November 30th, 2007, 08:28 AM | #196 |
Alex...
I have yet to see a factory signal flow block diagram that verifies your statement. Considering that there are over 1M pixels, to extract the analog signal from each and port it to the analog output would require massive computational power. I'm afraid every camera digitizes the signal coming from the sensor block in order to condition(amplify, color correct, apply bayer filter, etc) the data stream as digital rather than analog. As such, any output analog signal has been A-D then D-A'ed at least one time. Going to SDI output saves the last D-A conversion. And all this happens before digital compression to HDV format. |
|
November 30th, 2007, 08:36 AM | #197 |
Major Player
Join Date: Dec 2005
Location: Natal, RN, Brasil
Posts: 900
|
Maybe David Newman can shed some light on this? I believe he has stated in previous posts, that some of the cams do have component output that has not been compressed? Does anyone have a way to verify or confirm this?
It would certainly be of interest to those possessing those "certain" cams if that were true! Especially given the purpose of this thread... |
November 30th, 2007, 08:49 AM | #198 |
Trustee
Join Date: Sep 2003
Posts: 1,435
|
Bill, in 2006 I have built a PC for the purpose of capturing the uncompressed Component out from Sony FX1. (And Sony engineering department was very interested in how I did that at the time.)
Before that, there was a lot of discussions on this forum whether that Component out is before or after HDV compression. Two simple tests show that it is before (and that's what I said in my previous post.) 1. No delay. The signal has no discernible delay vs reality. This would've been impossible if it was mpeg2 encoded, since long GOP introduces about 1/2 sec. delay. Just compare to the FireWire output, which is mpeg2 compressed. 2. Quality of the signal. No artifacts at all on fast pans/movements out of Component, vs blocky/tear artifacts out of tape/firewire. I had side-by-side comparison of live capture posted online, but took it down recently as I thought the subject is already clear :) |
November 30th, 2007, 08:57 AM | #199 |
Alex...
Let me make the distinction between "compression" and analog to digital conversion, i.e. digitizing the analog video signal. Digitization is not compression, per se. It's the process of converting the analog YUV levels to 1's and 0's. Each time a conversion happens, there's some inefficiency that degrades the signal quality(dithering, etc) My point is that any analog output signal will never be as clean as a digital SDI signal, as long as an A>D and D>A process has occurred. I am at a loss to explain, otherwise, why the compressed m2t images out of my HD110 show higher resolution than the analog images viewed out the analog ports. I assume that the native sensor signal is 4:4:4 in digital terms, anyway. This is "trimmed" to 4:2:2 somewhere in the signal path. I don't beleive sensors detect in digital 4:2:2 natively. I, also, think we should move this discussion to another thread so we don't hijack Cineform's basic intent. |
|
November 30th, 2007, 09:13 AM | #200 |
Trustee
Join Date: Sep 2003
Posts: 1,435
|
I was replying to Ron Evan's post.
My point is that Yes, there would be a very big quality difference capturing from Component (in his case; as well as HDMI or HD-SDI on other cams) into the Cineform recorder box, vs recording on-camera. So Cineform recorder is not just a firewire capture device in league of FireStore, but rather Cineform provides (will anyway :) a totally different level of quality of signal. |
November 30th, 2007, 09:55 AM | #201 |
CTO, CineForm Inc.
Join Date: Jul 2003
Location: Cardiff-by-the-Sea, California
Posts: 8,095
|
To be clear, all live outputs from all HD/HDV cameras are pre-compressed. Some have been downsamples to 1440x1080 4:2:2 first, but none (that I've aware of) go down to 4:2:0 before their live output. This in not a gift from the camera manufactures, but a practical design issue. The compression engine in camera consumes a lot of power, compressing then decompressing, consumes more power. As these are battery operated devices, the cameras are all only compressing or decompressing, never both at the same time. As a result all live outputs are free of compression artifacts and free of delay. The recorder concept is suitable for all cameras, we just need the interface to make each of the camera types: HDMI, HDSDI and component.
__________________
David Newman -- web: www.gopro.com blog: cineform.blogspot.com -- twitter: twitter.com/David_Newman |
November 30th, 2007, 07:43 PM | #202 |
Inner Circle
Join Date: Mar 2003
Location: Ottawa, Ontario, Canada
Posts: 4,220
|
I think we have probably discussed this enough, but to clarify my meaning is in agreement with Bill. I did not mean MPEG2 encoding but the signal has to be encoded and scaled to come out the analogue ports.Significant delay only is visible when there is a GOP involved requiring a time frame for encoding ( the length of the GOP as a minimum). The DSP in the camera reads off the information from the imager, and encodes it for further processing. This could be decoding/scaling to come out of the analogue ports or encoding for recording to tape etc. If one draws the line after the first processing from the imager then there are several routes to follow. Encode the signal to come out of the HDMI port, encode the signal to come out the analogue ports, encode( in this case compress to reduce storage needs) the signal to record to tape( or other record medium). Depending on the camera the encoding could be ....(re-)scale to go out the HDMI port, (re-)scale and filter for the analogue ports and (re-)scale and compress for storage. I have not seen the signal flow from each of the cameras currently available with HDMI so have no idea where these processing elements appear for each model. I have tried to find a write up about the HC3 that I saw some time ago that inferred the output from the DSP was already 1440x1080. It was then upscaled for HDMI to 1920x1080, decoded for analogue and encoded to MPEG2 for tape recording. I will continue to try and find the reference and post if I find again.
Ron Evans |
December 1st, 2007, 08:44 PM | #203 |
New Boot
Join Date: May 2006
Location: Mexico City
Posts: 9
|
video assist for RED?
hi, this looks amazing!
I think this would be a perfect video assist for the RED using it's HDMI output. If the unit is rugged enough and LCD has sufficient quality it could also work as a portable director's monitor/video assist. It would be great if it could be powered via V or gold mounted batteries on the back of the unit. Add a way to properly secure the HDMI cable connection and a smart touchscreen interface and it would be a dream come true. Thanks for making it happen (and listening!) LB PS. "Solid" sounds cool |
December 4th, 2007, 12:57 AM | #204 |
Regular Crew
Join Date: Nov 2007
Location: Washington, DC
Posts: 78
|
Sounds good, but......
First of all, great concept, I hope you get it out the door soon. The world of HD videography is evolving very rapidly and is driven by technological advancements that happen so fast that, by the time solutions are worked through on the forums, completely new pathways are announced which render the workout as just an alternative. I can't believe that my first post will this long and complicated, but here are some initial thoughts on the Cineform SOLID solution:
1) I see your competition not just as the tethered 40lb PC or the Firestore-type device, but more so what I have been patiently waiting on for six months now: wireless uncompressed HDMI transmission. Phillips should have its wireless HDMI product out before christmas. Its priced at $299 but has a 33 feet limit (http://www.dailywireless.org/2007/01/11/wireless-hdmi/). Tempting, but what I am really waiting for is a different technology made by an Isreali company, Amimon, that can transmit uncompressed HDMI over distances of 100 feet (through walls) or 167 feet line-of-site (http://gizmodo.com/gadgets/no-more-h...rs-293937.php). The large A/V companies have invested a lot in this technology and devices with wireless HDMI chipsets built-in will be displayed at CES 2008 in January. I spoke personally with the company's VP who said that a dongle to retrofit devices will be available at retail in March '08. Motorola has invested heavily in this company, so its for real. I think this will solve many problems with a tethered setup, but there are limitations. Obviously you couldn't use it to go into the desert (or even go more than 167 feet from a car or temporary camp with a 40 LB PC setup inside), but tetherless short run&gun style, as well as studio and most location shoots should not be a problem. Moreover, it would allow for real time monitoring (no latency) using a production-style 24" or larger monitor, capturing on as big a hard drive as you can get into a PC, while still allowing the camera to roam around tetherless. Amimon has said that the wireless dongles will have a target price point of $200. So $250 for a BM Intensity card, plus the $200 wireless dongle equals a sweet untethered solution that competes with the SOLID and records directly to your editing computer. Of course, you don't get Cineform plug-in software with either the SOLID or the dongle/Intensity card, but it is nevertheless much cheaper than the proposed $2k Cineform price point (which I believe does not include the Neo or Prospect software, or am I wrong). The Cineform SOLID would be great for certain location shoots (like off road stuff) where electricity and extreme portability would be required. Other than that exteme usage ability, $450 versus $2,000 is a no brainer. So, adding in the cost of a Prospect license ($999) to the dongle/Intensity workflow, the SOLID, bundled with Prospect, is comparable to a wireless solution only if the SOLID+Prospect bundle is priced at $1450 or lower, or unless you need that off-road potential. Kudos to Cineform for tossing this solution around, but given that Cineform is thinking about $2k for the SOLID alone, its probably not a good alternative workflow for all people. In addition, the pair of 32gig CF cards are not going to be cheap. In all likelihood, I could probably buy a great editing/capture PC to use with the wireless/Intensity/Prospect capture system for the likely price premium of a SOLID/Prospect bundle + dual 32 gig CF cards. Or, since I already own one, put the $1500 or so back in the bank. But, and here is where I play devil's advocate with myself, I also use DV Rack's monitorig tools to analyze the signal. Of course neither the Cineform SOLID nor the the wireless HDMI solution allows for being untethered from the firewire that DV Rack's (On Location) currently requires. Wireless Firewire is on hold. Unless someone puts a product out that allows DV Rack to read video through an HDMI port, or converts an HDMI signal into a firewire signal, (strong hint) or unless Cineform adds a firewire output to the SOLID recorder, I will still be partly tethered, unless I want to spend a few thou on a production monitor that has built-in scopes. That actually would be my ideal setup if money were no object: Use the Cineform Prospect to capture uncompressed HDMI (either to the Cineform SOLID or the wireless HDMI/Intensity port), use DV Rack's scopes to monitor the luma, colors, etc via a tablet PC (using a firewire port on the Cineform SOLID or through the HV20's firewire port), and use an HDMI out to monitor the framing and focus, either to a portable 8" HD monitor (if using the Cineform SOLID) or to a remote 24" monitor (if using the wireless HDMI dongle - going out through the Intensity HDMI out port). 2) I capture sound via a dual channel Azden wireless system, using either wireless Azden lavs or Rode NTG-2 shotguns with an Azden wireless dongle. The output goes to my HV20's mini-stereo plug. If bypassing the camera's intake system will result in better audio quality, and still be in-sync with the image, then that would be a strong selling point for the SOLID. I really don't care if it is RCA, XLR, ministereo, etc. Its just a different cord from the Azden output. 3) I would definitely like to see a way to mount this (either on rails or to a 1/4 thread). 4) I would like to see a firewire out for possible use with DV Rack. 5) I would like to see an HDMI out for monitoring. 6) I would like the CF slots to be able to handle 32Gig cards (and potentially be upgradeable to 64gig cards -coming soon, no doubt) 7) If I record on the SOLID and download the avi or mov onto my editing computer, will I still need to buy Prospect or Neo in order to work with the files in Premiere CS3? Will I lose anything if I don't have the presets that come with Prospect. Will I be unable to output from Premiere CS3 using the Cineform engine (that's what happened when my trial version expired)? Thanks and good luck with this. It could be a great solution for many people. Last edited by Michael Panfeld; December 4th, 2007 at 01:46 AM. |
December 4th, 2007, 02:42 AM | #205 |
Regular Crew
Join Date: Nov 2006
Location: Cary, NC
Posts: 132
|
David -
Great job! It's so nice to see a good company listen to the community. Since we have to take what the huge camera conglomerates give us, it's companies like yours that allow us to take what they give us to the next level and get the absolute most we can out of it. I am currently working on a completely portable, free of tethers and 120v power and under 10lbs (incl. monitor), cineform capture system myself based around the HV20 and it's HDMI output. And i have to say... bravo. If this works like it's supposed to, and sells for sub 2k like the page states, you can definitely expect me to pick up a few. Everything on the spec page looks great. My one thing would be (and excuse me for repeating anything, as i haven't read the entire thread yet) to include a fairly high-res/contrast/brightness screen. Seems to me that if you're not going to do that, you may as well make the whole thing smaller (if it's the screen dictating the size) and put a smaller less consumptive screen. It would be a pain to have to carry that thing around AND a decent monitor, especially since your device HAS a monitor. I'm sure you've thought about this, and i know it's easier said than done, but that's a feature above most things i would appreciate more than most. Also, multiple mounting points would be a plus (Red) - and i'd probably make the device itself black. Everything else seems fine - no problems here with the RCA jacks (everyone in "pro" video is so stuck on "Pro" connectors). If you're really serious about your audio anyway, you'll be capturing audio separately on another device... Again, sorry if i'm being redundant, but any idea when this thing is going to market? Thanks and congratulations... Dale |
December 5th, 2007, 02:42 PM | #206 |
Regular Crew
Join Date: Feb 2003
Location: London, England
Posts: 126
|
Very excited about the prospect of recording high quality sound, which is directly linked to the video. Is that really what you are proposing with the "pro audio" inputs? That would make this device unique. I don't think there is currently any way of recording high quality sound for video without using a physically separate sound recorder, and then synching the sound in post.
But, realistically, how much of the sound system could you build in? High quality AD converters for 24bit 96k PCM stereo? Mic pre-amps? Phantom power? Could be a real breakthrough product, if you decide to go for it... Patrick Bower |
December 6th, 2007, 06:11 PM | #207 | |
Trustee
Join Date: Sep 2005
Location: Gilbert, AZ
Posts: 1,896
|
Quote:
My only concern is the HDLINK gets real warm. This tells me it probably draws a lot of current. I hope it's not to much a current demand. Any thoughts on this? I'm really looking forward to using SDI to a portable solution with Cineform. |
|
December 6th, 2007, 06:42 PM | #208 |
Major Player
Join Date: Sep 2003
Location: Solana Beach, CA
Posts: 853
|
A power tap is currently in the features list for the unit, but it is not shown in the illustration yet. It's on the list to dive a bit deeper into the converter specs from the various manufacturers to make sure that adding this feature is feasible.
|
December 7th, 2007, 04:50 AM | #209 |
Major Player
Join Date: Oct 2005
Posts: 376
|
I love this concept CF2CF (CineForm 2 Compact Flash) :)
How about to release 3 models: HDMI 8 bit codec for consumer HDV and AVCHD camcorders HD-SDI 10 bit codec for prosumer camcorders with HDSDI GIGe for camera heads like SI2K and Sumix to capture into Cineform RAW. |
December 7th, 2007, 10:28 AM | #210 |
Major Player
Join Date: Sep 2003
Location: Solana Beach, CA
Posts: 853
|
Serge, we do have a family of products planned of which the HDMI recorder is the first. Within throwing distance you've touched on some of the different versions we intend.
|
| ||||||
|
|