There could be more than several reasons for this to happen and difficult to give an accurate reply to without seeing the resulting video or at least a clip of the problem.
If you could provide a link to a clip of the exported video that would help a lot. Also, is the playback in the program OK?
In addition to @CubeAce comment we need more information. Assuming you mean the exported video flickers on playback in a media player:
Computer specification, Windows version and program version, see this topic for what is required and please quote processor and graphics card make/model in full, and also what monitor/screen resolution(s) your are working with if this is a laptop.
I would suggest you put this information in your profile signature so we do not have to keep asking for it.
Are the graphics card drivers up to date?
A screen shot of the Program settings, Display options or Device options tab would help.
What is the source video recording device ie camera, phone screen capture etc, and resolution/encoding?
Download and install MediaInfo and analyse one of the clips and post the results, see this tutorial on how to setup MediaInfo and analyse a video clip for all the data required.
In the Video section of the MediaInfo analysis if you see this: Framerate mode: Variable then the video needs converting to Constant framerate using a converter such as Handbrake or AviDemux
If the framerate mode is Constant, then copy and paste the MediaInfo into your reply.
Export preset your are using and did you make any changes to it?
Setting the frame rate to the same frequency as the local mains supply can be problematic as some older mains powered LED lights would go on and off with the voltage drop between cycles. Same with incandescent lighting but slower to react. Mixing both is worst scenario. More modern cameras now detect this and compensate or can be adjusted for in the settings menus. I believe this can now be fixed with Neat Video.
At the risk of getting away from the point of this topic, the wedding was in mid 2016 and I was using MEP 2015 if memory serves me well - not so many whistles and bells to play with. I've since re-edited it wirh MS2023 (including Neat) and it's much better even though I can still see the "scrolling" slightly.
When I say the video was unusable at the time that was not strictly speaking true, disappointing would be a better word.
To be fair to MEP2015, in the worst scenes I tried a hack that I read about somewhere involving duplicating the video onto another track and phasing the second video 1 frame forward (removing the audio onviously).
I think it appeared to decrease the scrolling visually but that may have been wishful thinking. It did also seem to reduce sharpness as well slightly.
All - Thank you for your responses and assistance. Here are the beginning of answers to your questions. I'm out of time to post the rest, as I have to get to work in a few minutes. Thanks in advance for whatever you can give me!
Playback of the clip in the program is excellent - as a matter of fact, after rendering, the clip is of a lesser quality than the original files used to render. That's something else I've been trying to determine how to fix. I'd welcome feedback on that too.
If the file is playing back and looking good in the program then the problem is probably in your export settings.
Are you using the 'Export Movie' option and then select the correct resolution / frame rate from the list of options in the a top drop down box?
Files for us to test would have to be put either on Dropbox or Google Drive or My Drive with a link to it here. I think Google is possibly the least hassle.
All - Again, I apologize for the delay. Life is busy! A few more answers to your questions: I installed MediaInfo and grabbed a snippet of the rendered video. Below follows the output of that:
For comparison, I've also included a picture of what MediaInfo has to say about one of the original video files:
For the export settings in movie studio, here are some pictures of those:
I'm also linking to a clip of the rendered video on my google drive, so you have a reference to what this appears like. The pulsing is most visible in the outside edges of the picture, but I can see it everywhere if I look for a few moments. This is a wedding video, so is about 30 minutes in length for the ceremony, so figure it would give people a headache after too long.
@Darynn-Ho Hi, You need to right click your file on Google Drive & change permissions to 'Anyone with this link'
I have a similar PC to yours, & an S23 Ultra phone (previously S21), you're going to get more comments about 'Variable' frame rate & you have to convert it to constant, there's no doubt constant works better but I very rarely have problems & if I get any issue it's playback in the Magix software itself, you've already said it plays well in MMS (Magix), All my footage is variable, sometimes greatly so,
These clips/files do flicker under florescent lights but that is easy to see in the preview, it doesn't just appear in the export.
Check your GPU (Graphics card) is up to date, Nvidia Geforce will help, that's what i use.
Additionally - it appears I do have the latest driver for the graphics card via a check in that GEForce software.
Former user
wrote on 6/17/2023, 1:48 PM
@Darynn-Ho I can't see pulsing but i can see blockiness of the colour in the sky, this is the bitrate that's too low,
It's not easy if your not comparing the same file but look at the Overall bit rate, Orig 73mb/s to the exported one 24mb/s, that's 1/3 of the orig,
Use one of the presets to export, you can change it to Best if you wish,
PS, I think I see what you've done, you've set the Bitrate to 72,000 thinking you'll get 72,000, The frame rate comes out at Constant 60 but the Bitrate is still variable, don't know how it works but the Max & Min have to be high enough that the exported average = the orig of 72,000, something like that.
I agree with Gids' analysis of the original file and your copy which has ended up too compressed.
These areas are losing the most information. If the original has blown highlights though that may not improve by much whatever you do.
However
Gids' graphics card is way more powerful than any other contributing member of the forum hence I feel has the least problems of anyone here when dealing with variable frame rate files. His 3090ti has the GA102 graphics processor which is a large chip with a die area of 628 mm² and 28,300 million transistors. It features 10752 shading units, 336 texture mapping units, and 112 ROPs. Also included are 336 tensor cores which help improve the speed of machine learning applications and better able to deal with the vagrancies of variable frame rate files.
I'm sure this is why Gid has much less problems with filling in the action between static frames where the frame rates are lower than most of the rest of us where our graphics cards would just repeat the same frame until the next change in movement occurred.
So for you I would suggest seeing if converting the file to a constant frame rate would help as well as setting your export settings to something that would give at least the same lack of compression to the video as the original.
These settings should work on that front.
You can type the settings in manually.
You can also try @Marc-Goder s comment on deselecting the anti interlace filter which personally I don't find makes a lot of difference but you never know. The HRD setting just gets the program to check what it is going to do twice.
The lack of compression will get rid of the blockiness but may not stop the juddery playback, in which case converting the original to a constant frame first and using that file with my suggested settings might do it. It may still not work where the frame rate is as low as 15fps.
You may still see over exposure with a lack of detail in brighter parts of the image or staccato movement in such bright sunlight as the exposer time for each frame may well be much less than the frame duration. That leaves blanks of time where movement would naturally occur between each frame. The slower the frame rate the more obvious this lack of movement becomes apparent to the eye. This is where if one was using a dedicated video camera one would consider using an ND filter to lower the amount of light hitting the sensor and keeping the shutter open for a longer duration during the exposure of each frame.
In future I suggest making sure if you possibly can, to select a constant frame rate when recording and the brighter the conditions the faster the frame rate you choose will help.
Personally I dislike variable frame rates as it prevents the natural flow of movement but so does not having the correct exposure for a given frame rate which is difficult to achieve for a camera with a fixed wide open aperture and no access to ND filters.
Ray.
Former user
wrote on 6/17/2023, 3:57 PM
@CubeAce Samsung phones don't have the option to record constant, don't think I've had a phone that does.
& not directed at you but I have to add - pretty much all my files have been phone clips since I started with MEP 20yrs ago, I don't have any of the orig early clips but the stuff from the last 5yrs has been variable, I've only had this PC 2yrs, my previous less so powerful PC's (i7700 GTX 1660) I didn't have a problem, in fact it's only in the last couple of yrs I've started learning more about variable, constant & converting, because although MP4 AVC variable will play ok converting to ProRes will give a much better playback in Vegas & Boris, I've never had to think about this with MEP/MMS (Not so in MMS with ProRes it stutters, )
Logic says a constant clip should play better than a variable clip but to my memory I've never had to convert a file in the past because it was variable, & I'm sure 'domestic' software like MMS where a huge amount of their users are phone filmers the software producers surely have had to make their software as variable friendly as possible.
I must probably have one of the oldest smart phones on the forum which I have never used for video as it is basically awful. It is possibly one of the lowest end Samsung phones out there. So I've just made a couple of videos in my kitchen and played with the very limited recording options. As you said and I was surprised, the files are variable frame rate with no option to turn that off. Second surprise was although my front facing camera has less MPs it gives the largest frame size of 1920 x 1080. The rear is less 1280 x 720 and no HDR with either camera. Under artificial light, the largest frame rate variation I could get was Minimum frame rate : 14.988 FPS Maximum frame rate : 29.980 FPS.
I even mixed the clips and exported and had no juddering or more blockiness than in the original using the following settings.
Media Info on the left is original file. Right exported file
It appears I can't currently upload any imges. I will shut down and restart my PC after posting this and try again as that normally cures that problem.
Yup that cured the upload problem.
So, could the following make a difference? Processor in the phone and it's compression ability?
Distance of main action and background differences?
Several clips being loaded maybe with different focus distances and exposures taken? (exposures being automatically set and not readable in the exif data).
I notice with your phone examples that it even records fine sawdust flying around and most of your videoing is done at closer quarters than the example given.
To make my project. I dragged the files into the program and accepted the programs offerings on the settings needed and used my own export settings which I will try to upload for this example.
[Edit]
What is your average Mb/s for 4K compared to @Darynn-Ho examples?
Ray.
Former user
wrote on 6/17/2023, 7:03 PM
@CubeAce Just found 1 of my old Nokia phones, might give it a clean & see if i can get i up n running, 🤣
These are S21 files, the Total Bitrate is pretty much the same as the Overall bitrate in MediaInfo.
These are S23 files, I don't have many that haven't been recoded in some way
This the slightly variable Hospital S23 vid, I might have shared this before 🤷♂️
What I would like to look at is the same clip BUT the original if possible. I'll explain why.
It (probably) is something to do with the variable bitrate and the fact that the original is identified as 60 fps - as discussed above.
What I'm seeing is, your encoded clip only really shows the "pulsing" in the high contrast areas - the grass and the tree leaves against the sky.
When I put your clip into MS2023 and then scrolled through it frame by frame virtually every other frame is duplicated - indicated by the mediainfo info(?).
I know this means nothing by using your encoded clip but I put your clip through Avidemux using the decimate setting which removes duplicate frames and the resulting clip has no pulsing.
This is why I would like to try this with an original clip if possible.
I generally have no problem with variable frame rate phone video - on occasion the audio drops out of sync if there are a few edit cuts - and when the camera or phone is moving then pulses or flickers tend to get lost. Your static high contrast shot highlights a problem I've not noticed.
The other interesting thing you could try, just to see if it makes a difference, is try setting the movie to 30fps and exporting at 30fps (or something intermediate) just to see if that pulses.
The bitrate blocking is easily remedied, removing enforced duplicate frames could be a little more tricky.
Are you talking about the blotching sky? If so, try an HEVC export at the same high bitrates. I've had bad blotchiness on exported S21 videos on sky and water using H264 ("MPEG-4" in Magix parlance).
. . . . Logic says a constant clip should play better than a variable clip but to my memory I've never had to convert a file in the past because it was variable . . . .
I do believe the example clips, you have posted from your phone, have a relatively narrow band of VFR, IIRC ~ ±10% maximum of target framerate, the MediaInfo data from @Darynn-Ho phone has an extremely wide VFR band of 15 - 60fps averaging 41.8 fps with a target of 60fps.
Definitely needs converting the CFR, the export settings recommended above I also agree with.
Greetings, just out of curiosity is the video you're editing a first generation straight out of the camera/phone or is it a re-encoded copy?
I ask because in the mediainfo you've provided for the original clip it gives the frame rate, then the max and min frame rates and then original frame rate and real frame rate. I've never come across that before. All my 1st generation videos do not show an original frame rate but they do after re-encoding.
It could just be the way Windows and subsequent programs show various bits of data. I have noticed in other posts where Ryzen users post their machine specs that the screen shots do not 100% match what I see from the same Windows page when I take a screen shot.
Having said that there are so many variables here to come to any firm conclusions and I agree an unaltered original clip would be best to see what can or can't be done tested on more than one system. The data from the clip you quote is only 13.7GBs in size so would able to be uploaded to Google drive within its 15GB limit. That is still a large file given a lot of Video cameras would have cut that into several 4GB chunks.
Although We don't have an original file to work with I have found differences and similarities in the files between my and Gid's example, apart from the resolution and frame rate differences.
First both of the files (Both from different Samsung phones) do not contain any B frames. Only I and P frames.
The amount of frames between I frames differ. Gid's file has them spaced further apart than mine by an extra third. Gids variable frame rate has less gaps in the frames of the slower parts of his clip than my own.
Gid's I frame spacing
My I frame spacing. Note the lack of B frames in both video clips but my I frames are closer together.
We could assume by this that @Darynn-Ho files may have yet another I frame spacing.
Where there are no frames or Data in my phone's files athe tholwer frame rates.
Gid's file by comparison. More condense but difficult to be precise with the limited views from this app.
Also note the amount of bit difference between the I frames and the P frames on the top images.
There is hardly any data depth in the P and B frames and looks really compressed. Even the I frames do not have enough Bits in them to represent the original file.
My conversion with my settings by comparison.
So yes, I too would like to see an original file to see what sort of gaps the program is trying to plug where there is no data to draw from and whether just upping the bit rate or another possible adjustment could make a difference.
Aaargh, and I thought video editing was supposed to be fun! Now I've got to think about Bs and Is and gaps and Ps (well, yes, maybe after a cup of tea 😂) as well