CRI test and Mcree weighted results

Johnnycannaseed1

Well-Known Member
Because HPS contains relatively little blue and lots of orange, it will score quite high against the Mcree curve potentially better than some or all of the cob samples. I haven't digitized any bulb spectrums but I'm sure that according to Mcree HPS should work well despite not having a swoosh. It's more a matter of where that light is falling in the Mcree shoosh.
So why do red bias Leds fall down especially compared to an old skool "green yellow" hps... is that not contrary to McCree data?

Not knocking you for testing or anything like that, in fact I commend you, but what I am saying is the test was flawed from the start because it lacks consideration and is relying on/using data from another experiment, whose results do not fit with what is currently known and proveable:peace:
 
Last edited:

Johnnycannaseed1

Well-Known Member
If only something would actually go to your head. Ah well, it's good to see you finally made the step from COB bashing to trying it for yourself. Now if you have some actual experience under your belt, perhaps you'll understand that spectrum means very little and photons are king.
You are so far off base with this comment that it is not even funny... Spectrum affects quality yield and finishing time, methinks you have been doing too much reading and not enough growing lol!
 
Last edited:

Greengenes707

Well-Known Member
@Johnnycannaseed1 Then please show us some growing. What results do you have?
You act like you are important or know something. But everything you have written suggest otherwise. Just like when you were running around thinking citi substrates has something proprietary that can't be touched...but couldn't seem to figure out aluminum nitride, one of crees biggest licensing techs. More like not touched by your "research". You literally don't even understand what you are saying or what McCree did.

So about the leaf section that was used, not a solution in vitro...
Photosynthesis measurements (Fig. l)
A section of leaf approximately 25 mm square was cut with a razor blade and placed, within a few seconds, between two sheets of wet household cellulose sponge. The sandwich of leaf and sponge was then clamped between two sheets of aluminum and screwed into a block in the base of the assimilation chamber. The sponge dipped into a pool of water in the bottom of the chamber. The chamber was closed and the air supply was turned on. For the standard test, air from a cylinder of compressed air (breathing quality), containing 350 __+ 20/d per litre of CO2, was humidified by passing it over a saturated solution of NaC1 in boiled distilled water (75 ~ r.h.). The flow rate was 200 + 10 ml min-1, measured with a Brooks Sho-Rate rotameter. The volume of the chamber was 125 ml. The air in the chamber was stirred with a fan rotating at 3,000 r.p.m. The temperature of the air was 28 _+ 1 °C, and the temperature of the leaf, as measured with a thermocouple pressed to the back surface, was within 0.5°C of the air temperature. The effect of changing the temperature (11 °-38 °C) and CO2 concentration (200-600/~1 1-1) was determined on selected samples. The difference in CO2 concentration between incoming and outgoing air was measured with a differential infra-red gas analyser (Beckman 315A). It was less than 10/d 1-1, and was measured to ___ 0.2/d 1-1. Since both air streams were humidified, the water added by the leaf had little effect on the differential indicated by the analyser (less than 2 ~ of the differential). The area of sponge exposed to the air was kept to a minimum to reduce exchange of CO2 by the water in the sponge. As an additional precaution, boiled distilled water was used in the sponge, and both were kept over KOH when not in use. The photosynthetic rates of leaf sections treated in this manner were surprisingly reproducible. After a period of up to one hour of adaptation in moderate illumination, during which the stomata were presumably opening, the rate of CO2 uptake under constant test conditions remained steady (_ 5 ~) for several hours. (BARTOS et al., 1960; NATR, 1970).

Or how about that "monochromatic" light he used...
Light source (Fig.2)
Monochromatic light was obtained from a Bausch and Lomb High Intensity Monochromator, fitted with a grating which covered the range 350-800 nm, a xenon arc light source, quartz optics, and variable slits. Higher orders of diffracted light were blocked with Corning CS 0-54 (350-575 nm) and CS 3-69 (600-750 nm) glass filters. A quartz cuvette containing water was placed before the entrance slit to reduce the radiant heat load on the grating. The wavelength calibration was checked with the mercury lines from a fluorescent lamp and found to be correct 1 2 i i i 11 10 9 f 7 6 5 4 14 13 12 L, Fig.2. Schematic diagram of equipment. 1 = compressed air supply; 2 -= humidifier; 3 = flowmeter; 4 = xenon light source; 5 = waterfilter; 6 = entrance slit; 7 = grating monochromator; 8 = exit slit; 9 = leaf sample in assimilation chamber; 10 = thermopile and thermocouple; 11 = C02 analyser; 12 ~ integrating sphere; 13 = photomultiplier; 14 = photometer; 15 = leaf sample in sphere for absorptance measurements. Agric. Meteorol., 9 (1971/1932) 191-216 PHOTOSYNTHESIS IN CROP PLANTS 197 within __+ 3 nm. Stray light was checked with Corning CS 7-54, 3-69 and 3-73 filters and found to be less than 2 ~. The width of the entrance slit was set to 6 mm, and the width of the exit slit was varied, from 1.5 mm up to a maximum of 4 mm, which corresponds to a bandwidth of 25 nm, according to the manufacturer's specifications. The irradiance at the leaf was found to be proportional to the slit width, at all wavelengths. The monochromator produced a spot of light about 22 mm in diameter, and this was masked down to 20 ___ 1 mm. The irradiance at the leaf surface was measured with an Eppley air thermopile, 6 mm square, which could be moved forward to the position normally occupied by the leaf. The irradiance was uniform to + 5 ~. The output of the thermopile was measured on a Hewlitt Packard DC Micro Volt Ammeter, Model 425A, calibrated to ___ 2 ~o with a Leeds and Northrup Millivolt Potentiometer, Model 8686.
http://www.inda-gro.com/IG/sites/default/files/pdf/ACTION-SPECTRUM-KJMCCREE.pdf


More to help you get it right. Real basic shouldn't be hard...
https://www.heliospectra.com/sites/default/files/general/What light do plants need_5.pdf

Wait, there's more...
Screen Shot 2016-09-04 at 9.24.11 AM.jpg
http://photobiology.info/Gorton.html

More...
CONCLUSIONS
Differences in radiation quality from the six most common electric lamps have little effect on photosynthetic rate. Radiation quality primarily alters growth because of changes in branching or internode elongation, which change radiation absorption. Growth and yield in wheat appear to be insensitive to radiation quality. Growth and yield in soybeans can be slightly increased under high pressure sodium lamps compared to metal halide lamps, in spite of greatly reduced chlorophyll concentrations under HPS lamps. Daily integrated photosynthetic photon flux (mol m-2 d-1) most directly determines leaf anatomy and growth. Photosynthetic photon flux levels of 800 μmol m-2 s-1 are adequate to simulate field daily-integrated PPF levels for both short and long day plants, but plant canopies can benefit from much higher PPF levels.
http://biology.mcgill.ca/Phytotron/LightWkshp1994/1.5 Bugbee/Bugbee text.htm


Did you forget Inada?? Years after McCree coming to near identical findings, and referenced almost as much too...
http://pcp.oxfordjournals.org/content/17/2/355.abstract

More data supporting McCree and RQE...
http://onlinelibrary.wiley.com/store/10.1046/j.1365-3040.2000.00563.x/asset/j.1365-3040.2000.00563.x.pdf;jsessionid=DE30785B056CC75202E2E5A8B92B46D4.f02t04?v=1&t=isrnkhsc&s=324c775889c7dd9575ff5c6972435f2562869554
 

Sativied

Well-Known Member
It's not that the MCCree curve is wrong, it has been confirmed, with minor deviations, especially in the blue range... for many crops in recent more advanced measurements. However, RQE of a leaf does not provide a complete picture, not even close. It also doesn't support 3500k, not at all.

You are so far off base with this comment that it is not even funny...
That sums up the entire dutch grow forum nowadays quite well... so far off base, it's not even funny. And me 'knows' he hasn't done enough growing, still trying to initiate flowering with P lol... Just one of the first who used cobs in the dutch forum, so people ask him questions, which made him believe he actually has the answers too. Classic one-eyed king scenario.

The Chlorophyl chart is in their datasheet ..... so ... ehm .... yeah I connected that "dot".

If only something would actually go to your head. Ah well, it's good to see you finally made the step from COB bashing to trying it for yourself. Now if you have some actual experience under your belt, perhaps you'll understand that spectrum means very little and photons are king.

McCree obviously also shows that blue and red are most readily used by the plants since chlorophyl is responsible for the bulk of the photosynthesis, so indeed it makes sense in a greenhouse to supplement with that. Moving a much lower red peak by a few Nm to the longer end of the spectrum is not going to overcome a huge deficit in quantum yield from CRI90.
Yeah, that sums me up so well lol... :rolleyes: Fucking delusional you are with your "experience under your belt" while growing with neither 2700k or 3500k... I've been growing a lot longer than you have kid. Touching a cob won't change its output or change what is an obvious more optimal spectrum. The "huge deficit" is not in the cri90 but in your level of intellectual honesty. Once something has gone to your head it will get stuck there forever, I've seen that over and over in the dutch forums. Like your true NFT nonsense... Problem is the same as usual, what you think you know isn't actually so. It's only what you wish to be true. Repeating your conjecture isn't going to make it any more accurate. The dog you have in this fight is that you just don't want to realize you're not all-knowing. That's just messed up, unlike Rhaz you are not even making money being so goddamn ignorant.

And I'm not "trying" cobs fool. That comment alone shows you are missing more than a few points. I never bashed high cri warm cobs either. I think you would get along well with Churchhaze, similar simplistic strawman logic.

"photons are king"... unfortunately for you it's not that simple. There's no black-and-white choice between spectrum and intensity, that's just bibled nonsense you are parroting. An artifact from people here trying to beat hps, like 3500K itself.


You guys just keep talking nonsense and shit and cherrypicking charts... pretending you understand shit about plants... will make 2700K 90 CRI (with a little extra red) harvest time even more enjoyable.

group.jpg (at 30% of 280W average per flower-hour)
 

Sativied

Well-Known Member
Because HPS contains relatively little blue and lots of orange, it will score quite high against the Mcree curve potentially better than some or all of the cob samples.
Funny how selective you are repeating the facts I posted... oh but that wasn't new to you was it :lol: The most common response I got when I laid out over several threads the higher cri warmer cobs are much better for horticulture...it's not all new to you all.... Showing so nicely what the problem is with you guys, more concerned with pretending how much you know, not getting the fact any expert that ever lived and will ever live does the opposite frequently. Try to reach a point where you realize how little you know and how much you still have to learn...

There are several photobiologists active on for example reddit, if you guys had any desire to "know", you would have contacted one long ago (who will tell you warm white and red for flowering for far more reasons than your silly quest for efficiency). Instead of still trying to pretend 3500K was chosen for any other reason than being 1000 lumens more, one bin higher, in the datasheet. I can probably still find a post where someone decided to go for 3500K instead of 3000K because what difference could 500K make right... Drop the confirmation bias, start from scratch, you too can design led lights with an optimal spectrum for flowering cannabis.
 

Uberknot

Well-Known Member
I don't know all about that.....I do know I researched before I built and added a 2700k because I wanted it. :P

The 3500k's I changed too because it was much cheaper, But I still kept the 2700k even though it cost me more...lol!

I was going to try a mix of 6500k-5000k, 3500k-4000k and 2700k. was tough deciding..

At this point I am going to use what I have and watch all these tests people do with interest.
 

Rahz

Well-Known Member
test was flawed from the start because it lacks consideration and is relying on/using data from another experiment, whose results do not fit with what is currently known and proveable
I think we all know there's no substitute for growing a plant. The math portion of the test set out to answer the question "What will the par readings for these 5 samples be, and how will the readings change when weighted against the Mcree data?"

That's it, and I was successful in those goals.
 

Shugglet

Well-Known Member
I think we all know there's no substitute for growing a plant. The math portion of the test set out to answer the question "What will the par readings for these 5 samples be, and how will the readings change when weighted against the Mcree data?"

That's it, and I was successful in those goals.
Not trying to over-extrapolate your data is a keen move ;)

It can be used to help make better educated hypotheses that will still need tested to gain any knowledge that can be readily applied to real world functions.
 

Rahz

Well-Known Member
Funny how selective you are repeating the facts I posted
That quote wasn't a reply to you. I called you out for suggesting this test was designed to make 3500K 80 CRI look good when it obviously didn't. You accused me of "reeking of bias". Seriously dude.

concerned with pretending how much you know
Yea, because people who know everything have no need to conduct tests. How do you come up with this stuff?

2700K 90CRI has the best spectrum according to the Mcree curve. Part of the question I wanted to examine using the Mcree data is how much the phosphor loss would negate the improvement... according to the Mcree curve. If you have a better data set for determining that, post it. Otherwise chill out and wait for the rest of the test to complete.
 

Rahz

Well-Known Member
So why do red bias Leds fall down especially compared to an old skool "green yellow" hps... is that not contrary to McCree data?
The numbers are what they are. I haven't profiled HPS against Mcree, but them doing well is likely supported by Mcree. I will throw a HPS spectrum into the mix and we'll see what the Mcree conversion factor is.

Important to keep in mind that cobs became popular once they started outputting more light per watt than the alternatives.
 

wietefras

Well-Known Member
You guys just keep talking nonsense and shit and cherrypicking charts... pretending you understand shit about plants... will make 2700K 90 CRI (with a little extra red) harvest time even more enjoyable.
That you even think that proves anything is even more laughable than the hysterical case of cognitive dissonance you keep displaying.

Sativied: I saw on Rollitup that 90 CRI is king. Everybody switch to this. I have never tried it and actually I have never even grown a single plant under COBs at all, but listen to me! I read a few posts recently so I know everything!!!
WF: There is actually much less light coming from 90 CRI for the same wattage.
Sativied: You parroting idiot, don't be delusional you are looking at lumens, lumens are for humans!!!!!!
WF: Ehm no, I'm looking at PAR. 9% more light from 80 CRI vs 90 CRI
Satvied: You are delusional, you one eyed king! PAR doesn't count of course! Pfft. 90 CRI also has tons of light above 700Nm. Tons of light I say! That;s where it hammers 80 CRI and you stoepid PAR doesn't measure that!
WF: Well actually it's only a few percent of the total amount of light and, as McCree shows, the plants do very little with that. So that shitty bit is only between 40% and 0% efficient so say 20% average. So divide that shitty bit of extra light by 5 and you are left with pretty much only what's in PAR range. That's why everybody uses that. In fact actual large scale grows have shown spectrum means quite little for overall yield. Quantity of light is what's king. Within reason of course.
Satvied: You cherry picking idiot, you don't know what you are talking about and neither does McCree. That shitty 1% extra McCree corrected light should be doubled because of the Emerson effect! That will produce huge yields. The spectrum is superior. I KNOW this! I will show you all when I finish my uncontrolled experiment with only one test set and no prior COB experience to my name.
WF: Ehm, but that's utterly useless since there is no comparison... Ah you know what, who cares. I've had enough of a laugh with this clown.
 

Johnnycannaseed1

Well-Known Member
@Johnnycannaseed1 Then please show us some growing. What results do you have?
You act like you are important or know something. But everything you have written suggest otherwise. Just like when you were running around thinking citi substrates has something proprietary that can't be touched...but couldn't seem to figure out aluminum nitride, one of crees biggest licensing techs. More like not touched by your "research". You literally don't even understand what you are saying or what McCree did.
Greengene think you need to lay off the haterade lol.

And wtf are you on about with regards to Citi they came up with ali substrates first... don't believe me then go ask them. fuck didn't they even invent the COB (but don't quote me on it lol)???

Judging from your post here seems like you have been rifling through mine, hmmm wonder why, seems people usually hate what they cannot understand or what they wish to be ....which camp do you fall into?

...or is it because you have been bigging up Cree for the past however long and now that people are on the Citi tip it's burst your Cree monopoly bubble and whatever deal you had going with them? ....have your sales been dropping is that the issue here??? Whatever the reason the Citi chips are king and offer way better value so get over it and quit whining.

As to your lame comment saying I do not understnd McCree experiment ROFL ...and just because you said it doesn't make it true, my posts on the matter (which you have no doubt been sniffing around) speak for themselves.

With regards to results I have done canna experiments in the past but did not record them, but if you are cool with tomato results I will make a thread which backs up what I have been saying including pics and timelapse, it's up to you?
 
Last edited:

ttystikk

Well-Known Member
The numbers are what they are. I haven't profiled HPS against Mcree, but them doing well is likely supported by Mcree. I will throw a HPS spectrum into the mix and we'll see what the Mcree conversion factor is.

Important to keep in mind that cobs became popular once they started outputting more light per watt than the alternatives.
I just heard this morning that DE HPS emits a fair amount of UVB. If true, it would go a long way towards explaining their performance in spite of their apparent spectral deficiencies.
 

Johnnycannaseed1

Well-Known Member
That sums up the entire dutch grow forum nowadays quite well... so far off base, it's not even funny. And me 'knows' he hasn't done enough growing, still trying to initiate flowering with P lol... Just one of the first who used cobs in the dutch forum, so people ask him questions, which made him believe he actually has the answers too. Classic one-eyed king scenario.
It isn't funny but your comment sure makes it seem funny Lmao:lol:
 
Last edited:

optzulu

Well-Known Member
I think its more a photon vs spectrum camp.
Funny fact the members of the photon camp have allot of real time hard data to present. Data about Weed!
While the spectrum camp is still figuring out what spectrum to choose and have allot of theory data. About tomato and Cucumis sativus(cucumber)
 

PhotonFUD

Well-Known Member
I think its more a photon vs spectrum camp.
Funny fact the members of the photon camp have allot of real time hard data to present. Data about Weed!
While the spectrum camp is still figuring out what spectrum to choose and have allot of theory data. About tomato and Cucumis sativus(cucumber)

The camps aren't mutually exclusive. Both are important variables.

Spectrum - What is the best spectrum and why? What do plants like best? <--- quality
Photon Camp - How many photons can we generate as efficiently as possible? What amount will saturate a plant? <--- quantity


An optimal spectrum allows the plant to be efficient. It enables any adaptations that take advantage of environmental factors affecting light by different plant species. Using a spectrum that supports the photosynthetic process most effectively will improve efficiency.

Providing a photon amount that saturates the absorption capacity of the plant provides the highest potential for photosynthetic output. That output provides resources for the plant. More photons increases growth.

As many photons at the optimal spectrum is the best of all worlds and should be the goal for all gardeners.
 

Johnnycannaseed1

Well-Known Member
I think its more a photon vs spectrum camp.
Funny fact the members of the photon camp have allot of real time hard data to present. Data about Weed!
While the spectrum camp is still figuring out what spectrum to choose and have allot of theory data. About tomato and Cucumis sativus(cucumber)
What hard data is this?

I have grown plenty canna back when I was a heavy smoker and I have tested different spectrums and I can tell you "Fact" spectrum makes a difference to the quality and finishing time.

If you had been paying attention in your own grow room then you would have noticed that canna grown under a white spectrum generally produces better quality bud than canna under Hps (for starters).

As I have stated earlier photon and spectrum go hand in hand, I have done the tests and in fact I will do them again and this time round I will document the evidence just to prove the point:wink:
 

wietefras

Well-Known Member
The camps aren't mutually exclusive. Both are important variables.
The point is in this quote from Greengenes:
CONCLUSIONS
Differences in radiation quality from the six most common electric lamps have little effect on photosynthetic rate. Radiation quality primarily alters growth because of changes in branching or internode elongation, which change radiation absorption.
http://biology.mcgill.ca/Phytotron/LightWkshp1994/1.5 Bugbee/Bugbee text.htm[/QUOTE]
In the Netherlands there have been several of these tests done with the same outcome. Spectral distribution hardly had any effect on yield. It affects internode length and number of leafs and such, but hardly the yield.

So that's the scientific reason to go for higher PAR values over slightly better spectral distributions.

Of course if PAR efficiency (and price) would be the same between two COBs, you'd choose the best spectrum. However, due to Stokes losses, the PAR levels for high CRI and warmer color temperatures will generally be less than on lower CRI and/or colder color temperatures.

Then again, this needs to be within reason, because if the morphology gets too bad it will start to affect yields. Which is why we don't flower under 5000K even if that would put even more light on the plants. But going 500K up from 3000K is not that big a deal and the extra 10% of light is.
 

PhotonFUD

Well-Known Member
McCree Curve represents a real world validation that the photosynthetic process is affected differently by each wavelength in the PAR spectrum. It supports the microbiology research findings explaining photosynthesis functions at a molecular level.

Photosynthesis uses 8 700nm photons for each reaction. Plants must condition absorbed photons of shorter wavelengths prior to the photosynthetic process. This allows plants to use all available light in the PAR spectrum.

Wavelengths closer to 700nm require less plant resources to condition and can be correlated to the bump in the red region on the McCree curve.

There is a lot more research out there that can be referenced especially that into photosynthesis. Much more still needs to be done especially since we have better technology allowing us to produce and measure light more precisely.
 
Top