Jung, I've already posted full size (1600 x 900 or 1600 x 1024 ) images without the 35mm adapter.
I'll make some more, but the weather is bad today and I'd like to shoot outside. I have the Elphel powered by a 12v battery from a hand drill now, so I can shoot wireless. I have to make some sort of stand for the laptop on two of the tripod legs.

Of course the 35mm adapter eats some resolution, but it's not the right to compare the two when I'm just at the first stage of developing this. The images on the TACX website are sharpened I guess. But don't forget that a normal video camcorder has an enormous amount of sharpening build in.
A well balanced sharpening on the Elphel footage in post works very well IMO.
'
I'll post new results when I have some more time.

Matteo, the chip much smaller than 16mm, so I need another lens to get a wider angle.

.
 

here's another sample clip

http://www.elphel.com/3fhlo/samples/333_samples/
i don't understand..why tacx-video color looks so good.. are they using diffrent CMOS model?

most of all the elphel 333 sample clip looks greenish.. expect tacx-video.

many thanks Jung
and for Andrey:
when did the 353 will be available?
can it record 1080p 24p format?
and about implementing sound in the theora files, did you have some more info?
and last I want to innstall linux on my pc
what distribution did you recommend me
or ...what distro did you support?
many thanks
Matteo Pozzi

Several things. The samples at our site were made with older sensor and all the code was focused on compressor - no precise color balancing. To say nothing about me being not so good at colors :-)

It is difficult fro me to answer these questions right now - I'm working on the camera design and not everything is settled up yet. But I'll try my best to make it happen sooner rather than later. Meanwhile it is not yet too late to add some wish list on our wiki.elphel.com

I'm not a Linux guru myself and we do not recommend any particular distribution. We are working with MPlayer team to add all necessary patches to the main MPlayer distribution so it will work with the camera stream natively.

Personally I use SuSE and like it, but I do not insist it is the best.

And we provide Live DVD with the cameras that includes all the software to run cameras on most computers. It includes modified version of MPlayer and scriptable MPlayer plugin for Mozilla Firefox browser.

I jazz up the colour in post a bit (lot :-) but you can now get the colour perfect for your own taste with the live-CD camera control/recording software. Here are some images from last weekends filming trip in Italy with the 333. http://www.tacx-video.com/images/HD2...ly/Gavia-Small

Hey Phil! Nice to see you on the board. Thanks again for all your info you gave me (some quoted on this forum) before I bought the Elphel camera.
Those are really beautiful images.

Phil, nice looking pictures. They certainly look a lot sharper and colorful than Oscar or Forrest's pictures, but can you tell us a little about them? Were they shot at SD res (binned?)? or were they shot at high res and resized in post? Also, they look like some kind of median filter has been used on them, did you do that or is that from the camera's built in processing? Did boosting the color make the images noisy which required a median de-noising filter? Thanks. And thanks for posting the pics

Dumb question, but can this be used with a mac? Or is it for PC's only.

I guess my 16mm lens is pretty sharp too... look at this frame grab with some more (bit too much) unsharp-mask and color balance in post.

http://bosshogg.logjam.com/oscar/Elphel/e3.jpg

.
 

HDR-hc1 also use 3mega cmos. it dose pretty amazing job with right codec in post.

http://hdv.blogdns.net/video/kitano.wmv

Hi everyone, sorry for not posting in a while. ive been watching everything going on. ive been doing different tests. and have come up with some various results.

Oscar great work, your images look great. they inspired me to take some stills at 1600 x 896 also and see what i could do. here are my results:

http://i11.photobucket.com/albums/a1...tti/lights.jpg
http://i11.photobucket.com/albums/a1...tti/window.jpg
http://i11.photobucket.com/albums/a1...ti/flowers.jpg
http://i11.photobucket.com/albums/a1...ti/guitars.jpg

I did focus my adapter onto a spinning gg with a 35mm lens attached and the results were good depth of feild but a major lack in sharpness as opposed to the c-mount lens only. Also, another major issue would be the light loss throughout the adapter. to get proper diffusion on the screen, you suffer light loss.

I did a video with a very close shot to show how sharp the camera can be. it looks nice and i will make stills out of it. (the video was shot at 7 fps).

As for the project in whole, i found a weird thing to be happening when i record footage in medium light. its sorta like compression noise, but much more noticeable and its over the entire image. i dont think its been doing this all along, it seems it started happening about 2 months ago. Im not sure, but it makes shots like those unusable. Its not a color issue, is more like a blurry/tv noise type pattern. kinda like when your vcr cables arent hooked up fully. not that horrible though.

anyways, the only other thing would be the rolling shutter. thats something you must work around. and perhaps the portablity of it all. A laptop would make that very simple though. but as for live preview, i am still searching for a good solution.

.
 

Phil..can you post hig-res 720p or 1080p footage.. when u have spare time.
i have seen it dozen times and still impressed.
i would compare your clip to SI-1920HDVR. this camera also use cmos but 2mega pix. color dosen't look good as yours ^^;

http://www.siliconimaging.com/Digita...y_footage.html

The true colour information is there with all these cameras I think, its obviously not mostly removed like with HDTV or DV tape. My original 1024p footage (I film 4:3 for our software based DVDs) has colour thats just the same as the others on here that have posted clips from the elphel or micron sensor. You just happen to like my wizard of oz like vibrant colour correction ;-). I think real movie film often has a lot of saturation removed, slight tints added to add warmth or feeling of cold. Sometimes even black & white to add that bit extra to the story line. Andrey has given enough control via the linux based software to get that look you want before you even start to fill up your hard drive.

Look at Panavision footage, it has a similar colour boosted look.

"A Man's Gotta Do" 2003 by Chris Kennedy, with John Howard (the actor) www.hopscotchfilms.com.au distributed by Roadshow entrainment (Warner Bros maybe in the states)

Is a good example, and a funny movie. They have a "making of" video that gives a good scene contrast to the Pana look.

Hello
 

I am a new member in the forum!
The last weeks i read the most articles about Elphel333 and cams with
MT9T001 resp. IBIS5A Sensor.
Because I look for a small camera for sports with wideangle lens, this should be one with global-shutter.
A big advance of the 333 is the MJPG comression, to save PC performance.
Now the question: functions the camera also with IBIS5A (found a PCB layout on the Elphel-developer page)

If not, then it has surely a reason, i think the IBIS5A fits only on 313 model.
With the quality of sensor I would be content, because i had a PixeLink A742 one week for testing. The global-shutter works very good.
About FPS: the higer the better.
thanks

They were shot at 1280x1024 & resized. I use a few filters on the footage in post. This is what the finished video looks like, it then plays inside our PC based software thats linked to a Tacx home trainer. The film needs to be as small as possible (low cpu load) & still look good at full screen.

sensors for 333
 

There is no difference for support of IBIS5 in 313/333 cameras - code is the same. The problem I've got with this sensor is that originally datasheet described 3 interfaces for the sensor control - serial I2C (2-wire), serial 3wire and parallel. Camera uses I2C - all the other sensors I've seen also use it, but it turned out that IBIS had a bug in the implementation, so they later removed it from the docs (not fixed the bug as I hoped :-( ). The IBIS5-1300 did not have a real snapshot shutter (seems like 5A has) and the image quality was not as good as with MT9M001 (they say I need an off-chip ADC, but I think CMOS sensor should have nice one inside) , price being 5x of that. So this is why I never came to modify the board for the 3-wire serial sensor control.

BTW - today I've got first images from the Micron MT9P001 with the model 333 camera) - 5MPix, 12 bit sensor

wow if you can ...post a link

Driver is not ready, sensor - dirty, lens - not fixed, back light through the open camera body (goes through the PCB and sensors plastic package) - so don't blame me for poor quality - http://webcams.elphel.com/5mpfp.jpeg

re: sensors for 333
 

Andrey Filippov , thanks for clarifying circumstances.
Then I want to concentrate on the new.
I hope that, the 5Mpix sensor profides a higher framrate at ROI.
My dream is 1280x720@50... with the remaining rolling-shutter distortion >=50fps i could live.
Please let me know, if you have experience about it.

Thanks , neat thread here...
Robert

Hi Andrey can you tell us how many MBytes/sec can transfer the next 353 model? (if you can) I'm looking for a notebook to use it so what I have to look for? 10/100 ethernet and 5400 hd is ok or I need more? the new macbook can run the live cd or not?
and about the software I've downloaded the live cd but without any camera I can't run the recording program! is there a way to view the graphic user interface (to see what can do by now and give you, if I or someone have,some ideas to make it better)? ....what I know are the pics on the wiki pages... I've look for the gui of the silicon imaging 1920... and I want to make something similar for the camera I'm planning to bild! so I've some idea for a touch screen user interface for a 7" monitor ...I don't know linux (for now) so what program I need, and to learn, to change the GUI of the elphel recording program? and is there someone that whant to help me?
Many thanks
and best reguards

Matteo Pozzi

What is going on!? no news from none of you!
can't wait to see new pics videos or news!

Busy with 353
 

5MPix boards tested, more sensors and production batch of the revised boards ordered.

Really busy now making PCB layout for the 353 that will among others accomodate IDE port (I mean - it seems to fit nicely on the board without increasing dimensions).

Come again, are you saying that you will have direct to disk recording (I am surprised)? Can you outline the features and performance of this camera?

I am considering the BBC open Dirac wavelet codec on FPGA, to see if it is a suitable target for your camera. I discovered something rather neat, and up my sort of compression alley, 3D Wavelet compression, that compresses over time as well as the screen. I have been wondering what this 3D wavelet stuff was. I understand they use wavelets in security camera applications. If it can be applied to bayer, as cineform did with their wavelet, then 100Mb/s could give 720p bayer lossless, and 1080p visually lossless. Very higher performance.

353
 

It uses newer/faster Axis ETRAX FS, slightly (20%) bigger FPGA (largest in the same size package), twice each memory (64MB video + 64MB system + 32MB flash), USB 1.1 (available in the Axis chip) and ATA-6 (routed to a high density connector so adapter will be needed to connect actual disks). System design makes use of actual bus switching in addition to pseudo-DMA channels of ETRAX - that can provide 5x bandwidth between the FPGA and the system memory.
Small 12GB HDD can fit into the current body.

wait wait...is it possible to record directly to hdd? what is the meaning of "USB 1.1 (available in the Axis chip) and ATA-6" ...we can connect a microphone for example or a video out signal through the usb is that the meaning ?!? thanks for the info
Matteo

Ata, Usb
 

Yes, that's the goal.

Also - yes - the easiest way to get audio in.

Don't think so - it is USB 1.1, not 2.0 and I don't know yet what will be the actual data rate from ETRAX FS available.

Awesome. Keep us posted here on dvinfo :)

Does not support S-ata, that is small connector.

But the actual performance data rate, or the compressor, and through the ata, and the drive, where do they all max out? Can all do 12.5MByte/s, or can it do more, like 20MB/s etc?

What is the current plan for this camera (seems portable) and what is current plan for compression engines?

This looks much more promising for video than current camera.

Have looked at ATA-6, needs DMA/UDMA to go above 16.6Mb/s if supported (upto 100MB/s) is this supported?

http://www.Seagate.com/training/ata/unit1/mt2p9.html


What we can look forwards to in the future, past hard disks. IBM is planning a Millipede SD memory card in the next year or so, uses atomic force microscope heads, 100GB ( I thought I read 150GB somewhere).

http://www.tgdaily.com/2005/03/11/ib...lay/index.html
http://blogs.zdnet.com/BTL/index.php?p=1143&tag=nl.e539
http://www.research.ibm.com/journal/.../vettiger.html

12GB, can be so small as to mount on the motherboard, and somebody is bound to put these things into a hard drive that uses a small laptop IDE interface.

Very good for me, as I plan game system and third world platform.

Well thanks, impressive.

I did not work with ETRAX FS yet (complete docs are available at http://developer.axis.com) so we'll see what it can provide. But compared to the current CPU it is twice faster and uses I/O coprocessor. On my side I design the hardware to use faster data transfer from FPGA to system memory (bursts of 400MB/sec, in 313/333 - 80MB/sec).

As for compressors engines - will use the same MJPEG and Theora, FPGA is just 1200K gates. I expect next year Xilinx will have something like Spartan 4 made on 65nm - then we'll be able to increase compressor performance.

Sorry, I thought you mentioned improvements to compressor engine for next model?

I am currently finding out about an alternative FPGA design, would that be of use?

It is a big project to develop FPGA code for compressor. Theora took me some 2000 hours, Dirac will likely take more and I don't have such a window in my near future.

Some more 333 images of Provence, France (compressed, resized & made Mpeg4 friendly) http://www.tacx-video.com/images/HD2...oux/Nesp-route The crash jpeg has the car on the left thats about to pull out into us & smash our door in!

With Dirac they are doing their own, their is one being done at Opencores I think. Their is stuff afoot, and Dirac might not be the only one out there. So only a complete design need be integrated with support logic in the core.

What do you think of this 3D Wavelet, that is being used in Surveillance and other industries, is it any good?

Here are some links, there is more going on I feel, then what is seen. I think, maybe, something usable for the next year or so:
http://dirac.sourceforge.net/
http://www.opencores.org/projects.cg...dirac/overview

Here are some others that might be of interest, but not FPGA:
http://en.wikipedia.org/wiki/Tarkin
http://en.wikipedia.org/wiki/Snow_(codec)
http://www.opencores.org/cvsweb.shtm...ssion_systems/ ;)


This for inspiration to beat, not such a great camera, but Bo's managed to get some good test footage out of it:
http://www.bophoto.com/HDV/video/

I have an HD1 & its absolutely nothing like the elphel quality footage wise but its one fantastic camera to have for filming the kids & taking about in your pocket. Ive even used it as a helmet camera. But its image stabilization is awful so its useless for us jerky impatient dads out there!

If you look at the HD1 view example video of the hill & grass you can clearly see the terrible artifacts in the back ground with all the trees smudged in the distance with the grass on the distant hill being clear. The camera has a low bandwidth so you have to think about how complex the scene is if you want everything to be clear. In the clip it seems like its able to dedicate the bandwidth to the foreground grass which looks nice.

Yes, but it still looks gorgeous so often in the right hands, that you can image how well an advanced Elphel could do? But, I would suggest look at the colour balance (when it is not oversaturated, see threads about turn saturation down, and off during playback) and response of the sensor under ideal conditions, this response does look pretty good compared to the Elphel footage I have seen.

I will look for some raw elphel footage to post.