Gamma shift with P2 to FCP vs. P2 to Avid Express Pro

Next question-- Is the FCP quicktime conversion of MXF files considered one generation of compression?

Cause if so, then grading/correcting/etc. would be the second and you wouldn't be working directly with actual raw footage. If compressing, even to the same codec, is still compressing... then it's one generation of compression right?

Because when I go to adjust the image to match the viewer, I see more noise in the smaller frame than the raw MXF in the p2 viewer.
 
This issue has been hashed a great deal before. There are more than one other post about this. If you're in FCP and are watching the footage on an actual high def professional monitor, then that's what the footage looks like. You can't judge off your camera's LCD, you can't judge off your computer screen, not for any NLE.
 
The quicktime rewrap is not a "generation", it's unmodified data.

I don't use FCP or Avid anymore, but I do recall there being a definite gamma shift between the way each decompresses the footage. The source footage is identical, yes, but the decompressors worked differently.

What I don't get is why Apple would be showing two different images; presumably it's using the same decompressor with the same settings for both the P2 Import and the actual timeline. But maybe not -- maybe there's some setting for the Quicktime codec that get used for the full-screen shot but get bypassed for the small preview window?
 
If you want to narrow it down, here's how I'd go about it: generate some color bars in FCP, and export to the camcorder. Then re-import the color bars, and compare against the original generated ones. If FCP is introducing a gamma shift you should easily see it.

Also, googling "FCP DVCPRO-HD gamma shift" brings up a whole lot of hits.
 
Id on't know how you're NLE's are set up over there, but my first guess is with Ulisses. It sure looks like the difference bewteen 0 & 7.5% black.

Most of us are getting exactly what we expect when we import into FCP - i.e looks the same as when we shot. If your's is looking much darker than i suspect something is off in your set-up in FCP
 
Lenilenapi said:
Most of us are getting exactly what we expect when we import into FCP - i.e looks the same as when we shot. If your's is looking much darker than i suspect something is off in your set-up in FCP
Click to expand...

What settings would you suggest? I'm running FCP 5.0.4 and the only setting that effects the image seems to be the Gamma Correction under system settings in playback control. The default is "accurate." When I put it to "approximate," the image does get less contrasty, but it still doesn't look as good as on the Avid. FCP gives you no options when importing off the P2 card.

I have an ATI Radeon 9650 video card and I've been thinking that FCP and Avid Express Pro utilize it in different ways, but when I export new uncompressed quicktimes out of both programs and open them in After Effects, they look the same as they did in their originating NLE's.
 
Barry_Green said:
If you want to narrow it down, here's how I'd go about it: generate some color bars in FCP, and export to the camcorder. Then re-import the color bars, and compare against the original generated ones. If FCP is introducing a gamma shift you should easily see it.
Click to expand...

Hey Barry,

That sounds like a great test. I'll try it tomorrow when I have time.
 
Jim2000 said:
What settings would you suggest? I'm running FCP 5.0.4
Click to expand...

With 5.1 Apple applied a whole new gamma handling scheme. As for P2 Import window vs. Canvas window, yes, they use different schemes for playback in both. Don't ever trust the P2 Import window to be a true representation.
 
Lenilenapi said:
Id on't know how you're NLE's are set up over there, but my first guess is with Ulisses. It sure looks like the difference bewteen 0 & 7.5% black.

Most of us are getting exactly what we expect when we import into FCP - i.e looks the same as when we shot. If your's is looking much darker than i suspect something is off in your set-up in FCP
Click to expand...

What version of FCP are you currently running? Do you have access to the P2 Viewer similar to my setup? I was ignorant to the fact that 5.0.4 did NOT include a p2 viewer, so are you using one and comparing what you see in the viewer versus the NLE when you make that judgement?

If there are settings for Final Cut, not Filters or correction but actual timeline settings that I am not aware of, please... inform me.
 
BenB said:
With 5.1 Apple applied a whole new gamma handling scheme. As for P2 Import window vs. Canvas window, yes, they use different schemes for playback in both. Don't ever trust the P2 Import window to be a true representation.
Click to expand...

So you're telling me to connect the camera directly to a professional monitor to judge what the image really looks like? Because otherwise, I would still be dealing with the way an NLE handles the footage, not the RAW MXF/DVCPROHD footage.

Is this what I should do to judge my footage?
 
No, connect the computer to a professional monitor. Computer monitors can NOT replicate the image properly like a professional monitor is made to do.

The RT settings will also effect the qualit of the image a great deal, as will the black level (0% vs 7.5%), and your camera's LCD does in fact lie to you, also, just like your computer monitor.

If you are still on 5.0.4, you REALLY need to upgrade to use DVCPRO-HD properly.
 
Well, let's hope that the other guy can get the color bar test working. I can't get the HVX LCD to show the color bars that are in the timeline within FCP 5.1.4.

Settings as follows:
FCP
720p60p Easy Setup
720p60p FCP generated Color Bars

HVX
MCR/VCR mode
Thumbnails off
720p60p
Connected via Firewire

No image on LCD when I scrub in FCP.

EDIT: Aragon told me the missing step-- SHOW ALL FRAMES under EXTERNAL VIEWER. That got it working.
 
Last edited:
Results:

Camera set to 0% Black Level to P2 = Red, Blue, and Teal channels were off by 1, 1, and 2 on RGB Scale

Camera set on 7.5 black level to P2 = Everything is exactly the same. Only difference is in P2 Viewer which actually shows the barely-there grey bar in the bottom right.

Next test: Go to Final Cut Bay with Pro Calibrated Monitors to see if the monitor is making that much of a difference.
 
My thought on this is that the difference is being introduced by Quicktime and consequently, the .mov format.

Consider the following: in all 3 cases where the image is appearing how you expect it to, you are reading the data out of the .mxf file. On the camera, in Avid and in the previewer. Quicktime won't be used in any of these cases.

The difference is introduced when you then toss the data into the .mov container, whereby Quicktime then has it's grubby paws all over it.

I bring this up because I've seen FCP do similar stuff to my DV footage until I changed my Quicktime playback settings upon which, my footage was back to normal. I'd check those settings and consider resetting to defaults and/or increasing the quality to "high"

Something you can try testing with is to export a similar clip out of AVID into a DVCPRO HD .MOV file and import that into FCP and see if the difference remains. If it's the same, then without a doubt, it's Quicktimes fault. Also, consider doing the reverse, FCP footage into AVID and seeing what happens.

I could be wrong but it's something I'd consider trying.
 
I'd consider that QT is doing something IF it was universal. But it's not. Many of us on FCP systems don't have the gamma shift problems you guys are discussing.

Any discription of how it looks on a computer screen is totally fruitless. If you are not on a calibrated professional production monitor, you don't have anything real to judge by.

I, personally, don't have these gamma shift problems on my FCP system.
 
Well, I did google what Barry green suggested (DVCPro-HD Gamma Shift) and apparently many of us are having issues with Final Cut.

Simply saying that having a real production monitor versus an LCD isn't going to help anyone. How does going from Camera > Import/Quicktime/FCP > Production Monitor represent the most accurate rendition of all facets?

Wouldn't going straight from camera to production monitor render the most accurate image portrayal?

And if my footage shows up how I expect it to be from Camera to a calibrated Production monitor, then is the camera lying or the production monitor?
 
I have to agree with BenB - to a point. What you are seeing on your computer monitor (the way you are doing it) is useless. However, a computer monitor can be used as a production monitor using the Matrox MXO. I use it every day and it has been very accurate when properly calibrated.

See a review by Shane Ross: http://lfhd.blogspot.com/2007_03_01_archive.html for a good perspective.

I don't have Avid on my Mac so I can't compare it to finalCut but what is most likely causing the gamma shift has less to so with QuickTime and more to do with ColorSync and the monitor profile you are using. QuickTime is ColorSync aware and is using the monitor profile to adjust how the image is displayed and output depending on how you did your screen grabs. Avid may not be this integrated into OSX color management system.

When in the P2 Import screen, the preview of the MXF footage apparently isn't affected by QuickTime and ColorSync until it is in FinalCut.

The only reliable solution if you must use a computer monitor is to get the MXO. Otherwise, get a production monitor if you have the cash. Short of these two options, you just cant be critical of the color you are getting from your computer monitor. You are just banging your head against a rock and in the end, it just isn't going to matter because without the right system and an understanding of how to set it up, what you see is not what you get.
 
For the record: I did connect my camera to a JVC Pro HD monitor (right downstairs) and I checked the footage out there. It was closer to the p2 viewer than the Final Cut NLE.

With that said, that sounds reasonable, when considering the color profiles of the monitor. I had no idea that Final Cut actually did that.

Thanks, Kevin.
 
You must log in or register to reply here.
Back
Top