CMOS vs. 3 chip

[Philip Brubaker]

Does a CMOS sensor look any better than a 3 chip camera?

[Boyd Ostroff]

The V1 and FX7 actually do have three chips. However they are CMOS chips instead of CCD's like the other Sony cameras use. The HVR-A1 and HC-3 use a single CMOS chip however.

If you do a search on CMOS in this forum you will find lots of discussion.

[Paulo Teixeira]

http://dvinfo.net/conf/showthread.php?t=86558

[John Bosco Jr.]

Let's start with the basics. Charge-Couple Device (CCD) is old technology that has enjoyed the top seat for sensors in cameras and other imagers for the past 20 years. In a CCD sensor, the charges on the first row are transferred to a read out register. From there, the signals are then fed to an amplifier. Once the row has been read, its charges on the read-out register row are deleted, the next row enters the read-out register, and all of the rows above march down one row. The charges on each row are "coupled" to those on the row above so when one moves down, the next moves down to fill its old space. In this way, each row can be read—one row at a time. The CCD shifts one whole row at a time into the readout register. The readout register then shifts one pixel at a time to the output amplifier.

Complimentary Metal Oxide semiconductor (CMOS) is a fairly old technology as well. In a CMOS sensor, each pixel has its own charge-to-voltage conversion. When introduced, CMOS was inferior to CCDs for higher end imaging because of higher noise and less sensitivity but have recently made a comeback primarily because of HDTV. Let me explain.

When video cameras expanded its resolution from standard definition to high definition at the necessary frame rates, it has been reported that the S/N ratio of CCD-based cameras degrades ~3 dB per octave (1). For example, a professional broadcast camera S/N ratio under standard illumination of 2000 lux at f/8 is at least 62 dB for NTSC (700x520), but degrades to 54 dB for HDTV (1920x1080).

Because of newer technology, CMOS devices for HDTV offer lower noise and better resolution while using around 10X lower power to directly generate 12-bit video. Thus, it is now possible for the CMOS sensor to migrate from the consumer market to the professional market. Bear in mind, though, most professional cameras still use CCDs as the sensor.

Part of this new technology for CMOS-based imagers is the use of pixel-based amplifiers in the imaging system-on-chip architecture. The pixel-based amplification can appropriately set the signal bandwidth.

Today’s CCD cameras offer the operator the choice of adjusting the image to trade lower noise vs. higher resolution. This option is not necessary with CMOS-based cameras because the noise-setting bandwidth is at the pixel amplifier rather than the video output amplifier. CMOS-based sensors reduce noise while eliminating the classic noise vs. resolution trade-off because the pixel-based amplifier’s bandwidth better matches the imager sampling frequency. The CMOS output buffer’s noise is usually negligible.

The past complaints of CMOS sensors have been sensitivity and fixed noise; the latter has all ready been addressed. The sensitivity of today’s best CMOS sensors now matches the sensitivity of comparable CCD devices since the amount of light collected at each pixel is now roughly equal. This is achieved because of a new technology for CMOS sensors called deep submicron processing. Deep submicron refers to the lithographic dimension where the smallest electronic circuit feature is less than 0.25 microns.

So based on the above, today's CMOS sensors have the technology to offer fundamental performance advantages including lower noise and higher sensitivity over CCDs. In addition, CMOS sensors offer other advantages over CCDs like lower power consumption, more compact, virtually no vertical smear, and, in my opinion the best sensor alternative for producing High Definition video.


1. K. Mitani, M. Sugawara and F. Okano, “Experimental Ultrahigh-Definition Color Camera System with Three 8M pixel CCDs,” SMPTE Journal, April 2002

[Stelios Christofides]

John that was an excellent description and explanation.

Stelios

[Brett Sherman]

I'm doubtful that the 1/4" CMOS sensor in the V1 is any less noisy than the 1/3" CCD sensors in the Z1 or Canon line. In fact, most people claim that the V1 is more noisy at low light than 1/3" CCD cameras. I don't know if it has to do with the size of the senors or some other reason. But it appears that lower noise is not an advantage with the V1. It could be growing pains, maybe the next generation of CMOS cameras will be better.

[Nigel Davey]

Originally Posted by Brett Sherman

I'm doubtful that the 1/4" CMOS sensor in the V1 is any less noisy than the 1/3" CCD sensors in the Z1 or Canon line. In fact, most people claim that the V1 is more noisy at low light than 1/3" CCD cameras. I don't know if it has to do with the size of the senors or some other reason. But it appears that lower noise is not an advantage with the V1. It could be growing pains, maybe the next generation of CMOS cameras will be better.

IMO it is noticably better (gain noise) when you sit it along side a PD170, as I did recently. True the 170 is about two stops faster and so can provide better exposure in the very low light conditions. However until the V1 runs out of iris and gain it fairs better as the gain is added. Since I don't have the Z1 or Canon A1 I can't comment on this side of things.