Chris Blair
Forum Replies Created
-
No reasons not to do it…but I do know that Adobe should fix their buggy products. We gave up on Premiere with version CS3, which also has quirky and sometimes crash prone Adobe Media Encoder issues. Thankfully, we have plugins that allow us to export to Procoder directly from the Premiere Timeline, which works 99.9% of the time and is rock solid.
But the only thing we use Premiere for is updates to old projects that were created on it. I also constanly run into quirky and just plain weird render and export bugs in After Effects CS3.
Anyway…Adobe Media Encoder is useless in my opinoin because of the bugs. If you can afford Procoder it’s well worth the money and will virtually eliminate the export headaches with it’s Premiere export plug-in.
Chris Blair
Magnetic Image, Inc.
Evansville, IN
http://www.videomi.com
Read our blog http://www.videomi.com/blog -
Why not just render that first graphic as a video file in whatever format you’re using on the timeline…then place that rendered file on the timeline instead of the graphic? Shouldn’t that in theory fix the problem if it’s the graphic causing it?
Chris Blair
Magnetic Image, Inc.
Evansville, IN
http://www.videomi.com
Read our blog http://www.videomi.com/blog -
There are DVD calculators available online to determine your compression settings. Google DVD bitrate calculate. If you’re going to standard def DVD, you’ll have to keep your total data rate (audio/video) below about 9MB/sec and if you want reliable playback across older DVD players, below 8MB/sec.
In terms of compression settings, the MPEG2 spec for DVDs is pretty tight. Your video MUST be 720×480 (4:3 or 16:9), and must use an MPEG2 codec using 4:2:0 chroma sampling. Audio can be PCM, MPEG1 Layer 2, or AC3. AC3 is the most efficient codec of the three and can give you more “headroom” for higher video data rates.
Make sure you set your encode to 16:9 so that flag gets set in the video. If you don’t do that, your video won’t letterbox on a 4:3 TV.
There are a bazillion MPEG2 compression options out there, from freebie programs to $49 programs, to professional products like Telestream Episode and Sorensen Squeeze. Apple’s compression app is kind of in the middle, good for most things and will likely be fine for what you’re doing.
High-end faciliites probably don’t use software at all, but instead use hardware based encoding solutions. Digital Rapids is one such solution. They have all sorts of hardware based color correction and noise reduction features built in. Although with the power of software based apps and the sheer speed of desktop computers, there are probably some that use software. However, it would not be out of the ordinary for a high-end facility to use Telestream products, which run the gamut in terms of price and functionality.
If you just do some googling you can probably find a ton of information to answer all of your questions. The biggest issue you’re going to face is getting a clean down-conversion from HD to SD MPEG2. Invariably, scenes that are gorgeous in HD can end up with weird jitter where there are thin lines or thin patterns, like say a rooftop, or patterned clothing on talent.
But…if your content is good, the average viewer doesn’t even notice these things. We’ve found that adding a slight amount of blur to an HD piece to soften thin lines or patterns can work wonders. Also adding a tiny amount of grain can help with color gradations and the like. One thing you’ll want to also do is convert your color space from 709 to 601 SMPTE. If you don’t, you might get color shifts and gamma changes. Again…google this. Most mid to high-end compression software has filters for this….at least on the Windows side (which I typically work on).
Good Luck.
Chris Blair
Magnetic Image, Inc.
Evansville, IN
http://www.videomi.com
Read our blog http://www.videomi.com/blog -
Upping the bit depth won’t help your final compressed video, especially web video. When you go to the final compressed resolution, you’re reducing everything, bit depth, color sub-sampling etc. All that extra information you added during creation gets thrown away. Same problem. You can watch network television and see this issue over and over on expensively produced, high-end television commercials and show graphics, where the designers uses a symmetrical gradient.
The only fix we’ve found is the grain plus irregular gradient tricks.
Chris Blair
Magnetic Image, Inc.
Evansville, IN
http://www.videomi.com
Read our blog http://www.videomi.com/blog -
To add to Bob’s point….will this setup work reliably when you’re working on deadline and you have drives that are 60-70% full?
We use an Apace vStor system with 4 edit stations attached via gig-e. When the drives are empty, we can layer 4 streams of uncompressed SD video, add multiple real-time DVE effects (our system uses hardware based DVE processing), set-up a minute-long looping sequence, and this timeline will literally play for hours looping continuously with no hiccups. We can even log and capture in a 2nd suite at DVCPro50 quality at the same time!
Fast-forward 6 months with drives now 75% full and three edit suites running simultaneously. We’re lucky to be able to layer 4 streams of 2:1 compressed SD video in each suite…(at about 10MB/sec each) with the same DVE effects before we start seeing playback issues. We can still edit 6-8 total streams of compressed SD across all the edit systems, but it WILL not handle uncompressed across all those once the drives start filling up.
Now each of these scenarious is ridiculous because they would NEVER occur in a typical editing environment. If we have 3 edit systems running at the same time….98% of the time the timelines are playing back 1 or 2 streams of compressed SD video (typically DVCPro50 quality). There is rarely even 4 or 5 total streams playing simultaneously across all 4 systems. In THAT scenarios, everything runs fine.
My point is, wait until you get into typical editing scenarios and see what happens. I bet you’ll start to see issues somewhere, somehow. As Bob notes, the ONLY way this is working is because of the relatively low-data rates. And to second his reply..if it DOES work for you…fantastic! I just know that in our place, we can’t afford to cross our fingers and hope something unproven works when we have clients sitting there paying $175-$200/hour.
Chris Blair
Magnetic Image, Inc.
Evansville, IN
http://www.videomi.com
Read our blog http://www.videomi.com/blog -
In the end…these are all just tools. I’m not a fan of the DSLR craze, mainly because of the numerous workflow issues involved that in my opinion add a tremendous amount of time and cost into using them…not to mention so-so audio and numerous hacks needed to use other professional tools with them.
But that doesn’t mean the DSLRs don’t have a use…for many they do. Same with the Flip. I’ve seen people use it and the video they get looks just like what your Uncle Zeb would shoot with his crappy little $300 DV camera. But I’ve also seen stuff shot with them that looks pretty darn good. The audio always sucks because there’s no way to control levels or add a good quality mic (at least not to my knowlege). Now why anyone with access to quality gear would use the Flip INSTEAD of that gear is a HUGE mystery to me. There doesn’t seem to be any value in that. But as an extra camera or as a camera that can get into places others can get….sure…it’s a decent option. But just like the DSLR, there are a dozen hoops you have to jump through to insure that you can use your other professional gear with it (if necessary), not to mention other workflow issues to use the video in an editing environment. Don’t forget it uses VERY high compression right off the bat and a codec meant for final delivery, not raw footage that’s going to be edited and likely compressed several more times.
I think part of the “craze” you’re seeing with corporate people using it is they think getting video of the client up on YouTube or Facebook trumps projecting a good image with high quality and meaningful content. I disagree. I think seeing poorly shot video with no lighting and hollow, auto-gain sound can make an otherwise impressive company look ridiculously amateurish and abashedly cheap.
Chris Blair
Magnetic Image, Inc.
Evansville, IN
http://www.videomi.com
Read our blog http://www.videomi.com/blog -
When I suggested added grain I’m talking about an almost imperceptible amount…like at a level of 1 or even .5 or somewhere in between.
It’s just enough to help eliminate the banding but not enough to ruin the smooth gradient. Another trick we use is to not use symmetrical gradients (if you can control it), meaning a straight (vertical or horizontal or even diagonal) or circular or oval gradient pattern. Instead, make the gradient occur across a wavy or irregular shape. We often take an extremely blurred animated image, make it B/W, then blend it with our color or gradient to make the transition from color to color be irregular. This and that tiny bit of grain will eliminate the banding issues, especially if it’s going to get compressed.
Good Luck
Chris Blair
Magnetic Image, Inc.
Evansville, IN
http://www.videomi.com
Read our blog http://www.videomi.com/blog -
Goodness, if ProRes using 4:4:4 color sampling produces banding, then something’s amiss in the video. One trick we’ve used is to add a touch of grain to scenes with smooth gradients and that often is enough to eliminate the banding. But if the video has a lot of these subtle color grads throughout, that might not be an attractive option.
We’ve found H264 to handle gradations quite well…and there are H.264 profiles that use 4:2:2 (and maybe even 4:4:4 color sampling…so maybe see if you can set it up that way.
Other than those couple of suggestions…I’m not aware of much else you could do. Are you sure the banding isn’t being produced by your media player upon playback….and NOT by the encoding or codec??
Chris Blair
Magnetic Image, Inc.
Evansville, IN
http://www.videomi.com
Read our blog http://www.videomi.com/blog -
There really isn’t any such thing as “European codecs” to my knowledge. There are international based video standards that differ from our NTSC standard, but the difference is in pixel resolution and frame rate, which should really be a non-issue from a codec standpoint if the original footage is transcoded properly to NTSC for editing in Final Cut.
I’m not saying you didn’t transcode it correctly, but tell us what you’re coming out of Final Cut with in terms of the files. What is the pixel resolution, what codec are you using, what format are you saving to (.mov, mp4?)….etc.
It would also help to find a media player or tool that can tell you (and us) the specs of the original files, meaning the info above, as well as the chroma sampling (4:2:2, 4:2:0, 4:1:1 etc.) since you mention weird trippy issues which I’m assuming are color issues. Trying to upsample from a reduced color space might cause some weird color issues (but really shouldn’t).
Chris Blair
Magnetic Image, Inc.
Evansville, IN
http://www.videomi.com
Read our blog http://www.videomi.com/blog -
Is the logo a static graphic keyed in the video or is it video with alpha? The only two possible reasons I can think of are:
1. The logo has some pretty thin lines, which are going to get crunched when you resize and compress. Not much you can do about that other than to blur the logo slightly or try to fatten up those lines a little.
2. If it’s video with alpha, does it match the interlacing of the project? If there’s a mismatch there (it’s interlaced and the video is progressive), you’d see the problem noted.
Other than those two things…I can’t think of a reason why it would look that much different.
It would also help to give more information…like the specs of the original video and timeline…how you’re keying the graphic (noted above), what you’re compressing to size-wise and codec wise etc.
Chris Blair
Magnetic Image, Inc.
Evansville, IN
http://www.videomi.com
Read our blog http://www.videomi.com/blog