..ill second what barry said. in my own words, the pics on the site are still an improper comparision for the consumer. reason being that if i were to compare the top speed of a VW diesel rabbit and a F1 racer, i wouldnt launch the VW out of a canon and show it blowing up and say.. see.. it cant handle the speed of an F1. rather i would show it going its top speed and then show the F1 going its top speed and let the viewer decide in clear terms which is better. the pics on the site do the preceeding.. the new chart would seem to accomplish the latter. the new chart is nice at showing what benifit the andromeda offers in terms of lattitude. I would love to get some hands on to run all sorts of testing myself sometime.. has anyone outside of the company done testing yet? i think that would really boost sales and squelch alot of the disbeliefe out there to get some third party seals of approval. anyway... great product youve got there! thanks for the comparision!
Get 9.5 Stops of dynamic range from your DVX100
- Thread starter pertierr
- Start date
The extra dynamic range is pretty obvious from the most recent footage, November Outside. Looks frakkin' beautiful.
heisest: The most compact(and longest recording length) setup is a Seagate Momentus 100GB 7200RPM notebook drive in a firewire enclosure. This type of enclosure takes the power from the firewire, so there is only one cable connecting the drive and the computer, and that's it. The drive + enclosure itself is only slightly larger than a cigarette box. You can record a bit over an hour of 4:4:4 Uncompressed footage on it, and yes you can record until the end of the drive because the Momentus never drops under the required bandwidth. It goes for a small fraction of what a P2 card costs, yet it records a LOT more footage, and is uncompressed.
Of course, you can use full-size drives with more than 100GB for even more recording time. Usually, however, full-size enclosures require separate power, so it depends on your setup.
Cheers,
Juan
Of course, you can use full-size drives with more than 100GB for even more recording time. Usually, however, full-size enclosures require separate power, so it depends on your setup.
Cheers,
Juan
Spartacus
Veteran
Doesn't´t the macmini also require separate power (or did I miss sth...)? I wouldn't´t mind having a small form factor computer with some firewiredrive under my controlling monitor anyway...pertierr said:
And did I understand correctly, that sculptorHD also has the monitoring function eg DVrack offers?
So a Powerbook (or obins minimac w/ screen) could be used as a monitor?
That is correct. Here is an image of what the uncompressed monitor window looks like, please note that this image is of a much older version of sculptor. It changes so fast it's sometimes hard to keep updated screenshots. For example, the monitor LUTs where not activated in this screenshot so this is why the monitor looks 'milky', since you are seeing the raw logarithmic output.
Also note that if you have an NTSC output, such as the S-video output on the powerbook, Sculptor supports external monitoring as well simultaneously, or you can use any other computer screen connected to the computer. Sculptor also has a histogram function.
Also note that if you have an NTSC output, such as the S-video output on the powerbook, Sculptor supports external monitoring as well simultaneously, or you can use any other computer screen connected to the computer. Sculptor also has a histogram function.
Yes, we will be adding a waveform monitor and vector scope, although they obviously will not be calibrated to a standard, but rather they will use raw RGB values. It's tricky because people are used to using NTSC waveform monitors, and the raw footage doesn't conform to any standards so it will be a bit different as far as the scales involved.
Cheers,
Juan
Cheers,
Juan
We have posted a greenscreen commercial shot by Obin on an Andromeda beta unit:
http://www.reel-stream.com/clips
It's the most recent clip at the very top of the page.
Cheers,
Juan
http://www.reel-stream.com/clips
It's the most recent clip at the very top of the page.
Cheers,
Juan
I've put together a simple paper which talks about dynamic range/latitude and compares film to digital in this aspect:
http://www.reel-stream.com/redirect.php?url=http://www.reel-stream.com/DynamicRangeAndLatitude.pdf
We've also just posted a new sample frame at 4:4:4 24-bit Uncompressed with an Andromeda DVX, and we had a 35mm SLR at hand so we shot the same scene using Kodak 400ISO still 35mm film:
4:4:4 Color Corrected and scaled-down (1280 wide for 720P box) Image:
http://www.reel-stream.com/betacap/rs_1115cc.tif
35mm Kodak Still frame:
http://www.reel-stream.com/betacap/rs_1115_35.jpg
4:4:4 RAW 8-bit log NOT color-corrected or scaled:
http://www.reel-stream.com/betacap/rs_1115.tif
My Andromeda color correction aimed to get the same tonality as the film scan, but you can try and color correct it yourself using the RAW image.
Cheers,
Juan
http://www.reel-stream.com/redirect.php?url=http://www.reel-stream.com/DynamicRangeAndLatitude.pdf
We've also just posted a new sample frame at 4:4:4 24-bit Uncompressed with an Andromeda DVX, and we had a 35mm SLR at hand so we shot the same scene using Kodak 400ISO still 35mm film:
4:4:4 Color Corrected and scaled-down (1280 wide for 720P box) Image:
http://www.reel-stream.com/betacap/rs_1115cc.tif
35mm Kodak Still frame:
http://www.reel-stream.com/betacap/rs_1115_35.jpg
4:4:4 RAW 8-bit log NOT color-corrected or scaled:
http://www.reel-stream.com/betacap/rs_1115.tif
My Andromeda color correction aimed to get the same tonality as the film scan, but you can try and color correct it yourself using the RAW image.
Cheers,
Juan
thisiswells
Retired
Why 400 Speed? It looks a little grainy. How was the film scanned? Also, did you use the full still negative or a "motion picture 35" cross section (being that a motion picture frame is much smaller than a still frame, it's helpful to know the specifics of this comparison)pertierr said:
Obviously, the Uncompressed DVX100 looks phenomenal, jaw droppping, etc.
Would you mind going into a little detail on your setup? What was the f/stop on the DVX and how was this lit? If you can, tell me about your method of "color correction" and what applications and processes work well with uncompressed. It looks really, really sharp. I would like to know what you did to accomplish this.
Sure thing:
I picked 400 in order to get close to the sensitivity of the DVX which is over ISO 400. I figured it would make sense to match what kind of film you would use to shoot the actual scene, if you wanted to shoot it on film instead of DVX+Andromeda. If this is of big concern I can setup another scene which is much brighter and shoot it at 200 or 100.
The 35mm image is the entire still frame, about twice the size of a motion picture 35mm frame.
The film was scanned with a negative scanner at the local camera shop.
The DVX was at F4.8. On the SLR we used a Tamron 28mm-300mm lens. We set the SLR on the exact same spot as the DVX was(tripod), and then I zoomed in to get the frame as similar as possible. The shutter speed on the SLR was 1/45 to get it as close as possible to the DVX.
The scene was lit using two 500W halogen lamps bounced off a reflector positioned over the items in the scene.
For color correction on the uncompressed footage, I just dropped the frame in photoshop and:
1.)Applied an 'S' curve with a wide toe region and slight shoulder.
2.)Performed one step of sharpening.
3.)Resized the image down from 1540x984 to 1280x908 which converts from NTSC pixels to square(HD) pixels, and the correct width for letterboxing to 720P.
One of the aspects that I am still working on and changes constantly is our HD demosaicing algorirthm. Nobody has really done this before, so we are breaking some new ground and the algorithm gets better all the time. The image will certainly look even better than this as new versions of sculptor come out.
Let me know if you have any other questions or suggestions.
Cheers!
Juan
I picked 400 in order to get close to the sensitivity of the DVX which is over ISO 400. I figured it would make sense to match what kind of film you would use to shoot the actual scene, if you wanted to shoot it on film instead of DVX+Andromeda. If this is of big concern I can setup another scene which is much brighter and shoot it at 200 or 100.
The 35mm image is the entire still frame, about twice the size of a motion picture 35mm frame.
The film was scanned with a negative scanner at the local camera shop.
The DVX was at F4.8. On the SLR we used a Tamron 28mm-300mm lens. We set the SLR on the exact same spot as the DVX was(tripod), and then I zoomed in to get the frame as similar as possible. The shutter speed on the SLR was 1/45 to get it as close as possible to the DVX.
The scene was lit using two 500W halogen lamps bounced off a reflector positioned over the items in the scene.
For color correction on the uncompressed footage, I just dropped the frame in photoshop and:
1.)Applied an 'S' curve with a wide toe region and slight shoulder.
2.)Performed one step of sharpening.
3.)Resized the image down from 1540x984 to 1280x908 which converts from NTSC pixels to square(HD) pixels, and the correct width for letterboxing to 720P.
One of the aspects that I am still working on and changes constantly is our HD demosaicing algorirthm. Nobody has really done this before, so we are breaking some new ground and the algorithm gets better all the time. The image will certainly look even better than this as new versions of sculptor come out.
Let me know if you have any other questions or suggestions.
Cheers!
Juan
scharky
Veteran
Juan, that simply looks amazing. i have long been impressed with the footage you were getting, and hope to be able to afford the andromeda mod when it becomes available. I do have to say though, that film scan isn't the greatest, and looks a little hot in the oversaturated colors, but does give a good comparison. I while back I remember mention that you might have an academic discount available. I was wondering if there was any more info on this.
Keep up the good work.
Stefan
Keep up the good work.
Stefan
Sorry for the late reply.
What I meant is that we haven't found any other video cameras that use pixel shift to obtain a higher resolution image than what each sensor alone can provide. There are many similarities between our approach, and the way a single color sensor works so we were able to use similar concepts to process the information.
This approach does have a big advantage in less noise and more dynamic range than an equivalent single color sensor because the pixel elements on the sensors are larger.
Hope this helps,
Juan
What I meant is that we haven't found any other video cameras that use pixel shift to obtain a higher resolution image than what each sensor alone can provide. There are many similarities between our approach, and the way a single color sensor works so we were able to use similar concepts to process the information.
This approach does have a big advantage in less noise and more dynamic range than an equivalent single color sensor because the pixel elements on the sensors are larger.
Hope this helps,
Juan
Very good looking stuff! Wow!
It's interesting to compare the effective resolution of the two. Notice for instance the small text and copy right symbol on the bottle. Granted, the film is actually a bit closer up. Looks pretty decent resolution wise!
I know you haven't wanted to post a resolution chart as your HD algorithm keeps getting better but is there any chance we could see where it's at right now?
It's interesting to compare the effective resolution of the two. Notice for instance the small text and copy right symbol on the bottle. Granted, the film is actually a bit closer up. Looks pretty decent resolution wise!
I know you haven't wanted to post a resolution chart as your HD algorithm keeps getting better but is there any chance we could see where it's at right now?
We've done some initial tests, and compared them with charts shot on the JVC HD100 and Sony Z1 at different apertures. Andromeda on the DVX100 at least matches the HD100. It beats by a lot the Z1 in CineFrame modes, and matches or beats the Z1 in 1080i mode horizontally, but the Z1 has an edge in vertical lines in 1080i. This is to be expected as the chips on the Z1 i believe have 1080 lines vertically.
Of course, one of the biggest differences that is immediately noticeable is the huge difference in dynamic range between Andromeda and all of these new HDV cameras. These cameras look like they have considerably less dynamic range than even the standard DVX, i think in the ballpark of 6 stops or maybe less.
Hope this helps,
Juan
Of course, one of the biggest differences that is immediately noticeable is the huge difference in dynamic range between Andromeda and all of these new HDV cameras. These cameras look like they have considerably less dynamic range than even the standard DVX, i think in the ballpark of 6 stops or maybe less.
Hope this helps,
Juan