Lingo, LiveCode, and Differing Results
Moderators: FourthWorld, heatherlaine, Klaus, robinmiller
-
- Posts: 48
- Joined: Tue Aug 30, 2011 2:24 pm
Re: Lingo, LiveCode, and Differing Results
Thank you for your assistance, Walt.
I'm not going to spend another second trying to figure out where I have to put the crutches and why.
Between versions 8 and 11.5, Director added a few new functions that I didn't know about. I'm eager to experiment with them.
I will keep an eye on LiveCode, and I may purchase the $99 license as an investment toward the future. But I lost most of the interest and enthusiasm I had for it a couple of weeks ago. Spinning your wheels for too long will do that...
Regards,
Luisa Klose
I'm not going to spend another second trying to figure out where I have to put the crutches and why.
Between versions 8 and 11.5, Director added a few new functions that I didn't know about. I'm eager to experiment with them.
I will keep an eye on LiveCode, and I may purchase the $99 license as an investment toward the future. But I lost most of the interest and enthusiasm I had for it a couple of weeks ago. Spinning your wheels for too long will do that...
Regards,
Luisa Klose
Re: Lingo, LiveCode, and Differing Results
It's not crutches, the imageData is meant primarily for display. If image processing without display processing artifacts is the primary requirement, the data in the image file itself has to be processed, for which LC has no specific features. As the documentation indicates, using imageData for that purpose will deliver uncertain results. LC only needs "crutches" if it is being used in ways outside its design envelope. Photoshop (among other apps) is ideal for that purpose.
BTW, thanks, I learned a lot in this thread
Walt
BTW, thanks, I learned a lot in this thread
Walt
Walt Brown
Omnis traductor traditor
Omnis traductor traditor
-
- Posts: 48
- Joined: Tue Aug 30, 2011 2:24 pm
Re: Lingo, LiveCode, and Differing Results
Ah, yes, it's the user's fault for not using it the way it was intended.
So, posterizing an image is outside of LiveCode's "design envelope", but blurring an image isn't.
Where does the documentation "indicate" that imageData will deliver uncertain results, (presumably with "display processing artifacts"), when used to process images?
Now this is becoming worse than a waste of time.
Luisa
So, posterizing an image is outside of LiveCode's "design envelope", but blurring an image isn't.
Where does the documentation "indicate" that imageData will deliver uncertain results, (presumably with "display processing artifacts"), when used to process images?
Now this is becoming worse than a waste of time.
Luisa
-
- VIP Livecode Opensource Backer
- Posts: 4003
- Joined: Sun Jan 07, 2007 9:12 pm
- Location: Bochum, Germany
Re: Lingo, LiveCode, and Differing Results
Hi Luisa, Walt,
To come to the conclusion that imageData is sort of a random data is in my view not correct, nor appropriate.
Luisa, if you give me the full algorithm of the posterization of Photoshop's posterization I am pretty shure that one can implement that. In my view the differences you see are rounding error of floating point operations. And since Red Green and Blue values of RGB are integers at some point one has to convert from floating point to integer. And as mentioned before that transition can be made at different points of an operation in different ways.
Anyways, during the short time that you looked into Livecode you did amazing things with it. Hats off to you. If you prefer other programming environments that is fine.
But to leave Livecode with the feeling that it 'can not count' would not do it justice. Of course Livecode has its limitations and quirks as discussed before. ImageData is not one of them.
I wish you all the best.
Kind regards
Bernd
In my experience imageData delivers exact and reliable color information for the image. What the documentation states is that imageData is not reflective of an original image in the sense that if you scale a referenced image, imagedData represents the scaled version, not the original version of the image. That is how I read it. On top imagedata has no idea of alphaData of an image, that is what alphaData is for. So here again imageData is not the total representation of the image. But you knew that already.Where does the documentation "indicate" that imageData will deliver uncertain results, (presumably with "display processing artifacts"), when used to process images?
To come to the conclusion that imageData is sort of a random data is in my view not correct, nor appropriate.
Luisa, if you give me the full algorithm of the posterization of Photoshop's posterization I am pretty shure that one can implement that. In my view the differences you see are rounding error of floating point operations. And since Red Green and Blue values of RGB are integers at some point one has to convert from floating point to integer. And as mentioned before that transition can be made at different points of an operation in different ways.
Anyways, during the short time that you looked into Livecode you did amazing things with it. Hats off to you. If you prefer other programming environments that is fine.
But to leave Livecode with the feeling that it 'can not count' would not do it justice. Of course Livecode has its limitations and quirks as discussed before. ImageData is not one of them.
I wish you all the best.
Kind regards
Bernd
-
- Posts: 48
- Joined: Tue Aug 30, 2011 2:24 pm
Re: Lingo, LiveCode, and Differing Results
Thank you, Bernd, for being candid, and not insulting my intelligence.
As I said, I will keep an eye on LiveCode.
And thank you for all your help over the last couple of weeks. I learned a lot.
Best wishes,
Luisa Klose
As I said, I will keep an eye on LiveCode.
And thank you for all your help over the last couple of weeks. I learned a lot.
Best wishes,
Luisa Klose
-
- Posts: 48
- Joined: Tue Aug 30, 2011 2:24 pm
Re: Lingo, LiveCode, and Differing Results
Greetings,
A final note on this matter for anyone who may be interested.
Look at the color values in the images of my cats. In the correct example, all of the RGB color values are combinations of 0, 51, 102, 153, 204, and 255. That is posterization.
Regards,
Luisa Klose
A final note on this matter for anyone who may be interested.
Look at the color values in the images of my cats. In the correct example, all of the RGB color values are combinations of 0, 51, 102, 153, 204, and 255. That is posterization.
Regards,
Luisa Klose
-
- VIP Livecode Opensource Backer
- Posts: 135
- Joined: Mon Jul 31, 2006 1:39 am
- Location: New Zealand, Australia, Germany, Japan
- Contact:
Re: Lingo, LiveCode, and Differing Results
I would check the color profiles being used and other threads regarding them in the forums.
Re: Lingo, LiveCode, and Differing Results
Recently, using Lubuntu 13.10, I was testing many different stacks
in LiveCode 6.5.x and discovered that the stack posted in this forum
thread by Luisa, back in 2011, now renders EXACTLY as the Lingo
and Photoshop example.
Test this by yourself. Download the file named "Image_Effects.zip"
from a message in page 1 of this thread and test the stack
using LiveCode 6.5.x.
I found no differences, using a Posterization of 6, with the
Photoshop and Director's image posted in page 1 of this thread.
If your results differ, please post back.
Al
in LiveCode 6.5.x and discovered that the stack posted in this forum
thread by Luisa, back in 2011, now renders EXACTLY as the Lingo
and Photoshop example.
Test this by yourself. Download the file named "Image_Effects.zip"
from a message in page 1 of this thread and test the stack
using LiveCode 6.5.x.
I found no differences, using a Posterization of 6, with the
Photoshop and Director's image posted in page 1 of this thread.
If your results differ, please post back.
Al