Need expert advice on new camera purchase
I am a member of the Documentary Institute at The University of Florida and thus wear two hats--one as a college professor and the other as a DP/editor. Because I am not in a position to do production full-time, I always fall slightly behind the technology curve, though I do my best to read everything I can get my hands on concerning the latest industry developments. But I must admit I am not the most knowledgeable concerning the lattest line of HD/HDV cameras. The bottom line is we are looking to purchase a high definition camera for the Institute and my colleagues and I are at a bit of an impasse. Two of us would like to go for a mid-range camera such as the Sony F350 or the Panasonic AG HPX5000. The third person insists that we should not spend more than $8-9,000 on a camera. He thinks we should go for the Sony HVR-V1U or the Panasonic HVX200. The 200 makes me nervous because of storage issues. We produce feature-length docs and often our shooting conditions are quite uncontrolled. So the thought of having to constantly halt production to download P2 data makes me uncomfortable. It looks like we will have between $30,000-$35,000 to spend on a camera and deck, so I am most interested in hearing what you would do in my position. I realize it is a bit like giving someone $30,000 and asking he or she to pick out a car! So I apologize, but I really respect your opinions and am in desperate need of guidance.
Why not buy 2 Sony Z1Us and a couple of Sony decks and you will have 12,000 left over for good audio gear.
Have not used the V1U, but we have 2 of the Z1Us and an A1U and they have been terrific. I am sure the Canon and Panasonic gear is good.....we just had a great experience with the Sony gear. We also have a PD-170 [standard def], Sony PDX -10 and a VX-2000. All have been terrrifcally reliable and more than paid for themselves. I also feel the Sony HDV decks are the most versatile.
Video is moving to a tapeless workflow. As an educational institution it is at least partly your responsibility to prepare students for the technology they will encounter. You will be best able to leverage your available funds without sacrificing quality by going with the Panasonic HVX. Learn the digital workflow, your students will need it, you will benefit from it and you'll all be able to make great feature length docs with an inexpensive system and be prepared for the future. And you'll be ready for the HPX 500 when it comes out...
I wouldn't use P2 if you gave it to me for free. It's a royal PITA and we've been on several shoots where a $2k per hour crew was standing in wait while files were downloading. Recording to an external device is another matter. But, remember this. Sooner or later, you'll need to store a bunch of footage. Think about it. Think about the future, not today. As much as we hate tape, it works for many years to come. I've had drives die because of static electricy or being dropped. My choice would be a tape recording unit or the Sony BluRay. The Sony has 1/2" chips also. That being said, I've seen some beautiful stuff come from the JVC cameras ans we just used an XLH1 Canon on a shoot in Mexico as a second unit to our HDCAM camera. Compared to our $80K Sony, the Canon was a beautiful thing at a lot less money. When we were done, I put the tapes in my pocket and away we went.
It's a dry heat!
High Definition is a very broad term, and--I believe--one needs to be careful when sorting through the marketing materials and advertising claims. Because you've narrowed down your search, why not rent those cameras and perform an experiment or two? In one weekend, your colleagues and you can test all the various claims from all the various marketing departments (sony, jvc, panasonic, canon, thomson...whomever). This can be done relatively inexpensively.
I shoot the Panasonic AgHVX-200, and I shot three straight hours with two 4 GB cards and one P2 store. The reason I stopped after 3 hours was because the show was over. I mention this because you seemed to think you'd have to halt production.
The reason I opted for the aghvx200 is because it was such a high data rate.
Another option to consider is the Canon HL X1 because of the HD-SDI interface. That's a lot of data to handle along with the expense of the special equipment (deck, wiring, drives, tapes, capture card).
The reason I opted for the aghvx200 is because it was such a high data rate."
Well...yes, when you shoot 60p, it's 100 Mbps...if you shoot 24p, there is a reason why you can get higher run times on P2 cards...you drop to 40% of that (40Mbps). XDcam's variable bit rate 35 Mbps should generally be in the same quality neighborhood (if not somewhat better), even at 1080. The aesthetic of the XDcam stuff is definitely different though...
We use an HVX200...the aesthetic is what is good about the camera...the sensor is basically SD, so the "look" is usually well-received by those who prefer the Panasonic 24p aesthetic, and the camera is relatively inexpensive. P2's disadvantage is that for continuous shoots, you need to add a person to the crew to wrangle P2 cards. Whether this is a good thing for students to be exposed to to prepare them...it probably isn't a bad idea in that sense, but in our experience, P2 has added time to our process.
In a student situation, the nice thing about a tape-based (or optical disk) workflow is that students can simply own their tapes and easily transport them. Students could also own drives, but data transfer adds some steps in that case as well.
All that said...HDV can be it's own world when it comes to optimizing post production. Something like the Canon camera being used to feed an NLE through the SDI can facilitate moving to another codec for post production, but in the case of the Canon, the last I knew, audio is not embedded in the signal, making ingest into an NLE a somewhat inconvenient process.
In the end...you have to pay sometime. At this point it seems like the less expensive cameras displace stress into other areas, cost or otherwise.
It will be a tough decision, particularly with the cost disparities...as you have alluded to, the downstream workflow is definitely a factor to carefully consider.
Creative Cow Host,
[TimK}"Well...yes, when you shoot 60p, it's 100 Mbps...if you shoot 24p, there is a reason why you can get higher run times on P2 cards...you drop to 40% of that (40Mbps). XDcam's variable bit rate 35 Mbps should generally be in the same quality neighborhood (if not somewhat better), even at 1080."
Just to clarify... Let me preface this by saying I have more experience with the DVCPRO HD (100Mbps) codec than HDV or XDcam and may therefor be biased toward the familiar:
Most video codecs deal with a 60 fps data flow. DVCPRO HD at 100 Mpbs should offer much less compression for each frame than XDcam's 35Mpbs, if both are recording 60 fps. The P2 card advantage shooting at 24Pn is because the data stream only has to deal with 24 actual frames per second, not 60 (and is why this advantage isn't available with the Varicam shooting to tape because it still lays down a 60 fps data stream even when shooting 24P). The relative merits of intra-frame compression versus long GOP compression aside, it appears to me that the DVCPRO HD codec still gives 3 times as much data as XDcam and 4 times as much as HDV.
Does anyone know if XDcam can use the full 35Mbps data stream to encode only 24p fps? or is the camera still recording 60 fps (with a 3:2 pull down for example) regardless of the shooting rate selected (24P, 30P, 60i). If XDcam records at 35Mbps with true 24P and no pulldown ("over 60")then the amount of data recorded WOULD actually be equivalent to the P2 24Pn.
That's how I measured it also. The electronics in the ag-hvx200 is always 100 mbs because it is 1080i/60 8 bit linear DCT. I also liked the sampling frequency of Y: 74.25 MHz, Pb/Pr: 37.125. The "200" is clever because as it records this signal it flags the appropriate fields and never records them to p2. The data rate therefore in the NLE (I use FCP) appears lower than what the in-camera sampling frequency actually is. You can imagine, then, trying to create an SD DVD from 720/24pN. It's missing some fields! I looked up the Sony XDcam info, and it has Y: 13.5 MHz; R-Y/B-Y: 6.75 MHz 10 bit (I couldn't find, but I assume it's also linear DCT). Red is the ONLY wavelet camera. Sony appears to be using DVCAM, so I guess/hypothesize it's laying all fields to "tape."
This month's American Cinematographer has two great articles on digital workflow, Finchner's film who shot on the Viper, and the new Red camera. Thomson also makes the Infinity, which ought not be overlooked.
[rmherd] "I looked up the Sony XDcam info, and it has Y: 13.5 MHz; R-Y/B-Y: 6.75 MHz 10 bit (I couldn't find, but I assume it's also linear DCT). Red is the ONLY wavelet camera. Sony appears to be using DVCAM, so I guess/hypothesize it's laying all fields to "tape.""
Red isn't the only wavelet camera...of course there is only wavelet compression, cameras are strictly visual capture devices...not to pick a nit, but the RED has several downstream recording options, including uncompressed if you can find a bit bucket fast and large enough.
The Silicon Imaging camera (of which there was a working prototype at NAB last year and it made it to the CML camera lineup for the second year this year...still no RED at the event) uses CineForm compression on RAW files, which is wavelet, and the RED compression is based on JPEG 2000 wavelet compression, the same as another wavelet based recording format, Grass Valley Infinity.
And of course, anything attached to a Wafian disk recorder running CineForm compression is a wavelet recording system...another method that has been in use for more than a year.
XDcam IS DVcam...XDcam HD is VBR at 35 and 18 Mbps settings, which offers an advantage over CBR of tape based formats. If you account for the fact that DVC Pro HD dumps 25% (1280 down to 960) at 720p and 33% (1920 down to 1280)at 1080, of it's image resolution upon recording, the spatial information lost vs. temporal losses in HDV or XDcam HD make any margins a bit harder to quantify in either direction.
The Varicam only does 720p and it looks the way it does much more due to the camera head than the compression system. I use one from time to time and really like the pictures it makes. Conversely, I haven't been impressed with the aesthetic of the XDcamHD stuff I've seen, but until I shoot with one, I have to reserve judgement.
What isn't useful in my opinion is comparing cameras based on image compression systems. We have an HVX200, and frankly I'd prefer to try to composite an HDV 1440x1080 image off an XL-H1's 1440x1080 CCD any day over the HVX 200's "higher quality" 960x720 image, subsampled from a 1280x720 image, formed from interpolating an image off a 960x540 sensor.
(And I have to say that I really LIKE the image aesthetic from the HVX200 camera...again, I just don't think you can use the recording format compression as the quality measurement for any particular camera any more...there are simply too many other variables.)
Creative Cow Host,
There's two ends to this data rate "problem." The first problem is the actual electronic gadgetry that sample and quantize the data. The panny electronics are always YUV: 74.25MHz : 37.125MHz : 37.125 MHz, (4:2:2). And it is quantized via 8 bit linear DCT. That rate is fixed at 100Mbps. The second problem is recording the data. In 720/24pN and 720/30pN, some fields are flagged and are NEVER recorded--although the first problem remains constant, so, when you bring it into the NLE, the data rate appears slower, but in reality, it just wasn't recorded, and the color data creating the pictures are very clean.
(The downstream workflow is its own mystery! It's absolutely crucial you know what your deliverable is going to be: BetaSP, DVD, EDL...so many to choose from.) The panny is very cool and clever but 24pN is not cool or clever for NTSC and SD DVD, in my opinion. I'm in a different position altogether where my material is screened in HD at a very high data rate. On those occasions when I am forced to make a DVD from 24pN, I have to ADD fields via compressor and motion correction because 24pN does not contain the fields necessary for 29.97. And it takes forever to render out.) It's why I recommend the Canon XL H1--HD SDI outs--for an educational program. Maybe next year, or the year after, the school can get an HDCAM SR deck.
(A note on content: By documentary I was assuming a Ken Burns style, and not a run and gun, news gathering style.)
At the moment the P2 technology is seem to be the best for the future it is true that these cards is very expensive at the moment, but with Panasonic and Sony are going to change over to the AVCH-Intra in a few years, I always question what do I do for shelve storage and I have come up with solution and that is to buy a DVCPro HD deck and transfere for shelf storage and take out a field deck for transfere during acquisstion. I was looking into purchasing a HDX 900 DVCProHD which uses tape and that is $30,000.00 with lens, if I pruchase a HPX500 that is $20,000.00 include lens and I could purchase some aditional cards plus a field recorder on both camera I still have to buy a deck, if I buy any of these camera, but if I buy the P2 technology I am looking into the future, I will not embrace the HDV or XDCamHD because all the majour players that make HDV and XDCamHD have sign on to the AVCH-Intra technology. And base on what I see so far that may just be the way to go.
[Edward G Downie] "if I pruchase a HPX500 that is $20,000.00 include lens and I could purchase some aditional cards plus a field recorder on both camera I still have to buy a deck, if I buy any of these camera, but if I buy the P2 technology I am looking into the future, I will not embrace the HDV or XDCamHD because all the majour players that make HDV and XDCamHD have sign on to the AVCH-Intra technology."
But does the HPX500 offer the AVC-Intra codec?
This month's American Cinematographer magazine reviewed the Red. The price point seems to be competitive.