[Matthew Keane]

Hi,

I found some tests a while back on this site: https://macperformanceguide.com/
that suggest that 3 modules can be better than 4, but they were more geared towards Photoshop than AE.

I don’t know whether Apple mention it on their site or not, but some MacPro models will quite happily accept 8Gb memory modules, so 3×8 would give you 24Gb.

Matthew

[Matthew Keane]

Hi,

I don’t think it’s really an alpha channel as such, more that PDFs don’t generally include a background, since a white background is usually supplied by the paper when printed. A workaround is just to precomp the PDF and add a white solid behind it, or ask the person who prepared the PDF to add a filled background object, or open the PDF with illustrator and add the background yourself.

[Matthew Keane]

Hi,

If you’re trying to spread parts of the same image out in z-space to give the illusion of depth, there’s a script called something like ‘psd2layers’ which does just that with the layers of a PSD, and applies an expression to the scale parameter so that you can slide elements around while maintaining their apparent size. Might be worth searching for, if that’s what you’re attempting.

[Matthew Keane]

An alternative, if you have a recent Mac, would be to install VirtualBox/VMware Fusion/Parallels to run a virtual install of Windows, and then download the free Windows Media Encoder from Microsoft to encode WMV files for your client. It’s a bit more hassle to set up, but it means you can also test playback on a PC before sending the files to the client.

[Matthew Keane]

Hi,

Things will depend on which particle system you’re using. Particular, for example, allows you to choose a 3D ‘obscuration layer’ which will hide particles passing behind it. You need to be careful though as particles can drift in 3D space and occasionally pop through the layer, but if your particles are following a motion path, that may not be a problem.

The other solution – or workaround – is to duplicate the Particular layer, and sandwich your 3D glass layer in between the two. You can then use the z-buffer settings in Particular to show only particles in front of, or behind, the glass on the appropriate layer. You’ll probably only want to do this once you’ve locked down your particle effect, to avoid having to redo the changes, although you can use expressions to tie the Particular settings on the two layers together, and I believe there is also a 3rd party script that can create the duplicate layers for you.

Hope that helps…
Matthew

[Matthew Keane]

[John Pale] “I have read that MP4s that are encoded by AME do not support progressive download (meaning they don’t play until fully downloaded from the web)”

I found that to be the case with AME CS4 (I haven’t checked in CS5.5 yet), but I found a small utility ‘QTIndexSwapper’ which adds the necessary header data to the MP4 file so that browsers can load the fie progressively. It does add an extra step to the compression process, but seems to work.

Matthew

[Matthew Keane]

Thanks for the feedback – glad the show went well.

You’re right – I just did some tests and, with MPEG2, encoding fails at 1080/60fps but succeeds for 720/60fps. With h264 though, I managed to encode a 1080/60fps file, so it may be a limitation of the MPEG2 format.

[Matthew Keane]

Ah, hang on… If you have an adjustment layer with colour correction effects in there – is it possible that it is affecting the gradient layer and causing the blockiness? Try previewing just the gradient and the colour correction layer and see how that looks.

[Matthew Keane]

Then there’s something strange going on with your project. I’ve just tried a quick test with both methods (solid + ramp & shape + gradient fill) and I’m not seeing blockiness like that. I even tried with a very small shape layer which I scaled up and with the layer quality set to draft, but it still looks OK.

To be fair, the test image I used was only about 2000px across, so smaller than your comp settings – is it possible it’s a memory problem when working with a larger image? Also, my tests were with CS4 – did something change in CS5?

[Matthew Keane]

I’m pretty sure that the solid layer with the gradient effect will need to be precomped for it to work as a gradient map. But it sounds like something is strange with the gradient on the shape layer.