Text size change in share YouTube format

[Jason Brown]

This is because you are using a non-square pixel sequence and a text generator filter.

I’ve been trying to figure this out for months. U are the only person I’ve ever heard experience this. Plz confirm u r using a non square seq and I’ll let u know my solutions…

[Jason Brown]

Dave,

This haPpens with the send to compressor command as well. I don’t use the share function.

[Jason Brown]

[Dave LaRonde] “I don’t trust that “Send to Compressor” thing.”

Dave…I’m losing faith in you! 🙂

I use that function literally – daily! It works perfectly (with this issue being an exception). It is ONE of my main 2 reasons I suggest upgrading to FCP7.

Yes, your workflow of exporting a same as source type of file in a QT container will fix the issue. The only time this issue occurs is when using non-square pixel comps (DVCPROHD, HDV, NTSC…etc) in conjunction with having an FCP generated font.

The fix – precomping into a square pixel sequence – or not using the text tool to generate any text…use motion or photoshop.

I haven’t tested…and would be interested to see if you have the same issue with a QT reference file…because that is basically what the “share” and “send to” functions are doing.

I love the send-to workflow…it is super fast, and you don’t have to create any intermediate files.

-Jason

[Jeremy Garchow]

Dave is right. Export a qt, then drag that to compressor. Less problems and faster as you can setup a virtual cluster in compressor. The share or send to function can’t use virtual clusters. If you’re scared of compressor, you can set it up once and create a droplet of your favorite setting(s). Then you can drop your export on the droplet and use the virtual cluster as well.

[Julie Gade]

Jason, yes I am using a non square pixel format (1440×1080). That explains it!

It stretches the text when I “send to Compressor” as well, so as I hear you there are 2 solutions:
1) Use PhotoShop or other files for text, or 2) export as QT and then import into compressor.

Thanks you all!

[Matt Lyon]

I also noticed this problem as far back as FCP 4.

IIRC, another workaround is to “force render” the sections of your timeline that contain text. Now those pixels will be “baked in,” so Compressor can’t screw things up when it renders.

I’m not sure if the problem is limited to text generators. I seem to remember “shape generators” had problems too, but I’m not positive. This issue always makes me think twice about using the “send to compressor” option. It’s definitely a good idea to spot check your quicktime after compressing this way!

Matt Lyon
Editor
Toronto

[Jason Brown]

Hey Jeremy,

Great information! I didn’t know this…MOST stuff I do is h264 and I use Matrox CompressHD…so virtual cluster isn’t something I use often…but good to know that “Send To” can’t take advantage of that!

-Jason