Highest quality video while editing

[Kelly Wilbur]

Hello all,

How can I make sure I am viewing the highest quality video while I am editing?

Right now, I am editing without any filters or effects, and I have the vieo playback through the Apple Firewire NTSC into my Canon XL2 and then out via composite cables to a standard CRT television.

The problem is that I am definitely seeing quite a bit of pixelization, especially in the darker areas (large pixels, not small noise) and I just can't believe that this is the uncompressed image that I captured. I am getting this same thing through the digital cinema full screen output to my 24" LCD, so I know it isn't the XL2/television setup.

This is some issue with not seeing the full quality video.

I've looked in every menu to try to make sure I have the highest quality selected, but nothing seems to help.

Any ideas?

Thanks,

Kelly

[Kelly Wilbur]

Just to give a little more information...

I have tried both RT extreme and RT safe. The only effect applied to this video is flip and it doesn't require rendering at this point.

The only other thing I can think of is...

In the sequence settings, under Quicktime Video Settings, you can select different settings on the dropdown by "compressor." Right now it is set at DV/DVCPRO - NTSC. I know from experience that I need that codec when I capture from the XL2, but do I need that in my sequence settings? What kind of compression does this apply?

I haven't tried putting "none" in that field yet, because it requires me to render my footage and I'm not even sure if it will work.

Any thoughts?

Is it EVER possible to view the final product in full quality? Or is my quality crap to begin with.

There is no doubt that this effect is more noticable when I am viewing footage shot in low light...I just find it hard to believe there is really that much large pixel noise in the XL2 for low light.

Thanks,

Kelly

[Kelly Wilbur]

Originally Posted by Kelly Wilbur

In the sequence settings, under Quicktime Video Settings, you can select different settings on the dropdown by "compressor." Right now it is set at DV/DVCPRO - NTSC. I know from experience that I need that codec when I capture from the XL2, but do I need that in my sequence settings? What kind of compression does this apply?

I haven't tried putting "none" in that field yet, because it requires me to render my footage and I'm not even sure if it will work.

I tried it with compression = "none" and that made it a lot worse. Something tells me this has to match the format I am using, DV/DVCPRO - NTSC.

I've checked some other things...

In the sequence settins, the QT video quality is 100% (in the advanced settings, quality is at best).

In system settings/playback control, video quality is at High and frame raet is at full. Gamma correction is at accurate.

Am I missing a checkbox? I can't find anything else that I think would alter my quality. Could I have messed something up in the capture?

More info:

I shot this at 24pA in 16:9. I captured it by removing the pulldown and checking anamorphic. The quality was set at 100%, the compressor was DV/DVCPRO - NTSC ("none" wouldn't work) and the digitizer was "DV Video" (I didn't try "none," should I have?).

Any help would be appreciated.

Thanks,

Kelly

[Zach Mull]

Kelly,
Have you tried playing back the source tape? That should give you your answer as to whether what you're seeing is in your raw footage or a problem with Final Cut. If you're shooting DV and capturing over firewire, you should just be using one of the DV NTSC or DV PAL presets, and you shouldn't adjust the compression settings. But I never do 24pA so I suppose it could be a problem with the inverse telecine.

[Kelly Wilbur]

Originally Posted by Zach Mull

Kelly,
Have you tried playing back the source tape? That should give you your answer as to whether what you're seeing is in your raw footage or a problem with Final Cut.

That is a great idea. Sometimes the simplest and most obvious solutions are the best. I'll check it out.

Thanks,

Kelly

How did the footage look when you viewed it. Is it on the raw footage you shot or is it FCP?

[Gabriel Patrick]

Very interesting thread. I have a similar setup XL2 and FCP5. When I was capturing footage from my xl2, The image in FCP was noisy and looked crappy. At first, I thought my camera was malfunctioning. Then, I ran a video out from the xl2 to my tv and it looked fine.

I am still trying to fix this in FCP. My capture settings were normal so I am not sure what to do.

[Alexander Ibrahim]

Originally Posted by Kelly Wilbur

The problem is that I am definitely seeing quite a bit of pixelization, especially in the darker areas (large pixels, not small noise) and I just can't believe that this is the uncompressed image that I captured. I am getting this same thing through the digital cinema full screen output to my 24" LCD, so I know it isn't the XL2/television setup.

Probably redundant, but you do realize that footage you capture with the XL2 is compressed using the DV25 codec at an approximate ration of 5:1.

It is entirely possible the codec has created the block you are talking about.

I have a similiar setup and routinely get full resolution full quality playback over firewire. My home machine currently has only 512MB so I sometimes drop frames, but rarely.

As a final note, if you are editing based on the CRT output I recommend using y/c (aka S-Video) cables for the connection. On my XL-1 the composite output is terrible, and the y/c is very much improved and hard to distinguish from component out via my deck. (That may be an XL-1 only issue, but check it out it could affect your XL2.)

That last recommendation goes double if you dare to color correct based on that output!

[Kelly Wilbur]

Originally Posted by Alexander Ibrahim

Probably redundant, but you do realize that footage you capture with the XL2 is compressed using the DV25 codec at an approximate ration of 5:1.

Quick question regarding your note about compression...

I realize the footage is compressed when recorded, but is this compression automatically removed when capturing with FCP?

Ultimately, I guess I am asking how is it possible to get the footage uncompressed so you are editing uncompressed footage?

I might just be missing some part of the puzzle here.

Thanks,

Kelly

[Zach Mull]

The compression stays. Most people do not edit uncompressed footage from DV source. You could either capture it over SDI (would require additional hardware) or render out uncompressed files from your FCP timeline or some other application. The performance requirements for uncompressed are much higher than for native DV, and you will need fast disks. I'm not sure how fast or whether it requires RAID for good performance. There is really very little advantage to doing this unless you want to use some kind of chroma smoothing process like the Nattress G-Nicer filter. You can see some comparisons here: http://www.kenstone.net/fcp_homepage..._nattress.html. But really, there is very little reason for most people to do this.

[Kelly Wilbur]

Originally Posted by Zach Mull

You could either capture it over SDI (would require additional hardware) or render out uncompressed files from your FCP timeline or some other application.

When I capture my footage with FCP from the XL2, it shows that it doesn't require rendering. Does that mean it is already uncompressed? If so, I assume it did it automatically during the capture. If not, then how do I "render out the uncompressed files" when FCP seems to show that they don't require rendering?

Just curious on this.

Thanks,

Kelly

[Zach Mull]

No, it means that you captured DV files and you are in a DV sequence. You can render uncompressed files using the Export>QuickTime Movie command and changing the render settings to 8-bit or 10-bit uncompressed. There may be a better application for doing this - I'm not sure. Before you spend any time doing this, you should probably read about DV codecs and the compression they use. This compression happens as you record, and with MiniDV cameras I don't think there's any way to avoid that. They also record in a 4:1:1 color space, and there's no way to get around that. Very few people record uncompressed. More commonly for "broadcast quality" or film out they use 4:2:2 color space with less compression than DV. This is the case for DVCPRO 50, DigiBeta, HDCAM and many other formats. If you're shooting DV, editing uncompressed probably won't do anything for you. I would bet there's a thread somewhere on here addressing that issue.

[Alexander Ibrahim]

Originally Posted by Kelly Wilbur

Quick question regarding your note about compression...

I realize the footage is compressed when recorded, but is this compression automatically removed when capturing with FCP?

Well, not such a simple question.

However you record your original footage is how your image stays, forever.

Video codecs come in two basic varieties: Lossy and lossless.

If I record in a lossy codec like HDV, DV or DVCPRO HD I will degrade my image quality permanently.

If I need the quality of uncompressed HD video I have to record either uncompressed or use lossless compression originally.

Now, if I am content with HDV recordings, I can still work with them as uncompressed footage if I like. I can either capture them as uncompressed over an interface like HD SDI, or I can capture HDV directly then render the footage to uncompressed.

What good does that do ? Isn't the image already degraded by the recording to HDV ?

The best I can do now is to say it gives you some more headroom to work in. Working in 4:4:4 colorspace allows me to make finer grade color adjustments and smoother effects. A lot of that will get thrown out again when I go back to a lesser format, like HDV, on output. The point is that if I create in high quality I have image quality above what I need to lose.

I have to add that while I think uncompressed HD is a huge leap above HDCAM and DVCPRO HD, those two formats are more than enough for many of our needs in the videography industry.

You can do great composites with DVCPRO HD and HDCAM, subtle color timing and you will be very happy with the results under some very demanding conditions.

I think images from HDV are fine for straight cut editing and other simple effects.

If your projects incolve a lot of color correction as a matter of course then you may want a better acquisition format.

If you do a lot of compositing, same deal. I think that DVCPRO HD is the standard codec of choice for professional videography. DVCPRO HD is very much like NTSC DV as far as post goes.

The jury is still out though. I haven't worked with the latest generation of HDV cameras. I expect improvement. Also, I expect that eventually someone will jump in with a 50Mbps HD format