跳到主要內容

Aptina/OnSemi MT9M034 vs Sony IMX224

from webastro forum ( in french), with help google translate:

Hi planetary imaging fans,;)

I've long dreamed about it, ZWO did it! The very promising sensor Sony IMX224 MC 1.27 Mpixels

Has finally found its place in an astro camera and I was lucky to be able to test in preview. Thanks Sam!

Before these first tests I was still a little skeptical, the announcement made by Sony concerning the IMX224 was quite thundering. And between marketing and reality, there are disillusionment ...

http://image-sensors-world.blogspot.fr/2014/10/sony-announces-high-sensitivity-sensor.html

After my first tests, there I must recognize that they are very strong at Sony! The IMX224 is keeping its promises, as announced: it is very fast, ultra-sensitive and the noise is really of a very low level. But I would come back.

The camera takes up the same aluminum case as the ZWO ASI 120/174. And Point Gray Blackfly is tiny to the side (and heats up too, the small boxes have unfortunately
cons ...).


http://www.webastro.net/upload/images/1359-1434579863.jpg


I did several sets of comparative tests between the new UPS 224 MC color, UPS 120 MM-S to Monochrome Sensor APTINA MT9M034 and Point Gray Blackfly BFLYPGE-14S2C Sensor Sony IMX104 color. The 3 sensors have pixels of 3.75 microns, so it makes it easier to
comparison.
 
It seems a bit unfair to compare a monochrome sensor to sensors
Color, a priori markedly less sensitive but the results are surprising ...

First note, the USB 3.0 output of the ZWO ASI224MC requires a PC to the height to absorb all the data flow, at least if it is used at full speed. With an SSD,
I get with FireCapture on 2-minute captures:: be:
152 images / sec in 1296 x 960
299images / sec in 640x480
In comparison the Point Gray ceiling at 60 images / sec, even windowed.


Then, the UPS is supplied with a clear glass in front of the sensor, therefore without IR-cut. And filter
With this sensor is a real advantage considering its potential. For classic RGB, it is necessary to add a luminance filter (Ir-UV Cut).
 
Equipped with an infrared filter Ir685 or an Ir742, it transforms into an infrared monochrome camera.
It can also be used without complex with a CH4 filter.)


Here are comparisons in images that make it possible to measure the respective sensitivities of each sensor. The gain is zero, 0 dB, to minimize the influence of noise.
 
The results are truly impressive:

1- In RGB, the IMX224 color is almost as sensitive as the Aptina monochrome and 2 times more sensitive than the IMX104 color! Who says better ?

1- Infra-red with Ir685 filter, this time the IMX224 is almost twice as sensitive as the monochrome Aptina !!

Sorry for the poor pictures of Saturn with the Newton T300 but the seeing was catastrophic when I made these catches between 2 clouds. I dropped the RGB, which looked like boiled (remains the tablecloth of the table ...). Ir happily limits the break. Note on brutes
In Ir the satellite above and to the right of Saturn and which easily emerges with the IMX224 then
That one can hardly guess with the Aptina in spite of a brighter image. And in spite of
Exposure almost 2 times longer with the Aptina!





http://www.webastro.net/upload/images/1359-1434579957.jpg


http://www.webastro.net/upload/images/1359-1434580097.jpg
 


http://www.webastro.net/upload/images/1359-1434580176.jpg



Then a comparison to gauge the noise. I tried to do the measurements under Isis with dark / offsets / flats but the results are weird (:? :) so I came back to the visual, more evocative. These are darks of 30 ms with 36 dB gain, the maximum for Aptina.

Aptina is very 'dirty' compared to the other two: noise line pattern more on vertical than horizontal. The IMX104 is much more presentable with a lightweight frame horizontal. Finally the IMX224 puts everyone in agreement, no frame, it's 'clean'.

 
http://www.webastro.net/upload/images/1359-1434580343.jpg


In the end, a small test run on Saturn at 100 images / second with 49dB of Gain! (The IMX104
Is limited to 40.5 dB and only 36 dB for the Aptina MT9M034)
It's very grainy I know, a little like a good old Toucam, but it freezes
Much better the very strong turbulence. And then in 240 seconds we have 24000 brutes! : B:
http://www.webastro.net/upload/images/1359-1434580419.jpg


There is also the post of Emil, who released a very honest M57 color with poses of only 100 ms!

http://www.cloudynights.com/topic/503061-new-asi224-camera-with-extremely-low-read-noise/


Sony does not seem to have a monochrome version of the IMX224 / 225
More than damage because it holds an exceptional sensor which 'pushes' without restraint the noisy Aptina and the oldest Sony sensors like the IMX035. : Confused:


LONG INSTALLATION:
------------------------------

I took exactly the same tinkering as this one to cool the camera, at least the camera envelope:

http://www.webastro.net/forum/showthread.php?t=120197&highlight=refroidissement+zwo+asi


http://www.webastro.net/upload/images/1359-1409999177.jpg

 
Here are darks with 10 seconds of installation, a gain of 30dB with the ZWO ASI224 MC.

I'm trying desperately to find a solution to recover 16-bit color images. For lack of better, I have to settle for the 8 bits.
I tested camera not cooled, cooled without insulating cover and finally cooled with insulating cover:;)
 
http://www.webastro.net/upload/images/1359-1434730630.jpg


If you use the infrared camera with filter, 2 important tips:

- oversampling the image so that the finest detail is sufficiently easy to extract of the frame. Personally I use a FD of about 30 (i.e. twice the resolving F / D of a monochrome sensor with pixels of 3.75 microns).

- an excellent 'truc seen on the Yahoo group FireCapture: display the debug image in FireCapture (the image is logically very red), then adjust the sliders 'Wblue' and 'Wred'
 
To obtain an image that is closest to a neutral monochrome image, at least in theory.
 
With the Ir685 the best balance is obtained with a red dominant so do the tests with your setup.
Once in equilibrium, the contribution of the red, green and blue pixels is balanced at best and the frame reduced to a minimum, almost invisible.
 
Do not forget to uncheck the debugging option before launching the captures!
 
Do not debilitate during the treatment (IR  condensation in IR = loss of resolution)

Alberic

留言

這個網誌中的熱門文章

越南香草

Ngo ~ "N-gaw" Mui ~ "Moo-ee" Ngo ~ "N-gaw" Mui ~ "Moo-ee" Ngo ~ "N-gaw" Mui ~ "Moo-ee" Ngo (N-gaw) ,  Mui  (Moo-ee )  Cilantro Coriandrum sativum, chinese parsley  Ngo Gai (N-gaw guy), Mui Tau (Moo-ee Tao), Ngo Tau (N-gaw Tao)   Mexican Coriander,  Sawtooth Coriander, Cilantro Eryngium foetidum    娥女帝(拼音), 刺芹   特徵:娥女帝是短株形的植物,氣味清淡,葉邊呈鋸齒形,十分容易辨認。來源地:越南。 功效:和白夏差不多,娥女帝亦有祛濕、解毒及驅風的療效。建議食法: Pho,  (Bánh Xeò) 越南煎餅, 炒菜,湯,咖哩 Ngo Gai ~ "N-gaw guy" Mui Tau ~ "Moo-ee Tao" Ngo Tau ~ "N-gaw Tao" - See more at: http://vietworldkitchen.typepad.com/blog/vietnamese-herb-primer.html#sthash.I9rzkzwI.dpuf Rau Ram (Rau Rahm) Vietnam Coriander, Laksa Leaf, "Vietnamese mint(actually not a mint)". Polygonum leaf Polygonum odoratum Peppery, quite spicy. In salad, soup Hung (Hoong), , Hung Lang (Hoong Lang) Spearmint.  Vietnamese coriander Hung Lui (Hoong Lou-ee), Hung ...

沖田博文 Hirofumi Okita 60cm F3.25 dobsonian telescope

  the making mirror from Mike Lockwood webpage   youtube uwakina bokura other ATMers in Japan blueforest anettai  

劣質洗衣機入水喉

上面白色是最易找到,$2x. 但漏水. 灰色, $4x, 是假冒 "MADE IN ITALY"  假冒 "MADE IN ITALY"  的標緻  左面是白色膠喉的喉頭, 右面是灰色膠喉的喉頭, 上圖左面是真正 好貨 ( MADE IN ITALY )灰色膠蓋.右面是冒牌 白色膠蓋. 膠蓋在安裝扭緊時爆開  上圖左面是真正 好貨 , 標了其他規格.右面是冒牌, 單單印了 MADE IN ITALY  好貨的膠蓋是可以下移, 露出喉頭及黑色軟膠墊 黑色軟膠墊是有坑紋. 質感較柔軟. 緊後可以"迫實"水龍頭 及喉蓋, 沒有滲漏 正板 MADE IN ITALY 賣 $4x, 價錢絕對合理. 冒牌貨在旺角新填地街買的, 也是$4x. 真是要小心!!! NB: MADE IN ITALY 是否真正 意大利制造實在無從考 証