Machine Vision standards have evolved providing defined models of how industrial cameras communicate to a PC allowing easier implementation of machine vision technology. Vision systems can be made up of cameras, frame grabbers and vision libraries from various manufacturers. The vision standards provides compatibility between the various manufacturers for easy implementation.
Machine vision applications require some basic tasks of finding and connecting to the cameras, configuring parameters, acquiring images and dealing with events to and from the cameras.
In order to provide cameras from various manufacturers to work together with 3rd party software and hardware from other manufacturers and provide the tasks above, a standard must be followed. “GenICam” is the basis for this standardization, providing compatibility using a Generic Transportation layer and Generic Application programming interface. These are referred to as “GenTL” and “GenAPI” respectively. GenTL provides the communication layer and GenAPI enables camera features to be configured by analyzing a compliant XML file for the camera.
Camera manufacturers however provide unique independent features providing various advantages from one to another. Creating these unique features blur the lines of the standard, not always making a camera fully compatible with another manufacturers software. For example, an industrial camera may use the GenTL layer to be recognized but may have special features making it unique as well.
While not an exact figure, we would estimate that about half our client’s problems with machine vision camera connections, dropped frames, etc. comes back to a cabling issue. This is especially true for USB and GigE cameras.
In most of these cases, the issue is that the user is using a poor/low quality cable that was not made for the high speed and/or long distance demands of the application. Most of the inexpensive cables available via mail order are not made for use in high speed highly reliable data transfer applications. If your phone isn’t transferring at the full USB3 bandwidth, you normally don’t care. You probably don’t even know. But when you purchase a high speed USB3 camera and you can’t achieve its full frame rate, or you achieve it intermittently, this becomes a big issue.
This is the reason 1stVision offers ‘machine vision/industrial’ USB3 and GigE cables. These cables are tested to specs, come with screw locks to prevent the connectors from falling out, use larger gauge wire, are over molding and have die cast aluminum shells. They are designed to be twisted and bent (somewhat) and are industrial!
Watch this 1-minute video to understand what we mean by “Industrial”
Signal amplitude (the voltage of the signal in the cable) is a function of distance and frequency for cables. For instance, Ethernet is specified to 100 meters. So your cable should work when each device is 100 meters away. However, without the proper cable, you will not maintain the full 1000 Mbits/s data transfer rate! You might only be getting 50% of the speed depending upon the distance without a high quality cable.
Finally, consider the cost if your machine vision camera is part of an instrument or product that is being sold to your clients. We see far too many clients who try to save $30 on the cable only to find out that it is costing them thousands of dollars to trouble shoot a problem that can be easily solved with the proper part. Not to mention the cost to their client when the system isn’t working, and a hit to their reputation of not building a reliable system.
Here is our advice:
If you are in an industrial setting, you are compromising the reliability and robustness of your system if you are not using an ‘industrial cable’. Even if you are not operating at maximum speed of the camera, you should have these cables. BTW, these cables are not that much more expensive mail order cables. They are in the 10’s of dollars, but not in the 1’s of dollars.
If you are using USB3 cables, you should really be using ‘industrial’ cables. Current ‘inexpensive’ USB3 cables are not reliable at over 2M, and only 1M for USB C connector types. If you are using USB3 specifically to get the higher speeds from this protocol, then you absolutely need to be using ‘industrial’ cables. Inexpensive cables are not reliable for high speed data transmission.
If you are in a lab environment, with the cable never moving, and only going a short distance, then a high quality ‘inexpensive’ Cat 6e cable will work. There is a difference between inexpensive Ethernet cables. The one that came with the security camera all folder up is NOT what you should use. A reputable mail order cable vendor selling high quality patch cables is OK.
These latest Nano models offer 2.4 MP (1936 x 1216) resolution with a GigE interface in color and monochrome with up to 102 frames per second utilizing TurboDrive.
What’s so interesting about the Nano M1950 and C1950 models?
2.4 MP resolution with the speed of the popular IMX174, but at the price of the IMX249:
Sony Pregius image sensors in a given resolution has created paired sensors, one being faster at a higher price and one slower at a lower price. The Nano M1940 / C1940 cameras use the IMX174 which is a great sensor and historically had the fastest speed at 2.4MP in GigE, but at a premium. We could opt for the Nano M1920 / C1920 cameras with the IMX249 at a lower price, but sacrificed speed.
Until now! – The latest Nano M1950 / C1950 models with the IMX392 provides the higher speed of the M1940 / C1940 cameras, but at the lower price of the Nano M1920 / C1920 cameras.
2.4MP resolution using a 1 /2 in sensor format, provides cost savings on lenses.
Thanks to the Sony Pregius Gen 2 pixel architecture, the pixel size is 3.45um, allowing the same resolution and eliminating the added cost of larger format lenses found in the IMX174 / IMX249 sensors which were 1 / 1.2″ formats.
When would you use the Sony Pregius IMX392 versus the IMX174 and IMX249 sensors?
The Sony Pregius IMX174 / IMX249 images still have an incredible dynamic range due to the pixel architecture found in the first generation image sensors. (Read more here on Gen 1 vs Gen 2). If you need dynamic range, with large well depths of 30Ke-, then use the IMX174 / IM249 sensors.
I’m so confused! Where can I get the specs on the new Nano M1950 / C1950, understand what sensors are in what cameras and get a quote?
The tough part today, is that there a ton of model #’s in the Sony Pregius sensors lineup and in turn camera product lines. Here’s a brief table to help with links to spec’s, related image sensors and a link to get a quote.
Machine Vision cameras continue to reap the benefits of the latest CMOS image sensor technology since Sony announced the discontinuation of CCD’s. We have been testing and comparing various sensors over the years and have always come back to using Sony Pregius sensors when dynamic range and sensitivity is needed.
If you’ve been watching the technology shift from CCD’s to CMOS, you have probably also seen a ton of new image sensor names within the “Generations”. Honestly, its hard to keep track of all of Sony’s part numbers, so we will try to give you some insight into the progression of Sony’s Pregius image sensors used in industrial machine vision cameras.
1 – First, how can I tell if its a Sony Pregius generation sensor?
2 – What is the differences in the “Generations” of Sony Pregius Image sensors?
Sony Pregius Generation 1: This primarily consisted of a 2.4MP resolution sensor with 5.86um pixels BUT had a well depth (saturation capacity) of 30Ke- and still unique in this regard within the generations. Sony also brought the new generations to the market with “slow” and “fast” versions of the sensors at two different price points. In this case, the IMX174 and IMX249 were incorporated into industrial machine vision cameras providing two levels of performance. Example being Dalsa Nano M1940 (52 fps) using IMX174 vs Dalsa Nano M1920 (39 fps) using IMX249, but the IMX249 is 40% less in price.
Sony Pregius Generation 2: The primary goal of Sony was to expand the portfolio of Pregius sensors which consists of VGA to 12 MP image sensors. However, the pixel size decreased to 3.45um along with well depth to ~ 10Ke-, but noise also decreased! The smaller pixels allowed smaller format lenses to be used saving overall system cost. However this became more taxing on lens resolution being able to resolve the 3.45um pixels! In general it offered a great family of image sensors and in turn an abundance of machine vision industrial cameras at lower cost than CCD’s with better performance.
Sony Pregius Generation 3: Sony’s has taken the best of both the Gen 1 and Gen 2 to create Gen 3! The pixel size increased to 4.5um increasing the well depth to 20Ke-! This generation has the fastest data rates, dynamic range and lowest noise. The family will expand from VGA to 7.1MP as well. We are just starting to see Gen 3 sensors in our machine vision camera lineup and expecting more to come through 2019+.
Sony Pregius image sensor Comparison Chart
Sony’s Pregius Product line up for machine vision industrial cameras below for reference as of April 2019
How can we help answer your questions about industrial imaging? We have used several terms you may not be familiar with and would like to help in your education. Below are some related blogs that will help. Or contact us via email or phone number as we love to help educate our customers!