A few weeks ago, I posted to this forum while I struggled to figure out how to equip myslef. In the end due to a combination of budget limits, but especially the uncertainty of my commitment ot video in combination with the uncertain though liely major changes to gear (seems big announcements every few weeks) with the moves to solid state and massive changes to storage and battery technologies I elected to eschew somehting really pro (love the Panny 170P) and in the interim go with tape in a form that would allow easy 24P, the HV40, so I could get into the game right away and started learning on the ground.
I'm a photographer and bring a modestly healthy knowledge of a lot of things technical to video. Nonetheless I'm still pretty new to it and there still seems a mountain range yet to climb. I've spent days searching the massive archives of this and other forums trying to up to speed in this new medium, (as well as days more just shooting and making a gazillion mistakes as I learn... and I am learning), but I remain utterly boggled by subjects variously surrounding codecs, formats, file types, resolution, compression (and the endless debate over frame rates and the effects of shutter-speeds), HDV vs Pro Res (in all its various flavours) and the true resolution of these cameras...etc etc.
From my reading of these forums it seems there is quite a lot of other confusion out there plus some really brilliant, generous souls.
My object is—within my meagre budget—to try to get the most of the tools I'm using. Much of what I'm coming to understand goes back to the basics of getting correct controlled focus & exposure but in particular a good steady shot and great mic placement, the latter two of which—as evidenced from my first attempts) remain hugely understated challenges that are much less discussed than all the pixel peeping plus the latest hacks and work-arounds and off course, gizmos (all of which I freely admit I can't get my hands on enough of) .
Assuming that I am gradually conquering those, (And that's a big assumption) I'm trying to get at some sort of definitive answer about workflow to get the most out of the capture, or potential capture.
So far I'm shooting Native 24p onthe HV40, recording to tape, capturing to FCP 6.06. via Firewire, but I suspect that is wrong and there is probably much else I could be doing to optimize my work on those rare occasions when I actually make a good shot.
That is, if I understand correctly the various forums, it would be better if I were not recording originally on tape, but via HDMI into some sort of hard medium to bypass the initial compression in camera, but besides the significant cost for added technology, that would very difficult ergonomically given the run and gun documentary nature of what I'm up to.
I've got a good pro shotgun that I agree in principle should be on a pole (my tests confirm that even a cheapo mic is better than a badly placed high-end mic) and indeed is on a boom stand for stationary interviews but sound crew with proper boom are not in my budget (I'd love to be on location with at least sound crew and better still also a line producer or to be that line producer/director with camera crew and sound crew) and so there's just li'l ole me.
The other thing is that I'm shooting this more like a still photog than a TV camera crew, a kind of in-your-face fly-on-the-wall. I'm on the move almost constantly, (movements being unpredictable making it hard to take advantage of teles) struggling to keep in front of my subjects, let alone keep up with them at all, yet have to be calm and unobtrusive enough both to get a steady camera shot but also not distract the subjects. Plus there are a bunch of my subjects who would just as soon run me over and kill me (literally)... and so the added rigamarole of more crew is virtually unworkable. Due to the struggle to keep in position and the aforementioned problem of anticipating camera position for good telephoto placement, to enhance mic placement I will be moving to lavalieres for a few critical subjects as soon as I can... s.
I'm armed with a host of stabilization devices: monopod, a DIY shoulder rig & "steadicam", and of course a tripod (though I'd really love a much better one). I find myself mostly partially stabilized with a shoulder rig which I simply need to learn to use better. When I can use a tripod, I do... though often I find myslef to have become separated from it.
I am working a hybrid double system for audio (external mic(s) to H4n to HV40) to retain the possibility of syncing the higher quality audio signal later or just using the more compressed audio in the tape.
Possibly worth considering HDMI collection for those stationary interviews. Guess I'd need some sort of adaptor to my Mac Powerbook G4 1.67 if such a thing even exists...?. (More gizmos, arghh...) Any suggestions?
Much of my A-roll is just audio collection as personally I prefer voice-overs to extended talking heads. This means I can focus on good mic placement and asking better interview questions, which BTW is no mean feat all by itself... I mean, there is a reason I became a photographer originally and not a writer
.
Other than the A-roll, my B-roll activity requires high mobility and unless there is an inexpensive, highly portable capture device for camera to storage via HDMI... will need tape.
For context, as I said above, I am a pro photog. As such, I do still captures only in RAW, work post in 16 bit files and layers for as long as possible until final output. And this is a strategy that has served me well notwithstanding the greater demands on processors, storage and my time inthe world of still photography.
What is the parallel workflow in video post?
Here is what I'm coming to understand from all that I've been reading. Please verify or correct:
1. the HVxx camera sensor sees full 1920x1080 but due to the bitrate compression limitations of writing to tape in real time can only record at 1440x1080 (the HDV codec)if recording to tape, which is an inherently lossy process.
2.That is, unless alternately one captures uncompressed data directly via hdmi port to a computer via eg. BlackMagic Intensity Pro (which BTW is not compatible with PowerMacG5s), or some sort of other portable capture device (what? does one exist? if so, is it affordable?) to grab as much as possible of the uncompressed 1920x1080 signal...
3. that this format of data is most closely approximated by the codec Apple ProRes 422 (or lately ProRes 422HQ?) if working in the Apple FCP environment, which is relatively lossless, but as a result consumes much more drive space. That is that one would either connect directly to the computer via some sort of HDMI cable/card interface collecting through FCP setup as some version of ProRes to capture the least compressed signal
4. that in photographic terms JPG is to RAW as HDV is too ProRes 422HQ though ProRes is notwithstanding also a form of compression
5. that if obliged to use tape to collect the data on camera initially, capturing to computer via FCP in HDV set-up (which he FCP saves out as .mov files) causes a second generation of lossy compression at the initial saving of that .mov file and that when further working with those .mov captures in FCP, rendering is once again a lossy process.
6. that somewhere in this process there is also additional compression necessitated by firewire which has a narrow bandwidth than HDMI in order to do the capture in realtime (or perhaps this only happens once at the time of compressing for tape in anticipation that it will be captured to computer via FW(?)... that anyway there is lossy compress happening right off the top...
7. and that therefore if obliged to use tape to collect the data on camera initially(already a second best choice to original realtime camera capture in HDMI), it follows that the best procedure would be to capture into FCP even from tape via HDMI cable/device into ProRes422(HQ?), that that second choice would be to capturing into FCP through firewire from tape via camera or deck.
8. that once the data is captured in ProRes, one continues to work on it in FCP in that format for as long as possible to facilitate the least amount of compression/recompression loss along the way... that in that format, data is not lost in rendering.
9. Apropos the foregoing, does working in ProRes put more or fewer demands on a computer's processor (bearing in mind I only have a PowerMAcG5 Dual core 2Ghz *GB RAM and e-sata connected 1TB externals, not some newfangeld 32core intellplus gaziliaxeon processor)?
10. or is it a case that there a division of of processor stress such thats editing in ProRes files makes fewer demands on the processor than HDV/Mov files, but rendering ProRes is harder on the processor than rendering HDV/Mov files... (this part is very confusing to me)
11. Or am I so underpowered on my post hardware side that I need to banish the notion of ProRes at all unless I'm willing to invest in the latest fasted processors, maximum RAM and RAID5 boxes?
12. Indeed, can I even work in ProRes on my Mac without a raid setup due to speed limitations of single drives, e-sata notwithstanding?
This knowledge curve is agonizing. I hope people won't mind my asking for this reality check. I realize that to people in the know, the preceding may appear utterly muddled replete with errors of assumption and comprehension. Please verify or correct/set me straight... or take me out back and shoot me (OK...not literally... please leave the killing to my subjects).
So, what is the best workflow... best marriage of frame-rates and codecs and capture and so on to get the most information/resolution? What are the reasons for the choices you suggest? What are the trade-offs?...
I'm a photographer and bring a modestly healthy knowledge of a lot of things technical to video. Nonetheless I'm still pretty new to it and there still seems a mountain range yet to climb. I've spent days searching the massive archives of this and other forums trying to up to speed in this new medium, (as well as days more just shooting and making a gazillion mistakes as I learn... and I am learning), but I remain utterly boggled by subjects variously surrounding codecs, formats, file types, resolution, compression (and the endless debate over frame rates and the effects of shutter-speeds), HDV vs Pro Res (in all its various flavours) and the true resolution of these cameras...etc etc.
From my reading of these forums it seems there is quite a lot of other confusion out there plus some really brilliant, generous souls.
My object is—within my meagre budget—to try to get the most of the tools I'm using. Much of what I'm coming to understand goes back to the basics of getting correct controlled focus & exposure but in particular a good steady shot and great mic placement, the latter two of which—as evidenced from my first attempts) remain hugely understated challenges that are much less discussed than all the pixel peeping plus the latest hacks and work-arounds and off course, gizmos (all of which I freely admit I can't get my hands on enough of) .
Assuming that I am gradually conquering those, (And that's a big assumption) I'm trying to get at some sort of definitive answer about workflow to get the most out of the capture, or potential capture.
So far I'm shooting Native 24p onthe HV40, recording to tape, capturing to FCP 6.06. via Firewire, but I suspect that is wrong and there is probably much else I could be doing to optimize my work on those rare occasions when I actually make a good shot.
That is, if I understand correctly the various forums, it would be better if I were not recording originally on tape, but via HDMI into some sort of hard medium to bypass the initial compression in camera, but besides the significant cost for added technology, that would very difficult ergonomically given the run and gun documentary nature of what I'm up to.
I've got a good pro shotgun that I agree in principle should be on a pole (my tests confirm that even a cheapo mic is better than a badly placed high-end mic) and indeed is on a boom stand for stationary interviews but sound crew with proper boom are not in my budget (I'd love to be on location with at least sound crew and better still also a line producer or to be that line producer/director with camera crew and sound crew) and so there's just li'l ole me.
The other thing is that I'm shooting this more like a still photog than a TV camera crew, a kind of in-your-face fly-on-the-wall. I'm on the move almost constantly, (movements being unpredictable making it hard to take advantage of teles) struggling to keep in front of my subjects, let alone keep up with them at all, yet have to be calm and unobtrusive enough both to get a steady camera shot but also not distract the subjects. Plus there are a bunch of my subjects who would just as soon run me over and kill me (literally)... and so the added rigamarole of more crew is virtually unworkable. Due to the struggle to keep in position and the aforementioned problem of anticipating camera position for good telephoto placement, to enhance mic placement I will be moving to lavalieres for a few critical subjects as soon as I can... s.
I'm armed with a host of stabilization devices: monopod, a DIY shoulder rig & "steadicam", and of course a tripod (though I'd really love a much better one). I find myself mostly partially stabilized with a shoulder rig which I simply need to learn to use better. When I can use a tripod, I do... though often I find myslef to have become separated from it.
I am working a hybrid double system for audio (external mic(s) to H4n to HV40) to retain the possibility of syncing the higher quality audio signal later or just using the more compressed audio in the tape.
Possibly worth considering HDMI collection for those stationary interviews. Guess I'd need some sort of adaptor to my Mac Powerbook G4 1.67 if such a thing even exists...?. (More gizmos, arghh...) Any suggestions?
Much of my A-roll is just audio collection as personally I prefer voice-overs to extended talking heads. This means I can focus on good mic placement and asking better interview questions, which BTW is no mean feat all by itself... I mean, there is a reason I became a photographer originally and not a writer
. Other than the A-roll, my B-roll activity requires high mobility and unless there is an inexpensive, highly portable capture device for camera to storage via HDMI... will need tape.
For context, as I said above, I am a pro photog. As such, I do still captures only in RAW, work post in 16 bit files and layers for as long as possible until final output. And this is a strategy that has served me well notwithstanding the greater demands on processors, storage and my time inthe world of still photography.
What is the parallel workflow in video post?
Here is what I'm coming to understand from all that I've been reading. Please verify or correct:
1. the HVxx camera sensor sees full 1920x1080 but due to the bitrate compression limitations of writing to tape in real time can only record at 1440x1080 (the HDV codec)if recording to tape, which is an inherently lossy process.
2.That is, unless alternately one captures uncompressed data directly via hdmi port to a computer via eg. BlackMagic Intensity Pro (which BTW is not compatible with PowerMacG5s), or some sort of other portable capture device (what? does one exist? if so, is it affordable?) to grab as much as possible of the uncompressed 1920x1080 signal...
3. that this format of data is most closely approximated by the codec Apple ProRes 422 (or lately ProRes 422HQ?) if working in the Apple FCP environment, which is relatively lossless, but as a result consumes much more drive space. That is that one would either connect directly to the computer via some sort of HDMI cable/card interface collecting through FCP setup as some version of ProRes to capture the least compressed signal
4. that in photographic terms JPG is to RAW as HDV is too ProRes 422HQ though ProRes is notwithstanding also a form of compression
5. that if obliged to use tape to collect the data on camera initially, capturing to computer via FCP in HDV set-up (which he FCP saves out as .mov files) causes a second generation of lossy compression at the initial saving of that .mov file and that when further working with those .mov captures in FCP, rendering is once again a lossy process.
6. that somewhere in this process there is also additional compression necessitated by firewire which has a narrow bandwidth than HDMI in order to do the capture in realtime (or perhaps this only happens once at the time of compressing for tape in anticipation that it will be captured to computer via FW(?)... that anyway there is lossy compress happening right off the top...
7. and that therefore if obliged to use tape to collect the data on camera initially(already a second best choice to original realtime camera capture in HDMI), it follows that the best procedure would be to capture into FCP even from tape via HDMI cable/device into ProRes422(HQ?), that that second choice would be to capturing into FCP through firewire from tape via camera or deck.
8. that once the data is captured in ProRes, one continues to work on it in FCP in that format for as long as possible to facilitate the least amount of compression/recompression loss along the way... that in that format, data is not lost in rendering.
9. Apropos the foregoing, does working in ProRes put more or fewer demands on a computer's processor (bearing in mind I only have a PowerMAcG5 Dual core 2Ghz *GB RAM and e-sata connected 1TB externals, not some newfangeld 32core intellplus gaziliaxeon processor)?
10. or is it a case that there a division of of processor stress such thats editing in ProRes files makes fewer demands on the processor than HDV/Mov files, but rendering ProRes is harder on the processor than rendering HDV/Mov files... (this part is very confusing to me)
11. Or am I so underpowered on my post hardware side that I need to banish the notion of ProRes at all unless I'm willing to invest in the latest fasted processors, maximum RAM and RAID5 boxes?
12. Indeed, can I even work in ProRes on my Mac without a raid setup due to speed limitations of single drives, e-sata notwithstanding?
This knowledge curve is agonizing. I hope people won't mind my asking for this reality check. I realize that to people in the know, the preceding may appear utterly muddled replete with errors of assumption and comprehension. Please verify or correct/set me straight... or take me out back and shoot me (OK...not literally... please leave the killing to my subjects).
So, what is the best workflow... best marriage of frame-rates and codecs and capture and so on to get the most information/resolution? What are the reasons for the choices you suggest? What are the trade-offs?...
Last edited: