Video Frame Rate Increaser

[Katrine Freggiaro]

Makingyour own animated GIFs is an excellent way to connect with your audience. GIFs are known to have a lower frame rate. Upload a video and lower its frame rate to convert it into a GIF. Use VEED to cut, trim and edit any video footage to remodel as GIF. You can also add text, stickers, emojis and drawings to give your animated GIFs a personal touch. VEED lets you download your videos in the GIF format, so you can share it on social media and instant messaging.

Essentially, videos are made of lots of rapidly moving images. In every second of a video, there is a fixed amount of images it flips through. The more images, or frames, it flips through in a second, the higher the frame rate. Higher frame rates lead to smoother viewing. Lower frame rates make videos appear choppy and jumpy.

It depends. If you want to make your video easier to download and share, a lower frame rate is the way to go. Lower frame rates mean smaller file sizes. Movies and films are typically played at 24fps or 30fps. Higher frame rates like 60fps are perfect for high-quality HD videos, but they tend to have larger file sizes.

I tested the other codecs and it seems like VP9 provides a little better framerate but not great. See video below- the recording software I used made the framerate a little lower than normal, but still without recording software it is around 20 FPS.

Thereotical OAKD-lite's stearo pair maximun frame rate is 200fps. But the actual tested result is about 80100 fps, is there any tuning method to increase the frame rate? I am using stereo images only and currently for capturing raw images!

Hello frankshieh ,

Could you share the full pipeline/code you are using? I believe depth could reach 200FPS without any special modes postprocessing filters enabled, at 400P resolution. Thoughts?

Thanks, Erik

Hello frankshieh ,

I think the main way to improve fps is to set pipeline.setXLinkChunkSize(0)

I have tried with OAK-D-Lite and I get about 115FPS on each mono camera with this script. Mono image sensor (OV7251) currently supports max 117FPS at 400P. Could you try it out?

THnaks, Erik

erik Sorry for necroing (idk if that's an issue here), but what is the FPSHandler class you import from the SDK in the script you linked? I can't find it in the docs and I get an error if I run it with the most recent version (1.13.1) of the SDK installed.

Hi, I tried and I can get 113 fps using your script but this is still far behind 200 fps. If I want to add a 200P cropped image mode with your source codes to increase the fps, where shall I start? Or the 120fps is limited by internal pipeline and nothing I can try to increase it for current internal firmware?

Hi frankshieh , currently that's the limit by the sensor config. I believe the current bottleneck is ISP speed (3A algorithms) - as that's combined 240 FPS. I believe we plan to update the ISP so 3A algos will only run once every eg. second or third frame, so this won't be bottleneck anymore and we can achieve higher framerates of these mono cameras (OV7251).

Thanks, Erik

Hi, when running high frame rate, we usually do not need the ISP 3A. The gain and shutter speed are usually fixed. So maybe you can have an option to disable ISP for users like us. We need higher fps and we do NOT need ISP 3A, and ISP 3A will surely disturb our application. When we need higher fps, we need extra LED lights in a short period. And we would like to set the sensor's gain and shutter speed our own manually and NOT by the 3A algorithm. Because these extra LED lights are not always on due to the high power consumption, and we only turn LED lights on when we detect the object is passing and falls in our view area of camera. And the 3A algorithm must have to be disable in our application. Otherwise the gain and shutter speed calculated by 3A will make the images to bright when we turn the LED light on for a short period(maybe onlt 60 frames@200fps)

Please help me out. I have a video of 0.04 frame rate and 02:46:40 duration, and I want to increase frame rate to 25fps and decrease duration to 10s-13s. I have used these commands for fps and speed respectively:

I have a MacBook Pro with a 2.2 GHz processor 8 GB of ram and an AMD Raedon HD 6750M 1024 MB Intel HD Graphics 3000 512 MB Graphics card. After playing League for a year, I've always had low frames, around 20-25. However, now League runs constantly at 15-7 frames and I don't know why. The game starts at a good 30 frames, but when minions spawn, it drops to 15. All of my video settings are at their lowest but what confuses me is that when I jump into a custom game with bots, I run at 20-30 frames the whole time.

The possible reasons that this is happening to you may be of patch 6.22. Many people have been complaining (here) about this, saying that changing the settings does not affect the frame rate drops, with people saying that the cause is due to the alpha client and the fix is as follows:

I am not associated with the service, but I tried it, and it really can be a solution.

The service is in Beta, you have to get a key from various websites.

If they are all out of keys, just wait a few weeks.

The new interface with a even more powerful machines will launch in 2 weeks or so.

Ps.: I had a friend who wanted to play on very low end machine (several times, different machines), and modifying the settings left and right did not do much. If the machine is just too weak, LoL will always run slow. And that game really needs 60 FPS, you cannot really cast spells and such with stutters.

If you live in the US, you might want to try "Geforce NOW" too, once it comes out. ( -now )

Also a cloud gaming service, though it's a bit (a lot) more pricey, and restricted to the US. However, it uses Nvidia architecture, and will be very fast - so they say.

I have a long video with a frame rate of 30 FPS that I want to convert into a 200x time-lapse with a frame rate of 60 FPS. My only problem is that avconv is unnecessarily duplicating every other frame in the output, making the 60 FPS output effectively 30 FPS. I want every frame to be unique. At a 200x speed increase and 2x frame rate increase, there is no reason to duplicate frames.

It turns out avconv isn't the better replacement to ffmpeg that I thought it was when Linux Mint, IIRC, removed ffmpeg from the official repository in favor of avconv. Anyway, ffmpeg is back and I installed it and found the equivalent command that doesn't duplicate frames:

Have you tried selecting 'High Speed' mode in Firecapture's camera setup. This uses a lower ADC conversion bit depth like 9 or 10 bits so gives a faster fps. As 8 bit mode is usually used in planetary this is fine. I assume you have 8-bit mode selected, as 16-bit will usually halve your frame rate. Also don't have gamma enabled, or leave it at 50 which is the same as disabled. Disabling preview debayering may give a few more fps on a lower spec computer.

Allocate around half your RAM to Firecapture's heap space on it's setup options. The first few seconds of capture are stored in RAM buffer and when full, starts writing to disk. If your frame rate drops only after the buffer space reaches near zero this means it's the slow disk writes causing your fps drop.

Try creating and using a virtual ram disk using the free software. Just remember that when the PC is powered down you loose all the data in the virtual disk so transfer it to the hard drive before switch the PC off

This gives you camera settings where you can set High Speed and USB Traffic. For a small ROI you can normally set the USB Traffic to 100% with no problems. If you start losing frames or get no picture turn it down a bit.

For info you can set the heap size on the Firecapture settings 'General' tab. This is the amount of RAM the program can use. I always have the RAM buffer setting enabled too as it gives a nice readout of the frame buffer and how many frames are available during capture. I think on later versions of Firecapture the RAM buffer automatically follows the heap size, so doesn't need to be set separately.

I believe AE and premiere use the same basic code for interpolation, but in AE, you can use the matte feature to help reduce artifacting if manually using timewarp instead of timeline mode. Also, there is no right or wrong answer because the framerate ratio depends on many factors such as how fast you are panning, what motion is in the frame, how fast the camera is moving in relation to the obect, and how smooth you want the image to be.

In my testing (aka fiddling) I've found 50-80% the sweet spot for pre-comping, then timestretching the same amount to reverse interpolation.This gives you access to smoother frames while maintaing access to AE's matte feature. If you go this method, it's important to use whole numbers as truncating a fraction may not match everywhere and you may introduce a slight time change.

I didn't even notice you were using a plugin, but adobe's and twixtor's are really close in quality. If you look at the comparison videos, I'd almost wager they were from the same original code. I honestly don't know if twixtor would be different in Premiere vs AE but I don't see them changing the core code.

I like using AE's timewarp because you can control how much warping is going on. Twixtor's warp amount is proportional to the output FPS you want, so may be overkill in some instances. And there is a twixtor Pro that supports matte splines for 3d rendering.

Matte Layer: The layer to use as a matte for defining the foreground and background areas of the image. White areas in the matte represent the foreground, black areas represent the background, and gray attenuates between foreground and background.

3a8082e126