Compressing / re-sizing image files

[craig_stowers]

 After users select an image from their native photo library I quickly re-size the image container and add to the screen before capturing the display object, saving it to the documents directory and finally removing from screen. It all happens before the screen renders which is nice.

This works great for creating thumb images, however I also like to keep the larger versions of the files… but not so large as some of them may be. For instance if the user selects a huge 5mp image… I would like to compress that down to 2mp. Is there any way to do this?

[mpappas]

The only compression the sdk will do is built into the display.save / capture functions, and there’s no user controls on the actual compression level. I believe you can change the basic compression scheme by naming your file .jpg or .png (png is lossless in Corona I believe, jpg has some default compression set), but that’s about it.

The only other lever on the filesize I know of is based on the pixel dimensions of your saved image. You will need to pay attention to this, because if you are using content scaling - the capture still happens at device resolution. This means that if you capture fullscreen on an ipad3 (retina) the saved image will be HUGE. But if the same code captures on an old iphone 3gs, there would be a lot less pixels (only 320x480).

Note that if you are using content scaling, there are some things you can do to adjust your image relative to the devices actual resolution - in particular, adjusting your image size based on the display.contentScaleX,Y variables might be helpful to get more consistent across platform image size…

Another note - the reason the sdk didnt render the next screen frame is because the thread your code was running on was still active. If you put a timer.performWithDelay in there to break up your code a little, for even 1 ms, the sdk enterFrame/render code might have kicked off (if it was time for the next frame).

Basically, your code to capture and save and then clean itself up could take 5000 ms (5 seconds straight), and the sdk wouldn’t fire it’s next renderFrame until your code exited from it’s processing (in your current code, until your photo code hits its final closing brace/return).

It’s just the way the sdk works - it won’t render in the middle of ANY of your code, until your code thread exits (pause your code with a timer.performWithDelay, and boom, the sdk has the chance to render).

[mpappas]

The only compression the sdk will do is built into the display.save / capture functions, and there’s no user controls on the actual compression level. I believe you can change the basic compression scheme by naming your file .jpg or .png (png is lossless in Corona I believe, jpg has some default compression set), but that’s about it.

The only other lever on the filesize I know of is based on the pixel dimensions of your saved image. You will need to pay attention to this, because if you are using content scaling - the capture still happens at device resolution. This means that if you capture fullscreen on an ipad3 (retina) the saved image will be HUGE. But if the same code captures on an old iphone 3gs, there would be a lot less pixels (only 320x480).

Note that if you are using content scaling, there are some things you can do to adjust your image relative to the devices actual resolution - in particular, adjusting your image size based on the display.contentScaleX,Y variables might be helpful to get more consistent across platform image size…

Another note - the reason the sdk didnt render the next screen frame is because the thread your code was running on was still active. If you put a timer.performWithDelay in there to break up your code a little, for even 1 ms, the sdk enterFrame/render code might have kicked off (if it was time for the next frame).

Basically, your code to capture and save and then clean itself up could take 5000 ms (5 seconds straight), and the sdk wouldn’t fire it’s next renderFrame until your code exited from it’s processing (in your current code, until your photo code hits its final closing brace/return).

It’s just the way the sdk works - it won’t render in the middle of ANY of your code, until your code thread exits (pause your code with a timer.performWithDelay, and boom, the sdk has the chance to render).

[helena9]

@mpappas

The only compression the sdk will do is built into the display.save / capture functions, and there’s no user controls on the actual compression level. I believe you can change the basic compression scheme by naming your file .jpg or .png (png is lossless in Corona I believe, jpg has some default compression set), but that’s about it.

The only other lever on the filesize I know of is based on the pixel dimensions of your saved image. You will need to pay attention to this, because if you are using content scaling - the capture still happens at device resolution. This means that if you capture fullscreen on an ipad3 (retina) the saved image will be HUGE. But if the same code captures on an old iphone 3gs, there would be a lot less pixels (only 320x480).

@mpappas could you expand on how to compress files?

I have an app that allows users to take/choose a picture, and it returns it as a jpg. Later in the app, the picture along with inputed text is sent to the server and eventually blasted to social media. It works, but the user often has to wait minutes before the post can be confirmed. I have been racking my brain for a solution and finally found this thread…

You mentioned that IF i use content scaling, it will be full resolution. Is there a way to turn this off? What are the consequences?

@craigstowers You mentioned that you quickly resize the image - is that changing the actual size of it, or the size it is displayed? If it changes the actual size, could you share with me how that was accomplished?

Thanks in advance!!

[helena9]

@mpappas

The only compression the sdk will do is built into the display.save / capture functions, and there’s no user controls on the actual compression level. I believe you can change the basic compression scheme by naming your file .jpg or .png (png is lossless in Corona I believe, jpg has some default compression set), but that’s about it.

The only other lever on the filesize I know of is based on the pixel dimensions of your saved image. You will need to pay attention to this, because if you are using content scaling - the capture still happens at device resolution. This means that if you capture fullscreen on an ipad3 (retina) the saved image will be HUGE. But if the same code captures on an old iphone 3gs, there would be a lot less pixels (only 320x480).

@mpappas could you expand on how to compress files?

I have an app that allows users to take/choose a picture, and it returns it as a jpg. Later in the app, the picture along with inputed text is sent to the server and eventually blasted to social media. It works, but the user often has to wait minutes before the post can be confirmed. I have been racking my brain for a solution and finally found this thread…

You mentioned that IF i use content scaling, it will be full resolution. Is there a way to turn this off? What are the consequences?

@craigstowers You mentioned that you quickly resize the image - is that changing the actual size of it, or the size it is displayed? If it changes the actual size, could you share with me how that was accomplished?

Thanks in advance!!