Input appreciated re differences in Paseq vs. History session rendering

[Dee Flouton]

How to explain the differences in these two images? The first was generated with various paseq presets along with Paint paper and canvas textures as an exercise. The 2nd was generated from the History session which was saved in tact, exported and then brought back into the canvas for a re-do. It looks like many of the steps were left out (not processed) from the saved history session. Why? I did not save the Paseq as it was filled with many "select-canvas" steps which were from trying to "region select" areas (it's not always easy to select exactly what one wants the first time) . I wanted to erase many of the (unneeded) action steps in the Paseq but could not find a way to do it except for highlighting each step and hitting the delete key, which I did several times, but there were too many "select-canvas" steps to deal with. There doesn't seem to be a way to select/highlight several action steps at a time to delete them. I did mark some of the steps as "off", assuming this would only affect the PASeq recorded, and not have an affect on the History session being recorded. Everything should have been recorded in the history to play back and look like image #1, but that is not the case. Any ideas/explanations would be appreciated. Thanks.

cheeses1.jpg

cheeses2-historyversion.jpg

[Synthetik]

Without looking at the actual PASeq and History sequences and what they specifically do it's hard to comment on any differences in the processing.

You are correct that you can only select one action step at a time.

[Dee Flouton]

sorry i did not save the PASeq, but will try to duplicate as much as possible and save both for future reference in case it happens again. thanks.

[Synthetik]

You could build paseq or history sequences that start processing with what's already in the canvas. If this was the case then their output would depend on what was in the canvas when you ran them.

The other thing to watch out for is mask or layer playback. if that's turned off in the paseq editor parameters for a specific paseq then the playback might appear different since the mask activation or layer changes would not happen.

[Dee Flouton]

I just sent you a note and zip file before I read this. I started with a clean canvas. As for the masking, it was on and off many times while building the original....only one layer was used.

[liveart]

in addition to processing what is on the canvas there are presets that use start functions which can be affected by:
1) being random; 2) prior paths, 3)path memories -- I have had gallery runs that seem to skip presets, often because these presets are set to follow the pen. Basically John is correct without having the PASeq and History to compare side by side is is nearly impossible to tell what particular step was behaving differently.

From your two images though is does appear that one thing going on is a substitution or update (or failure to update) of the background texture function.

[Dee Flouton]

what you say makes sense. i did try to replicate the situation and sent john a zip file with the original creation, the paseq and the history. I'm attaching here as well.

cheese.zip

[Synthetik]

We meant the actual paseq preset you used. There's only jpeg images of the output in the folder i downloaded.

[Dee Flouton]

Sorry, thought it was included. This should be it.

cheesePasq+Hist.zip

[Synthetik]

You have mask playback turned off in the paseq parameters. So when you play it back any masked action steps you may have been working with when you initially generated the paseq are not going to playback masked.

[Synthetik]

Prior paths is a good one to bring up. If you build PASeqs that generate paths without erasing the bezier path layer first, then they may build up over time since the old paths never get removed from the path frame.

Version 4 has a new path : path layer record option flag called 'erase before action'. Turning this on will automatically erase the path frame before an action that can generate new paths.