-
An experiment for a CS5 / 5.5 user?
Hi,
I’ve already used up my “try before buy” CS5 installation so has anyone got CS5 / 5.5 and a few minutes spare?
Could you try:
1. Make a 1920×1080 psd file (with some smallish text).
2. Make a low res 16×9 prem pro sequence (eg 160×90)
3. Put the psd file onto that timeline and scale it down to full-frame.
4. Export the timeline using Adobe Media Encoder as a 1920×1080 h.264.
How does it look?I was hoping that this workflow would “do the math” and produce a nice crisp export of the 1920×1080 psd file, instead of looking like it’s scaled it down to 160×90 then scaled it up again. In CS4, the Export preview window *looked* like it was going to do what I need, but the MOV I ended up with was all blurred and jaggly. 🙁
I get batches of movies in a variety of formats/frame rates/sizes/volumes/field-settings/letterboxing etc, to which I have to add a fade-to-black slide at each end. I’d hoped to make a native-size sequence for each movie, then add a before and after sequence onto each, then export at native size, compressed, DV and 1080 TIF sequence, with as little image degrading as possible, and making it easy to update a slide and re-export the whole lot if needed.
The other option is to blow up the smaller films to match the 1920×1080 slides, then export to smaller sizes, but I think this will degrade the film quality by expanding them and shrinking them again.
So I think my QuickTime Pro workflow (where I have to make lots of slide movies of different sizes) still seems the best option.
Any suggestions gratefully received.
Thanks
– Paddy -
3 Replies
-
Hi Paddy,
I really don’t understand your workflow here.
Can you set for us the actual output of what you want the final output to be and we can test this for you in a realistic and sensible way.
To have something 1920×1080, then put it in a 160×90 timeline, only then to blow it back up again at the export to 1920×1080 is just going to look like rubbish.
Granted you may have stumbled onto a bug in that the interface thinks it will come out looking clean, but it’s just a trick, and wouldn’t reflect simple mathematics of an image being told to be small and then blow up the pixels to fit a new, 10x dimension.
I want to help you find the best workflow, but lets be realistic about the needs first.
Cheers JB
Jon Barrie
Adobe Video Solutions Consultant ANZ
Jon’s YouTube Tutorial Page
follow Jon with twitter -
Hi Jon,
Thanks for your reply.
Here’s my workflow:
i) I get a pair of 1920×1080 slides to add to the start and end of a bunch of films with a fade and black.
ii) The bunch of films are a mixture of PAL, NTSC, and various other files of different codecs, sizes, aspect ratios, frame rates, letterboxed or not)
iii) I need to output 1080p TIF sequences (for DCP), DV-PAL (or MPEG2) for DVD production (anamorphic for the widescreen ones) and 640px wide h.264s (for web).What’s worked best for previous projects has been to
i) crop letterboxing off the movies in QT Pro
ii) create a before.mov and after.mov slide movie for all the different sizes of movie I’ve got
iii) Paste them onto the cropped movies in QT pro and export (deinterlacing if needed).This gives the best quality, but if (when!) someone spots they’ve made a mistake in the slides, I’ve got to pretty much start the process from scratch.
The thing about the 160×90 sequence: if I received a tiny 160x90px film (for the sake of argument), I’d want to add the 1920×1080 slide to that in a 160×90 sequence and export it at its native size, but also export it at 1920×1080, which I think should use the slide at its native resolution and just upscale the tiny movie – of course the movie would look awful, but at least the slide shouldn’t.
I hope that makes sense.
– Paddy -
Let me see if I understand the outcome here.
You get slides, which are always 1920×1080, that top and tail video content, which could be a variety of formats and dimensions.
Why are you getting video content with different dimensions for starters. Perhaps everyone can gain better outcomes with a strict format to work to, nobody watches anything below 360 wide and that’s only for mobiles these days, which get sized automatically by the flash media server in Youtube or something similar.
160×90 has clearly been tampered with and was never shot or originated from this dimension. Blowing this up will always results in horrible results. Best Idea is to keep the pixels accurate and have it live in a boxed frame when going up to another resolution. Even doubling it is mathematically going to results in some pretty blurry and horrible results. Especially going to even DV dimensions.
I would look at speaking to those that provide the media content and explaining the value of keeping everything at the highest possible quality at recorded stage.
There must be a point where clients need to understand what they are forfeiting in being behind the times and the users expected quality of experience.
Where is the internet content being hosted/managed?
Most used dimensions these days are 1280×720, that goes for HD, which can upres to 1080 quite well, and down to DVD nicely too, but can stay as 720 for web.
Streamlining is where I would be looking and making strong points of delivery experience. If the content you are getting is old, and there are no more higher quality or dimension versions of it then you are stuck with the lowest quality experience and anything you get that is lower than the lowest export dimensions I’d consider not increasing the scale by any more than 25% and leave it in a black bordered box.
Perhaps your clients are not willing to shift their thinking or future proof themselves, but I’d not expect the old saying to be any different these days “rubbish in, rubbish out”. There are some expensive ways to upres, but even they require an SD dimension to get something decent out for HD.
How do you feel a conversation with the client/s go around supplying higher quality video?
JB
Jon Barrie
Adobe Video Solutions Consultant ANZ
Jon’s YouTube Tutorial Page
follow Jon with twitter
Reply to this Discussion! Login or Sign Up