't want to drive.

[aplman]

As far back as February 2010, I went through a lot of pain coming to an understanding of keyframe scaling. I put forward a suggestion as to how I would like to see it work, but apparently my long dissertation struck a chord only with Peter (fh1805).

I have resubmitted it today, hoping to get more interest.

There's just one complication -- the abominable KFSD concept has been introduced meanwhile, and led to much confusion. The two subjects are horribly interlinked.

Please see "Scaling of keyframes/keypoints" in "Ideas and Suggestions"

Ken T (APLman)

[davegee]

Ken,Please remember that KFSD was introduced in response to a specific problem highlighted by amember of this forum.You are not forced to use it?Wrt scaling of KFs I am not sure what Igor's intentions are about non-KFSD.I have suggested that separate calculations are applied.DG