Perfect Image

[andy_starbuck]

Tutorial to help you sort out the quality issues (Part 2)... as requested.

Tutorial to help you sort out the quality issues (Part 2)... as requested.

(...part 2 of the tutorial, continued from previous post).

10) When you create a new project, Premiere adds a default sequence called "Sequence 01" into the bin area. Click on this sequence and make sure its name is revealed on the tab in the timeline. Then drag a short clip from the video you captured and drop it on the timeline.

If everything is set correctly, then the video should drop into place with NO colored line above it in the timeline. This means that the audio and video in the sequence and project are set the same as the video that was recorded on the camera. If you see a colored line above the clip in the timeline, then something is wrong. Some part of the clip has to be rendered in order to be previewed. Rendering means that Premiere will create a special preview file that converts the video from whatever mode it was recorded in to the mode defined in the project preset and in the sequence settings. You shouldn't need to render standard DV footage in a standard DV project.

If you see a red line above the clip, go back and check all your settings in the camera and in the project to make sure they are identical.

11) Preview the sequence to verify the quality.

Press the "Play" button. In the upper right corner of the screen you should see a tab with the title "Program: Sequence 01" on it. Press the right triangle "play" control beneath the window to preview the sequence.

There are two controls that may be important. The first is the pull-down menu at the center of the window that probably says "Fit". This tells Premiere what size to view the video. Depending on the size of your monitor, you may only have enough room to view the whole image at 25% of its normal size, or you might be able to view a small part of the image at 100% size. Since we are looking at the video for quality purposes, rather than for editing, you will probably want to play the video at 100% so you can see everything.

Also, right click on the viewer window. A pop-up menu comes up. There is an item on the menu that says "Quality", and when you move the mouse over this, a fly-out menu shows "Highest quality, Draft quality, and Automatic quality." This is the playback quality. For our purposes we will want it set in highest quality.

Does the video quality look good?

12) Time to output the sequence.
a) Select [File]->[Export]->
Export
Video and audio
Entire Sequence

Filetype
Microsoft DV AVI

Video Settings
Compressor: DV NTSC
Frame size: 720h 480v (0.900)
Frame rate: 29.97 frames/second
Pixel Aspect Ratio: D1/DV NTSC (0.9)
Fields: Lower Field First

Audio Settings
Compressor: Uncompressed
Sample rate: 48000 samples/second
Channels: Stereo
Sample type: 16-bit


c) Now click on [Settings...] to open the Export Movie Settings dialog box.

d) Verify under the "General" settings that "Microsoft DV AVI" is selected.

e) Click on "Video" settings and verify that "DV NTSC" is the compressor being used.

Click "Okay" to return to the main export window.

f) Now you can choose a file name and click "Save". For this example, use "testing.avi". The movie will be rendered.

13) Verify the result.

Use Windows Explorer to locate the output video file "testing.avi". Double click on this file to launch Microsoft Media Player, which will load and play the file. You should be able to see a good quality result. If the video that is output doesn't look good, go back and verify all the settings for Export Movie to make sure that they are set correctly.

14) Now look at the bin area in Premiere where your clips are. Premiere has added the movie you just output into the bins. There is nothing to distinguish this output footage from video that you captured from the camera, even thought this rendered video is now 2nd generation and should not be used in subsequent edits.

a) Right click in the bin area. A pop-up menu opens. Select "New bin". Give it the following name: "DO NOT USE IN EDITS".
b) Now click on the movie file name, in my case "testing.avi", and drag it into the "DO NOT USE IN EDITS" bin.

15) Nesting sequences

In Premiere you can embed or nest one sequence in another.
a) Click on [File]->[New]->[Sequence] to create a new sequence named "Sequence 02". The new sequence opens in the timeline. From the bin area, drag "Sequence 01" onto the timeline and drop it on "Sequence 02".

This is how you break up your work in Adobe Premiere. When you go to render out your movie, the nested sequences will all go back to the original files that you captured from your camera and give you the best quality output.



If you've gotten good quality up to now, then your workflow is functioning.
But let's go a few steps further to see how you can screw up the quality of your video even from this point forwards.

Now, let's review what NOT TO DO.

Open the bin called "DO NOT USE IN EDITS" and drag your output file "testing.avi" to the timeline and drop it on "Sequence 02".

You now have two copies of the same video in "Sequence 02". But are several differences between the two copies.

16) Nesting sequences versus using rendered movie files.

Verify that you have two copies of the same video in "Sequence 02" by playing the video. They probably look identical at this point. But they are not. The first one, which is a nested link to "Sequence 01" preserves your edit decisions and preserves the original quality of the video. The second one, "testing.avi" no longer has a relationship to the original video on the tape and is also a generation of quality removed from the video you recorded. Using "Export Movie" and then subsequently using the rendered movie in a sequence is a common cause for loss of video quality during an edit.

a) Let's make a change in "Sequence 01" and see what the results are. Click on the "Sequence 01" timeline tab. Select the razor blade tool. Then click in the middle of the video on the timeline in "Sequence 01", dividing the video in half. Select the pointer (arrow) tool. Click on the second half of the video and press the [delete] key on your keyboard to get rid of it.

b) Now click on the "Sequence 02" tab in the timeline and let's see what happened.

The change in "Sequence 01" has been propagated (carried forward) to "Sequence 02". The second half of the video that you just deleted in "Sequence 01" is also gone in "Sequence 02".

HOWEVER, take a look at the video in "Sequence 02" from "testing.avi". It is unchanged.


17) Differences between nested sequences and rendered movie files in an edit.

The nested sequence preserves the relationship between your project and the original tape. So you could output an EDL (Edit Decision List) or save your Premiere Project and use that to recapture the video from the original tape and reconstruct the entire edit. However, you can't do this when you export a movie file and then re-import it, such as the example testing.avi. Remember that testjng.avi was rendered from "Sequence 01" before we made changes to it. Then we cut part of the video out of "Sequence 01". Premiere no longer has the "recipe" for making testing.avi !!! The export movie "testing.avi" represents decisions at a moment in time. So you can't use the information to recapture and reconstruct the project. Some of your edit decisions have been lost by exporting the movie and then using THE EXPORTED MOVIE instead of nesting "Sequence 01".

Also, testing.avi is has been through decompression and then re-compression when the movie was exported. So it really contains second generation video rather than first generation video. It is like a photopcopy of a photocopy; there is generational loss. This is only second generation video. If you keep going, and "lock in your edits" by exporting a movie every once in a while, and then use that exported movie instead of the the original sequence, each time you do that it is another generation of quality lost. It is really easy using this technique to end up with 7th or 15th generation video in your final product. And perhaps worse, since this method destroys the continuity of your editing decisions, you can't easily re-assemble the video from first generation clips. If you are using a non-DV codec for the export of the movie, such as a compression intended for use over the web or to play back video in a media player, then the quality drop will be worse than the generational loss from the DV codec.
" type="application/x-shockwave-flash" pluginspage="http://www.adobe.com/go/getflashplayer" width="576" height="324" allowFullscreen="true">

 

[Brandon Rice]

Wow, great, great post! This should be a sticky in the Premiere forum!

 

[Kyle Stebbins]

andy starbuck, you deserve to be a mod. -- and whitewood, dude... you get way to fed up with people trying to help you its almost like you dont appreciate anything. -- i use premiere 1.5 and let me tell you one thing... if you want high quality, small files... dont export compressed, export UNCOMPRESSED avi's THEN bring them into Quicktime Pro and export h.264... have you tried this?

film -> capture -> edit -> export DV AVI -> open in QT Pro -> export h.264

 

[Kyle Stebbins]

the .avi will look awful in QT, but ignore that... it wont turn out that way once QT's had a chance to turn it into a .mov file.

 

[dailyrushes]

Wow, that's a lot more thorough than I was thinking, but nice job, Mr Starbuck! I'm even saving this to my own computer for future reference!

 

[WhiteWood]

Can't someone just give him the exact menu sequence of settings you use for capture, editng and export? It would be so easy...my Premiere is on another hardrive (requiring reboot/same computer/no internet), or I'd do it right now...if that doesn't happen, I'll take the time to do just that.

Whitewood, if that doesn't bring joy, it's not that hard to just import/edit/export a short clip to see what you're doing wrong. Experiment, try things, learn! You can't depend on others to give you all the answers or you'll never get ahead, because in the rush of daily living, they don't always have the time to understand what you're asking.

But an exact sequence of what worked for someone else usually will solve the problem. I doubt seriously it's a Premiere problem, but just a settings problem.

Anyone?

man, I don't mean to sound like a pissed off dependant. lol. I was just REALLY mad. I filmed this short JUUUUST to test my settings, check it out:

http://www.youtube.com/watch?v=zU0csIAlx3o

You see how it looks like ****!! That's how ALL my videos come out. I made SEVEN-7 different versions yesterday! I toggled with compressed, uncompressed, mpeg, wmv, quicktime, avi. I've converted the video to mpeg 1-2, Quality 100, quality 0, bits...EVERYTHING...I don't knwo WHAT setting it is...

 

[WhiteWood]

#1) The word "sh*t" does not communicate much about the results that you are getting, making it difficult to identify the source of the problem. A screen capture and some actual description of the problems you are noticing would help a lot. You began this thread by asking for recommendations about the best settings. If you want specific advice, please get specific about the problem and provide examples.

#2) You say the video is capturing just fine. First and foremost, you need to be using a DV codec and working exclusively in a DV codec like Microsoft DV. Many of the codecs are intended for low quality high compression output for use on the web or on a DVD or CD, and these "publishing" codecs are not suitable for use in editing.

#3) The timeline of the project in Premiere must match the format that the video was shot in.

#4) I've seen a lot of quality problems from Premiere when people "render out" their intermediate edits and re-import them. Premiere makes it easy for people to misunderstand editing workflow because it always adds the link to the rendered file back into the bins. A lot of people think that you are supposed to "lock in" your edits on a sequence by rendering it out to disk, and then using that rendered copy of that part of the video in your edit. Don't do that! When you render out the video it goes through another compression. Every time you re-compress the video that way there is another generation in loss of quality. If you do that seven times, then the seventh time you use it in a sequence it will be seven generation video and look horrible. Plus if you mix it with other video of different generations then the quality will jump around. The correct way to use this program is to embed sequences into other sequences, not to render out and embed the newly rendered clip in other sequences.

#5) If you are misusing Premiere by rendering the video out and using the rendered clips instead of rendering only the whole project at completion, and especially if you are rendering clips out using publishing codecs instead of editing codecs, then you are going to get really ugly results from Premiere. In my experience these are the two most common errors that turn good footage into useless crud. Of course maybe you are doing something completely different. Without any other information all I can do is take pot shots at the problem.

Here are the stills:

http://i36.photobucket.com/albums/e31/JAIEILLA/girls1.jpg
And here is the video:
http://www.youtube.com/watch?v=zU0csIAlx3o

 

[WhiteWood]

andy starbuck, you deserve to be a mod. -- and whitewood, dude... you get way to fed up with people trying to help you its almost like you dont appreciate anything. -- i use premiere 1.5 and let me tell you one thing... if you want high quality, small files... dont export compressed, export UNCOMPRESSED avi's THEN bring them into Quicktime Pro and export h.264... have you tried this?

film -> capture -> edit -> export DV AVI -> open in QT Pro -> export h.264

Tired of people saying that. I am appreciative. Don't know how to word it unless I type "I'M APPRECIATIVE OF YOUR HELP!!" Sorry to say, but nothing has worked for me so far, so when it does, I will be sure to type it out plain and big for the world to see.​

 

[andy_starbuck]

Great Whitewood! Thanks. I watched the video and looked at the stills. Let me tell you what I think is going on.

Exposure

First, you need more light. To get the images that dark, you will need a ton more light and then turn down the iris. That will get more separation between the foreground and background and get more modeling on the actors.

Second, if the autoiris is making the image too dark, which is likely, then turn the iris to manual and open it some.

Part of what is happening here is that you can't get sharp focus on the original video because the light is too low and the aperture is set too wide. The smaller the iris hole, the more focus there will be -- sort of like a pin hole camera. If the iris is wide open, then the image tends to be less sharp. And it may be if you are using autofocus that it doesn't have enough light to operate properly, so its not getting focused. Recommendation: add a lot more light and then adjust for too much light by shutting down the iris a bit, and this will sharpen the image a lot.

As someone said, "The camera eats light. Feed it and it will be a faithful companion. Starve it and it will bite you!"

Contrast

Okay, another problem is the position of the actors relative to the brightly lit monitor. Video cameras have a limited contrast range. Maybe 1/10th the range of film and 1/100th or less the range of the human eye. We can stand inside a dark room and look outside to a brightly lit day and see detail on the lawn at the same time as we can see detail inside the room. Film can only do that under very special circumstances, and video can't handle that range of difference.

So the problem is that you have a bright light source right behind an actor with a dark complexion. You could get a lot better image if the PC monitor was sideways to the camera so that its light was helping to illuminate the actor's face rather than contrasting against it. I would either turn the monitor off, put up a really dark contrasting color, like a navy blue, or maybe put a white sheet of paper on the monitor to generate a lot of white light, and position the monitor so it doesn't show light to the camera -- only reflecting off the actor.

When I photograph actors with a darker complexion, I tend to use a lot more light and to let white objects blow out a little bit in the shot, and then correct with the iris or even the neutral density filter. Here is my reasoning: the most important thing in the scene is the emotion and expression on the actor's face. So you want to get really good modeling of light around the face, showing off the roundness of the muscles. It is the volume of these muscles in motion which is recognized by the brain and processed to recognize expression and from that to infer the emotion of the character. Since video has limited contrast range, make sure you are getting the actors expression first, and then deal with lighting the rest of the scene. You might even have to do 3-point lighting on the main actor or actors and then bring in smaller lights to light up the set behind the actor so that it doesn't just fall into black or get blown out into white. Ideally, get good modeling of the actor's expression and then add enough light in the background to get some detail and some sense of the location.

You can do all this with cheap lights. And you can use 75 watt daylight bulbs.

Focus
I think the focus was failing you because of low lighting. Just a guess.

White Balance

Make sure that you have all your lights in position and turned on before you do a manual white balance. Don't catch anything in the scene except a white card or a cloth or whatever -- no hands or anything in the frame when you set the white balance. The camera needs to know exactly what white looks like in this light so that it can get the colors right. The presets are not going to work in this environment, especially with the light pointed down at the red pool table. Its dumping TONS of red light into the room which is screwing up the coloring of everything. By setting the white balance every time you change position of the lights or the actors or big furniture in the scene -- anything that changes the color of the light bouncing around -- then you'll get much more faithful color and it won't look all red or yellow or orange. That color is typical of the kind of warm light that light bulbs generate. Get some of those "daylight" bulbs and replace the light bulbs in your practicals (the lights that are in the scene) with higher wattage daylight bulbs. They'll show more blue color and provide a better color rendition in the video.

Interlacing

Are you shooting in 60i ? 30p ? 24p ? 24pA ? I would guess that you are shooting in regular video 60i mode. The difference between interlaced and progressive video is that interlaced video divides each frame into a top field (all the even numbered lines) and a bottom field (all the odd numbered lines) and displays one and then the other every 60th of a second. What happens is that when there is quick motion, the object that is moving will move between the time that one set of line is recorded and that the next set of lines is recorded. The result is interlacing artifacts, or "comb teeth". To fix this, we have progressive video. 30p and 24p. Progressive video records the video lines from top to bottom in order, rather than field at a time. The result is that when something moves rapidly, it blurs instead of turning into "comb teeth".

30p is cool because it gets rid of the interlacing artifacts and you can intercut it with 60i footage and work with it on a standard 29.97 timeline. However, it doesn't look exactly like film.

24p is intended to work exactly like film. Here is the difference. At 30p, when someone is walking and swinging their hands, the ends of their hands and wrist will probably still be in focus. But at 24p, their hands will blur out more like in a movie. So motion blur is directly a result of choosing 24p or 30p or 60i. If you choose 24 frames per second, then there is another problem, and that is something called "strobing". Basically, if the camera pans too quickly, the illusion of continuous motion will be interrupted and it will start looking like separate frames of video instead of a moving picture.

If you were shooting in 24p instead of 60i, then the problem is pull down removal. What happens is that firewire requires 30 frames per second to transmit video from your camera to your computer. The geniuses at Panasonic figured out a couple of different ways to “top up” the 24 progressive frames per second to 30 frames per second. These extra frames have to be removed before you can edit the original 24 frames on a 23.98 timeline. Sometimes programs will screw up the pulldown removal, resulting in some of the filler frames getting through. And then the frame looks all interlaced . But since the interlacing artifact show up when people move, I think its probably the 60i/interlacing problem rather than anything with pull down.

Computer Monitor in the shot
Another place where interlacing artifacts are showing up in your video is when the actor moves in front of the computer monitor. Now this isn’t just motion interlacing artifacts. Its also because of the monitor. See, the monitor also has scan lines and is also flashing an image up several times a second. So it is basically a strobe light. Only computer monitors tend to blink at a different speed then the 60i of a video camera. The result is that you have a strobe light behind your actor that is adding to the motion artifact problem. When he moves, the motion is getting caught in the strobe effect, and that is making the “comb teeth” around him even when he is moving just a little bit. Again, nix the monitor. In fact, I’d suggest moving the monitor to the side and then put a light behind the monitor so that the camera doesn’t see it, and point it at his face. A regular light operates at 60 cycles per second – same as the camera. So it won’t strobe. That will get you the appearance of the computer light, extra light on the face of the actor, and eliminate the strobe problem all at the same time.

Composition
Finally, the camera is pretty much locked down in a wide shot. You should take the opportunity to use that zoom lens and get some close-up shots of the people’s faces. Maybe even some hand-held shots. By varying the distance and the angle of the shots you are taking you’ll have a lot better option when you go into edit. And some of the images will be better lit than others and show better detail than others. So for example, you probably won’t be able to show a good gesture at the same time as you get a good shot of people’s expressions. Especially if the video is going to show on the “small screen” online, you will want to vary the distance. So get a good shot of gesture and then get a closer shot of expression, and edit between them. That will increase the expressive range of the video and also you won’t be stuck with bad lighting or a bad shot. You’ll have more options available.

Wow. Now I’m looking back over what I wrote, I don’t see a single thing about Premiere in there. It could all be camera technique.

Good luck Whitewood!

Hope you get the shots you are looking for. I know it can be frustrating to not get the results out of the camera that you were hoping for.

At your service,

Andy

 

[WhiteWood]

YOU ARE A FRIGGIN GENIUS

I mean, I haven't TRIED anything you've suggested yet, but as soon as I do, I will put ALL of this to use and come to YOU with the result. Alot I didn't understand, but I know I will as time progresses.

Ok, I shoot ALL the time in Progressive. 30, 24 and 24advanced. I rarely EVER use 60...EVER. I think the problem MAY have been that when I took it into the editor, I chose a diff option that wasn't 24P, but rather for 60...I don't know. I'll check it all out later. But I'll mos DEf be updating you tomorrow, as I'll be editing something new, Im going to tryit your way.

 

[andy_starbuck]

If you are using Premiere 2.0 then you are pretty much safe with any 24p setting. It does a good job of recognizing and removing pull down. If you are using Premiere 1.5 it has a bug with 24P and won't work with it correctly. It only handles pull down correctly for 24pA.

In 2.0, after you import the footage into your bin area, right click on it and choose "Interpret Footage". That's where you'll find the settings to change the frame rate and to remove pull down.

 

[dailyrushes]

More good advice, Andy. I have been really pissed about 1.5 not handling 24P properly, and it's nice to know it does handle 24PA! I didn't know that.

Whitewood, Starbuck just gave you a lot of the things you should have learned in film school (OK, so maybe you didn't go) but at least now you can apply them and get much better results...

I mean, I used to wonder why you couldn't shoot a feature film in 2 hours! Until I got my first video camera and tried to shoot something! Boy, was I surprised! (I went to film school after that)

 

[andy_starbuck]

Another thing about lighting...

Another thing about lighting...

The graphic below illustrates the inverse-square law.

Here is the deal: Our eyes adjust so well to light that they fool us. In the images above we are moving a light back from a subject and looking at how bright the light is on any single square.

The image on the left shows the light on one square, and it is really bright. Just about 100% of the light is on that square.

When we move the light back so that it covers four squares (center image), our eyes and our intuition will tell us that the light is about half as bright. But we would be wrong! The same amount of light is now being shared by four squares -- so each square is getting about 1/4 of the light. The light on any single square isn't half as bright -- its 25% as bright.

If we move the light back further so that it covers nine squares (right image), our intuition is that the light is maybe 1/3rd as bright as it was when it covered only one square. Wrong again! The same amount of light is now shared by nine square, so each square gets only 1/9th the light. It's not 33% as bright -- its about 11% as bright.

Fortunately, it works the other way too. If our light is positioned far away, as in the image on the right, we can get the light a lot brighter than we might imagine, just by moving the light closer to the subject.

This means that you can make do with a lot smaller and less powerful lights if you push them much closer to your subject. So if you can't get a good exposure, move the lights closer. And if necessary, re-frame the scene by moving the camera closer or using the zoom.

It also means that its a *lot* harder than you think it will be to light a large area.

We are used to thinking about changing the brightness of a light using a dimmer switch. Move your lights and watch the autoexposure (iris number) change on your camera display. Let your camera teach you how it sees.

 

[WhiteWood]

OK, I followed EVERYTHING you told me to do. EVERYTHING, and THIS is the result...
http://www.youtube.com/watch?v=F5CdvOhtLCY
ALL of my videos come out looking like this blurry, Grainy crap. My camera does NOT shoot like this. YOU know what a DVX does...this terrible quality should be impossible.

BUT, here are the settings that I was given the option to set that you didn't cover and I was lost as to what to do, so I had to guess, as I always do:

1-Image Quality?

2-And I thought 'Widescreen' and 'Letterbox' were the same thing? What is 'Squeeze'?

3-How many encoding Passes should I allow for the video and Audio. And what does Encoding pass even mean?

4-Bitrate Mode?

5-Maximum Bitrate?

6-What is interlaced Processing?

 

[Ogrus]

AMAZING AMAZING AMAZING MR STARBUCK! :thumbsup::thumbsup::thumbsup:

I do have a question...
When i output to DVD i usually have like a million options
vbr low pass/high pass 4mbr 7mbr etc etc etc
i dont have a damn idea which to use???
what is the best quality for PAL?

 

[dailyrushes]

Hey, WhiteWood, I like your content, and your editing choices of music/clips! Well done, funny stuff!

On the video, was the camera set to auto focus and auto exposure? Because it sure looks that way.

On focus, the camera is sometimes focused on the sign in the background, and rarely on her. Or maybe you had it focused on the sign the whole time? It's hard to tell for sure because of the quality of the Flash video encoding from YouTube, but I'm half-way through the video (on dialup here) and I haven't actually seen her in focus once!

If you do use auto-focus, keep in mind that the camera auto-focusing circuitry tries to focus on straight lines and if that's the majority of the shot, (probably especially near the center of the screen) then it will focus on that instead of the softer image (the girl) close to the camera.

As for the tip-off on auto-exposure, when she moves in front of the sign, she gets lighter as the camera is trying to expose for the average exposure overall, which is now darker as she's partly covering the sign with her head.

So, my suggestion is: Learn to control your camera manually. By this I mean five things: focus, F-Stops, shutter speed, gain, and white balance. There are lots of other wonderful settings and tricks you can do with that camera, but these are the main things to get total control over! So find how to set these all manually and experiment with them to get total control over them.

24p or 60 is just setting the camer to look like film or like, for instance, TV news. That is how many images you're capturing per second, and is a creative choice you'd make in addition to the above five things.

Another thing is that videos on YouTube are re-encoded to Flash video from the format you upload them in, so you lose something there, too. I've read a lot of complaints about how fuzzy a perfectly good video looks on YouTube versus MySpace. So if you're intent on quality, not content, try uploading to Myspace instead.

 

[Huy Vu]

2-And I thought 'Widescreen' and 'Letterbox' were the same thing? What is 'Squeeze'?

Letterbox simply puts two black bars on the top and bottom of your 4:3 video, so you basically lose resolution using this option. Squeeze is compressing the video on the LCD screen but it'll display widescreen on the comp. I believe you also lose resolution using this method but I can't remember why at the moment. There's a lot of debates about which option to use so do a search and you should find it.

How many encoding Passes should I allow for the video and Audio. And what does Encoding pass even mean?

Always do 2 encoding passes for video, audio doesn't matter so much.

 

[WhiteWood]

Hey, WhiteWood, I like your content, and your editing choices of music/clips! Well done, funny stuff!

On the video, was the camera set to auto focus and auto exposure? Because it sure looks that way.

On focus, the camera is sometimes focused on the sign in the background, and rarely on her. Or maybe you had it focused on the sign the whole time? It's hard to tell for sure because of the quality of the Flash video encoding from YouTube, but I'm half-way through the video (on dialup here) and I haven't actually seen her in focus once!

If you do use auto-focus, keep in mind that the camera auto-focusing circuitry tries to focus on straight lines and if that's the majority of the shot, (probably especially near the center of the screen) then it will focus on that instead of the softer image (the girl) close to the camera.

As for the tip-off on auto-exposure, when she moves in front of the sign, she gets lighter as the camera is trying to expose for the average exposure overall, which is now darker as she's partly covering the sign with her head.

So, my suggestion is: Learn to control your camera manually. By this I mean five things: focus, F-Stops, shutter speed, gain, and white balance. There are lots of other wonderful settings and tricks you can do with that camera, but these are the main things to get total control over! So find how to set these all manually and experiment with them to get total control over them.

24p or 60 is just setting the camer to look like film or like, for instance, TV news. That is how many images you're capturing per second, and is a creative choice you'd make in addition to the above five things.

Another thing is that videos on YouTube are re-encoded to Flash video from the format you upload them in, so you lose something there, too. I've read a lot of complaints about how fuzzy a perfectly good video looks on YouTube versus MySpace. So if you're intent on quality, not content, try uploading to Myspace instead.

Yeah, that was a while back. I noticed the day that they changed their uploading and had better quallity videos. Earlier this year, like snow outside early, their videos made things look like TRASH. Now they look almost just as good as you upload them. As good as or better than MySpace. Everyday that site imporves in some way. just added these new players 2 days ago.

Let me show you something. This is the most POPULAR YouTuber:
http://www.youtube.com/watch?v=8pE9uwYqbQs
All of her videos are CRISP CLEAR. And they don't even look better than my DVX! I want to export my footage to look as good as I import it...Any idea how this YouTuber gets this quality?

 

[andy_starbuck]

dailyrushes, jade8472 -- good advice and information!

Whitewood, where are you being presented with these question? Is it in Permiere when you are getting ready to ouput? Tell me about your output process. Are you using Adobe Media Encoder and going straight out to a WMV or are you creating an AVI or QT first and then using that DV output in a separate sequence to go out to file for uploading?

I think your video looks a *lot* better. And I can see the softness in it now.

Outta time at the moment. I'll explain the terms in the next post ... a little later. A little context about when you had to make those choices will help tho.

 

[WhiteWood]

I'll Screen Capture my options in the order that they appear...

IMG1

IMG2

IMG3

IMG4

Now what's squared in Red is what you DIDN'T advise to me and could possibly make the difference between a good and bad quality video. See, I don't think you're understanding. You're commenting on my footage BEFORE it's exported and uploaded. But I'm telling you that everything is okay to me and DVX SHARP when it's on the camera. Even in PREMIERE..it's just after I EXPORT it that it comes out looking as it did.
​

Ok, in IMG1, I always pick cable Modem DSL. Because well...it's better and faster then the rest I presume.
In IMG2, As you can see, I dable in the 'custom' alot, attempting perfection. Of course it never works. But if any one of the presets works better than anything I could customize, let me know.
And as I explained above, everything in Red, I know nothign about. The settings that you see them set to are the settings I used to export that last YouTube video I sent you.​