Which still format will render quickest?

[Russell Lasson]

What computer are you using? What codec are you using on your timeline?

I’ve never seen a difference in rendering speeds of different type of still image formats. The big difference comes in when the still image files are huge. Adding the extra resolution sometimes only slows you down.

Do you ever change the scale to 60-70% or higher on any of the images? If not, you could probably be find making the images smaller in Photoshop or iPhoto before you bring them in.

-Russ

[Mark Maness]

Yeah… I agree.

Asking which codec is the fastest is a completely relative thing. This will vary depending on your computer and included hardware. BUT, the resolution of photos make a real difference in render times.

If you are just doing static shots, then scale them to your sequence size in Photoshop. If you are doing motion shots, render times will be much higher but, the cost of doing quality work is more important that render times and charge clients accordingly.

_______________________________

Wayne Carey
Schazam Productions
http://www.schazamproductions.com

[Chris Poisson]

I should have mentioned two things as pointed out, one, I’m mostly working in 8bit SD on a G5 dual with 4.5 gigs of RAM. The other is I am well aware of resolution and sizing issues, I am one of those anal people who make EVERYTHING 72 dpi at no more than twice my frame size, so my tuff is never too big.

I brought this up because a timeline with a preponderence of jpegs seemed to render slowly, and I wondered if FCP is having trouble rendering such a compressed source on an uncompressed timeline. Also someone on Ken Stone’s site aluded that jpegs could be my render problem, so I thought I’d bring it up here.

[Russell Lasson]

I like to use PICT files.

I’d suggest you take 10 minutes and run some time tests of the different formats.

-Russ

[Chris Poisson]

Good point. Think I’ll make a couple of identical sequences with pict, tiff and jpeg and put a timer on the renders.

[Steve Radley]

[Chris Poisson] “Think I’ll make a couple of identical sequences with pict, tiff and jpeg and put a timer on the renders.”

Let us know what the results were. I just did 2 jobs with stills I animated. I use tiffs, at 72 dpi, sized to around 1500 x 1200 pixels or so. It does take a long time to render on a G5 Dual 2.7 with 8 GB of RAM.

Steve Radley
Digitec
Orlando, FL
https://www.digitecinteractive.com

[Chris Poisson]

It must be crazy being in research, you get a wild hunch and test away, and nothing much happens. What a crashing bore.

Anyways, I took 11 jpegs of around 1440×900 or so at 72 dpi, and I did a save as to tiff, pict and png. I put :05 clips of each into a timeline, each in it’s own sequence. I did a little move and scale on each, exactly the same on all. Rendering each on a stopwatch gave the following results, after which I tossed all the render files and did it again, same results.

Pict: 3:22
Tiff: 3:13
Jpeg: 3:15
Png: 3:15

Adding the default transition to all took a bit longer to render but the results were consistent with the first round:

Pict w/trans: 3:50
Tiff w/trans: 3:37
Jpeg w/trans:3:32
Png w?trans: 3:35

The answer? Pict files seem slower to render, the rest seem about equal. Sorry this wasn’t more exciting…(yawn!) Oh well…

[Steve Radley]

Thanks for your indulgence in this highly scientific experiment. TIFFs rule! I sure wish there was one format that had much faster results.

Steve Radley
Digitec
Orlando, FL
https://www.digitecinteractive.com

[Russell Lasson]

ABSOLUTELY FACINATING!!!! (not)

Just kidding. Great job on putting a test together. It’s developing those kind of skill that will save the day during a middle of the night project when the cow has gone to bed.

Okay, the cow doesn’t really sleep.

Great job though. What were the differences in files sizes of the photos?

-Russ

[Chris Poisson]

Russ,

All were around double the frame size, give or take a few pixels if they were vertical, but all were about the same. I always work that way, about double my frame size, unless I need to do a really big move on one.