Thursday, July 26, 2007

Canon HV20 - 24p or not?

Yes, it is 24p.

Got that out of the way. It seems that there have been a bunch of forum threads attacking Canon, saying that this awesome little camera doesn't really shoot 24p. Not that misinformation is unusual for the internet, however, these posts often quote me or CineForm as backing this position. Neither myself nor CineForm support these posts or claims.

The problem arose when I did state that there can be a subtle issue for chroma keying when using any 4:2:0 24p signal encoded into 60i. Some users took that and ran with it. I had seen some of this in customer footage, nothing I have shot. I probably wouldn't have mentioned anything other that it is a another selling point to using the HDMI output from these new cameras (which is damn cool), and I'm a video geek. Since then I'm not even sure this issue exists outside of MPEG compression artifacts, as some more recent burrowed footage looks great. I have so little HV20 footage to work from that we shoot ourselves -- Canon, we need a camera longer than a few days -- I want one to take home :).

Basically, the 24p signal is good, and the CineForm pulldown from 1080i60 HDV tape works perfectly. That is my position.

Here is some 24P extracted footage from a friend's HV20 as she was documenting some behind the scenes footage for our 48 Hour Film Project shoot -- the camera pictured is a Silicon Imaging 2K. We intercut the HV20 footage into the credits of this movie and presented it at 1080p24 on a Sony 4K projector. The linked clip is a 110MB 1440x1080 CineForm AVI, so if you need a CineForm decoder for your PC (Mac version coming), you can download one for free from CineForm.com.

P.S. Other HV20 misinformation : when recording the Canon HV20 to tape, the image is 1440x1080, that is the HDV standard used. It is not 1920x1080, you only get that out of the HDMI port, and even then the image is likely upsized from an internal 1440x1080 image (which is still very nice.) The 1920x1080 native image is only available in the still camera mode.

19 comments:

Anonymous said...

I dont think it was 'some users ran with it,' I think was just myself. I quote a credible source since everyone reacts so defensively when I bring up the infamous '4th frame issue.' Its downright bizarre.

Have been trying for weeks in vain to figure out why I am the only person who sees the HV20 24p footage as poor. HV20 fans get mysteriously riled, & support gets evasive over the weak chroma problem that you have clearly described as resulting in some residual interlacing in the HV20's tape recorded 24p.

You asked me to post an example on DVinfo which I am mysteriously not allowed to post there. Everyone reacts so conspicuously about this topic(?). What is everyone so defensive and afraid of? Just take a look. Tell me where I am wrong?

Here is a clip from my HV20(6.5mb) before 24p conversion.
http://www.box.net/shared/static/b7qayhgfqa.mpeg
Here it is after-
http://www.box.net/shared/static/va357ns51c.mov
My frame 4 still has interlacing-
http://www.box.net/shared/static/6anr20frau.jpg
What does your frame 4 look like after 24p conversion?

Have yet to figure out how any amount of interlacing can be construed as progressive.

David said...

Your links don't work, so I still can't show you where you are going wrong.

David said...

Got the downloads, and converted the MPEG to 24p using CineForm tools with no issues. If seems your tool for extracting pulldown isn't working correctly.

Anonymous said...
This comment has been removed by a blog administrator.
David said...

Please comments on subject. Reviews for other cameras is not appropriate particularly when I'm not even reviewing this one.

Thanks.

Anonymous said...

IMHO, both the "24p" and "cinema mode" bits on the HV20 are nothing but simple marketing gimmicks by Canon. The output leaves LOTS to be desired.

I am personally yet to be impressed by any HV20 footage I've seen so far, 24p or 60i alike.

Don't know what all the fuzz over the HV20 is about. Footage starts looking good at the XH-A1 level on up.

Anonymous said...

Isn't the format 1440x1080 simply because the pixels are rectangular on the CMOS?

Also, the link to the NEO player on cineform's site is broken.

David said...

1440x1080 is because that is the HDV standard, 1920x1080 is not supported as an HDV format no matter what the source pixel shape is (which I think is square in the camera.)

Thanks for pointing out the broken link -- now fixed.

Anonymous said...

The problem arose when I did state that there can be a subtle issue for chroma keying when using any 4:2:0 24p signal encoded into 60i.

Instead of weird 23,976 (not 24p!) in HV20 wrapped to 59,94i (not 60i !),
I prefer to use real 25p (HV20 PAL) without pain and issues.
Just hate these pulldown, drop frame, non drop frame...inches, foot...
Inches is a diametr of penice of an english king? ;) Weird metrics...

Jonas Hummelstrand said...

Just too clarify your original post; the live (not from tape) HDMI output is true 1920x1080.

David said...

Yes, the data format out of the HDMI port is 1920x1080 4:2:2 YUV 8-bit. What happens early in the processing chain is not known outside of Canon -- I expect the internal processing is 1440x1080 just like it is with the Sony V1.

Anonymous said...

23.976p seems to be continually the problem, doesn't it? Perhaps the person with the issues tried a 2:3:3:2 inversion as opposed to a 2:3 inversion?

A lot of the HV20 news I've seen is the typical problem when new features get introduced at a consumer price point - confused consumers with no idea about pull-downs, drop frames and interlace. Not that there aren't confused pros...

Anonymous said...

sound like to me everyone that bought xha1 is probably pissed off that the hv20 that is under 1000 dollars really rivals it,and I have seen amazing footage made from it, people that paid $4000 for that camcorder is in denial about the truth period!!!and honestly I would be too if i own one....

DavidNJ said...

David,

Could you elaborate how the 4:2:0 chroma subsampling is stored with B-C split frame in 2:3 pull down of AAB-BB-BC-CD-DD.

Thanks,

David

P.S.
Does it affect the newer HF S10/100s also?

David said...

While the results are progressive the encoding is interlaced, so it doesn't matter whether the chroma is 4:4:4, 4:2:2 or 4:2:0, the chroma for each field correctly matches the luma for that field. As each of the fields for luma and chroma are separable, pulldown can be fully reversed.

DavidNJ said...

Now I'm confused. Didn't you say there was an artifact every fourth frame? What causes that artifact? It would seem to be tied to the 2:3 pulldown.

David said...

What artifact?

DavidNJ said...

" CineForm Aspect HD does support automatic extraction of 24p from HV20's 60i stream. However, Ridlen is correct that the 60i encoding of 24p is not completely reversible, whereas the in 24F in the XH-A1 and XL-H1 is it. The reason the lack of completely reversibility is the 4:2:0 60i encoding which has only chroma value for pixels over two scan lines -- these scan lines may contain data from different frames, cause chroma cross-talk. The luminance is perfectly correct. Luma is encoded at 60i, and chroma at 30p, you can extract 24p from luma but not fully from chroma. If the pulldown is extracted correctly, the chroma cross-talk will only appear on every fourth frame. As our eyes are far more sensitive to luma, it is very hard to see this in motion. The only time this can be an issue is when keying, for that I recommend using HMDI and shooting live via the Black Magic Intensity card (using the new 1.5 drivers.) All CineForm tools support real-time pulldown removal using that card from the HV20. When you do a live capture the 24p in 60i is encoded as 4:2:2, with has separate chroma values for each scanlines, now the pulldown is completely reversible.

David Newman"

David said...

And I recorrected that in later posts and in many forum discussions on the subject. The mistake made was the confusion of 4:2:0 progressive sampling versus interlaced sampling. The fields are completely separable and therefore the pulldown is reversible