How long and how much does it take to develop this kind of a program and a system?
Hello.
I need some programs. But you know that I am not a programmer. I have given up even displaying Hello World, and fixing the eternal bugs of GIMP by myself.
I would like to hire programmers.
I need this kind of a program and a system for the moment.
- Can broadcast live TV's screen to the internet. And people can watch the screen with their devices.
- It is practically assumed that people use mainly tablets or smart-phones to watch the live video, and do not have particular interest in free software.
- People can watch it free of charge and without subscription or such kind of troublesome things, ideally.
- The live stream must be very very stable. Ideally, even million people can watch it without stress.
- I assume that a specific App would not be necessary (from App Store, Google play to watch the live stream). But App form might be better than visiting the broadcasting website for some reason. (e.g. for the spread of GPL, or for including VPN as a function, or to watch the video it needs GnuPG etc)
Something like this.
Would you tell me the market price of hiring ethical developers? and how long and how much it is taken to develop the system do you estimate at?
Thank you.
PS: Of course if you could provide the program and the system, I would appreciate it very much. Let's talk about it with email or something.
name at domain
GPG Finger Print: 342B 4471 8075 C50B D250 62CE 1ACE 9A10 8057 E54F
GPG Public Key: https://gunchari.net/links/publickey.html
Jami ID: 719f9f85f735440540b6d920cdc57eaad6831d3c (desktop)
Of course I would be very happy if you provide them free of charge eternally because of your something I do not know well about it. Yeah very happy.
Contribute to PeerTube: https://joinpeertube.org/roadmap
Step 4 to have version 3 is about live streaming. To have a million people watch simultaneously, you want a Peer2Peer distribution (as in PeerTube).
I have serious difficulty in reading the writing of them because of the big popup. How to close that?
I merely took a look at some of them but 30 seconds to 1 minute lag is not practical. I would not like even 3 seconds lag. Lag spoils my purpose very much. At most 1 second. If we use it in LAN, can it achieve 1 second lag? It would be hard to say that 1 minute lag is live streaming. Or alternative?
Edit: Typo
As far as I understand, a single server for your stream would require a bandwidth that is proportional to the number of viewers. If you have a VDSL2 connection, with a maximal upstream rate of 100 Mb/s, and if you stream is 1 Mb/s (that is not HD!), you can only reach a maximum of 100 concurrent viewers. For thousands of concurrent viewers (fewer if you want HD), you need either the infrastructure of a GAFA (what most streamers use; they suffer censorship; their users suffer tracking; etc.) or peer to peer (and PeerTube seems to be the most promising solution; you can host your own instance or use an existing one).
I don't understand well. How to share the live video in LAN (assume 10000-80000 people) with PeerTube less than 1 second lag? I use it for an event which is held in a soccer stadium. Can I host my own PeerTube instance just for the LAN? How to connect all of the people's devices who are in the stadium with the instance? with WIFI? or Bluetooth? Wired? So from the beginning, can I use PeerTube for LAN anyway? I would ask the dev of PeerTube directly later though.
For outside the LAN, 30 seconds - 1 minute lag is acceptable. No problem.
Within local networks, users enjoy higher bit rates. But 10k+ users simultaneously watching a video stream looks like a lot. I have no idea what is the network infrastructure of the stadium. Those administrating it should be the ones you talk to.
Why not using the large screens the stadium certainly has?
> Within local networks, users enjoy higher bit rates.
OK, that's good news.
> But 10k+ users simultaneously watching a video stream looks like a lot.
> I have no idea what is the network infrastructure of the stadium. Those administrating it should be the ones you talk to.
? So that it depends on the network infrastructure of each stadium? I also have no idea. It is likely they have at least WIFI. OK, it seems that I have to confirm that first anyway. I think that they have not assumed such a situation. So that must be a weak infrastructure for the purpose. If so, what can I do? Preparing and bringing several servers? Ugh.
> Why not using the large screens the stadium certainly has?
My friend also said that but I have seen those screens in some stadiums and it did not seem useful. I did not see the screen almost at all because those were too small from my seats. Besides, people would not tolerate to keep watching the screen for 2 hours. It would differ from watching the pitch. And I would like to stream sound too. If there are 30 inch displays in front of the every seat in the stadium, it would be the best. And a retractable roof over the seats. But there is no such a stadium in Japan. Germany seems to have some good stadiums though. So I think that using customers's own tablets would be the best. Or other good idea?
+10K Pyras and ad hoc.
So that it depends on the network infrastructure of each stadium?
It does. It is apparently called "High-Density Wi-Fi". For instance, https://www.scalablewifi.com/datasheets/Xirrus_SolutionBrief_HighDensity.pdf promises "Reliable wireless service for over 45,000 onsite attendees" but I doubt it holds if all of them concurrently watch a video stream.
It may be much simpler and more reliable to lend more large screens and to settle them all around the playing field.
> but I doubt it holds if all of them concurrently watch a video stream.
I see. We would have to test, if they have that High-Density Wi-Fi.
> It may be much simpler and more reliable to lend more large screens and to settle them all around the playing field.
I'm a bit not sure about your exact imagination but I would not think so. But it depends on customers's preference. They might prefer it. I'll think about it. Anyway I have to ask stadiums for the question first. It was a quite more difficult work than I thought. Thank you.
I am a member!
> I have serious difficulty in reading the writing of them because of the big popup. How to close that?
In Abrowser, you can switch off style by pressing Alt+V (for View), then Y (for Page Style), then N (for No Style). Of course, the letters may be different in other languages, but I assume they are the same in Japanese as in English.
(If you know of a way to make "no style" the default, please let me know; I don't think about:config provides one.)
Ctrl+Alt+R or a click on the sheet icon on the right of the URL provides a cleaner result.
I am a member!
Oh right, thanks, that's very helpful as always, I forgot about "reader view" because I never use it, because I prefer lines of text to run from edge to edge, but "reader view" does have its advantages over "no style".
I am a member!
Apparently even "reader view" does stretch from edge to edge when the zoom level is high enough. Nice.
It disappeared. Thanks.
I am a member!
PSA:
Sorry to keep derailing MS's thread, but Mozilla disappoints yet again: I just found out that it looks like switching to "reader view" causes the Firefox-based browser to disobey the UBO extension's settings, at least by loading media that should stay blocked. (After switching "reader view" off again, the media were still rendered, despite UBO's media symbol being crossed out.) Now I wonder if "reader view" makes the browser disobey everything every extension does, causing it to execute potentially harmful scripts as well. And I don't know whether to blame Mozilla or UBO.
> Sorry to keep derailing MS's thread,
No problem at all. It is far better than no reply. But I do not understand well what you are talking about though. Anyway it's OK. Forget it.
> I forgot about "reader view" because I never use it, because I prefer lines of text to run from edge to edge
The purpose of Reader View is to make articles easier to read. Text is most readable when there are between around 55-75 characters per line.
> I just found out that it looks like switching to "reader view" causes the Firefox-based browser to disobey the UBO extension's settings, at least by loading media that should stay blocked.
> I don't know whether to blame Mozilla or UBO.
I would first file a bug with uBlock Origin. If after some investigation it turns out that this can only be fixed on Mozilla's end, then file a bug against Firefox.
I am a member!
>I would first file a bug with uBlock Origin.
Thanks, I might try that later.
And just now I noticed that for some webpages, my outdated browser version doesn't offer the "reader view" option.
>Text is most readable when there are between around 55-75 characters per line.
But without auto-hyphenation, such narrow columns cause a higher occurence of either undesirable big differences in line end position (if text is left-aligned, like "reader view"), or (if justified) undesirable large spaces between words. This is more noticeable and hence more annoying in languages with a higher occurence of long "words" (in the typesetting sense), such as German. Anyway, it's good that we have the different options.
Secondly, when I wrote that "I prefer lines of text to run from edge to edge", I mistakenly assumed "reader view" to not adjust the column's width for increased font size (or rather, higher zoom level), which would result in columns looking very narrow and quite annoying to read.
[EDIT: When compared to "no style", another disadvantage of "reader view" is that you only get one column even if you have a wide screen that could easily accomodate more. Having a single column does not make full use of screen real estate unless zoomed in. This may be easier to read, but it also means you reach the lower edge of the screen sooner, requiring more frequent presses of space or whatever you use to scroll down. This is bothersome when you have food on your hands. The government should do something about this.]
I wonder what the optimal line length is in Chinese script with its absence of ascenders, descenders and spaces (except for the white space above other punctuation like 。), hence less for the eye to use as a waymark.
What do you think about TriCasters? For example, A TryCaster Mini HDMI Advanced R2 rents for ￥22,000- per a day. The highest grade rents for ￥66,000- per a day. Is that a necessary? or useful? or needless? Priceless? can use with PeerTube?
I just got to know the existence of such a device. A stadium might have this kind of things, though, how to handle freedom issues with those devices? can isolate it in the LAN? Would anyone know something about this kind of devices?
Can a Trisquel system on a X200 work as a broadcasting system in a football stadium?
Again, I am not an expert at all when it comes to networks. Anyway, as far as I understand, your computer will send the stream to a router that will broadcast it to wireless access points. The public will connect to them to download the stream. The problem should be with the downstream bitrate of the network as a whole. For tens of thousands of people to watch the video stream at the same time, there must be a need for many access points not interferencing with each other. Your computer uploading the stream should not be a problem.
If you manage to live record the video, I see no reason why there would be a problem uploading it live. Most streamers use OBS Studio. It is in Trisquel 9's repository and much documentation is available on the Web: https://obsproject.com/wiki/
Wow we have useful free software. Good good. And it seems that X200 is enough capable of sending the stream. But it sounds a bit strange because, given that there are ten thousand audiences in the stadium. Maybe the X200 would send the same stream to each device of the audiences. There must be ten thousand destinations. Is X200 really such capable? The X200 must send the stream simultaneously. It must be time lag even if each time lag is under 1/1000 sec. I mean, I cannot imagine this poor X200 is capable of doing such a hard task. Throughput.
And about the access points, I just a bit understood why you said that it depends on the internet infrastructure of each stadium. The Levis Stadium has 1200 access points and The U.S. Bank stadium has 1300 Wi-Fi access points. Those would be the latest gorgeous systems. The latter stadium spent $60,000,000 for the technology of the stadium. The tech team of the U.S. Bank stadium says that theoretically, all 66,000 attendance can use Wi-Fi. But as you doubt I don't know if the attendance in the U.S Bank stadium can watch the live stream of a match in the stadium without lag or other network troubles with the latest infrastructure. Did you get that?
I am confident that the stadium which I am going to use as the grounds has not such a latest infrastructure. That is a relatively quite small stadium. I have not received their reply yet though.
But renting and bringing thousand access points seems not practical. Or would it be possible?
Is there a way to construct such a reliable network? You said that:
> It may be much simpler and more reliable to lend more large screens and to settle them all around the playing field.
What do you mean by "large screens" and "them"? and where is the playing field you say? the green pitch, eh? How big the large screens are? So you mean that on the seats of the stadium, there are no people? All of them are on the green pitch?
If so, I should say I think that it is important not to make such difference from a normal football match day watching. People sit on the seats and watch the game and cheer from the seat. I think it must make the same sound effects by their cheer. I would love it. There are other important reasons, too. Look at this picture of Camp Nou. Would not it explain everything, eh? Is not there a way to construct a good live streaming system...? Seems impossible? Surely, even the U.S. Bank stadium might not be capable of sending the live stream of the match to 66,000 people's devices without network troubles such as time lag or interference or something... if so, sadly maybe I have to let them sit on the green pitch... ****! But if the local broadcasting system specializes in sending only the live stream, possibly would it be possible? No?
> What do you mean by "large screens" and "them"?
I think the giant size sceens would be facing the public watching from the stands. You would replace the pitch by a huge wall of screens. You could also play BDSM documentaries in 3D to a huge cheering audience during the breaks. Some extra people might join specially to enjoy the breaks.
People would still look at the pitch, but would see a huge screen instead of the pitch itself. That way, no need to simulaneously live stream for 80k people through wifi. You feed the screens and the screens feed the eyes of the audience. All you are doing is expanding the concept of big screen.
> You would replace the pitch by a huge wall of screens.
Wow that's great. But how?
> You could also play BDSM documentaries in 3D to a huge cheering audience during the breaks.
We should talk about it later.
> how?
Jumbotrons.
That is not what I expected... disappointing. I expected that I would be able to use the pitch itself as a big screen with 3D hologram movie system or something... But I might have to use this screen by compromise. I might ask if I can borrow it from them. ****!
Maybe the X200 would send the same stream to each device of the audiences.
Not directly. Again: the required bandwidth proportionally grows with the number of spectators. The stream would be duplicated until reaching the Wi-Fi access points.
The tech team of the U.S. Bank stadium says that theoretically, all 66,000 attendance can use Wi-Fi.
The problem is: what do they mean by "use"? 1300 access points for 66,000 users means more than 50 users per access point. Can 50 people simultaneously connected to a same access point all watch HD videos? With IEEE 802.11n (or better) and its 600 Mbits/s, that looks possible. But, again, I am no expert in networking.
But renting and bringing thousand access points seems not practical.
Not only the access points but also the rest of the infrastructure, to duplicate the stream up to the access points.
What do you mean by "large screens" and "them"?
I am thinking of video walls commonly seen in concerts organized in stadiums: https://en.wikipedia.org/wiki/Video_wall
where is the playing field you say? the green pitch, eh?
Yes. I believe pitch is UK English and field is US English.
How big the large screens are?
Maybe 4 meters by 3 meters. The stadium probably knows companies renting such equipment, if they ever organized concerts.
So you mean that on the seats of the stadium, there are no people? All of them are on the green pitch?
I was thinking of video walls oriented toward the seats. Nevertheless, the reverse may be better for viewing quality (but worse for the maximal possible audience). Or a mix of the two, with both orientations and only using seats close to the field?
OK, let me organize my thoughts please.
There are two ways to provide the live stream to the customers in the stadium.
1. is solved by constructing the stable wireless LAN. It is not impossible, but very difficult. And a lot of troubles would be expecting.
2. is solved by big screens. My concern was that those jumbotrons were too small to watch the live stream, especially if you are on the farthest seats of the stadium even if how the screens were big. But if we have the AT&T stadium's jumbotrons in the stadium, it seems sufficient to solve the problem. Because it is on the center of the stadium and very big (around 48 meter width) But probably we have not such a jumbotron at least in Tokyo, Kanagawa even in Japan. Those big jumbotrons in our area are all not practical for the purpose.
Or a mix of both 1. and 2. solutions.
About 1., I started to think that this is not practical. Human beings's technology is too poor to meet my needs. The aliens must manage such a job easily.
But how about if we have +10 thousands Pyras and connected all of Pyras ad hoc? Or does anyone have a better idea, s'il vous plait?
About 2., I started to think that this might be better than 1. if we rent and install big balloon screens on the athletic field. Surely it would be more simple and reliable. I found that there are some companies which rent big balloon screens (the biggest one is 500 inch, as far as I studied. One of the companies seems to be able to make a bigger balloon screen though). Those screens and projectors are used for mainly drive-in theaters. But even 500 inch does not seem useful because again I think that the audience on the top seats would not be able to watch the projector clearly, from my actual experience. That stadium was a soccer specific stadium, the stands are very steep, but a player seemed very small.
But I just thought that 2. would be better for the first event because I am not going to borrow a so much big stadium... maybe even from the top seats, 500 inch might do the job. I am not sure though... It is hard to imagine.
Something called digital projectors would do the job. If I think next, establishing the stable LAN solution will be useful though...
By the way, there are the digital projectors. Can we use OBS for it? Ask them by myself?
Anyway, so
> Maybe 4 meters by 3 meters.
Too small.
> Nevertheless, the reverse may be better for viewing quality (but worse for the maximal possible audience). Or a mix of the two, with both orientations and only using seats close to the field?
I think it spoils the beauty of Camp Nou. From the beginning, how many video walls do I have to prepare? It should be at least a few hundreds I suppose.
Anyway I am going to go with 2. with several big balloon screens for the present. Thanks thanks.
If you got a good idea, please inform me.
I had never heard of such a device. It certainly comes with proprietary software.
No doubt about it. It is OK if a X200 and OBS Studio do the job.
So for the present, that is like this picture.
Please replace the people around the center champions league's ball flag who hold the round flag with the players of the game event.
The white objects are the balloon screens. If it has 30 metre width, it seems sufficient. The projectors of them can install behind the screens and project from there.
I got a estimate for the balloon screens rental.
4 screens rental for 1 day and for a test = around $40,000.
Can I somehow construct the stable Wi-Fi system under $40,000 please?
The estimate does not include projectors rental. Only balloons!
Surely, those are 300inch balloons, width 8m × height 6m. Very big balloons. But balloons are balloons. And it is rental. It is not that I buy them.
At the same time I would like to ask you about a slightly different application.
It is an application for all operating systems (iOS, MacOS, Andoriod, Windows, GNU/Liunx, etc.) whose purpose is to collect all users' voices in the application and consolidate them.
For example, it uses a microphone to collect the voices of all users watching a soccer game, and then the voices are combined and delivered as sound to all users. It's as if all the users watching an unattended game on the Internet are cheering for the game as if they were in the stadium, and the collected cheers are distributed in a batch.
In this case, the time lag is probably not so much of a problem if it's only a few seconds, but if it's 10 seconds, it's a bit of a problem. One minute is out of the question.
But the problem is still time lag. Assuming that there are people all over the world using the application, and each user has a very stable Internet connection, such as 4G, what is the expected lag time?
Can you also deliver the video and audio separately?
Or rather, is there any free software that allows you to batch the voice of all users and make a big cheer?
Assume that there are more than 10,000 to 100,000 users who also use it at the same time.
Ask NSA?
I believe both the downstream bitrate of your connection and the CPU horse power to merge the many sounds would be problematic: you would need an infrastructure that would gradually merge not so many sounds, themselves to be merged together.
What exactly do I need to do that?
With one single server, the download bitstream rate of the Internet connection would have to be greater than the sum of every recorded streams. Even with a constant bitrate of 16 kbit/s for every stream, multiplying it by 10,000 such streams gives 160 Mbit/s. And I am not sure it is even possible to simultaneously receive from 10,000 different IP addresses.
The CPU issue relates to mixing many sound streams in real time. Mixing consoles, which are dedicated to that tasks, only handle a few dozens of inputs, I believe. Not tens of thousands. So you would need hundreds, if not thousands of computers, each mixing a few dozens streams. Either an original stream or the streams output by computers processing original streams, before reaching the last computer doing the last mix.
Keeping all that synchronized and with little lag looks hard. But maybe possible.
But again, I am not expert with either networking or sound processing: maybe I miss something.
Is that something that a rich big European club can do when it wants to let its players hear the cheers of its supporters in the stadium in an unattended match? Or definitely impossible?
edit
> With one single server, the download bitstream rate of the Internet connection would have to be greater than the sum of every recorded streams. Even with a constant bitrate of 16 kbit/s for every stream, multiplying it by 10,000 such streams gives 160 Mbit/s.
I think that it is the opposite. Does the constant bitrate of 16kbits/s mean the bitrate of one recorded cheer?
If so, it does not have to be so much high quality sound, I suppose. Perhaps very low quality should be OK, or rather, very simplified sound should be OK. The noisy sound is bad though. Simplified, I suppose.
About the mixed sound too, it should not have to be so much high quality. But the noisy sound is bad. I feel if it is the simplified, clear, mixed sound, it would not have to be so much high bitrate. Of course I am merely a bit experienced user, I have no idea how the 16 kbit/s or 16 Mbit/s or 16Gbit/s sound quality sound though.
> And I am not sure it is even possible to simultaneously receive from 10,000 different IP addresses.
How do I test it? Do I need several servers?
> The CPU issue relates to mixing many sound streams in real time. Mixing consoles, which are dedicated to such a tasks, only handle a few dozens of inputs, I believe. Not tens of thousands.
I think I understand what you mean.
> So you would need hundreds, if not thousands of computers, each mixing a few dozens streams.
At the same place? or geographically separated computers can do the job? And what kind of computers do I have to prepare? Do hundreds X60s do the job?
Or do I need more powerful computers such as TalosⅡ？
> Either an original stream or the streams output by computers processing original streams, before reaching the last computer doing the last mix.
I think I understand what you mean.
> Keeping all that synchronized and with little lag looks hard. But maybe possible.
I definitely should test it. I seem to have to prepare computers first. Which the best software is for mixing? Then install it to all the computers... find 1000-10000 people first for the test. I need the application too which collects the voice. Maybe here forum's people would be cooperative to install the application and for the test, I guess. And what kind of a server should I prepare? Would a X60 do the job? or TalosⅡ?
I feel we are talking about only sound, but I need to send the live streaming video too to their devices. But I guess that sound and video do not have to be integrated with one application. Or is it better to be integrated?
Does the constant bitrate of 16kbits/s mean the bitrate of one recorded cheer?
Yes.
At the same place? or geographically separated computers can do the job?
The more network equipment to traverse, the greater the lag.
And what kind of computers do I have to prepare? Do hundreds X60s do the job?
I guess you could try to play more and more sounds at the same time (and save the output mix to the disk) and see at what point the systems fails to do so in real-time. The limiting hardware may actually be the sound card rather than the CPU.
I seem to have to prepare computers first. Which the best software is for mixing?
For low latency, composers use the JACK Audio Connection Kit on top of a low-latency kernel. At least that is what I understood. I have no experience whatsoever with computer music.
I feel we are talking about only sound, but I need to send the live streaming video too to their devices.
That goes in the other direction, right? It is what we discussed earlier.
OK, so what I have to do are:
1. Prepare computers. It seems that naturally it would be better to have better computers, with professional sound cards, for reducing lag. I am going to ask some hardware manufacturers. But again and again, I have no money.
2. Prepare servers. A ditto. But I have no money.
3. Ask JACK Audio Connection Kit developers whether they can make a specific version or add-on or something to reduce lag. I have no money though.
4. Ask Trisquel devs whether, I have no money though, they can make a specific low-latency kernerl something like Ubuntu Studio has. Or Libre-linux has the kernel? Libre-linux has the kernel?
5. Test it. But how? really. I am not such a big shot, have no money. I have not so many henchmen. Maybe I have to prepare some pawns first.
Anyway thank you. I got what I have to do, have no money though. How? Consider Masaru, how?
> That goes in the other direction, right? It is what we discussed earlier.
? It was about in LAN, right?
I ought to quit.
Can several TL2WK2s with 22-core POWER9 v2 CPUs
https://secure.raptorcs.com/content/TL2WK2/intro.html
and one TL2SV1
https://secure.raptorcs.com/content/TL2SV1/intro.html
handle the task? There does not seem to be mention of the professional sound cards.
Or like this? It seems to be Youtube's servers. I wonder if every single one of those all in the flat spaces are all servers.
Still the question is how long and how much does it take to develop this kind of a system.
I have to estimate the cost. But before that, I need to know if it is possible or not, though. Maybe possible, since there are such games. So given that it is probably possible, the problem is time lag. Even if it is estimated even more than 1 minute, I think that we will be able to improve it.
So even if it is not like Youtube's servers, I have to estimate the least reasonably practically expected cost to raise the funds. Would someone be able to estimate the cost to construct such a infrastructure even if vaguely?
IP multicast...