Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Ikonoskop DII vs 2/3" interchangeable - a very speculative comparison

Peter Osinski

Well-known member
Joined
Jun 5, 2009
Messages
370
Reaction score
0
Points
0
Location
New York
http://www.ikonoskop.com/blog/back-on-track/

they are set to deliver in July with a limited pre release in April. Looks like they are on a close schedule to that of the 2/3" cinema

The Ikonoskop comes at a price of about 9600 american dollars and includes the following:

Camera w/ pl mount
Battery
9mm Lens
1 memory catridge

I would find that comparable in general purpose to

2/3" Cinema - 2750
Side CF Card Module - 500
RED 16GB CF Card - 300
Side Handle - 950
8mm RED mini prime - 950
2.8" LCD - 900

Total 6350

but lets consider the fact you'd pretty much need to buy a PL mount if you ever wanted a zoom on this thing because RED has said over and over they aren't making a 2/3" zoom

$6850 for cinema vs. $9600 for ikonoskop to get the following:

RAW recording, power, control, small form monitoring, 15 min worth of recording media, a wide angle lens, a PL mount for interchangeable lenses (there are multiple options available for both companies but I am using this for the example)

That seems like a pretty big difference in price especially when you consider the following:

RED's often confusing and complicated workflow is still more solid than the somewhat nonexistent workflow of cinema DNG, and that Cinema DNG seems like it may be an Adobe only thing for a while leaving out FCP users until and if apple decides to pick this up

Cinema has higher resolution, and higher frame rates with smaller file sizes

more recording format options (3k, 1080p, 720p), versatility in its ability to be built up or down (compared to the d II which is limited to a smaller battery, one type of recording media, only records 1080p, and has limited I/O options down the road),

generally more features such as gps and what not

On the other hand I give ikonoskop a point because it's "complete package" won't involve any stray wires for monitoring options and its longer depth/ shorter height body could be beneficial in certain circumstances.


In europe this camera will be going for 6950 euros. It would seem to me that the increased price for americans because of the exhange rate alone could potentially be a deciding factor for someone in the states picking between the two cameras, as compared to someone in europe who isn't affected by the global value of the dollar when choosing his camera.

I have been trying to follow the development of both cameras as closely as possible but I may have gotten some of the details wrong. Also, I of course realize that both of these cameras are unreleased and many aspects can change still before they are released.

I would love to hear any comments, or criticisms of my comparison (which is done in the context of an budget conscious indie/doc filmmaker doing product research). I know this is the scarlet subforum but by all means if someone wants to throw in a few more reasons to go with the ikonoskop then do so and we'll just say you were playing devil's advocate
 
DNG workflow

DNG workflow

somewhat nonexistent workflow of cinema DNG, and that Cinema DNG seems like it may be an Adobe only thing for a while leaving out FCP users until and if apple decides to pick this up

While full support for CinemaDNG may not quite be there, there is support for the Acam dII through my software using TIF files made from the DNG, and through Cineform. Cineform has a DNG to Cineform convertor program, you have to ask them for it because it seems that it is not a retail program yet.

Other workflows using DNG to DPX can be done about like any other DPX job.

SI-2K and Kinor-2K both have DNG output options as well, and all Camera types no mater what their internal format can be edited in my software as long as some program can convert them to compatable 48bpp TIF.

I am working on my own RAW to DNG/DNG to TIF convertor program that will also do TIF to TIF for RED cameras image processing and color corrections. That should be out this summer. Until then DNG_validate.exe, DCRAW (tm), and Adobe Lightroom as well as other de-Bayer programs that read DNG Bayer packed bit 12 bit data can be used.

The results from the Kodak sensor look good, I am tuning my de-Bayer now for better finished results.

The main issue is 3K compressed vs. 1920x1092 uncompressed. The RED de-Bayer programs do a very good job, so Acam is going to need to look at the de-Bayer issues quite a bit and how K value and EI/ISO changes impact the noise. I think those are a bigger issue than the price for real filmmaking.
 
The design of the DII is beautiful.

Spec wise, Scarlet clearly wins... but until we can see test plates/samples from both, it is impossible to say which performs better.
 
I think that it is safe to say that scarlet will out perform the DII... seeing as how the scarlet is scaled 1080p from 3K... and for 10K one can get a lot more from the scarlet icluding not one but a few lenses...
 
The design of the DII is beautiful.

Spec wise, Scarlet clearly wins... but until we can see test plates/samples from both, it is impossible to say which performs better.

like i said, speculation...but i do agree with you that the DII looks great and the form factor looks like it would be very comfortable to shoot with, not to mention amazingly lightweight

dan,

1. I am slightly more interested in a native workflow using RAW. I have tried the adobe cinema dng in cs4 and while ppro didnt really work I was really psyched how it worked in AE. The only problem I found though was that I couldn't scrub through the clip while adjusting whitebalance, exposure, etc. I remember specifically because the clip started off out of focus and I wasnt able to get a view or it in focus from later in the clip for adjusting those settings. I am sure it is an easy fix though.

2. are you making an an NLE or more of just a straight conversion program. I haven't looked at your website too in depth but it all seems very interesting

mark, I'm not quite sure what your post means

any other feedback on how much you think the euro/dollar exchange rate will affect the choices of american buyers? I am curious to know how other people think.

I am not trying to talk down about any camera system, I admire both, I simply started this thread for the sake of discussion because I don't really have anyone to shoot the sh*t with when it comes to digital cinema
 
Have you ever seen a good image from the DII? They are using old skool sensors that just aren't very good.
 
That Ikonshop camera doesn't have nearly enough fanboys, so it probably isn't very good. It looks really cool though. I'd like to buy one just to try it out. I wonder how realistic their shipping dates really are though.
 
NLE/CC

NLE/CC

dan,

1. I am slightly more interested in a native workflow using RAW. I have tried the adobe cinema dng in cs4 and while ppro didnt really work I was really psyched how it worked in AE. The only problem I found though was that I couldn't scrub through the clip while adjusting whitebalance, exposure, etc. I remember specifically because the clip started off out of focus and I wasnt able to get a view or it in focus from later in the clip for adjusting those settings. I am sure it is an easy fix though.

2. are you making an an NLE or more of just a straight conversion program. I haven't looked at your website too in depth but it all seems very interesting

A1) I am working on a de-Bayer program that will have several "process methods" and work in batch mode going from source to result files, since it will work with both RAW/DNG data and RGB/TIF data input it can be used with both RED and Acam/SI-2K/Kinor-2K and other cameras.

To do the color correction you will still need to open the CAD program and use its color correction, you can pick any frame in a shot you want to use for the "keyframe" and you can "clone" the KCC (kinema color correction) file from one shot to another to touch up rather than need to start from scratch each time.

One bonus for starting with true RAW Bayer data is that it is about 1/4 as big as TIF files, so the conversion from the RAW data to workprint frames is much faster, I am getting more than 1fps using 2592x1102x12bit source frames, and with ACAM files being smaller it should go faster. On a better computer with RAID SSD you might get 5fps for the workprint de-Bayer.

Once you make the PIX frames (size depends on the computer speed used for editing) you can start editing, then later do the full de-Bayer cc on the full size images as needed.

You don't need to do shot by shot CC on the workprint to start, you can just re-use a KCC file for the K, ISO, sensor that worked before, and do the shot by shot cc at a later time, my NLE will automaticly conform the edits to both the TIF full frames and workprint PIX frames. The PIX frames for every frame in the project stay on disk so you can open the CAD program to edit without needing to "load" anything if the PIX are done before or over night.

I developed my DI system before CinemaDNG and my file system works well for making feature type films where you need 12 resolutions of each frame and 512 audio tracks per shot, so I am just going to import other formats into my system.

Anyway, by mid-summer I should have RAW workflow working faster, I have in interest in this because I will be using the programs myself.

The programs are not as "interactive" as most people would like, maybe, but if you get into the flow of the workflow and know what you want to end up with I think it should work out as far as I know, they are in "beta test".

A2) The conversion program can work with files in the NLE file directory structure, so under Windows you can have both programs open at the same time to go back and forth. The is a graphics display in the convertor that lets you see the test frames before you start conversion. The NLE is needed since that is where the CC interactive screens are, so you need to convert one frame to TIF to do the CC, the use that KCC file to finish the shot.
For TIF to TIF CC that is built into the NLE, the convertor is not needed for that, but may run faster.

The de-Bayer convertor is in 32bit code and uses about 2GB, while the NLE is 16 bit code and uses 16MB. For speed I needed to write the de-Bayer program outside the CAD program, so its a bit awkward for now, but it works with the cheep computer render farm concept. I might be able to port the de-Bayer program to UBUNTU at some point, and it can run under ME rescue boot floppy, so the render farm computers can be cheep like $200 computers.

Its not quite the mess it sounds, if you want to talk more you can PM or email me. The quality of the end results is what matters, not what the program looks like that gets you there.... A long team goal would be to update all the code using a better compiler, but there is a problem with the tools that I can use under windows so I don't know when that might happen, its over 300,000 lines of code about.
 
CCD

CCD

Have you ever seen a good image from the DII? They are using old skool sensors that just aren't very good.

It uses a CCD rather than CMOS.

The issue in not so much the sensor but the lenses used in the example shots.

Most C mount lenses, even megapixel ones, are not up to 3K resolution in all respects and that is a problem not just for ACAM but for Scarlet also.

To get the maximum the sensor can put out you need the right OLPF, a very good lens (not just an expensive one), and the right software.

RED has a time lead in their "de-Bayer" software development, so unless Acam can get software that works well they will have image issues. I have been working with the sample Acam raw data, but its hard to see how good the images can get due to the OLPF and lens issues. Some of the sample images look nice and usable with my processing, a bit different that the RGB samples on Adobe web site maybe.

Because true RAW data can be processed in many ways the camera has no basic look, its more a matter of exposure and processing, the sensor only is a smallish part of the end result for a given pixel size and count.

RED is a "closed" system so they can control the look of the end results, while "open" true RAW data cameras do not have that control over how the images are processed, so you can see bad results if its not done well...
 
NAB 2010

NAB 2010

Maybe they will get it together. Who knows? It's not there yet.

I just got a newsletter from Acam dII about their being in Vegas during NAB 2010 to show off their cameras and footage, they are asking for RSVP, so maybe you need to reg to see things?

Quote: [Welcome to the A-Cam dII preview at NAB 2010

We would like to invite you to a hands-on preview of the A-Cam dII at the Mirage Hotel in Las Vegas on the 13th of April 2010 between 4 PM and 8 PM.

We will show a limited number of pre-series A-Cam dII, newly shot footage, answer your questions and listen to your feedback.]
 
Back
Top