Canon 5D Mark III -IPB contains more detail and has less artifacts than ALL-I

Hi all, now another 13(!) months later... is the verdict now in on all-i vs IPB? I just picked up the 5d3 after 5 years with the Mkii (and we have the C100/Ninja and GH/1,2,3s...). Using it mostly for sills, but imagine we'll also grab some video catch as catch can as well. Lovely low light!
 
All-I is better if there's motion in your footage, IPB is better for static shots. I would shoot All-I all the time. Except I would shoot ML RAW video all the time, of course.
 
I'm glad you brought this old thread back up. It needs a revisit. I find ALL-I to be the better codec for all situations. The only reason I would use IPB, is to save CF card or disk space.

When the 5D3 first came out in March, 2012 the current verison of Adobe Premiere Pro, After Effects, and Media Encoder was CS5.5. Adobe CS5.5 is seriously messed up when it comes to handling 5D3 footage, and it was NEVER fixed. The bug fix was to buy CS6 or CC, thanks ADOBE. So... CS5.5 doesn't know how to handle the 5D3 and takes it through QTserver (thats the Apple Quicktime Processing Engine) instead of handling it natively like other Canon DSLR video. You want to AVOID Quicktime like the plague, especially on Windows editing systems. Any codec that uses Quicktime should be an automatic non-starter. Many people still don't know this, and make erroneous judgements on the quality of IPB and ALL-I based on using it in CS5.5 and / or transcoding in Quicktime dependent applications using Quicktime codecs. Avoid 5DtoRGB, MPEG Streamclip, ffmpeg, ffmbc, Prores, and the free AVID Quicktime verison of DNxHD if you use Windows.

Note: This does not include the new DNxHD codec in mxf format that comes with Adobe CC. This one does not use Quicktime and it is one of the best codecs I have ever tested in Windows. The 10-bit 4:2:2 175x variant is basically indistinguishable from uncompressed 10-bit 4:22.
 
Last edited:
Someone correct me if I'm wrong. I thought the only point of ALL-I was to make it easier to edit, and that to achieve the same quality in ALL-I that IPB has, you need to double the bitrate, hence the files that are twice as big?
 
IPB saves bandwidth when contiguous frames are similar to each other. When they are not, it gets starved and issues appear.
 
I ve never edited footage from a 5Dmk3 but Im editing all the time from 6D and 70D which are using the same codecs, so I suppose there isnt any difference. I have to admit that I ve never made any detailed test comparing IPB vs ALL-I, but from a few shots that
I ve tested when I just received my 70D I dont remember finding any noticeable difference. Probably ALL-I is better, but if I cant see the difference why to sacrifice so much more card space? I dont know how much motion is needed to make the IPB codec break, but in normal situations (people dancing for example) I ve never had any issues. As for Adobe bugs...I m sure there are many, but Im still using premiere CS5 (not even 5.5) and have no problems with IBP.
 
Thank you all for the feedback. I'm still on FCP, what are you doing for transcoding to Prores? I've been using 5dtoRGB with the first files and find the black look best with rec 709 and Full Range.

Overall, the video is a let down for sure from the C100. I'm now used to a LOT of detail!!!
 
The 5D3 is never going to have the detail of the C100 in its standard h.264 video. However, it doesn't have to always look soft. Sharpness and detail are not the same thing. The common wisdom on the Internet is to always shoot with sharpening at zero and then add about 15-30 in the sharpen filter in post with Adobe or FCP. To put it mildly, this advice sucks. Sharpness all the way down at zero is bad idea on the 5D3. It is a major fail, shame on those preaching it and everyone else for just going along with it blindly like sheep. I also do not like the sharpen filter in Adobe. Very low sharpening in-camera plus modest better quality sharpening in post is much better for the 5D3.
 
Last edited:
Hi Celluloid Dream. I would be grateful if you could provide some more info with regards to your last statement/sentence. What in-camera sharpening settings are you using and what post sharpening method/settings?
I have been using the cinestyle preset with in-camera sharpening at zero and sharpening in post wit usm but would be interested in alternatives also. Thanks.
 
Celluloid Dream said:
The 5D3 is never going to have the detail of the C100 in its standard h.264 video. However, it doesn't have to always look soft. Sharpness and detail are not the same thing. The common wisdom on the Internet is to always shoot with sharpening at zero and then add about 15-30 in the sharpen filter in post with Adobe or FCP. To put it mildly, this advice sucks.
Click to expand...

Wrong. Post sharpening the 5d is better than the camera algorithm.
 
AlexOlson said:
Hi Celluloid Dream. I would be grateful if you could provide some more info with regards to your last statement/sentence. What in-camera sharpening settings are you using and what post sharpening method/settings?
I have been using the cinestyle preset with in-camera sharpening at zero and sharpening in post wit usm but would be interested in alternatives also. Thanks.
Click to expand...

My recommendation for sharpening the 5D3 is set the camera to 1 and then sharpen in post with USM.
For the USM settings in Adobe, Amount 70 to 100, Radius .08 to 1.0, Threshold 1

Try it for yourself, compare sharpness 0 and sharpness 1 with the same subject and decide which one looks better to you.
 
maarek said:
Wrong. Post sharpening the 5d is better than the camera algorithm.
Click to expand...

maarek, That is rather rude of you. My thoughts on sharpening 5D3 video are a subjective, artistic opinion. Do you think it is AT ALL possible that I may have a different idea than you of what looks good? In your mind, am I allowed to think differently than the hive mind, or is that some sort of violation of the "code of conformity" that offended you personally?
 
We used to use +1 sharpening in camera with the 5d2 as well, as it was just TOO soft otherwise and I often felt I was permanently loosing detials. That said, I've seen Bloom's tests at 0 sharpening and it looks like a lot 'comes back' with post sharpening.

On a side note, what delightful grain the camera produces with stills at around 4000-5000 iso. And thank goodness I can now use manual mode with auto iso not locked at 400.
 
Last edited:
I also put some sharpening in-camera back in the T2i days, I think I set it to 2. It did create some extra aliasing/moire, but you're not going to have that problem with the 5D3.
The biggest problem with sharpening in post is that it brings up all the issues from the codec: it's not a big problem if you're shooting standard, but it's an issue to avoid if you're shooting with something like CineStyle (at some point you'll start to get macroblocking in the midtones, including skin) or Flaat_12 (macroblocking in the shadows).

Also, regarding how much movement you need for the IPB codec to break down: I haven't tested this extensively, but I'd say "a lot". My NEX-5N has the same issue (or maybe even worse), and I've never had it show up in real footage. Because I don't move my camera like a maniac.
 
Celluloid Dream said:
maarek, That is rather rude of you. My thoughts on sharpening 5D3 video are a subjective, artistic opinion. Do you think it is AT ALL possible that I may have a different idea than you of what looks good?
Click to expand...

Yes it is possible. I just think it's wrong. I guess sharpening could be interpreted as an artistic decision but that's a decision that's best left to post unless you have no control of the material afterwards. The sharpening algorithm of the cam itself is just not very good.

Rude? You made a statement :"To put it mildly, this advice sucks."

I just made a counterstatement.
 
Do we have a language translation problem here? I never said that in-camera sharpening was better than post. Please re-read post #52? I recommend a COMBINATION of very light in-camera and post sharpening. I've found that if you don't do that with the 5D3 there is some detail that you can never recover in post.
 
After more tests, I'm now on all i and no in camera sharpening. Amazing what 25-30 sharpen in FCP can bring out of this seemingly very soft footage.
 
Anymore thoughts on the topic in the light of the fw updates? Some people have claimed that they got slightly better video out of their 5D with the new fw but I am still on 1.1.3 for ML compatibility...
In the few unscientific tests I have done, I have been unable to see any difference between ALL-i and IPB... when I start grading the footage heavily, both seem to show heavy macroblocking ...
Also, I am still a little confused as regards to picture profile settings. People seems to disagree quite heavily best practice and the few tests that I have had time for have been inconclusive :)
 
You must log in or register to reply here.
Back
Top