|
|
Quote
I suggest you report the issue directly to CL tech support and see if they can get things working properly. You can reach them here.
I've never used the feature so I can't easily test this, but you might try uninstalling PhD and then manually delete all the related 14.02 folders you can find. When you reinstall PhD, it's possible that it will properly process the data in the 14.01 folders this time.
I did not even have to re-install Photodirector. I tried removing all references to the 14.02 version as you said and imported the files from version 14.01.
Apparantly the newest version of Photodictor converts the old files well, as I have all my faces back!
Now I am importing the photo's taken since the date of the previous conversion to 14.02.
Face recognition is a tedious process and I have put a lot of time in it.
You saved my day, optodata! Thank you very much.
|
|
|
Quote
I suggest you report the issue directly to CL tech support and see if they can get things working properly. You can reach them here.
I've never used the feature so I can't easily test this, but you might try uninstalling PhD and then manually delete all the related 14.02 folders you can find. When you reinstall PhD, it's possible that it will properly process the data in the 14.01 folders this time.
Thanks, optodata, will do.
As you served me with excellent advice before, I will try to re-install Photodirector 365 first.
|
|
|
I lost all information about facial recognition during the update from Photodirector 365 version 14.01 to 14.02.
In the backups, I see that the files have been reduced in size to almost zero:
- In the \14.0\Noud_13.02_14.01\Noud_13.02_14.01_faceme folder, the file fm_thumbnail.db is 553MB in size, while in the \14.0\Noud_13.02_14.01_14.02\Noud_13.02_14.01_14.02_faceme folder, it is only 20kB.
- In the \14.0\Noud_13.02_14.01\Noud_13.02_14.01_faceme\CyberLink\FaceMeSDK\db folder, the faceMe.fdb file is 96MB, while in the \14.0\Noud_13.02_14.01_14.02\Noud_13.02_14.01_14.02_faceme folder, it is only 48KB.
I tried restoring old files exclusively from the Faceme folders.
I realize that there may be a mismatch with the other files in version 14.02 and indeed it doesn't work.
The folder structure has also been adjusted, making it difficult to determine which file should be where.
Who can help to transfer the facial recognition from an old file of version 14.01 to 14.02?
|
|
|
Quote
The blue arrow is just to go to the top of the current page.
The forum has several threads with messed up "solution" controls. I saw several threads get hacked when CL fixed a previous forum issue. You can see one such thread here: https://forum.cyberlink.com/forum/posts/list/83087.page#post_box_343521 optodata post is listed as a solution to his own post, usually not possible. Many other threads like that too.
If the mark is offensive to you, contact forum moderator and see if they will correct.
Just verify that PD is doing what you want after "Produce" to common profiles and results are good, audio too. Several issues with all "I" frame source from certain cameras and support aware of them with: "Cyberlink tech support responded that they've been able to reproduce the issue in their lab and have forwarded it on to engineering." I've not tried ffmpeg all "I" as source to know if the same issue persists. I've not seen a fix yet.
Jeff
Thanks for clarifying this. So unintended behavior of the blue triangle.
I am not bothered by the wrong solution vote, but it might be confusing for other forum users looking for the solution.
I hope Cyberlink will support all modern codec and bit levels soon as I see many having problems with them.
|
|
|
What is the blue triangle in the top right-hand corner next to each message?
I was curious and clicked on it. It than marked the answer as solution and I cannot remove it.
If this is the intended behavior a pop-up with some explanatory text would be appreciated.
How can I remove my ‘solution vote’?
|
|
|
Quote
FYI - I've now also submitted the following to PD:
Thanks, Nathan, for bringing this to the attention of Cyberlink. Your explanation is far better than what I could have written. Probably has more impact too, coming from an experienced beta tester!
I cannot wait for the results 😊
|
|
|
Quote
Looks like you are spoilt for choice on your XF705!
- As you are in EU Stick with 25/50fps : If you ever need to do a project for someone in a NTSC country you can always change everything to 29.97/59.94 .. but just set, forget (and check) you are on 25/50fps
- Looks like Filmic Pro (and a bunch of others) is available for the iPhone that lets you change your FPS which is good news
Some other things to try (if you are bored): Having a better play trying to import various formats into PD, the biggest limitation in PD is the formats it will accept in 4:2:2 sub-sampling is only ProRes and V210. PD seems rather happy to accept 8 or 10-Bit content. Many will argue 10-Bit is more important than 4:2:2 for consumer production (as there is no 4:2:2 consumer content anyway) as you get 1024 levels for 10-Bit over 256 in 8-Bit. Perfect for HDR. The next limitation is that unfortunately PD does not seem to support producing in (HDR) 10-Bit for any format that I can find (but I'm sure it will come at some point). Even so, 10-Bit could still be of interest to you if you want to do any grading of your footage.
So at present I'd suggest
- If you want a "Quick" to produce method in PD, I'd set your XF705 to shoot UHD, Wide DR, in AVC (8 Bit 4:2:0) @ 50fps and convert to MKV (for 4ch Audio converted to Stereo)
- If you want the "Best" quality and want to grade you footage in PD, I'd set your XF705 to UHD HEVC (10-Bit 4:2:2) @ 50fps. At some point PD will natively support this format but, in the interim you could use the "MKF HEVC to Pro Res" script that will take the Long GOP HEVC 10-Bit 4:2:2 footage from the XF705 and convert it to ProRes HQ all I Frame 4:2:2 that PD can edit. I've updated the util to make it (a bit) quicker and also changed the Audio Layout to match your preference. This Pro Res format will give you much more flexibility to grade, apply LUTS etc in PD. When PD adds a (HDR) 10 Bit output, you could then open up your old projects and simply produce in this new profile. I have no idea however when PD will add all of this. Only other option (for now) is to use another prog like Resolve.
- If your XF705 could shoot in a UHD 10Bit 4:2:0 profile (either HEVC or AVC) then it would be a great middle ground as you could just quick convert to MKV and I'm pretty sure it would import into PD. Unfortunately, I don't think the XF705 has this option.
Personally, I'd probably go with #2 (assuming #3 is not possible). It will be a pain to transcode to ProRes in the interim to use in PD but your original as shot footage will be kept in the best possible format the XF705 can produce while you wait for PD to support importing the 10:Bit 4:2:2 HEVC clips and allows for the production of (HDR) 10:Bit projects.
Thanks
Nathan
Hi Nathan,
The XF705 has so many options to choose from to tailor it to your needs. A wonderful camera! A bit of a learning curve in the beginning, but once you get to know the menu’s it becomes easier. On top of that I like the programmable buttons on the outside that let you choose the most used options instantly during shooting.
Thanks for the good advice on shooting in 8/10 bit etc. The XF705 does not shoot UHD in 10 bit 4:2:0 indeed. For now, I will stick to your option 2. I used your tool to convert my HEVC footage to ProRes. The resulting MKV files are way bigger than the ones from camera, as you already warned me for. It does the trick but is more demanding for my storage capacity. And that is a shame, because the XF705 produces relatively small files by using H265 in camera. I hope that Cyberlink will honor my suggestions to create real support for the Canon MXF-files.
Enjoy your XF405!
Noud
|
|
|
Quote
Hi NoudvR
Here is some more info to consider:
- PD Time Line: Looks like your timeline is in 25fps. Got to the Gear Icon --> General --> Timeline Frame Rate --> Change to 50 FPS (PAL)
- MXF from 4 Mono to 2 Stereo Alt MKV: I've attached a modified script that now puts Mono Track 3&4 to Stereo Track 1, and 1&2 to Stereo Track 2
On the matching Frame Rate options, there is really only 3 ways this is done:
- Get all devices (Display device, Cameras etc) on the one FPS or clean multiple (eg 25/50/100 etc)
- Speed Up/Down: Works well of the FPS of the two clips are a close multiple/fraction. Eg, If you have 24 fps material, you could do "PAL Speed Up". Speed up your 24fps on the timeline by 4%. This will make the 24fps play a bit faster but will stop the need to do an uneaven pull down (frame drops or repeats). If on a 50fps timeline, your speed up 24fps material to 25fps, PD will then also just repeat each frame twice. Works very well and is used commerically to move "Movie" to "PAL". If you also want to keep the Audio on the 24fps material, you may need to do a Pitch Correction by 0.7067 of a semitone.
- Other Pull Down Patterns: By Repeating/Dropping frames you can match two very different FPS Clips on the timeline (PD Does this Automatically). Downside is you get a repeating judder that (too me) look terrible. Eg putting 30fps material on a 50fps timeline
- Interpolation: I don't know if PD has this option, but you can interpolate from one FPS to another, but the results (to me) look weired, kind of like going in and out of focus, as the sofware tried to make "new" frames at the new FPS by blending some frames from the original. FFMPEG can do this but I don't like the look (and it takes time as you have to render the clips) so I don't bother
The Real Answer is: Getting all devices on the same FPS:
- Smartphones: The built in "Video" apps normally have very little control. On Andriod (no idea about iOS), you can get 3rd party apps that lets you set the FPS (as well as a bunch of other more "advanced" options) for shooting. On my LGV50, I've can use both FILMIC Pro / Cinema 4K that lets me select UHD/25p that works well in PD (the Filmic Pro seems a bit off 25fps but seems OK).
- Drone: It too can probably shoot in 25 or 50p but you will need to check witht the MFR
- Canon XF705: You "may" be able to switch your XF705 to be able to select NTSC Frame Rates (eg 59.94/60fps) but Canon have "crippled" the firmware (at least in my XF400) so you can NOT see the option to select NTSC Frame Rates (and vice versa for thouse sould in the USA). Certaily on my XF400, I can pay for a SW update to expose these other settings in the menu but...... then I'd have to switch my TV etc to NTSC frame rates to enjoy judder free playback. Just not worth it.
Thanks
Nathan
To avoid confusion: the clip in the project was 25 fps as was the timeline, so in perfect sync I would assume.
Thanks for the change of channels!
I had not thought of using the speed option to correct for frame rate, so I am going to give that a try.
Excellent that you gave me the pitch correction in a number also.
I have an iPhone and will look for other camera apps that allow for 25/50 fps. My DJI drone is no problem because the settings contain all frame rates. I have learned to check these after every firmware update, because they can be reset to the default values afterwards.
In my XF705 I can change the system frequency which enables me to select ‘NTSC’ framerates. I still have not figured out whether there any real problems using this e.g. with my TV. Filmmakers and Canon advise against it in Europe. Why, I do not know.
|
|
|
Quote
I don't think that PD really support MXF well due to the issues about the exposing and subsequent mgt of the multiple Audio Tracks.
It is something that CyberLink needs to look at if it really wants to claim it can support MXF files (or that matter even other formats that do the same thing, eg on my XF400 I can also use MP4 to store the same 4 x Mono Tracks as the wrapper). The workaround about mixing the tracks in my utility, remuxing to MKV and then using "languages" in PD is a kludge. It is even more akward if you keep the 4 x mono tracks in the MKV remux then trying to manually deal with each track in PD to create say a Stereo or Surround mix.
Thanks
Nathan
I totally agree. Does Cyberlink have a kind of voting by users on all suggestions?
|
|
|
Quote
".....1-2-1-2-1-2 etc."
This is a Pull Down pattern and used to covert video from one FPS to another. You want to avoid this where possible.
Make sure that you set your time line for your project in PD to the same FPS as the majority of the media you import. So if your media is 50FPS, make sure your timeline is also 50fps and that you produce at either 50fps or 25fps (which is clean pull down of dropping half the frames). For Slo Mo you would use 2x on the time line (clean pull down that just doubles each frame). The situation to try to avoid is mixing 23.976 / 50 / 59.97 etc fps media in one timeline. This can be hard to do if you also have footage from another source (say a phone) where the FPS is not (or can not) be set to the same frame rate.
Both the clip (according to MediaInfo) and the timeline (according to Preferences) are 25 fps.
I have uploaded the .PDS-file in the folder Stutter in timeline PD19, so you might check it out for yourself. I sent this .PDS file to Cyberlink Customer Support and they acknowledged the problem and escalated it to their engineering team.
It is also my experience not to use footage from my smartphone (30 fps) or my drone (30 fps after a new firmware update ☹) in a 25-fps timeline. Are there any tools (like ffmpeg perhaps) that to a better job than just leaving out frames?
|
|
|
Quote
I'm really glad you were able to get the audio working in Resolve, and even better to have the specialized tools from jmone which seem to have gotten everything straightened out, so I'm going to assume that we don't need to worry about any more MXF conversions.
The audio conversion problem has indeed been solved. I hope Cyberlink will incorporate native MXF-support containing more audio channels in PD19 soon.
Thank you very much for your continued support!
|
|
|
Quote
I'd suggest with such a great camcorder you should look at shooting in 50fps all the time (I do on the XF400), as it not only gvies you smoother pans but you can then do 2x slo mo in post and still get 25fps. I presume the XF705 has a UHD 50P AVC profile and the util will remux the 50p AVC to MKV as well.
Also let me know if you want the order of the 2 Stereo streams swapped. At present the 1st is from Ch3/4, and the 2nd is from Ch1/2 but if you normally use Ch1/2 then it may be easier to have that as the default.
So many people advised me to use 50 fps that I will continue to do so (after some tests I even made the first shots of my second grandchild in 50 fps!).
I can (as you probably know as a XF405 user) select which audio inputs I want on which channel.
I have put my external mic on 1+2 and the internal mic on 3+4.
Swapping the two video streams would be nice, because it is more intuitive.
|
|
|
Quote
I thought I reproduced all the steps you mention correctly by panning channel 1 in Audiodirector to the left and channel 2 to the right. The result however is mono, whichever track I mute or play both. I have attached a screen dump from Audiodirector and the produced audio. Please advise what to do.
18-10-2020: Thanks for your efforts! Jmone has solved my problem today. Now I get 2x2 stereo, which can be edited in PD19.
Your post however revealed that I want to study and experiment more with Audiodirector!
|
|
|
Quote
No Probs: Here is a few more batch files to play with:
- MXF from 4 Mono to 2 Stereo MKV: Will create 2 x Stereo tracks form Ch1/2 and Ch3/4 with AVC in MKV that shoud work in PD
- MXF HEVC to Pro Res Stereo and 2 Mono WAV Files: This will transcode (and take some time) the HEVC to ProRes with a Stereo Track in MOV + 2 x Mono.WAV files for the other audio tracks. These should all import into PD and work fine but be warned the MOV file will be 10x the size of the MKV but will keep your full 4:2:2 Video Steam in a PD compatible format.
- MXF HEVC to Pro Res Stereo and 1 Stereo MKA File: This will transcode (and take some time) the HEVC to ProRes with a Stereo Track in MOV + 1 x Stereo.MKA file for the other audio tracks. These should all import into PD and work fine but be warned the MOV file will be 10x the size of the MKV but will keep your full 4:2:2 Video Steam in a PD compatible format.
Thanks
Nathan
PS - let me know what combo you need and I'm happy to knock up a custom script.
The 2x2 stereo conversion suits my needs 100% and it works perfectly!
I will experiment with the 4:2:2. This will enable me to use 50 fps which diminishes the stutter in exports (only visible with fast panning or rapid moving objects).
|
|
|
Quote
No Probs: Here is a few more batch files to play with:
PS - let me know what combo you need and I'm happy to knock up a custom script.
The world would be a better place if people help each other like in this forum. And it is Corona proof! I rather wear a headphone than a mouth cap. Thanks for all your efforts. Noud
|
|
|
Quote
FYI - I've updated a simple FFMPEG based utility that will (lossless) re-mux MXF to MKV so audio will work with PD19 (for AVC). More Info at https://forum.cyberlink.com/forum/posts/list/83622.page Works well for me on my Canon XF400 MXF clips (AVC based). This will not help with UHD HEVC 4:2:2 MXF support in PD19. If you need that you will have to look at other NLE (like Resolve). This "fix" however will make your 4 x Mono Audio tracks in a UHD AVC 4:2:0 MXF clip useable by remuxing them to MKV.
Thanks
Nathan
As Optodata said: Wow! Quite impressive, as it works very well, amazingly fast and works in PD19!
In your other post where you uploaded the software you were quite modest saying it is easy to adapt the .BAT files. For you as a professional that might be true, but not for me.
Would you be so kind to upload a .BAT file that generates 2 times stereo from channels 1+2 and 3+4?
My internal mic is a back-up in case I forgot to switch on the power of my external mic (bit of a flaw in this mic that it does not turn itself on when recording).
|
|
|
Quote
To be certain that a 4:2:2 clip is converted correctly for editing in PD I absolutely need video content. Record the view from your window, water in the sink, your lawn, anything with content that can be assessed for quality issues after converting.
I now see that there are 2 clips in the HEVC folder, but for some reason the XF Utility is only showing me the 3 AVC clips:
Looks like there's still more work to do on my end...
I have added a new clip to OneDrive as requested. It is in the HEVC folder. Clip date 18-10-2020.
The XF Utility shows all clips in one ‘mediaroom’ per memory card, so mixed XAVC and HEVC.
Another cause of apparently different MediaInfo might be that there is a setting which generates new clip names with sequent numbers. The clip name can no longer be compared to the original ones.
Last night I found out that importing the clips in Davinci Resolve and using the panning tool in the fairlight tab did the trick. I can export this to PD. It was too late to late you know immediately 😊.
I am going to read the new posts now, as there seem to be other solutions as well.
I hope that I can use my (renewed yesterday) subscription to Cyberlink with Canon MXF XAVC and HEVC files natively soon without stutter!
|
|
|
Quote
I have tried to look for the manual of Audiodirector, but the ? and ‘user's guide’ at the top right only brings me to the download of the software itself.
The manual for Audiodirector 11 is probably not finished yet. I found version 10!
|
|
|
Quote
On the positive side, the XF app was finally able to see the 3 clips when I copied the CONTENTS folder onto a USB drive. The bad news is that they're all displayed as black/empty video, even though I have the Canon XF-AVC and XF-HEVC Decoders installed - so I still have more work to do.
The other issue is that all 3 clips on the SD card (the ones that I can't play) are 25p AVC 4:2:0, not the 50p HEVC 4:2:2 clips I was expecting. Can you clear the SD card, change the settings and record at least one clip then upload the contents again? Make sure the lens cap is open, just in case 
Sorry for the confusion, but to prevent images of my hardware and desk etc. on the internet I choose to keep the lens cap closed. I thought you would not need the video. Using headphones, the panning from right to left in front of the audio source can be heard clearly, so no video needed. If you do not hear a difference: that is my problem ☹ it is mono. Playback in Canon XF Utility clearly reveals stereo sound because it can switch on and off each of the four channels and attribute them to left or right. I already asked Canon to have the option of exporting the resulting video/audio combination.
The clip in de HEVC_001 folder is 50 fps according to MediaInfo:
Video
ID : 2
Format : 0E15000402100001-0E15000500013000
Format settings, wrapping mode : Frame
Codec ID : 0E15000402100001-0E15000500013000
Duration : 11 s 520 ms
Bit rate : 152 Mb/s
Width : 3 840 pixels
Height : 2 160 pixels
Display aspect ratio : 16:9
Frame rate : 50.000 FPS
Color space : YUV
Chroma subsampling : 4:2:2
Bit depth : 10 bits
Scan type : Progressive
Bits/(Pixel*Frame) : 0.367
Stream size : 209 MiB (97%)
Color range : Limited
Color primaries : BT.709
Transfer characteristics : BT.709
Matrix coefficients : BT.709
The other two files are for those who want to join in using 4:2:0 H264 files.
|
|
|
Quote
All I did was work with the two audio files from the mkv in stereo. Removed channel Left from "Audio 1" source and copied "Audio 2" source to Left channel of "Audio 1" source and produced as stereo. So in the end one has stereo output file with:
Channel R is from "Audio 1" mkv source
Channel L is from "Audio 2" mkv source
Jeff
I am afraid that I must ask you for some more guidance because I am new to his kind of stuff in Audiodirector.
I can see the two language channels in PD. Only one at a time can be selected. When I transfer this file to Audiodirector it only sees a mono track. I cannot see any option (even greyed out) that would remove a left channel or to copy Audio source 2 to left, as suggested.
I have tried to look for the manual of Audiodirector, but the ? and ‘user's guide’ at the top right only brings me to the download of the software itself.
|
