C100 & 60i..

GaryinCalifornia

GaryinCalifornia

Veteran
Its been so long since I thought about this... I wonder if someone will make a plug-in to de-interlace the footage so you can do slow-mo...
 
Last edited:
The 30P in the cam is a 30pf so 30P in a 60i container like the XA10 and Canon's AVCHD consumer line. So 60i has to be proper 60i as the sensor can scan at least 60P from what I understand.
 
I have a feeling that one point or another the 60i will be 60p via firmware upgrade, it would be stupid not too lol, But we still want one lol!
 
I can't remember (Don't hold it against me) but isn't there an external recorder out there that takes the 60i and turns it into 60p during recording?

I'm probably wrong. >.<
 
GaryinCalifornia said:
I was wondering if any C300 owners have converted the 1920x1080: 59.94i to 24p yet and if so could post an example...
Click to expand...

I've not done this, but I am curious how it would compare to the 720p60 footage. I'll try to do a brief comparison over the next couple of days and post results.
 
cowpunk52 said:
I've not done this, but I am curious how it would compare to the 720p60 footage. I'll try to do a brief comparison over the next couple of days and post results.
Click to expand...

My prediction is that the results would depend on the codec used. On the C300, you might actually get slightly better results from deinterlacing 1080i60->60p for slow motion than with upscaling 720p60 to 1080p. Even assuming the worst case for loss of vertical resolution (that being an effective drop to 540) there is still much more horizontal resolution in the 1080i image than in 720p.

However, using the internal avchd codec on the C100 would be slightly different. The C300 records 4:2:2 internally, but the C100 is only 4:2:0. Luma resolution would be the same, but chroma resolution is devastated when 4:2:0 is applied to interlaced footage. On a hypothetical C100 with 720p60, you'd probably be better off recording 720p and upscaling as 4:2:0 handles progressive chroma better than interlaced.

For the C100, this is probably one of those scenarios where an external recorder with 4:2:2 would be of use, even if it doesn't do 1080p60.

But of course, that's all theoretical.
 
Wesley Byram said:
My prediction is that the results would depend on the codec used. On the C300, you might actually get slightly better results from deinterlacing 1080i60->60p for slow motion than with upscaling 720p60 to 1080p. Even assuming the worst case for loss of vertical resolution (that being an effective drop to 540) there is still much more horizontal resolution in the 1080i image than in 720p.

However, using the internal avchd codec on the C100 would be slightly different. The C300 records 4:2:2 internally, but the C100 is only 4:2:0. Luma resolution would be the same, but chroma resolution is devastated when 4:2:0 is applied to interlaced footage. On a hypothetical C100 with 720p60, you'd probably be better off recording 720p and upscaling as 4:2:0 handles progressive chroma better than interlaced.

For the C100, this is probably one of those scenarios where an external recorder with 4:2:2 would be of use, even if it doesn't do 1080p60.

But of course, that's all theoretical.
Click to expand...

I can do the test with the 35Mbps 4:2:0 codec setting to approximate a more even playing field with the C100. Theoretically, of course.
 
cowpunk52 said:
I can do the test with the 35Mbps 4:2:0 codec setting to approximate a more even playing field with the C100. Theoretically, of course.
Click to expand...

Cowpunk52 shot some footage for our team at 4:2:0 35MBS with his C300. Our editor loved the footage, and she is a television network editor so she deals with broadcast footage. His 4:2:0 35mbs footage was much better than our FS100 footage. From this shoot we released a short event video where the speaker had to be cropped from about one quarter of the frame and even though it was filmed at what I think was 6400 ISO, not sure, the result was acceptable even after color correction.

I really look forward to Cowpucnk52's test. He is the man when it comes to optimizing the C300 and I eagerly anticipate what he can do with the C100. Canon get him a camera,


 
Last edited:
philiplipetz said:
Cowpunk52 shot some footage for our team at 4:2:0 35MBS with his C300. Our editor loved the footage, and she is a television network editor so she deals with broadcast footage. His 4:2:0 35mbs footage was much better than our FS100 footage. From this shoot we released a short event video where the speaker had to be cropped from about one quarter of the frame and even though it was filmed at what I think was 6400 ISO, not sure, the result was acceptable even after color correction.

I really look forward to Cowpucnk52's test. He is the man when it comes to optimizing the C300 and I eagerly anticipate what he can do with the C100. Canon get him a camera,
Click to expand...


Just curious, but wast that 35Mbps footage progressive or interlaced? 4:2:0 holds up pretty well with progressive footage. Shooting interlaced is when it jacks up the chroma. That said, I've gotten decent results converting 1080i60 HDV footage from an HV20. You just can't push it in post at all.
 
Hey, guys - just finished doing the comparison, and I'm uploading my results to vimeo now. I shot 1080i60 @ 35Mbps (to approximate AVCHD codec performance) and converted it to 1080p60 before slowing it down to 1080p24. For comparison, I shot 720p60 @ 50Mbps, upressed it to 1080 via Instant HD, and slowed it down to 24p. Both processes were done in After Effects CS5.

It wasn't the most precise or scientific tests (just my hand flipping through some magazine pages with probably a bit too shallow DOF), but I did come away satisfied that the 720p60 slowmo gave me significantly better results than the 1080i60 conversion. I just couldn't get rid of all the interlacing artifacts from the 60i footage, particularly in areas of red color. Someone else may have a better conversion process than me. Here's what I did:

1. Drop your clip into AE. Select the clip. Select File > Interpret Footage – Set to 23.976 and set “Separate Fields” to “Upper Field First” for HD sources.
2. Drop the clip on the New Composition droplet in the Project window.
3. Select Composition > Composition Settings and set the frame rate to 23.976.
4. Put the clip into your Composition. Select the clip in the timeline and select Layer > Time Stretch from the menu bar. Enter 200%. This will take each field (half of a frame) and turn it into a whole frame. Remember the number that appears as the new clip length. If playback seems to jump back and forth, double-check that you selected the correct field dominance in the Separate Fields box inInterpret Footage. This has turned your 60i footage into 60p. Since both your timeline and Interpret Footage are set for 23.976, the footage will now play in slow motion.
5. Go back to Composition > Composition Settings and double the duration of the Composition (enter the number that appeared in step 4).
6. Export through the Render Queue. Make sure Frame Blending is OFF, Field Render is OFF and your frame rate is 23.976.

I'll post a link to the video once it's finished uploading. If anyone has a better method for converting 1080i60 to 1080p60, I'm all ears. Also, if anyone is interested, I can probably find a way to host the native 60i file so you can have a go at it. I will say, thought, that after upressing the 720p footage with Instant HD, I'm totally comfortable cutting it with 1080 material. It looks really nice.

Here's the vid:


Download the source file to see best quality comparison.
 
Last edited:
@Brian ^

Just saw your footage and I must say, 60i coverted to 60p looked pretty good. If the C100 60i to 60p conversation (if possible) can look any where near as clean and smooth I will be very happy!

Thanks for doing that test!

Edit:
One thing I thought of was, do you think that because you used a pretty shallow DOF, that maybe it covered up some of the artifacts that might normally show with a process like that?
Did you try it with out such a shallow DOF? And if so, were the artifacts more noticeable?

I'm not trying to dog your test at all, I'm just trying to get a feel for what I might have to deal with if I get a C100.

Again thanks for the help.
 
Last edited:
Tim Nitsch said:
One thing I thought of was, do you think that because you used a pretty shallow DOF, that maybe it covered up some of the artifacts that might normally show with a process like that?
Did you try it with out such a shallow DOF? And if so, were the artifacts more noticeable?
Click to expand...

I didn't get a chance to do anything with less shallow DOF, unfortunately, but I feel like there's enough detail in the pages in the areas that are in focus to be indicative of what you can expect.
 
I think it looks fine with the 60i converted. I guess it depends on how much you need slow-mo for your work. IDK, from COW's test I really can't see a major difference or at least anything to cry about. I'm considering the C100. Just waiting for all these new cams to hit peoples hands to feel more footage out of each and get feedback. Would be nicer if the C100 were under the $6000 mark but I'd pre-order at $5200 or so.

Anyway, thanks for the test COW.
 
Hi Brian,
I'd love to get the original 60i file to play with it. Let us know if it's possible. Thanks again for taking your time and posting this test.
 
You must log in or register to reply here.
Back
Top