Operations in Scratch that affect the screen are automatically limited to a speed of 30 times per second.

For example, if you draw with the pen you can see the pen move because each “move” block takes 1/30th of a second to happen.

If you draw with the pen using a define block routine without screen refresh, then you will not be able to see the pen move because the drawing will just suddenly appear.

That's because Scratch did not have to slow down to 30 times per second. It could do what it is capable of, which is thousands of moves per second.

footsocktoe wrote:

Operations in Scratch that affect the screen are automatically limited to a speed of 30 times per second.

For example, if you draw with the pen you can see the pen move because each “move” block takes 1/30th of a second to happen.

If you draw with the pen using a define block routine without screen refresh, then you will not be able to see the pen move because the drawing will just suddenly appear.

That's because Scratch did not have to slow down to 30 times per second. It could do what it is capable of, which is thousands of moves per second.

In other words, it will basically treat it as if it were one block instead of a ton. If it were a ton of different blocks, it may take a lot of time for your different things to all happen. However, using “Run Without Screen Refresh” will make it all happen a whole lot faster, and usually instantaneously if your custom block is not too big.

But also be careful because any blocks that are supposed to be delayed will lag if run by a “run without screen refresh” block. That includes things such as:

wait (1) secs
broadcast [this v] and wait