4 ms·

> Every minute of 4K ProRes LT footage (which is a very lightweight format, compared to RAW) is 3 GB of space Do you really need to save all of that footage? I

> Every minute of 4K ProRes LT footage (which is a very lightweight format, compared to RAW) is 3 GB of space Do you really need to save all of that footage? I would think keeping the pro res footage for the current projects on the workstation and reencoded archive video on a NAS would be sufficient. I'm not a video professional, but I suspect it's easy to fall in the trap of thinking that you need to save everything in highest possible quality in case you need it later, but what are the realistic chances of that? If you end up needing some old footage again, AV1 coded 4K or even HEVC 1080p would probably be just fine. The final result are Youtube videos after all. I know he mentions editing from it, but that's enough space for more than a week of pro res video.

rektide 4y ago
Personally I kind of fright from the "have you considered throwing away your best copies" as a sales strategy. On the flip side, I think it would help a lot a lot a lot if there was more discourse out there about compressing high-dynamic-range content. AV1 supposedly has some capability to do a good-ish job with HDR. The idea of taking reels of raw video and spending a couple days squishing it to 1/10th the sizes but preserving the quality/flexibility very very well is a value proposition that I dont think is clearly attainable, even though it seems technically perhaps within reach. Right now, I think the general feeling is, raw is raw & everything else bakes in a vast amount of assumptions & constrictions. Some advocacy that compressed video can be as flexible, as dynamic, as capable need to be more present, elaborated, & proven before anyone's going to be comfortable throwing away the bits. These are people's life's works & a couple hundred or thousand dollars a year more in storage isnt a real factor for such integral, near work.
- sorenjan 4y ago
  I don't have a sales strategy, I'm not selling anything. I'm questioning the need of buying needless storage for data hoarding you're never going to use. Have you seen Jeff's videos? They're mostly him talking to the camera in what I think is his basement, or closeups of various PCBs. If you're running a stock photo company I get wanting to keep footage at max quality, but once your Youtube video is published, do you really need to save all those alternate takes and cut out bits? I just don't see them being very useful in the future, and if you do end up needing some of it for B roll or whatever, would a couple of seconds of compressed video really make a drastic difference for the project as a whole? Especially since the final result ends up on Youtube and consumed on a phone screen or a TV two meters away. Much of the content that's produced today is not trying to be timeless classics with endless rewatch value, it's ephemeral and only really relevant for a short time. If you can reupload your videos to other video hosts in the future that's probably good enough. Nobody is watching reviews of three year old Raspberry Pi add on boards for the low noise in dark parts of the image, or the accurate color reproduction, or even the 4K instead of 1080p resolution. Kill your darlings!
  - vineyardmike 4y ago
    For many people it’s an art. Just because it’s a raspberry pi video doesn’t mean people don’t want to be proud of the video quality and color accuracy and all that jazz! Why comment your code or name your variables well if the project will be over soon? No one is buying your product for the clean code behind it. If you need it later just use a reverse compiler. It’s good enough for a project you may never need! People do things for the art and to be proud of the quality. And you don’t want to throw that out!
smilekzs 4y ago
Not a video pro, but a counterpoint: "lossy encode, then process" is inferior to "process, then lossy encode". Reasoning: Some processing steps may amplify the noise introduced by lossy encoding. Other processing steps may be sensitive to SNR. Such steps chained together in a pipeline can make the otherwise small extra noise unacceptably big.
- sorenjan 4y ago
  Note that I suggested reencoding footage for archiving, not processing. Although again, the final result ends up on Youtube and gets reencoded by them and viewed on all kind of devices.
dylan604 4y ago
> I'm not a video professional I am a video professional, and you always keep your originals. However, the part being left out of the conversation is that as a storage pool for editing goes, this isn't deep storage. Content is typically only left on the edit storage while the project is active. Even the TFA mentions he goes all the way to cloud cold storage. Edit storage has always been expensive, and maintaining capacity was always a juggling act. Just because terms like NAS are being used, one should not think of this type of storage as a dump it and forget it type of storage. There are many levels of professional. On the high end, the footage from the camera is copied to multiple hard drives on set. These are the backups, and us old timers still use the terms camera originals. These are as sacrosanct as film negatives, only there's magically multiple copies. This data gets transferred to the editor's storage. Once the edit is completed, the edit session and other content used in the edit may be transferred to the camera original drives for archival. On the other end of professional, you take the SD card out of the camera and transfer directly to the edit storage. In these situations, woe be unto thee that doesn't make proper backups. At that point, it's really more pro-sumer than professional, but hey, if they're getting paid, they're professional enough.
- sorenjan 4y ago
  So lets say you release a 10 minute Youtube video. You now have 60 minutes of raw footage with multiple takes, you sitting in front of the rolling camera going over the script, ums and ahs, and maybe a take where the phone rang in the middle of a sentence. What are you going to use that footage for in the future, and is this hypothetical use case absolutely dependent on having raw 4K video? I understand that your answer is from a general video production standpoint, but my question was more in the line of "is this really necessary for this kind of content creation?". And yes, I did see the part about editing, but are you really editing nine days worth of footage at once? Would it make more sense to put less but faster storage directly in the workstation?
  - dylan604 4y ago
    Again, you're limiting to just YouTube creator styles, and I'm including full on editorial for things like corporate training, episodic, feature, music videos, etc. That's my world. Obviously, the longer the final piece, the more footage necessary. However, something short like a music video can require a huge amount of footage that you might not consider. All of those quick edits of just a few frames to seconds of a shot wasn't that short from the camera. Some of those types of shots might not have ever been intended to be used and be something in between takes. It's all creative, so there are no rules in music videos. Doing that kind of editing does require access to all of that footage at any moment. If you're doing supervised sessions with the client sitting next to you means that they on a whim could decide they want to work on a totally different part of the content. When the client is footing the bill by the hour, you don't get to waste time "loading" content. It's really one of those things that until you've walked a mile in another person's shoes, it's best to not go blindly suggesting "better" based on one's limited knowledge. Asking questions for a better understanding is a totally acceptable way of learning, and I'm all for it. We're danger close to the former. Hopefully, I'm not sounding like an asshat leaning towards the latter.

rektide 4y ago

Personally I kind of fright from the "have you considered throwing away your best copies" as a sales strategy. On the flip side, I think it would help a lot a lot a lot if there was more discourse out there about compressing high-dynamic-range content. AV1 supposedly has some capability to do a good-ish job with HDR. The idea of taking reels of raw video and spending a couple days squishing it to 1/10th the sizes but preserving the quality/flexibility very very well is a value proposition that I dont think is clearly attainable, even though it seems technically perhaps within reach. Right now, I think the general feeling is, raw is raw & everything else bakes in a vast amount of assumptions & constrictions. Some advocacy that compressed video can be as flexible, as dynamic, as capable need to be more present, elaborated, & proven before anyone's going to be comfortable throwing away the bits. These are people's life's works & a couple hundred or thousand dollars a year more in storage isnt a real factor for such integral, near work.

sorenjan 4y ago
I don't have a sales strategy, I'm not selling anything. I'm questioning the need of buying needless storage for data hoarding you're never going to use. Have you seen Jeff's videos? They're mostly him talking to the camera in what I think is his basement, or closeups of various PCBs. If you're running a stock photo company I get wanting to keep footage at max quality, but once your Youtube video is published, do you really need to save all those alternate takes and cut out bits? I just don't see them being very useful in the future, and if you do end up needing some of it for B roll or whatever, would a couple of seconds of compressed video really make a drastic difference for the project as a whole? Especially since the final result ends up on Youtube and consumed on a phone screen or a TV two meters away. Much of the content that's produced today is not trying to be timeless classics with endless rewatch value, it's ephemeral and only really relevant for a short time. If you can reupload your videos to other video hosts in the future that's probably good enough. Nobody is watching reviews of three year old Raspberry Pi add on boards for the low noise in dark parts of the image, or the accurate color reproduction, or even the 4K instead of 1080p resolution. Kill your darlings!
- vineyardmike 4y ago
  For many people it’s an art. Just because it’s a raspberry pi video doesn’t mean people don’t want to be proud of the video quality and color accuracy and all that jazz! Why comment your code or name your variables well if the project will be over soon? No one is buying your product for the clean code behind it. If you need it later just use a reverse compiler. It’s good enough for a project you may never need! People do things for the art and to be proud of the quality. And you don’t want to throw that out!

smilekzs 4y ago

Not a video pro, but a counterpoint: "lossy encode, then process" is inferior to "process, then lossy encode". Reasoning: Some processing steps may amplify the noise introduced by lossy encoding. Other processing steps may be sensitive to SNR. Such steps chained together in a pipeline can make the otherwise small extra noise unacceptably big.

sorenjan 4y ago
Note that I suggested reencoding footage for archiving, not processing. Although again, the final result ends up on Youtube and gets reencoded by them and viewed on all kind of devices.

dylan604 4y ago

> I'm not a video professional I am a video professional, and you always keep your originals. However, the part being left out of the conversation is that as a storage pool for editing goes, this isn't deep storage. Content is typically only left on the edit storage while the project is active. Even the TFA mentions he goes all the way to cloud cold storage. Edit storage has always been expensive, and maintaining capacity was always a juggling act. Just because terms like NAS are being used, one should not think of this type of storage as a dump it and forget it type of storage. There are many levels of professional. On the high end, the footage from the camera is copied to multiple hard drives on set. These are the backups, and us old timers still use the terms camera originals. These are as sacrosanct as film negatives, only there's magically multiple copies. This data gets transferred to the editor's storage. Once the edit is completed, the edit session and other content used in the edit may be transferred to the camera original drives for archival. On the other end of professional, you take the SD card out of the camera and transfer directly to the edit storage. In these situations, woe be unto thee that doesn't make proper backups. At that point, it's really more pro-sumer than professional, but hey, if they're getting paid, they're professional enough.

sorenjan 4y ago
So lets say you release a 10 minute Youtube video. You now have 60 minutes of raw footage with multiple takes, you sitting in front of the rolling camera going over the script, ums and ahs, and maybe a take where the phone rang in the middle of a sentence. What are you going to use that footage for in the future, and is this hypothetical use case absolutely dependent on having raw 4K video? I understand that your answer is from a general video production standpoint, but my question was more in the line of "is this really necessary for this kind of content creation?". And yes, I did see the part about editing, but are you really editing nine days worth of footage at once? Would it make more sense to put less but faster storage directly in the workstation?
- dylan604 4y ago
  Again, you're limiting to just YouTube creator styles, and I'm including full on editorial for things like corporate training, episodic, feature, music videos, etc. That's my world. Obviously, the longer the final piece, the more footage necessary. However, something short like a music video can require a huge amount of footage that you might not consider. All of those quick edits of just a few frames to seconds of a shot wasn't that short from the camera. Some of those types of shots might not have ever been intended to be used and be something in between takes. It's all creative, so there are no rules in music videos. Doing that kind of editing does require access to all of that footage at any moment. If you're doing supervised sessions with the client sitting next to you means that they on a whim could decide they want to work on a totally different part of the content. When the client is footing the bill by the hour, you don't get to waste time "loading" content. It's really one of those things that until you've walked a mile in another person's shoes, it's best to not go blindly suggesting "better" based on one's limited knowledge. Asking questions for a better understanding is a totally acceptable way of learning, and I'm all for it. We're danger close to the former. Hopefully, I'm not sounding like an asshat leaning towards the latter.