[Les Galusha]

Concerning the degrading of footage when Scaling/Dropshadow/Interlacing in FCP…
…Our company’s engineer got a response from Apple.

The Apple rep said they ARE aware of the footage quality degrading when scaling interlaced footage in a Sequence with field dominance.
Although Apple is not calling it a “problem” because there IS a workaround…They call the workaround “Motion”.

In my opinion, that’s not a good answer…but there IS some good to it, I guess…

The technician that spoke with us said that Apple FCP does NOT plan to fix this issue, but suggests exporting the footage into Motion when working with footage in this way.
Similarly, AVID used to suggest exporting to After Effects when doing processes like this; however, AVID “Fixed/Updated” this quality issue at least 4 years ago.
The FCP developers plan to focus these kinds of developments in Motion, not in FCP. This makes rumors of a “Graphics/Editing package” for the future more valid. Similar to the AVID-DS I assume BUT I expect FCP’s to be much better.
Unfortunately for FCP users, this does not help this issue for the time-being.

The Apple Rep did not provide any other suggestion other than using Motion. However, It seems there is a solid workaround, so it is not a total disappointment.

If interested then read below:

Start with setting your edit-sequence with a field dominance to NONE.
Make sure all your captured footage has field dominance set to NONE. This should be true even for captured interlaced footage.
These settings should take place BEFORE adding any clips to your timeline or it will not work — If you change the field order in the Sequence setting AFTER clips are already dropped in then it will appear blurry…(at least in version 6.)

The quality difference when scaling and adding dropshadows is noticeably different, and “so far” has not caused problems with broadcast dubs sent out on D5 and Beta SP. Typically, you will see shuttering in noninterlaced fast movement. I assume if it causes a “Field” issue with NTSC playback then the Sequence could be sent to Compressor or Motion to render a file with a set field-order.

Again, if you are working with progressive footage in a progressive Sequence then this is not an issue, at least in the 720p test we have done.
And Typically taking clips to Motion or After Effects may be the method to get the best results. However, when a “simple task” as scaling and adding a shadow can not afford the time to send to a 3rd party app, then this appears to be a workaround. If you experience problems with this, please post…

Once again, this information came from a single Apple tech via our engineer, so there may be more reliable sources or methods. But this is the best info we’ve received after 3 weeks of painfully trying to find a solution.
Sorry this post rambles but here ya go.

[Les Galusha]

OK, I did a test using DV Footage. I thought maybe the reason Rafael was getting clean results was due to importing and working with DV in FCP. Our tests so far have been with Beta/film captured at 525 Uncompressed 8-Bit rather than DV. We thought that since ‘back in the day’ FCP used to be native DV, maybe it had something to do with that, a longshot but….

I captured at 720X480 via KONA3 card. I could not do a test capturing via firewire at this point. I dropped the shot in a sequence the same as the clip settings 720X480 Lower Field.
I compared the 26% scaling and dropshadows side-by-side to a WHITE SOLID created within FCP over COLOR BARS.
Again, The results were the same as I had before. The FCP shadows Generated under teh SOLID look great, but the shadows generated under the Footage looks bad. Even blurring the shadows at high levels was no good.

Unlike a few users who are only experiencing poor shadow quality, we are still having tearing in the interlaced and scaled footage. I have verified this scaling/interlacing of footage with 3 other people now.

Outputting to our Apple Cinema Display, Toshiba Plasma TV, JVC NTSC TM813SU broadcast monitor, and sending to a composite television is displaying both problems.

Anyone else have any “luck” replicating this?

[Les Galusha]

Specs on this system is:

FCP 6.0.2
OS X 10.5.2 Leopard
QT Version 7.4.1
2 X 3.2 GHz Quad-Core Intel Xeon
8 GB 800 MHz DDR2 DB-DIMM
Kona 3 HD with AJA breakout box
Apple 30″ Cinema Display and out to 55″ Hitachi Plasma

I can check the other system later, but had the same problems. They are all running the same versions of OS, FCP, and QT as the system above.

[Les Galusha]

I have been swamped today so have not been able to run any more tests — Hope to try a few more tonight with some progressive footage and then go the opposite direction with DV and a DV sequence.

Yes, Rafael I’d be interested in your test results too.
On another website’s Forum I’ve had some feedback with a user replicating the tearing of interlaced footage and shadows as well. I need to verify what is footage was shot on.

[Les Galusha]

Kevin, thanks for this info — I will cross reference your post when I get back to the office in th AM to some follow up tests.
If I remember correctly, I have tried the things in your post; however, not sure of the “View” menu settings you mention. I will check these as well as talk to our veterin FCP editors about a few new developments.
I am shocked they have now stumbled across this before as it was one of the first things I noticed when picking up the program. And yes, I agree, thankfully there are editors that these “anomoles” slip past, giving the success and opportunity for higher-end houses without the larger-budget software. Anyways, I will check that out and post again Monday.

[Les Galusha]

This is MORE info on concerning the last post — I failed to give all the info on the files within this link:

Here is a link to the FCP Project and 8-Bit Source file.

https://www.jonesinc.com/tech/

Unfortunately the 1.5 seconds of footage is still 46 MEGS as I wanted to maintain the exact same resolution I am working with. Thank you for your patience in downloading.

Also included are 2 files to compare what I am seeing on my system.
One file is the rendered from FCP in a lower field sequence. (FromFCP.mov)
The other is using the same car footage clip in After Effects, scaling and rendering. (FromAEandFCP.mov)

My point in showing these 2 files is to display the degrading of the scaling AND shadows in FCP. Processing the clip from AE looks better, as expected HOWEVER, I would think that FCP could do a better job at it than these animated files show.

I can explain more about these 2 clips if this info is too vague…

Have fun with these and again, thanks for the diagnosis on this. I am anxious to see what you all find in this…

[Les Galusha]

In case anyone is interested…
I posted this LINK on some other forums across the net, and so far have come to the conclusion that this issue is a FCP ‘bug’.

Feel free to look at the FCP Version 6.0.2 file and captured clip.

https://www.jonesinc.com/tech/

This vehicle running footage was shot on 35mm film and transferred to Beta SP.
We have tried the same tests with film transferred to D5 , Footage shot on Beta SP and also have tried footage on DVCPro. All have the same look.
We just brought the footage to another Suite with FCP version 5 rather than version 6. Also with the same results.

I’ll check the filter that Rosie mentions…See if I can locate that (Rosie, for some reason I could not get the link you provided to work…Thanks for the info.)

My goal is to see if: 1) it is something with our captured footage 2) A User error with a FCP setting that needs adjusting, although we’ve tried about every option, I think 3) We are over-analyzing FCP, assuming it can be ‘pushed’ to do things that After Effects, Motion or other graphics/compositing programs are meant to do in its place — although I’d expect it to handle this task of scaling-down footage beyond 60%.

About 8 years ago we ran into a similar issue with AVID where scaled footage was getting degraded and they recommend doing these motion effects in After Effects.
Several versions of Media Composer later could handle these tasks so I’d expect this to work these days. I’d understand how “Motion” would be used as a complementary program for FCP, but in this instance it feels like something is not right. Especially since the quality drops after rendering.

[Les Galusha]

Thanks Tom for the response, any insight is helpful.

As an AVID editor for years, I have always setup for “lowerfield dominance” for NTSC-Broadcast. When I did not work in a fielded environment, the ‘issues’ I would involve strobing of fast moving objects — such as simply moving a box across the screen.
In the AVID Media Composer software that we used before switching to FCP, there was not a good “motion blur” to help blend any strobing artifacts. BUT we never has this ‘quality anomole’ so it was not a problem.

However, FCP has a motion blur tab — but I do not know how well it will hold up for professional broadcast work? Will this work in “most” cases, I do not know — Again, I am fairly new to FCP, but NOT to broadcast production so Maybe you have some insight on this?

Also, I am afraid that much of our footage would lose it’s integrity if it was edited in FCP in a sequence without fields. If you have experience in FCP outputting for broadcast, then that might shed some light for us, thanks!

On another note — When we render animations out of Maya or Max, we typically dont have to render to fields, but motion blur will suffice….for the most part.

[Les Galusha]

Following up on the last post..about render sequence settings…

I created a new sequence and verified that Render Settings Precision is High Quality and Motion > Best.

Still looks the same. The fact that we are getting the same results on 3 different systems concerns me.
One of the Edit Suites does NOT have a capture card, but when output via Firewire DV 720X480 we get the same unacceptable results.

Any other suggestions? Thanks!

[Les Galusha]

In my original post, I failed to mention that we have been certain all position are on whole numbers. Even, whole numbers to be exact. I double checked and that is not the problem.

If that WERE the issue, why would shadows look bad on just those shots???

I notice you mention rendering in High Precision and Motion>Best. Humor me and tell me exactly where you do this. I want to make sure that I have this setup. Our FCP editors and our engineer has looked at this too, but I want to follow up with this, thanks!

I am also seeing the same results when bringing in ‘progressive’ footage and scaling the progressive footage in a 525 8-Bit project, it seems. However, if we capture at 720p and drop in a 720p comp, we do not see the bad shadows etc.

Using a deinterlace filter in FCP does no good, etc.