| Author | Message | ||
     David E Honigs (dhonigs) New member Username: dhonigs Post Number: 3 Registered: 4-2012 |
Hello friends, First, let me stipulate to the fact that whatever anyone sells is the best thing for doing whatever is under discussion. Next, I would like to tell a story. There was this man who told his wife he had to be out of town. The wife went shopping and came back home only to find her husband and his secretary in bed together. Flustered, the woman was trying to deal with what was right in front of her. "But, you are supposed to be out of town", was all the woman could say. Her quick thinking husband replied, I am out of town. Which are you going to believe, your husband or your eyes? In any of these issues about what does light do in a material, I have been taught to just look. One learns so much more by simply candling (looking through) a sample using red light or a near-infrared night vision system than guessing what is going on. When one looks through objects like a corn kernel it is obvious that like the linked image of the statue there are areas where the light comes through brightly and others that are quite dark. When one looks through a scattering medium like milk it is easy to put something black in the middle of the glass and see it doesn't change the transmission much at all. Light does not propagate through all parts of an object with equal density or distribution. In conclusion, I would recommend that someone simply look at what is going on and ignore what experts say. But then, who are you going to believe; Me or your own eyes? David Honigs | ||
| Gabi Levin (gabiruth) Senior Member Username: gabiruth Post Number: 68 Registered: 5-2009 |
Hi David, It is absolutely true that light in transmission through solids is not travelling in straight lines - it is also referred to as forward diffused transmission - but in the case of corn it is even helpful - as it helps to represent the volume of the seed even if the light "doesn't completely cover the whole dimension of the seed". True, corn is tough even in transmission, and this is why it is best done in motion with small size illuminating spot - allowing to average the whole volume while moving. Gabi Levin | ||
| David E Honigs (dhonigs) New member Username: dhonigs Post Number: 2 Registered: 4-2012 |
Dear Art, The truth is one. best regards, David | ||
| Art Springsteen (artspring) Senior Member Username: artspring Post Number: 46 Registered: 2-2003 |
Hi David, Emma Thompson in "Angels in America"? I never thought of light scattering in a corn kernel that way! | ||
| Art Springsteen (artspring) Senior Member Username: artspring Post Number: 45 Registered: 2-2003 |
Hi David, Emma Thompson is 'Angels in America"??? I never thought of corn scattering that way! | ||
| David E Honigs (dhonigs) New member Username: dhonigs Post Number: 1 Registered: 4-2012 |
Hello Friends, Corn is difficult to measure because the germ and other parts are quite different in composition. This is true in transmission or in reflection. Many people think that light goes straight through in transmission. The truth is that the light behaves more like this:http://joshwills.com/rendering/lucyss.png Many people solve this problem by measuring all the kernels on a single cob as they all have the same genetics. The other way to solve this problem is to collect a small pixel image in transmission or reflection and analyze that. | ||
| Lidia Esteve (veiva) Member Username: veiva Post Number: 12 Registered: 9-2006 |
Marijana, as William says you should contact a Perten representative and ask for it (i.e. David Honigs from the US division). You don't need to switch to transmittance mode, as I said and agreeing with Gabi, there are a lot of instrument characteristics that matter - and from what I have seen the measurement mode transmittance-reflectance is not the most relevant-. From what I have done and what I have been told from colleagues, Perten 7200 works fine for single kernel analysis.I attach a picture I took of the single kernel sample cup.  | ||
| William Wade (williamwade) Junior Member Username: williamwade Post Number: 6 Registered: 6-2010 |
"I have only two cups - the smaller is with a 75 cm diameter. How can I get those small ones? " Direct from Perten. They will likely let you try one out for a week before you choose to purchase if you ask. | ||
| Marijana Maslovaric (vidra) Member Username: vidra Post Number: 12 Registered: 5-2011 |
Thanks very much to everyone! I will take all of your comments into account. @Lidia Esteve, @William Wade: I have only two cups - the smaller is with a 75 cm diameter. How can I get those small ones? Is reflectance mode appropriate in this case, or I have to try transmittance too? But then I would need another instrument... | ||
| Gabi Levin (gabiruth) Senior Member Username: gabiruth Post Number: 67 Registered: 5-2009 |
Hi guys, Corn is one of the worst kernels to analyze on a single seed basis due to the very large degree of non uniformity of the constituents in the kernel. Almost all of the oil is located in a "pocket' on one side of the kernel. In reflectance the degree of success depends on how "deep" your light penetrates and 'retruns" to your collecting optics. This will vary with the size of the illuminating beam, the size of the kernel, etc. Based on experience with an instrument called Seed Meister, even with slow motion of the kernel under an illuminating beam which is small enough not to cover the whole kernel at once, and working in transmission, the results may vary depending on number of factors. The main advantage of the Seed Meister is that you can measure while they move, and analyze a large number of seeds - something like 15 per minute or even a little more. The measurement under motion gives you the ability to scan the whole kernel from start to end, averaging the spectrum from all different parts of it. As I said, even with all this it taakes a very careful work and most of all it also requires very good reference data on a signle kernel level, something that few labs can reall achieve. Gabi Levin | ||
| Lidia Esteve (veiva) Member Username: veiva Post Number: 11 Registered: 9-2006 |
Hello Marijana. Perten has a concave mirror single kernel sample cup for DA7200. I have not tested it for calibration myself for that instrument, but I know it works fine. I used it for discrimination with that device, and I was satisfied with the results. I know that some people use the instrument with the small sample cup or even other sample holders, and they get good calibrations too (sorry I don't have numbers from them, that's what I have been told). The trick is placing the seed always on the same spot on the cup (center). As mui says, you have to take into account few aspects if your seed is heterogeneous (the case of corn). You may have to decide which side you analyze. If you focus on oil in corn kernels, placing the germ up will work best. I do not agree 100% that transmittance would give you better results, because transmittance is very sensitive to sample positioning and light pathlength, so what you may gain in having light traveling across the whole sample, you loose it adding extra variability. From the literature, it is not clear that transmittance measurements work better than reflectance. I have not seen that myself either.It depends on a lot of other factors and instrument characteristics. So my advises to you would be: (1) place the kernel always on the same spot, the sample holder does not seem to matter so much (2) work on averaging several spectra to increase your signal and (3) according to my experience with other instruments, the spinning capability of Perten DA should improve your calibration accuracies. You can try scanning your samples with and without spinning, and if you are consistent on sample positioning, the spinning mode should give you better results. | ||
| Sirinnapa Saranwong (mui) Member Username: mui Post Number: 13 Registered: 10-2006 |
There are various instruments that allow NIR single kernel measurement. The most important thing is the mode of measurement, if the component of interest has different distribution between the external portion and the core of the kernel, then you will need to perform transmittance measurement. | ||
| William Wade (williamwade) New member Username: williamwade Post Number: 4 Registered: 6-2010 |
We have a DA7200, and although we do not use it that way, I know that it is possible. (Albeit much better to use larger samples.) There is a small cup that Perten offers that can handle 1-6 kernels and another one that is about 15-30. I cannot speak on the quality of the results however. | ||
| Marijana Maslovaric (vidra) Member Username: vidra Post Number: 11 Registered: 5-2011 |
Hello to everyone! I haven't stydied NIR for some time, my unit was broken, now I'm back after a long time. I read about single kernel analysis, maize samples. I'm wondering if there is a possibility to use Perten DA 7200 and other units which do not require milling for grains analysis, cause I would like to examine some maize hybrids (chemical composition) for the selection purpose. Regards, Marijana |