It's there in the Mirror menu (didn't see it on first test), under Shoot 5.
Last edited:
5D Mark III firmware with clean HDMI is ready right now.
- Thread starter Celluloid Dream
- Start date
It's there in the Mirror menu (didn't see it on first test), under Shoot 5.
On the 5D3 with the new firmware I saw zero increase in resolution by A/B comparing the exact same footage recorded internally at 420 vs. external 422 (180Mbps on a Nanoflash).
The C100 provides higher resolution while the 5D3 is full frame. Most would argue the C100 provides higher quality footage. My solution for > 5D3 quality and full frame was the FS700 + Speed Booster. The C100 provides slightly higher resolution than the FS700, however the FS700 with Speed Booster gains a T-stop of light performance as well as full-frame FOV. Sony's firmware and hardware upgrade for the FS700 may also bring the XAVC codec to the FS700 (if what they stated in their marketing video is true) in which case an external recorder won't be needed in most cases (not interested in the 4K upgrade at this time).
legrevedotcom
Veteran
So far I'm hearing no audio and only 24p and 50i? What a major cock-up... sorry, but that is a f***ing Canon joke... again.
Atleast Sony was wise enough to make the FS cameras world cams, because anything else is just plain stupid.
Atleast Sony was wise enough to make the FS cameras world cams, because anything else is just plain stupid.
marvinhello
Well-known member
No audio, but there is 24p, 50i for 25p and 60i for 30p.
This is what they took 6-9 months out to release? What a joke.
They deserve to be smoked by Magic Lantern for this.
marvinhello
Well-known member
$100 says they won't.
Is there no audio period? Or does 60i with pulldown give audio?
I don't buy that for a second.
There is something disengenuous about this whole affair to me.
The fact that there is virtually no difference in resolution, sharpness, etc from a compressed internal codec at 1/4-1/2 it's bit rate is highly suspect. People are now going around saying that's normal. No, frankly I don't think it is. I have never seen this small a difference before and that is honestly just way too convenient for Canon.
Seems more likely that they spent 9 months finding a way to stall ML and inhibit an output which could possibly compete with their higher end products.
Last edited:
marvinhello
Well-known member
Hehe. the difference is drastic, depends on how you use it. Judging from your comment, you probably don't need this firmware, or even 5D3 itself.
marvinhello
Well-known member
Canon HDMI doens't output audio
I do a fair amount of grading and I regularly see a great deal of difference from what people have done bringing in an HDMI output vs. the internal codec. To be honest the internal stuff is often just garbage by comparison, 8bit or 10bit notwithstanding.
For myself, I shoot Sony F3 or RED, so yes on one level I don't need the 5D, but on another level it sure would have been nice to have some quality B-Roll. Whatever behind the scenes processing that they are doing to control the moire and aliasing (because they don't have an OLPF optimised for Video) is still in the mix. So this might be clean in terms of no overlays, but some part of that processing is IMHO still impeding the quality of this output.
Last edited:
That's because the internal codec used at ALL-i is good. It really is. For example the codec on the FS700 is quite a bit worse compared. Shooting some flames on the FS700 and you will get macroblocking. Those same flames will be completely clean on the 5dmarkIII internal recording. I find it odd that someone wouldn't consider the 5d as a good b-roll cam. I've shot b-roll several times with and some of those shots have been used next to RED and FS700 material.
There is a strange kind of mosquito like noise in the ALL-I at low ISO's. IPB is actually better for many things that don't have movement and overall I think it suffers from having a variable bit rate. The image is soft, and yes I can sharpen reasonably well in post, but it doesn't give it the kind of sharpness that I am seeing coming natively off of many cameras.
nikon_obsessed
Active member
andrew reid supporter, kudos. he's the first one to discover this problem
greenheron
Well-known member
The ALL-I codec is the best on the Canon 5D III. Don't use low ISO settings!! We have used the 5D III as a b camera for the Red Epic and the Canon C300. I don't understand what all this hand wringing is about. This is for productions for national network TV as well as regional commercial spots. Get away from the forums and internet and go out and produce something. Jeez!
When stating that ALL-I or IPB is better, it's helpful to show proof: http://www.dvxuser.com/V6/showthread.php?278746-ALL-I-vs-IPB-IPB-Wins (see the two frame grabs). In this particular case, IPB is clearly better than ALL-I, where ALL-I shows less detail and significant macroblocking. Here is (was) an example case showing ALL-I better than IPB for fast motion: http://forums.planet5d.com/threads/120057-ALLI-vs-IBP-for-fast-motion (unfortunately the images got lost when they moved the forum).
The general guideline:for low motion, IPB produces the best results (and smaller files), for fast motion, ALL-I can be better. For the 5D3, 720p60 looked better in ALL-I regardless (from my A/B tests- sorry no frame grabs).
Not much point arguing which is better for those with the camera: shoot both and compare yourself. If you can't see a difference, use IPB and save disk space.
Regarding the new 1.2.1 firmware: I did a quick test and saw IPB break up shooting a model up close and moving around slightly (hands, face). This was live playback and not pixel peeping, though had auto ISO on and may be the result of global lighting changes (which has the same effect on a codec as fast motion). Other folks report that the new 1.2.1 firmware's internal compression is improved, so it's possible the codec (or parameters) have been changed and is thus important to test ALL-I and IPB in various conditions to see how they perform.
The general guideline:for low motion, IPB produces the best results (and smaller files), for fast motion, ALL-I can be better. For the 5D3, 720p60 looked better in ALL-I regardless (from my A/B tests- sorry no frame grabs).
Not much point arguing which is better for those with the camera: shoot both and compare yourself. If you can't see a difference, use IPB and save disk space.
Regarding the new 1.2.1 firmware: I did a quick test and saw IPB break up shooting a model up close and moving around slightly (hands, face). This was live playback and not pixel peeping, though had auto ISO on and may be the result of global lighting changes (which has the same effect on a codec as fast motion). Other folks report that the new 1.2.1 firmware's internal compression is improved, so it's possible the codec (or parameters) have been changed and is thus important to test ALL-I and IPB in various conditions to see how they perform.