Why? Do you really need all the read speed and P2 capability rather than just use dual record on normal SD cards ? Of course microP2 will not work quite as well as the V90 cards. You could get this card with twice the capacity and a faster write speed for less money https://www.bhphotovideo.com/c/prod...es_ddsduhs2128g_128gb_cinema_uhs_ii_sdxc.html Write speed is what is important for recording in the GH5 not read speed.
GH5 ALL-Intra 4:2:2 10bit recording mode with 400 Mbps
- Thread starter Vardalos
- Start date
Why? Do you really need all the read speed and P2 capability rather than just use dual record on normal SD cards ? Of course microP2 will not work quite as well as the V90 cards. You could get this card with twice the capacity and a faster write speed for less money https://www.bhphotovideo.com/c/prod...es_ddsduhs2128g_128gb_cinema_uhs_ii_sdxc.html Write speed is what is important for recording in the GH5 not read speed.
I suspect that it is a data rate issue which could also explain why All-Intra 422 10bit 4K recording is limited to 23.98p and why higher frame rates on the GH5 are recorded using the Long GOB codec. ProRes HQ at 4K/24P would require cards capable of reliably recording at 754Mbs which is almost twice the data rate of GH5's All-Intra 10bit 4K and ProRes 4444 XQ would increase the data rate to 1131Mbs. Either ProRes flavor would limit the options for recording media and significantly drive up costs for GH5 owners even if such media were available. Of course it's possible Panasonic could have gone with one of the lower data rate ProRes options but I don't think most shooters would chose a firmware update with ProRes LT or ProRes Proxy recording over one with Panasonic's apparently superior All-Intra codec.
Assuming that Panasonic's All Intra 422 10bit 4K 400Mbs update matches the quality of ProRes HQ I think they have done GH5 owners a favor. Many shooters want to acquire at the highest quality possible and if you want to record internally on on the GH5 All-Intra looks like the winner. If ProRes is important at least Panasonic allows external 10 bit 4K recording. On the C200,which costs around 4 times as much as the GH5, Canon has limited external 4K to 8 bit.
Of course this discussion assumes that GH5 would be capable of recording ProRes. Even Panasonic's flagship Varicam limits internal ProRes recording to 2K resolution so it wouldn't surprise me to find out that GH5 hardware might not be capable of doing what you want.
Cary Knoop
Veteran
I do not think that is true, according to this link from Panasonic 23.98p, 24p, 25p and 30p are all supported.
http://www.panasonic.com/global/consumer/lumix/gh5_firmware.html
It depends, long GOP has a far better compression ratio than All-Intra with the same quality. It remains to be seen which one gives a better quality and under which circumstances.
One word about ProRes, some folks, people not neccesarily on this forum, seem to think ProRes is some gold plated codec that is far above in quality compared to other codecs, perhaps they even believe their footage will become better just by transcoding it to ProRes.
It's seems Apple marketing did take those folks in really well!
I think it is nonsense, with a similar bitrate H.264 All_Intra provides a comparable quality to ProRes.
stoneinapond
Veteran
It depends, long GOP has a far better compression ratio than All-Intra with the same quality. It remains to be seen which one gives a better quality and under which circumstances.
One word about ProRes, some folks, people not neccesarily on this forum, seem to think ProRes is some gold plated codec that is far above in quality compared to other codecs, perhaps they even believe their footage will become better just by transcoding it to ProRes.
It's seems Apple marketing did take those folks in really well!
I think it is nonsense, with a similar bitrate H.264 All_Intra provides a comparable quality to ProRes.
I personally don't know anybody who transcodes to ProRes in their workflow believing it improves quality. I'm glad you added the "not neccesarily on this forum" qualifier as I have to say it's really tiresome to be told over and over that a choice I've made to stay within the Apple ecosystem because it is generally more unified, might be because of some zombie like behavior on my part.
The reason to use the codec is that it is rigorously maintained by Apple and there are no ugly surprises. It is also very easy on CPUs, making my workflow progress smoothly. A case in point. If I want to create a file for use in Pro Tools to do some audio work, an H264 file will bog down the processor as the software tries to decode the video for playback while simultaneously recording and editing audio. I can stay within ProRes knowing there is less degredation, and only going to H264 for final output.
I was referring to DCI 4K and should have been more clear. While All-I Intra 25p and 30p is apparently possible in UHD it's noteworthy that even at 25p the GH5 can't do DCI All-Intra 400 Mbs. It's just a guess but I think the frame rate limit may be the result of a hardware limitation.
One word about ProRes, some folks, people not neccesarily on this forum, seem to think ProRes is some gold plated codec that is far above in quality compared to other codecs, perhaps they even believe their footage will become better just by transcoding it to ProRes.
It's seems Apple marketing did take those folks in really well!
The only gold plated ProRes I have seen recently was ProRes 4444 XQ shot on an Alexa Mini. It was beautiful but I don't generally think ProRes is the pinnacle of codec technology. I also haven't met met anyone who thought their footage was improved after it is transcoded into ProRes. For me the beautify of ProRes is that it fits into existing editorial pipelines with relatively low CPU overhead which means it can be edited on almost Mac released this decade. The director of the aforementioned UHD ProRes 4444XQ Alexa Mini project edited a rough cut on a several year old dual core 1.6 ghz Macbook Air.
This isn't a religious decision for me and I am not trying to convert anyone to Macs, ProRes, or Final Cut Pro X. It's just the best choice by far for me and workflow. Again for the record I think Panasonic was right to choose All-I as an acquisition codec instead of ProRes given the GH5's target audience and recording media. All-I looks like it will be a beautiful codec but I would still likely let Final Cut Pro transcode to ProRes when editing.
Cary Knoop
Veteran
By the way DCI does not support 23.98p, it supports 24p exact!
Cary Knoop
Veteran
Here we go again with taking stabs at H.264 and claiming ProRes is better.
While that may be true in the strictest sense it seems a bit pedantic to argue about it. I was referencing the DCI frame size independent of the frame rate which something even Avid and Blackmagic do in their product documentation.
http://resources.avid.com/SupportFiles/attach/WhatsNew_MediaComposer_v8.4.pdf
https://www.blackmagicdesign.com/products/decklink/techspecs
stoneinapond
Veteran
You're fixating. And perhaps I have been a little less clear than it seems you need.
I am talking about an end to end workflow. H264 in, H264 out. In the middle I want the most malleable codec available to avoid any recompression artifacts. To my eyes and to my CPU muscle, ProRes satisfies those criteria.
Cary Knoop
Veteran
Well if you want to avoid recompression artifacts I would suggest you don't transcode at all!
Again you are claiming that Prores is better to your eyes and CPU. H.264 All-Intra with a similar bitrate should look comparably to ProRes and have a similar CPU load.
I am saying ProRes is great but so is H.264 All-Intra.
stoneinapond
Veteran
Cool.
Jarek Zabczynski
Veteran
Because I already own a bunch of MicroP2 cards...I didn't buy them for my GH5...
Back in my AG-HPX250 days a 64GB-P2 card at B&H was $639.00 today a Micro P2-64GB UHS-11 card is $299.00 I’m sure Panasonic has tested them with the GH5 All-Intra 10bit at 400Mbps, and that they work, but I note in the B&H ad they have a UHS-11 designation meaning they may be a new version. https://www.bhphotovideo.com/c/prod...nic_aj_p2m064ag_64gb_microp2_fast_memory.html
Early testers say the only cards that are reliable are V90’s, this one at 128GB’s looks like it might be a much better value at $229 https://www.bhphotovideo.com/c/prod...es_ddsduhs2128g_128gb_cinema_uhs_ii_sdxc.html
"Same quality" is questionable. There is at least one video up on YouTube from an early tester showing improved quality for All-I. (At least when pixel peeping.)
I actually think it's an important distinction to make. 23.98 and 24 are not the same, and trying to mix them in the same timeline will cause issues.
I think there's no point in making a "similar bitrate" argument between codecs. All that gets you is a similar file size. If you want "similar quality" between different codecs, you really have to ignore bitrate.
And in my experience, H.264 will always consumer more CPU horsepower than ProRes/Cineform/DNx. Even in ALL-I.
filmguy123
Veteran
thoughts on times to NOT use 400mbps ALL-I?
ie portrait shots, or shots without a lot of movement?
Or, if you have the memory card space/budget to do this, always better to shoot ALL-I?
ie portrait shots, or shots without a lot of movement?
Or, if you have the memory card space/budget to do this, always better to shoot ALL-I?
Looks like the older Panasonic Micro P2 cards might not be fast enough for 4K All-Intra 10bit 422 @ 400Mbps. New V90 cards are coming probably to coincide with release of the version 2.0 Firmware. They will be pricy...........http://shop.panasonic.com/cameras-a...ccessories/sd-microsd-cards/RP-SDZA128AK.html
Cary Knoop
Veteran
There is absolutely a point in comparing results between different codecs using similar bitrates.
For instance H.265 will provide a far superior result compared to H.264 and H.264 will be superior to ProRes unless we use H.264 all-intra in that case the result is comparable, H.265 should still be superior even if used in all-intra mode but I have not tested that.
You can test codec performance in various way beyond visual inspection, an objective method is to compare SSIMplus or PSNR scores
Last edited:
