Personal View site logo
How do you deliver content? The bandwidth topic.
  • This is a continuation of a discussion that started in the Final Cut X topic: http://personal-view.com/talks/discussion/comment/57810#Comment_57810

    The subject of tapeless deliverables was brought up, which was very timely for me, since I was in the middle of a job that was going crazy with uploading deliverables. Curious to know, how are you delivering final masters to your clients? Is it still tape? Do you see a decline in tape and an upward trend in digital delivery? How do you bill for upload time?

    This is a topic for us to discuss the future of delivery in terms of bandwidth. Ideas, recommendations on best practices in the work environment. I'm very curious to know what your personal standards or your company's standards are for delivering digital content. Where do you see the need for bandwidth in small business and homes over the next few years - and will it be affordable? How do you organize the vast swath of files and content that's created?

    Let's keep the NLE debate out of this conversation and focus on the realities of our day-to-day needs in delivery. Please post in the Final Cut X topic if you've got "beef" (and you know I do!).

  • 22 Replies sorted by
  • Rather more on documentation, work practices and storage, I suggest Filemaker Pro, which will store all your individual video files within a container field - or references to a video file on another drive. Its size is nominally limited to 8TB but really only limited to disk size.

    Each file can have a number of fields to describe, say, the effects you have applied to it. A quick time file can be played from the database by clicking; other filetypes must be set to open with default applications.

    A Filemaker database of this kind is quickly set up. But most importantly, it's completely customisable - unlike turnkey solutions which might come with an editing suite.

    Watch an interesting historical overview of FMP in moviemaking given by Academy award-winning Walter Murch at http://www.filemaker.com/downloads/quicktime/customer_murch_interview_high.mov

    A more detailed demonstration of Murch's own database is at http://www.filemaker.com/downloads/quicktime/customer_murch_demo.mov

  • @5thwall - it seems that this problem will grow exponentially with RAW data, such as in the new Black Magic Cinema camera. I don't see how we can avoid physical media. When we start talking about petabytes of data, even film begins to look attractive as a storage medium.

    @Roberto - I am an experienced Filemaker developer, and yes, it is almost infinitely adaptable to your workflow, and with Filemaker Advanced, you can create a runtime database for one Mac or Windows machine that clients can view without having Filemaker. However, the keywords in FCPX go a long ways toward solving the problem of referencing metadata in scenes. For other NLE's, Filemaker is a good way to go, but it is very easy to get trapped into adding more neat functionality, simply because you can. When that happens, Filemaker can take over and suck up all your time. Murch's procedures go way beyond keeping track of what you have - his solution actually becomes a tool for managing costs and resources, a valuable thing, no doubt, but not exactly where I, at least, want to go as a filmmaker.

  • @CurtisMack Hi! I also develop systems in Filemaker - and I agree with your pragmatic approach. As always, the system should save time, error, omissions and clutter - so as to bring the user to a more comfortable and creative relationship with the job. As you will have noticed, the very people whose workflow is so disorganised that they fail to function to their potential are also often those who fear a system would take away their spontaneity.

  • @CurtisMac @Roberto Do you think DLT tape backup is the solution for long-term storage? It's something I've considered, but ultimate, I'm able to store all my historical projects (final quicktimes, project file, reference graphics) I need on a couple of hard drives.

  • My background is in enterprise IT support, not delivering content to broadcast clients (at least not since the quad days), but my strong feeling is that tape is a dead end. One read error and things go bad quickly. To store video on DLT, you will need a tape library that holds several DLT drives on a fiber network, and expensive proprietary backup software. You're looking at $20K or more. I have managed several ugly incidents caused by storage array or tape library failure - this is a place no individual practitioner or small shop should go. For client delivery, some form of SSD seems ideal, since it is relatively inexpensive, and its inherent degradation over time would not matter, and it eliminates the expense of high bandwidth. The only sound method for archiving digital material is to imitate DNA - turn it over as the media decay. If we start processing RAW video, that means a storage array for post, and smaller arrays for archiving, which must be redone every few years as the media degrade. That's why I mentioned film. If you plan to make features, black and white separation negatives would be the ideal archival choice if cost were not an issue. Under ideal conditions, properly processed black and white will last for centuries. For those of us in the real world, backing up to inexpensive disk drives and moving the data to new disks as the old ones fail is a more realizable choice, but it does require paying attention.

  • @CurtisMack @5thwall

    Ditto! I'd still add some sort of commonplace (ASCII) cataloguing system. (If not for yourself, for your next of kin).

  • @CurtisMack Thanks for the info on DLT. I get a lot of sales guys calling, saying that's the only true archival method. I know the high-end post facilities use DLT as it works in their price-range and it gives them added security. But at my level, seems like it's not a necessity.

  • So, to take it back to the bandwidth question, it seems to me that that small companies who have access to the fastest bandwidth will have an edge - at least in the short term. Does anyone know of a way to get reasonably-priced access to fast upload (over 1Mb/s) and download speeds?

    In L.A., my current options are T1 services (too expensive), Cable (prohibitively expensive for fast speed and comes with long-term contracts), Uverse (reasonably priced, up to 3mbps upload, but don't know how reliable service is and service changes based on your zip code), and Clear for wireless 4G. Maybe I'm overlooking something?

    I use Clear, realizing that I'm very lucky that I get a strong signal and haven't had problems, yet - it's been 1 year. That said, I am worried about the company's long-term growth. But also excited by it's expansion into LTE. With Clear, I'm leveraging low price and relatively good service with the risk of the service going down due to a number of reasons.

    Years ago, if my internet went down, clients understood. But now, it's such a crucial part of doing business. I just don't want to spend over 300/mo for the "privilege" ;).

    Thanks!

  • @5thwall I can't believe the options and prices you get in L.A. In northern CA (specifically SAC) you can easily get 25Mbps fiber for some $62/month and 50Mbps for a few dollars more. Speed are exactly the same for up/down. Any other provider is just miles behind.

    Too bad the company only serves the Sac area and Kansas City

  • @pdlumina Just got off the phone with Time Warner Business Class Cable. It was outright highway robbery. 250/month for 2Mb up/2Mb down with a THREE-YEAR contract! Maybe I'm missing something, but I was actually offended by that. :P

    What company are you using in NoCal?

  • I live in eastern Canada, around here we can get 30mbps down and up for 70$ a month. 70 down, 30 up for 100$. 170 down / 30 up for like 200$.

    No monthly bandwidth limits here either.

  • I found that, for my neighborhood, it looks like the best up/down for the price is Uverse. FIOS is the best at 25 up and down for 75 or 80/month, but AT&T has a hold on my area. It takes me back to the days of Ma Bell. I guess it's really the same thing.

  • In keeping with the title of this thread "How do you deliver content?" it's becoming extremely relevant to ask what format as well.

    For example, some Blu-Ray and DVD players will play your 60p wedding video, others won't. Someone said this week,"Don't go buying shares on a DVD rental business." Then, 24P is strangely a safe investment, because there'll always be some sort of way to play all the existing films.

    TV stations still have spec sheets determining what formats and media they'll accept for broadcast, so no problems there.

    But weddings? Short films? Right now is a hard time to try to future-proof our content. Whatever we think in quality/ease of delivery terms, the market will decide. Sadly, Joe Public will probably go for the fastest, cheapest download (if that's the way we're headed).

  • I have sent well-packed 1TB SATA USB drive, together with a return sticker (so they can send it back when they receive the stuff). Works like a charm and you can't beat bandwidth/price ratio ;) Small stuff can easily be sent over FTP, reasonably compressed to MPEG-4.

    Archiving? That's another story...

  • @Grunf What format are you sending your video data in? [ Edit: besides Mp4 as final product] I'm not talking about archiving myself, but what clients will be able to view for a reasonable time into the future. Has anyone ever asked for a 60p wedding video?

  • Regularly send and receive 1-2 minutes of footage over FTP. DNxHD takes the least space keeping the highest quality since material gets edited/archived. Is not ideal, but appears to be the easiest way so far. I've also received material from dropbox accounts.

    Prores takes too much space, followed by DVCProHD. Still researching a better compression method/smaller file size that does not discard as much information as H264 for online transfer.

  • I deliver content as well as exchange it w/ colorists and editors using the time tested formula: Fedex + Harddrive. Primary exchange formats for me are RED RAW and Cineform Film Scan1. Nice thing is that the colorist can usually just email the metadata back.

  • @5thWall For my personal account with Charter cable in Pasadena, I pay a little under $100 a month for a "100 mpbs" connection. Here is what I got at the moment via SpeedTest.net off the Wi-Fi (note that the Wi-Fi itself is limiting to 52 mbps).

    40.75 mbps download 5.19 mbps upload

    Coming in via Ethernet, I often get download speeds in excess of 60 mpbs.

    As far as wireless our options in the U.S. are quite limited compared to back in Sweden. Last year 3G in Sweden often outperformed 4G in the U.S., while 4G in Sweden had maximum sustained transfer rates higher than many cable solutions here. But it was also rather expensive. I haven't kept up on it.

    As far as deliverables, it depends on what the client asks for but my preference is Cineform in Higher, Film Scan 1 or Film Scan 2.

  • I believe the future is with broadband as so far as delivering content. And it will be rather via cloud-based storage.

    In the US, the fastest "pipe" depends on where you are (and where your customers live), so it could be Fios, U-verse, docsys cable or wireless 3G/4G. The US is not an easy scenario for deployment because we are a geographically diverse nation with great distances. Having originally launched broadband in Korea, Japan and elsewhere, I say that if you live in metropolis, it is a lot easier.

    Unfortunately, pricing of very high speed broadband in the US is among the world's highest. I used to laugh, that given the prices I am selling at (sometimes, just a bit more than 1 trip to a McDo meal), my customers are always looking for a 6-month positive business case for their broadband services.

    It is my view that 3G/4G and U-verse or Fios are reliable services. They mostly share the same backbone network.

  • @thepalalias wish i could get the same speeds you are seeing. a 5 fold increase for my uploads would greatly increase my productivity. have researched every company and nothing in my neighborhood in my price range for around $100. so, have to wait to see what's next - hoping on 4G LTE. thinking about getting a second 4g modem so I can have separate systems uploading.

    @ghkqn probably in about 10 years, i have a feeling we won't be using physical drives much at all. everything we do will be on some cloud storage server somewhere and the connection speeds will be good "enough." perhaps highly specialized work will still be done on local drives. point is, it will "feel" like the drives are local. good thing about it is that our files will be accessible anywhere. start editing a project on set, work at home, finish at the office without having to carry drives around. but yes, most likely in the metropolis only for that type of work.

  • @5thwall The good news is that speeds have improved rapidly. The datarates I get on good days now are literally about 1,000x times faster than when I started high school (remember searching for an ISP with "56kbps" support in one of the two incompatible formats?)

    Like you said, only a matter of time before the drive "feels" local. Of course in the meantime, for smaller files we can use things like DropBox to eliminate "the upload step" as things automatically start to upload once they are saved. Still too expensive to use it for all the files (at least for many of us at this point) but it is a good step in the right direction.

    Sorry if that is less than succinct - I am a little tired from carting 60 lbs. of camera gear up and down Griffith Park to shoot the "super moon" tonight. But I got some great shots to share. :)

  • Yes, I started out with a 1200 baud modem! Does that date me?

    I was big into BBS's back in Florida. I've always felt the internet, as a whole, is just a natural extension of that world of hackers and weirdos.

    Look forward to seeing your shots. I love moon photography. Which forum are you going to post it?