At relase time for Snowmix 0.5.2.1 (ultimo 2025), installing required libraries and packages for for compiling Snowmix on Linux Mint Cinnamon 22.2 result in package conflict for some packages. The Linux Mint Community has fixed the issue and Snowmix can now be compiled on Linux Mint Cinnamon 22.2. And Snowmix still compiles fine on Linux Mint Cinnamon 21 and 22.3. Best regards Peter Maersk-Moller
New version 0.5.2.1 Released. Support for macOS and OS X missing for Snowmix 0.5.2 has been enabled again. Snowmix 0.5.2.1 has been tested with macOS and OS X. More details on compile and install for macOS and OS X are available. The page on requirements also have more information on this. Fixing issues with some path preventing GUI scripts such as snowaudio.tcl, snowscene.tcl and more to work. This was mainly an issue for macOS/OS X but in some cases also for various Linux distros. Adding native...
Python Testers Wanted. I'm looking for somebody wanting to try out using the SNowmix embedded Python interpreter. It is currently alpha code and needs to be tested out to mature the implementation . So if you want to write some Snowmix configuration with some Python scripts included, please contact me. I can help with the Snowmix part. Best regards Peter
Welcome back Peter, I honestly thought you'd were done with Snowmix and had found a better project to work on? Unfortanely, I've continued to try and try again in developing and learning code old C for my DATV repeater project and have been year in to understanding the GStreamer Framework. I have even written my own GStreamer plugins to do the transitioning between video and audio feeds using SIMD which has a been a long and rewarding learning cure. Anyway. Good to have Snowmix 0.5.2 out and look...
Thanks Peter It was a while ago but I did try pretty hard to get in touch with you. I remember I sent you a message on linkedin. If you're replying on here then I suppose that's enough. I don't have the details for the improvements I made at hand right now, and I'll need to take a look at the new version and work up a patch. I'll just open a new ticket on here. Thanks
Hi Antisol. The project has no public available Git repository. For now you can post here in the "General Discussion" or you can sen DM to my account here. If you have anything more substantial, DM me and I can send you my email. That would usually require I know who I am sending to. I am not flashing my identity but also not hiding it and with my name, I am easy to find on the Internet. Bug reports are more than welcome and fixes too.Suggestions for features are welcome too. Best regards Peter ...
Awesome! i'll be very interested to check out the new version and features. But you didn't answer one of my questions about how to get access to a repo with e.g git history? I don't see any way to access such a thing. How should I contribute fixes to snowmix?
Hi On the outlook for somebody with macOS newer than Monterey for testing Snowmix. Please contact me if you have an Intel based Mac and can and would like to test Snowmix. M1, M2etc all based on ARM is to be tested later. Let meknow if you have the hardware and can and would like to test. Best regards Peter Maersk-Moller
Sorry for the very delayed reply. Will look into it. Snowmix 0.5.2 is now released and you can now specify a loaded image ( from a PNG file) as dead image for a feed. Regards Peter MM
Testing Snowmix-0.5.2 on macOS has revealed a bug in the bootstrapd/bootstrap-OSX files. The following part needs changing from this # Until 1.5.0 the package type was universal, after that x86_64 if [ $GSTMAJOR -lt 5 ] ; then PKGTYPE=universal else PKGTYPE=x86_64 fi to this # Until 1.5.0 the package type was universal, after that x86_64 until 1.20.0 PKGTYPE=universal if [ $GSTMAJOR -gt 4 -a -lt 20] ; then PKGTYPE=x86_64 fi Otherwise Snowmix was installed on a cleann ew fresh and updated install...
New Snowmix version 0.5.2 is available for download. This version Snowmix adds a number of new frameworks and new functionality to Snowmix. Some of this is: Asynchronously reading and writing images. Chroma keying (green screening). Features to handle images and feeds with transparency correctly. Embedded Python interpreter Indexed images for animation. Input support for various raw video formats Access to keyboard and mouse events from embedded interpreters Various new features to assist internal...
Snowmix is back with a new release. Have a nice day. peter
Hi Robert. Sorry for being silent for years. There were reasons. Snowmix is back with new release. I know you have moved, but thought you'd like to know. Have a nice day. Peter
Greetings to Peter :-) at least in LinkedIn he (his profile...) made some likes lst week hopefully your life is good :-) I would refuse to say this project is dead! It is a very special piece of software, almost feature complete :-) exellent, rock stable what else we need? developed over much more then 10years, a year of silence here is nothing to worry about - look above Of course if someone has skills to contribute, i'd love this too, but due to the nature of this very special thing i'm hoping...
Hi, I abbadoned SNOWMIX long ago and looked at using GStreamer framework for my project. The core code in snowmix is brilliant. I'd suspect somebody would take what Peter had and break it up into a API. The methods with scripting is convoluted and could much simpler. Rob On Wed, 6 Aug 2025 at 07:04, C J egftechman@users.sourceforge.net wrote: @Robert Are you willing to share what you wrote? I really liked the simple API for Snowmix, but now looking at writing to websocket API on OBS. Where is Peter?...
Before I start to rewrite all my applications to use OBS websockets, checking to see if there are any forks of snowmix that are actively maintained...
@Robert Are you willing to share what you wrote? I really liked the simple API for Snowmix, but now looking at writing to websocket API on OBS.
Hey guys, I was using snowmix for a while however, as I become familiar with Gstreamer, I ended up writing my own AV mixer and ditching snowmix. I had a lot of scripting to setup the ini file and then use the socket for communications once snowmix was up and going to do stuff. I never liked how snowmix used a file descriptors for audio! The shared memory used for video would have been a far easier approach. Anyway, I hope Peter is ok?? On Thu, 4 Jul 2024 at 4:15 AM, C J egftechman@users.sourceforge.net...
It appears to be a dead project at this point... Are there any forks of this project before I look at doing so?
Is this project dead? I use it for our main video production pipeline, and wondering about the future as I start replacing hardware, and wondering if we need to start looking at different software, or start looking at this source more in-depth to see if we can maintain it. Anyone interested in helping manage a fork?
Hello, Thanks for snowmix. It's really great! I'm using it to mix realtime video from multiple security cameras onto a monitor. I find the use of a simple tcp server and nc to send it commands to be super flexible and useful - great design, and your bash scripts and examples have been excellent resources for hacking together my own solution. I wanted to have a custom idle images, but I wasn't able to make The Gimp's "raw export" feature write out the raw images in the correct BGRA byte order, so...
i posted it without login in? My appoligies. sourceforge is rather clunky and doesnt remember me .
How do I get may hands on the tcl code?
Just an update, Getting audio playing still alludes me. I've used both input2feed.sh and av_output2screen.sh scripts to remove any issue I have with my python version of the bash scripts. Checkout attached images. I'm sure I have a simple issue which I'm yet to expand my knowledge around. I found feeds and mixer feeds muted although I have them set to unmuted in my ini script? Question Peter, Getting audio in and out of snowmix seems to be a black art. Video is the easy part.. Is there any appetite...
Hello Peter, I'm struggling getting audio into snowmix. Yes I'm a bit thick. Video is working fine with simple stack for the video feed. I'm using the av_output2screen from snowmix to get audio from mixer 1 I'm using script below to input into snowmix as a camera and audio test source. Video is working, no audio. #!/bin/bash # These settings are from the Snowmix ini file for the video feed with test audio # for feeding into snowmix CONTROL_PIPE=/tmp/towercam-video width=1920 height=1080 framerate='25/1'...
For audio, you can copy the ini file $SNOWMIX/ini/audio and edit it to suit your requirements. Note there is a guide to audio on the website under guides. Don't forget to have the dummy audio sink 0 and to start it to start the mixer it is sourced from. Explained in the audio guide. For how to send audio into Snowmix and extracting audio, see the script scripts/input2feed and scripts/output2screen. If you have problems, just ask here and please post your ini files and scripts so I can replicate whatever...
For audio, you can copy the ini file $SNOWMIX/ini/audio and edit it to suit your requirements. Note there is a guide to audio on the website under guides. Don't forget to have the dummy audio sink 0 and to start it to start the mixer it is sourced from. Explained in the audio guide. For how to send audio into Snowmix and extracting audio, see the script scripts/input2feed and scripts/output2screen. If you have problems, just ask here and please post your ini files and scripts so I can replicate whatever...
Something like this ?
Hello again Peter. Please email if you could. I understand you time is valuable. I happy to pay for your input. vk3eht (at) gmail (dot) com I've been researching the long list in General discussions to find suitable answers to my project. snowmix is very powerful and I have already basic parts working however, I'm using the simple 'stack' command and have started to add audio to switch same time as video for the same feeds. My project has 8 video and 9 audio feeds with 3 sinks including the local...
Hello again Peter. Please email if you could. I understand you time is valuable. I happy to pay for your input. vk3eht <at> gmail <dot> com. I've been researching the long list in General discussions to find suitable answers to my project. snowmix is very powerful and I have already basic parts working however, I'm using the simple 'stack' command and have started to add audio to switch same time as video for the same feeds. My project has 8 video and 9 audio feeds with 3 sinks including the local...
Hello again Peter. Please email if you could. I understand you time is valuable. I happy to pay for your input. vk3eht <at> gmail <dot> com. I've been researching the long list in General discussions to find suitable answers to my project. snowmix is very powerful and I have already basic parts working however, I'm using the simple 'stack' command and have started to add audio to switch same time as video for the same feeds. My project has 8 video and 9 audio feeds with 3 sinks including the local...
Hello again Peter. Please email if you could. I understand you time is valuable. I happy to pay for your input. vk3eht <at> gmail <dot> com. I've been researching the long list in General discussions to find suitable answers to my project. snowmix is very powerful and I have already basic parts working however, I'm using the simple 'stack' command and have started to add audio to switch same time as video for the same feeds. My project has 8 video and 9 audio feed with 3 sinks including the local...
Hello again Peter. Please email if you could. I understand you time is valuable. I happy yo pay for you input. vk3eht <at> gmail <dot> com. I've been researching the long list in General discussions to find suitable answers to my project. snowmix is very powerful and I have already basic parts working however, I'm using the simple 'stack' command and have started to add audio to switch same time as video for the same feeds. My project has 8 video and 9 audio feed with 3 sinks including the local...
Hi Peter, Thanyou for the reply, I'm sure your time is valuable for other tasks. Thanks for the ini file? I have a login for Sourceforge and have been logging in to post on the snowmix discussion forum. Not sure why they continue to be "anonymous" I find Sourceforge clunky and old and not much has changed., Could you look into using github for 5.2 and on which provides a better workflow for discussions and allows others to contribute more easily. Just a thought and maybe you have considered Github?....
Hi Rob. Two empty messages. Not sure what it is. Regards Peter MM
Hi Rob. I'll have a look at it and come back. Regards Peter MM
Hi Rob. When posting anonymously, I have to manually approve the posting before it is shown in discussion, which may take time if I am busy. Furthermore you don't get notification when I answer. That said posting anonymously is allowed :-) Next thing. The ini file posted does not end on a newline or carriage-return,newline or just a newline so the last line is not executed (image place command). I assume it is the editor you are using that is the cause. I cannot recommend Windows Notepad, if that...
Hello, Peter, Its been over a year now.. Is there a version of 5.2? I'd like to further intergrate with Python? I'd like to give 5.2 a spin. 5.1.1 still running fine with GStreamer 1.18 on ubuntu 20.LTS
Hello, I've recently revisited using snowmix and are using python to wrap pipelines in and out of snowmix. Snowmix has all the parts like transitioning and shape wipes without needing to construct another massive gstreamer pipeline to do all that. I'm working through add piece by piece and I can't get any text or overlay showing on any stack? the snowmix demo works, and I've looked at examples, there possibly a thing or two i've overlooked. The text place, text align fails on snowmix which I've commented...
Hello Peter, Do you have any examples in how to implement a horizontal bar graph? I'd like to place an overlay frame with transparent cut-outs to insert some solid moving bar from left to right. Use TCL or python, Python preferred. 3 bars with additional text fields which can be updated dynamically while the overlay is visible and then fade in and out the complete overlay, bars and text on the top stack? Here is example attached. I'm sure it is possible with overlays and shapes possibly. Rob
On Sun, 30 Apr 2023 at 3:53 pm, Robert Hensel vh3eht@users.sourceforge.net wrote: Hi Peter, Not sure if you pressed send prematurely? Yes, only feeds audiomixer, could use a second mixer to do indent only. I guess there is many ways to skin a cat. The indents is played every time the repeater is receives a new feed from 1-4 to indicate repeater ID as per regulation. Hope this is clearer. Rob On Sun, 30 Apr 2023 at 3:44 pm, Robert Hensel vh3eht@users.sourceforge.net wrote: Hello Peter, I understand...
Hi Peter, Not sure if you pressed send prematurely? Yes, only feeds audiomixer, could you second mixer to do indent only. I guess there is many ways to sink a cat. The indents is played every time the repeater is receives a new feed from 1-4 to indicate repeater ID as per regulation. Hope this is clearer. Rob On Sun, 30 Apr 2023 at 3:44 pm, Robert Hensel vh3eht@users.sourceforge.net wrote: Hello Peter, I understand it has been a long time since we last spoke. I occasionally look to see how and when...
Hello Peter, I understand it has been a long time since we last spoke. I occasionally look to see how and when snowmix 0.5.2 is advancing. In my initial email I’ve given you a basic outline on my project (club project) developing a DATV repeater which operates as a tiny TV station all in software which receives DVB-S/S2/T/T2 signals from Amateur operators, decodes their transmitted DVB signals either one of the 4 to raw AV adds overlays and then encodes back to h264 video audio aac back into a mpeg...
Many ways to do it. If you want to execute a single TCL command you will use this: tcl eval rest_of_line_is_some_tcl_command If you want to execute a macro containing several TCL commands including defining functions etc. you use this command: tcl exec some_macro_name_usually_ending_with_dot_tcl If you want to in Snowmix want to execute a Snowmix command from within Snowmixs TCL interpreter, you can use this command. tcl eval snowmix parse some_snowmix_command tcl eval snowmix parses some_snowmix_command...
still dry swimming :-) loading the images would be ~~~ command create loadmyani.tcl for {set i 0} {$i < 251} {incr i} { set loadnr [format "1%03d" $i] set fnam [format "ccla-%04d" $i] set filename ${fnam}.png snowmix parse image load $loadnr path/$filename } command end ~~~ and run with tcl exec loadmyani.tcl Maybe better would be to load one image each x frames from snowmix instead of loading them all at once. Yes, i'd look into the manual (RTFM) and see howto use loop or loop 250, postone with...
still dry swimming :-) loading the images would be ~~~ command create loadmyani.tcl for {set i 0} {$i < 251} {incr i} { set loadnr [format "1%03d" $i] set fnam [format "ccla-%04d" $i] set filename ${fnam}.png snowmix parse image load $loadnr path/$filename } command end ~~~ and run with tcl exec loadmyani.tcl Maybe better would be to load one image each x frames from snowmix instead of loading them all at once. Yes, i'd look into the manual (RTFM) and see howto use loop or loop 250, postone with...
OK, i have to take care if i'm logged in :-) To show the image i need to count up the image id i'm assuming the following will not work, as the tcl for-loop might be not executed as one step per frame: command create myani tcl eval for {set i 0} {$i < 251} {incr i} {set loadnr [format "1%03d" $i] tcl eval snowmix image place 1000 $loadnr 0 0 loop 250 command end mabe something like this will work command create myanipreset tcl eval {set i 0 } command end command create myanishow tcl eval {incr i}...
OK, i have to take care if i'm logged in :-) To show the image i need to count up the image id i'm assuming the following will not work, as the tcl for-loop might be not executed as one step per frame: command create myani tcl eval for {set i 0} {$i < 251} {incr i} {set loadnr [format "1%03d" $i] tcl eval snowmix image place 1000 $loadnr 0 0 loop 250 command end mabe something like this will work command create myanipreset tcl eval {set i 0 } command end command create myanishow tcl eval {incr i}...
Hello, Thanks for response, indexed images sounds charming. I will try. Replacing the greenscreen with transparent in imagemagic is easy. adjust fuzz to get rid of the (anti?)-aliasing blurr around Text and objects - not to much then you lose your content for f in ../images.greenscreen/*.png; do i=$(basename -- $f);convert $f -fuzz 40% -transparent \#00fd00 $i; done i'll swim dry now: I think i should create 2 snowmix commands, one for loading the images (if i want in the middle of our show a different...
See other message for reply.
There are at least two ways to do the animation with logos. First way is to load is logo as image with transparent background, and then create a macro that handles all the scaling, placing, movement and turn them off in the end one by one. Second way is convert your animation into images and load them as an indexed image and have a macro enabling, forwarding the index pointer and and then just overlay the indexed image. Now this solution has a smaller problem because indexed images is not supproted...
Hi Peter, I just use the stack command to switch between the cams, nothing else. (i send "stack 4" with netcat to snowmix to switch the output to cam 4) We have 5 cams in FullHD we switch thrugh as need, some Text's and some images that i am changing/placing with commands via netcat to snowmix. Here i put a sample and a greenscreen video from what i like to do (the blurred background is not need to implement, it is just to see the animation better) https://drive.google.com/drive/folders/1GItdgjwBQhb-MvHGATzQfPXyipcNrW0s?usp=sharing...
ups , i forgot to log in so my last post is Anonymous...
Hi Jens. I am not sure I fully understand what you want to do. You can "stack" mutiple sources on top of each other, but I guess that is not what you want or need. You also define virtual feeds with sources being feeds or images and, then overlay these, but that may not what you want (I'm not sure what that is). Can you share an example of the animation you need (I assume need to overlay a camera feed)? What is the geometry size of the animations and how many frames does it have in total? What kind...
Yes that can work. However you don't need ffmpeg in this example. So what do you want to do since you seem to be wanting to need ffpmeg as video source? Just wandering because you are making it harder than it has to be. The "-stream_loop -1" did work with the older version of ffmpeg I tried it with. Depending on hardware and video stream geometry and framerate, I can recommend gstreamer queue elements inserted between elements using a lot of CPU such as scale and convert and I would remove one convert...
Hi Peter, I'm using my setup to simply switch between several cams and having some Text and pictures on my output. I'm using the simple stack 7 command to switch to cam 7 and so on. See here https://www.youtube.com/watch?v=iPSSVx6r0x8 Now i would like to have a few seconds of a transparent? (green/blue screen?) video with animation where you see the background picture of the active cam behind. All cams (8) are Full HD, the animation as well. I think i have to use virtual feeds? but how to use virtual...
Hi, a bit late, but on my first steps with snowmix i've looped real videos from ffmpeg via stdout to gstreamer stdin. i used this for testing ffmpeg -re -stream_loop -1 -i /path/to/test.mp4 -c:v copy -f matroska - | \ gst-launch-1.0 -q filesrc location=/dev/fd/0 ! \ decodebin name=decoder ! \ videoconvert ! \ videoscale ! \ videoconvert ! \ video/x-raw,format=BGRA,pixel-aspect-ratio=1/1,interlace-mode=progressive, width=1920, height=1080 ! \ shmsink socket-path=/tmp/feed5-control-pipe shm-size=2...
Hi, a bit late, but on my first steps with snowmix i've looped real videos from ffmpeg via stdout to gstreamer stdin. i used this for testin ffmpeg -re -stream_loop -1 -i /path/to/test.mp4 -c:v copy -f matroska - | \ gst-launch-1.0 -q filesrc location=/dev/fd/0 ! \ decodebin name=decoder ! \ videoconvert ! \ videoscale ! \ videoconvert ! \ video/x-raw,format=BGRA,pixel-aspect-ratio=1/1,interlace-mode=progressive, width=1920, height=1080 ! \ shmsink socket-path=/tmp/feed5-control-pipe shm-size=21...
Hi Raul. I don't really understand the pipelines you have posted. The pirpleines contains spaces in places that would make the pipelines not work, if pasted into a file and executed. Does that mean the pipelines are not copy-and-pasted from a working setup, but rtaher handwritten to the best of your memory serves you? If it the latter, copy-and-paste is more reliable to convey exact and precise information. Regards Peter MM
Hi Raoul. Again Snowmix itself uses the same amount of CPU instructions for doing a specific setup no matter the content within the individual streams. In fringe cases, this the same amount of CPU instructions can lead to two different measured load percentages. The fringe cases are when Snowmix for executing the instructions needs to wait more than average for data to load into the CPU cache or to wait longer than average for CPU cache to be written to main memory (RAM). Such fringe cases occur...
Hi Peter, some updates on the tests in progress. In fact the snowmix workload seems to depend heavily on the different gstreamer pipelines. A typical example: removing "n-threads = 4" from videoconvert from all pipelines, everything seems to be much more stable at the workload level, with fewer ups and downs. With 5 1080p30 feeds more or less all the same, for example: /usr/bin/gst-launch-1.0 -v v4l2src device = / dev / video0! video / x-raw, width = 1920, height = 1080, framerate = 30/1! videoconvert!...
Just to be clear, Snowmix load does not depend on content of a video stream. So you can for test just use "input2feed" as simulation for any input video stream. Then when all work logically, you can go back and test with your real streams which may load your system (not Snowmix) so much, that Snowmix eventually gets affected. As an example of the latter. If your gstreamer decoders og encoders ends up using all available memory bandwidth (CPU Cache to Memory, you can't really see it anywhere or with...
Hi Raoul. We need to clear a misunderstanding of Snowmix I believe you have expressed. The CPU load and memory bandwidth load and the memory usage is constant no matter what is on video streams. Snowmix will for every frame do its job mixing frame after frame at a constant framerate reading frames from shared memory, mixing them according to the macro set with "overlay finish" and writing the mixed frame shared memory. What is not constant, is the load decoding the incoming streams will have depending...
Hi Peter. Some more information from the latest tests. My video sources are "real" (taken by PTZ surveillance cameras) and above all, they are in constant motion (camera tour). I noticed that with constantly moving pictures (especially on fast movements) snowmix reaches (dangerously) peaks of 100% in the same period. In the case of more static images (for example by blocking the camera motors) the CPU load returns to quite manageable values (50/60%). This is valid for the feed displayed in the background...
Hi Raoul. I may have a fix, but I need to understand your setup first. I have used your ini file, added the vfeed configuration and run 5 times "input2feed" to feed 101 to feed 105 feeding vfeed 2 through 6. I also run output2screen and my overall CPU usage for all on my 2012 i7 budget PC is little more than 50%. This mean I have not yet executed the change (redefinition) for vfeed 1 yet. Are you having the same result (meaning no problems) or do you already have problem so far? According to what...
Hi Peter! Yes, sorry, my snowmix_test.ini file is a bit messed up. The commands I added before are executed directly in telnet, after loading snowmix with snowmix_test.ini. I confirm that on Ubuntu 22.04 and i7 10700 I also get a 15% cpu load with the tests you asked me to do and there are no particular problems with only one input feed. My problem is more complex: I have 6 (six) gstreamer sources at 1080p30 and my own output2screen script (similar to your original). As long as I keep the feeds scaled...
Hi Raoul. If I take only your snowmix_test.ini file and run it in Snowmix 0.5.1.1 and add the vfeed command you list in your message and then run the two scripts in two windows: $ output2screen $ input2feed 1 test/big_buck_bunny_720p_H264_AAC_25fps_3400K.MP4 Then the load on my test PC an i7-3770 (Old old old stuff) is showing 10-15% and I get a 1920x1080 window with black background and I see a moving bunny in quarter size window in the top left corner of the 1920x1080 window. Now if you execute...
Hi Raoul. I will look into your performance issue. First a couple of questions and remarks. The snowmix_test.ini you provided seems doctored and incomplete. Always nice to distill to a minimum for clarity. Just make sure you have all with it to make your point. Next thing is the i7 desktop CPU. While it is an okay desktop CPU and you can do more single thread stuff per core than most Xeon, this only applies, if you do not saturate CPU<-->Memory capacity. Decoding N number of FullHD streams will eventually...
Hi Peter, after a period of inactivity we started working on snowmix again, investigating better and using all your advices in this and in other forum threads. We have minimized the configurations and settings in order to have a simpler picture of the situation. We understand that what you said in the previous answers happens. The snowmix process (we don't see multiple threads, but a single process) reaches 100% cpu load and at that moment some gstreamer processes start to get out of alignment. We...
Hello Peter, How can i get my hands on Snowmix 0.5.2 ? I like to play around with Python scripting between app and Snowmix.? I see there is a Node.js Javascript wrapper for snowmix. https://github.com/matthew1000/node-snowmix A python one would be great. Many Regards Rob
I am not the ffmpeg expert on ffmpeg interface types, but I think the answer may be no. You can try ask in the gstreamer dev mailing list and also asks gstreamer pythons questions there.
Hello! I was wondering if there is an ffmpeg equivalent to gstreamers shmsink? I am writing an interface to send video streams to snowmix using python but the python gstreamer bindings are esoteric and not well documented. Python ffmpeg bindings are very well documented on the other hand. Thank you in advance :)
Yes, I had found the youtube example a few days ado. It looks neat. Looks more than enough for what I require. Ill draw up an idea and let you know. Regards Rob On Tue, 12 Apr 2022 at 12:48 am, Peter Maersk-Moller pmaersk@users.sourceforge.net wrote: Powerpoint is fine. An example is available here https://www.youtube.com/watch?v=GeZhGMdjLZA I know it is not very exiting, but I made it long ago to test sequences and viewer experience, so it would not made sense to speed it up. You can jump in it....
Powerpoint is fine. An example is available here https://www.youtube.com/watch?v=GeZhGMdjLZA I know it is not very exiting, but I made it long ago to test sequences and viewer experience, so it would not made sense to speed it up. You can jump in it. Anyway the point is that everything animated is made with Snomwix shapes and a script running embedded in Snowmix. An external script is connecting to Snowmix and feeding simulated rocket altitude data and then the embedded script calculates everything...
Hello Peter, It's been awhile since I've again looked into Snowmix. Covid etc.. I'm now back on my DATV repeater project. Forget about the PNG question. I thought images had to be all BGRA however, PNG is supported. Don't know what made me think PNG was in the pipeline for 0.5.2 when plainly the last few versions support PNG files? I have been using gstreamer pipelines in bash scripts also embedding many gst-launch-1.0 pipelines into python scripts for a number of parts of this project development....
I'll put together some ideas with powerpoint and forward to you. Open to suggestions btw, Donation is in order.. Regards Rob On Mon, 11 Apr 2022 at 18:58, Peter Maersk-Moller pmaersk@users.sourceforge.net wrote: It is quite easy. Show me (a single) example of a text and graphics you would like, and I will tell you how it can be done. You can then extrapolate from there. Just draw something with a drawing program or do something in Powerpoint/Draw. Specify colours and transparency as well. Once you...
Hello Peter, It's been awhile since I've again looked into Snowmix. Covid etc.. I'm now back on my DATV repeater project. Forget about the PNG question. I thought images had to be all BGRA however, PNG is supported. Don't know what made me think PNG was in the pipeline for 0.5.2 when plainly the last few versions support PNG files? I have been using gstreamer pipelines in bash scripts also embedding many gst-launch-1.0 pipelines into python scripts for a number of parts of this project development....
It is quite easy. Show me (a single) example of a text and graphics you would like, and I will tell you how it can be done. You can then extrapolate from there. Just draw something with a drawing program or do something in Powerpoint/Draw. Specify colours and transparency as well. Once you get the hang of it, adding graphics and overlaying it is rather easy and you can pretty much draw anything you can imagine. And if we miss a graphic primitive or a drawing technique, we can always add it. But most...
Hi Rob. Yes, yes and yes. What's the issue with PNG? Send me a personal message either via here or via Linkedin etc with your email. Tell me also your expected Linux distro and version so I can verify installation. Regards Peter Maersk-Moller
Hey Peter, Is there any change I can get a link to your prerelease of 0.5.2 to work with PNG files please. In addition, I understand you have been adding python support? is this true? I'd prefer python over TCL. Regards Rob
Hello Peter, What's involved in producing a groovy dynamic OSD (On Screen Display) to show status of a incoming signal? Signal report monitoring for what I'm doing. Sat feeds from DVBlast via Femon and PID reports. Example :- Display SNR, BER , LOCK.. and both VID and AUD PIDS? horizontal bars with a fancy borders box around them. below is a text version. FREQ 1255Mhz SYS DVB-S VID PID 256 AUD PID 257 LOCK [X] SIGN [===========|=======] SNR [==========|========] BER [======|============] SIGN [===========|=======]...
the rtpjitterbuffer element has solved the issue! Thank you so much for your help. I have attached my Audio and Video output script for anyone else that stumbles on this thread.
The thing is, if you take audio samples (on the Raspi) from Snowmix (or directly from audiotestsrc), and then play it out via autoaudiosink, you have a jitterbuffer or playoutbuffer of absolutely zero. This means that if Snowmix or the unnamed unix pipe used is delayed by some ms (just a few or more), you will have a gap. On the other hand, if you encode it and save it into any containerformat or even stream it for decoding and playing, you are all good. This is significant aand important. and I...
VERY large development. I used a separate machine to connect to the Snowmix instance running on the Raspberry Pi using netcat. That audio played back very clearly with only minor clicking. This leads me to believe that the issue is on the Audio Output pipeline. The following is the script I used on the remote device to capture audio output from Snowmix running on the Raspberry Pi. Just some additional information: the remote device is running OS-X 12.0.1 Monterey and using an Intel x86 processor....
VERY large development. I used a separate machine to connect to the Snowmix instance running on the Raspberry Pi using netcat. That audio played back very clearly with only minor clicking. This leads me to believe that the issue is on the Audio Output pipeline. The following is the script I used on the remote device to capture audio output from Snowmix running on the Raspberry Pi. Just some additional information: the remote device is running OS-X 12.0.1 Monterey and using an Intel x86 processor....
My laptop uses an Intel Dual-Core i7 x86-64. I will attempt to reproduce issue on it today. It is much much faster than a Raspberry Pi but perhaps I can overwork the processor to simulate a slower environment. I could also attempt to overclock the Raspberry Pi CPU however it doesn't appear that CPU loading is an issue because the issue persists even when no video mixing is happening. I apologize for replying in the middle of the night sometimes, I am in New York, USA.
I have my Raspberry Pi connected to an HDMI video capture card. The Raspberry Pi's main HDMI output is the sink for video and audio output. When the test source is played like so gst-launch-1.0 audiotestsrc ! autoaudiosink there is a clean tone with no stuttering or gaps. Attached are my snowmix_config.ini and my avoutput2screen.sh script I use to get Snowmix output to autovideosink and autoaudiosink.
My laptop uses an Intel Dual-Core Intel Core i7 x86-64. I will attempt to reproduce issue on it today. It is much much faster than a Raspberry Pi but perhaps I can overwork the processor to simulate a slower environment. I could also attempt to overclock the Raspberry Pi CPU however it doesn't appear that CPU loading is an issue because the issue persists even when no video mixing is happening. I apologize for replying in the middle of the night sometimes, I am in New York, USA.
My laptop uses an Intel Dual-Core Intel Core i7 x86-64. I will attempt to reproduce issue on it today. It is much much faster than a Raspberry Pi but perhaps I can overwork the processor to simulate a slower environment. I could also attempt to overclock the Raspberry Pi CPU however it doesn't appear that CPU loading is an issue because the issue persists even when no video mixing is happening. I apologize for replying in the middle of the night, I am in New York, USA.
For the last clip, the clip with audiotestsrc, how do you play the output from snowmix? Details, scripts and setup. I assume we should hear a continuous uninterrupted tone without distortion (assuming no clipping) and without stuttering and no micro gaps. Is that correct?
Thanks. Will look at it tonight. I wonder if you can reproduce the issue on an x86 architecture. Maybe on an older slow laptop to not give to much advance to x86. I have a couple of times been using Raspi for video mixing and often ended up having to replace it with x86 for sustained reliability . It worked with Raspi ... most of the time, but not all the time ... which is frustrating.
I updated the Snowmix config to put a delay of 100ms on feed 1 like so: audio feed delay 1 100 Improved the situation slightly but issue still persists. Attached is a video of Snowmix being fed audiotestsrc is-live=true do-timestamp=true as well as Audio and Video from the video capture card. I had to record these with my phone so that I did not introduce any further CPU load to the device with a screen record. I also included screenshots of the audio delay from snowaudio.tcl. The feed and mixer...