H.265 VS H.264 - D-Cinelike comparison

Video recorded using H.265 is not smaller than H.264 -- they both consume 100Mbps so the file sizes are pretty much the same for a minute of video. The theory is that with H.265 being more efficient it can record the same detail in a smaller file OR more detail in the same sized file. Again, the file size will be the same for a minute of video. You will not save any drive space using H.265 with the DJI drones -- not a byte!


Brian

You cannot do C4K60 with H.265 with the Phantom 4 Pro V2.0. Also, rendering in H.265 takes Significantly longer than H.264. HEVC is cool to save storage if you keep a lot of big files but it's problematic.
 
You cannot do C4K60 with H.265 with the Phantom 4 Pro V2.0. Also, rendering in H.265 takes Significantly longer than H.264. HEVC is cool to save storage if you keep a lot of big files but it's problematic.

But with the DJI drones the H.265 files are not one bit smaller -- they are 100Mbps just like H.264. In DJI drones you do NOT save any storage at all -- NONE!


Brian
 
But with the DJI drones the H.265 files are not one bit smaller -- they are 100Mbps just like H.264. In DJI drones you do NOT save any storage at all -- NONE!


Brian

I've checked the file sizes by testing H.264 and H.265 in flight. The H.265 is indeed small and the detail in the trees is less than H.264.
 
I've checked the file sizes by testing H.264 and H.265 in flight. The H.265 is indeed small and the detail in the trees is less than H.264.


Please put some numbers up so we can what your going on about. I always shoot in 4K30 MP4 and have tested both H.265 and H.264 and find zero difference in file size for a given video duration. With the settings as described the P4P closes the current file and opens another file when the file reaches 4GB, actually about 3.81GB. Each file is exactly 5:27 in duration so if I record for exactly 20 minutes that would result in 3 files of 3.81GB and 5:27 duration and one file, the last file, of 2.55GB and 3:39 duration. It makes no difference one bit if I shoot H.265 or H.264.

So, once again, you need to put up some numbers and list the resolution etc so we can see what your saying. Again, the P4P records at 100Mbps when shooting at 4K30 and there is zero difference between H.265 and H.264 as far as file size, file duration, and bitrate.

Tell me if you think this video, shot in 4K30 MP4 and H.265, is lacking detail...



Brian
 
Please put some numbers up so we can what your going on about. I always shoot in 4K30 MP4 and have tested both H.265 and H.264 and find zero difference in file size for a given video duration. With the settings as described the P4P closes the current file and opens another file when the file reaches 4GB, actually about 3.81GB. Each file is exactly 5:27 in duration so if I record for exactly 20 minutes that would result in 3 files of 3.81GB and 5:27 duration and one file, the last file, of 2.55GB and 3:39 duration. It makes no difference one bit if I shoot H.265 or H.264.

So, once again, you need to put up some numbers and list the resolution etc so we can see what your saying. Again, the P4P records at 100Mbps when shooting at 4K30 and there is zero difference between H.265 and H.264 as far as file size, file duration, and bitrate.

Tell me if you think this video, shot in 4K30 MP4 and H.265, is lacking detail...



Brian


I was shooting 2.7k60 with say a 5-second clip at H.265. Switched to H.264 to compare. Looked back and forth and noticed a difference in both trees and storage. The other finding was rendering with AE or Premiere at H.265 HEVC vs H.264. The H.265 HEVC at 240 Bit rate everything Max would freeze my 36 thread 2x 1080 32GB RAM NVMe PC (not the PC but the program). It's HEVC for a reason. Like I said, if you are shooting above 30fps you have to be in H.264.
 

Attachments

  • 265264.png
    265264.png
    27.9 KB · Views: 283
I did the comparison difference of 2.7k60 H.265 vs H.264 on a Mavic 2 Zoom. Those specs are off a Phantom 4 Pro V2.0. Also, another thing worth taking into consideration is the bit rates are Higher in H.264 vs H.265 at the same resolution and frame rate by around 20%. I did not notice before.
 
Last edited:
H.265 most definitely is harder to work with than h.264 and even simple playback of raw video in an Windows box will be an issue unless you have a system with latest generation GPU and/or CPU that natively handles h.265 files. The bitrate used does depend on resolution and with 2.7K60 the bitrate is limited to 80Mbps with h.265 but 100Mbps with h.264. When shooting at 4K30 it is 100Mbps no matter if you use h.265 or h.264.

That's why I asked you to put up some numbers so I could know what you were doing. I have a 4K camera in the P4P so I shoot in 4K and if I need a lower resolution for some reason I downres to whatever I need. You can upres as well, but it is NOT going to give you the look and detail of actual 4K. The only upside to 2.7K I can see is being able to shoot at 60fps -- the P4P is spec'd as being able to do 4K60 using h.264, but its a BS claim as anyone that's tried 4K60 can attest.

Interestingly, the much newer flagship of consumer drones from DJI, the Mavic Pro/Platinum, does not list a frame rate greater than 30fps until you drop down to HD resolution and it's max bitrate is 60Mbps. I'm surprised at how DJI went from pushing the boundaries when threatened (GoPro) and have since, in the last couple years, made no net improvements for consumers. We've been speculating about the Phantom 5 which now looks like it will never happen so the Phantom line appears dead -- sadly, the best of the Mavic line is not really equal to the 2.5 year old P4P and one wonders when or if DJI is going to advance beyond a 2.5 year old design for consumers -- ever!


Brian
 
H.265 most definitely is harder to work with than h.264 and even simple playback of raw video in an Windows box will be an issue unless you have a system with latest generation GPU and/or CPU that natively handles h.265 files. The bitrate used does depend on resolution and with 2.7K60 the bitrate is limited to 80Mbps with h.265 but 100Mbps with h.264. When shooting at 4K30 it is 100Mbps no matter if you use h.265 or h.264.

That's why I asked you to put up some numbers so I could know what you were doing. I have a 4K camera in the P4P so I shoot in 4K and if I need a lower resolution for some reason I downres to whatever I need. You can upres as well, but it is NOT going to give you the look and detail of actual 4K. The only upside to 2.7K I can see is being able to shoot at 60fps -- the P4P is spec'd as being able to do 4K60 using h.264, but its a BS claim as anyone that's tried 4K60 can attest.

Interestingly, the much newer flagship of consumer drones from DJI, the Mavic Pro/Platinum, does not list a frame rate greater than 30fps until you drop down to HD resolution and it's max bitrate is 60Mbps. I'm surprised at how DJI went from pushing the boundaries when threatened (GoPro) and have since, in the last couple years, made no net improvements for consumers. We've been speculating about the Phantom 5 which now looks like it will never happen so the Phantom line appears dead -- sadly, the best of the Mavic line is not really equal to the 2.5 year old P4P and one wonders when or if DJI is going to advance beyond a 2.5 year old design for consumers -- ever!


Brian

I mean, the sharpness is quite unbelievable at C4K60 though. It's a lot better when it's straight off the drone. I had to edit through Premiere to cut then to AE and this was at CBR 240Mbps bit rate too on each. Still, some loss happened.

Here's the original but I'm not sure if Youtube compresses.

You are saying at C4K60 it's not full C4K60?
 
Last edited:
I mean, the sharpness is quite unbelievable at C4K60 though. It's a lot better when it's straight off the drone. I had to edit through Premiere to cut then to AE and this was at CBR 240Mbps bit rate too on each. Still, some loss happened.

Here's the original but I'm not sure if Youtube compresses.

You are saying at C4K60 it's not full C4K60?


There have been many complaining about dropped frames and other artifacts using 4K60 -- I tried it a while ago and I don't recall the exact issues except to say I decided to stay at 4K30. Also, given the limited bandwidth of 100Mbps, using a video format that could double the data rate required would result in greater compression than at 30fps -- it's not a linear relationship but I'd expect the required data rate for comparable quality 4K60 to be 30%-50% greater than 4K30. Even at 4K30 I get problems with image tearing that appears to me to indicate the compression hardware or firmware is marginal at 4K30. The problems others have reported at 4K60 would further suggest the hardware and/or firmware is marginal at 4K period. DJI could have had another purchase from me if they'd released a Phantom 5 with usable 4K60 and a bitrate in the 150Mbps range or a Mavic model with improved camera and 4K60 with the higher bitrate in the 150Mbps range. That's my issue with DJI, as I mentioned in my prior, they killed off the biggest threat in GoPro back in 2016 but have not released anything in more than 2.5 years that actually improves upon the camera of the P4P.


Brian
 
...Again, the file size will be the same for a minute of video. You will not save any drive space using H.265 with the DJI drones -- not a byte!
Yes I have found this to be the case. The file cuts off at about 4GB and starts a new one. But the 4GB file has the same number of minutes for both codecs.

What I have also learned some time ago is that the P4P has a bug in the H.265 codec. Just after the new file is created, a couple seconds into it there is a video freeze of, I forget, maybe a second or so. I’ve documented this in detail elsewhere on this forum. I’m not sure if a fix for this has crept into any more recent firmware releases (I’m on the 509 version). But it’s something I’d look out for. I stopped using H.265 because if it.

Edit:: note, I was shooting 4K30. And these were not dropped frames. It was a freeze hold for maybe 30 frames or so and then jump forward.
 
There have been many complaints about dropped frames and other artifacts using 4K60 -- I tried it a while ago and I don't recall the exact issues except to say I decided to stay at 4K30. Also, given the limited bandwidth of 100Mbps, using a video format that could double the data rate required would result in greater compression than at 30fps -- it's not a linear relationship but I'd expect the required data rate for comparable quality 4K60 to be 30%-50% greater than 4K30. Even at 4K30 I get problems with image tearing that appears to me to indicate the compression hardware or firmware is marginal at 4K30. The problems others have reported at 4K60 would further suggest the hardware and/or firmware is marginal at 4K period. DJI could have had another purchase from me if they'd released a Phantom 5 with usable 4K60 and a bitrate in the 150Mbps range or a Mavic model with an improved camera and 4K60 with the higher bitrate in the 150Mbps range. That's my issue with DJI, as I mentioned in my prior, they killed off the biggest threat in GoPro back in 2016 but have not released anything in more than 2.5 years that actually improves upon the camera of the P4P.


Brian

There are no dropped frames if you view it on the latest smartphone. On a Windows PC, it looks like there are dropped frames but it's our PC, not the video. It's rock-solid smooth if viewed w/ the latest Apple. C4K60 issues are probably on defective P4Ps or P4Ps not V2.0. I got the latest drone with 07/18 mfg. date. Latest firmware and using it on a Note 9. I have the latest Evo 64GB Samsung microSD. If the microSD is bust it can artifact too I assume.

I mean, there can also be dropped frames when rendering w/ cheap software and a not so powerful PC. I have an i7-2696v3 which is 18-core 36-thread. 2x 1080 so I can do other things too. Also NVMe. 32GB of RAM this way it won't hiccup at 100% use somewhere and render as intended.
 
Look at my post.

Is the H.264 better than H.265 even if the bit rate is shown to be 20% higher? Is the quality of H.265 at 100Mbps bit rate eq. to H.264 at 50Mbps bit rate? Is this how this works? I've ran my own test w/ the Mavic 2 and the H.264 had more detail in the trees vs H.265.
 
I have a M2P and started with H.265 and quickly realized my computer hated it and I wasn’t skilled enough to edit it. I switched back to h.264 and do not regret it!
 
In January, I upgraded my dual Xeon workstation to Windows 10 Pro. I did this after building another general purpose PC that acted as a backup NLE. The new PC required Windows 10. That's when I discovered that Windows 10 Pro can play HEVC with no strain, while Windows 7 would just see the player freeze.
I upgraded the NLE to Windows 10 Pro and found that I could also play HEVC with no efford (under 1% CPU). Then I upgraded Premiere to 2019. I've been using NVENC to hardware encode HEVC for about a year. HEVC is my Youtube upload format.
The P4P v2 records at 100mbs regardless of CODEC chosen. The quality is notably better in HEVC, which I started using this spring. Zoomed in to 400%, even dense forest remains clean, looking more like Prores out of my cinema camera than heavily compressed footage from a drone.
The workstation with the latest software and updates handles it well, encoding 4K HEVC from 4K HEVC footage in about realtime speed.
Oddly, Premiere plays HEVC more smoothly than the h.264 I had been using.
I shot this castle using HEVC. In 4K, you can see the quality is quite good:

 
In January, I upgraded my dual Xeon workstation to Windows 10 Pro. I did this after building another general purpose PC that acted as a backup NLE. The new PC required Windows 10. That's when I discovered that Windows 10 Pro can play HEVC with no strain, while Windows 7 would just see the player freeze.
I upgraded the NLE to Windows 10 Pro and found that I could also play HEVC with no efford (under 1% CPU). Then I upgraded Premiere to 2019. I've been using NVENC to hardware encode HEVC for about a year. HEVC is my Youtube upload format.
The P4P v2 records at 100mbs regardless of CODEC chosen. The quality is notably better in HEVC, which I started using this spring. Zoomed in to 400%, even dense forest remains clean, looking more like Prores out of my cinema camera than heavily compressed footage from a drone.
The workstation with the latest software and updates handles it well, encoding 4K HEVC from 4K HEVC footage in about realtime speed.
Oddly, Premiere plays HEVC more smoothly than the h.264 I had been using.
I shot this castle using HEVC. In 4K, you can see the quality is quite good:

Do you have any trouble in the first seconds of a new file after you cross over the 5 minute mark (4GB file limit) and a new file is created?
 
Really? That’s your aircraft firmware?

I’m on 1.03.0509. And that’s fairly old. And I have the problem every time. Tried different cards too.

I'm pretty sure. There are several other numbers below it, but they are very similar values. The last time I updated it was about a year ago. It flies perfectly, so I don't want to risk upsetting it with a possibly buggy firmware.
 

Recent Posts

Members online

Forum statistics

Threads
143,095
Messages
1,467,610
Members
104,981
Latest member
Scav8tor