Brightness dipping when fading layers

[ping141]

Let´s say I make the letters A-B-C as a Photoshop file with tranparent background, and I want the letters to fade up one by one next to eachother. If I lay A on layer 1, and A-Bb on layer 2, then the A will dip darker in the middle of the crossfade. I can fix that by putting A in layer 2, and fade up A-B on layer 1, but that is not an ideal solution, as you´ll soon run out of layers if you have more than a few letters.

This can be worked around by changing files and snapping with a bunch of extra ques, but is there a way without workarounds? Simply fade something up on top of itself (on a higher layer) without the brightness dimming down in the middle?

[Mr_P]

I assume that your letters are rendered on a transparent background. I think this s just a programming issue...

Cue 1 - Fade up layer 1 (A)
Cue 2 - Fade up layer 2 (B) - [layer 1 is still at FULL from cue 1]
Cue 3 - Fade up layer 3 (C) - [layer 1+2 are still at FULL from cue 1+2]

There are no cross-fades - but you do use 3 layers

Diples crossfading is a function of most lighting desks - but most dont do it very well as most lightsources dim in different ways!- try changing the xfade curve in your cue on the lighting desk, or delay the fade out cue slightly

If you have a lot of letters (ie a sentance) then you should make a movie file with the correct fading (building). Then you can play this on 1 layer and just play between the 'in' and 'out' points if each letter needs to be cued. (else, just let the movie runif your timing is correct in the edit)

A number of video editing softwares have text tools built in, so it would be fairly simple to create.

Simon

[SourceChild]

I've done this for many presentations. Simon has the best and simplest way to do it. Just stack layers leaving bottom layers up.

Be careful though and do not use a black background as there might be transparency issues if your letters are anything but pure white. Instead render the images as PNG or TIFF with the background made of transparent pixels.

At one time Richard had mentioned creating a play mode which would work with index positions in Keynote so that a user could export a Keynote presentation as a Quicktime movie and then initiate a cue which would automatically halt at each index position.

To do it the way Simon mentioned, your best bet would be to just create the the movie with the freezes a specific distance apart (4 seconds for example).
In cue 1 of the list initiate play once forward with outframe set to 100 frames (100 = 4 seconds in 25fps). In cue 2 set outfram to 200 and the file will continue to play for 4 more seconds. Continue this etc.

[Mr_P]

Re dipless x-fades...

theres a setting I learnt about today - which i was un-aware of...

You can change the way x-fades work in mix windows by selecting either 'Glitch A' or 'Glitch B'.

X-Fading works slightly differently in a mix window due to 'double transparency' - so theres a 'B' setting to compensate for this.

Try selecting 'Glitch B' from the 'Mix Type' options and see if this helps.


Simon


Cant take credit for this nugget of information - thanks goes to RB...

[SourceChild]

Actually, doing it with Glitch B would mean you can simply create opaque images and fade one up over the other.

[ping141]

Glitch B darkened my images significantly, but it did take away the brightness dip... What does Glitch B actually do, do you know?

[samsc]

double transparency when rendering into mix buffer -then again onto screen.

i couldnt find a solution to this - that worked both ways - last time i looked at this -and i was under a lot of pressure to solve it-

[SourceChild]

double transparency when rendering into mix buffer -then again onto screen.

i couldnt find a solution to this - that worked both ways - last time i looked at this -and i was under a lot of pressure to solve it-

What if you were to add a gain enhancement to the mix? Something where the gain is proportionate to the opacity?

[samsc]

What if you were to add a gain enhancement to the mix? Something where the gain is proportionate to the opacity?

no. its an alpha channel problem.
gain messes with colour.