Alien Conquest Quality

Rainstorm26 · May 2011

Anyone know where the Alien Conquest sets are made. Are we relegated to the cheap Chinese plastic again?

georgebjones · May 2011

I have bought a couple sets and there are a couple of the softer plastic, but for the most part, they are high quality, hard ABS.

Andrewdash · May 2011

The sets are made in Denmark, but there are numerous lego production facilities around the world

Huw · May 2011

AC sets, and all other *proper* sets, for that matter, are made in Denmark or other LEGO-owned factory. The alien heads are made in China but are of a high quality.

Extended lines like collectable minifigs, magnets and keychains are made in China by a 3rd party and tend to be rubbish but with proper, boxed, sets, you're safe.

LEGO_Nabii · May 2011

You're safe where ever LEGO is made, it's the same plastic in the same molds with the same colourant and the same quality demands. There were a few issues with some early mini-fig line parts, this is solved.

There has never been any 'cheap Chinese plastic' the issue was with colour quality and mold polishing, the phrase annoys me, why link 'cheap' and 'Chinese', no one says cheap Mexican, or cheap Danish when there is an issue with plastic parts from those locations. The reason LEGO has a factory in China is capacity, we needed more, they had more, cost was never the reason.

As for Alien Conquest, yes some parts are China made, try to pick out which ones.

Huw · May 2011

^ IMO, series 4 is as bad as series 1 in terms of translucency of plastic, surface finish etc. of the legs and torsoes in particular.

Chinese does not automatically mean cheap and nasty, of course, I agree, as it's all down to the quality control and materials used which can be the same wherever in the world things are made.

From my observations, the parts made in China are those with -B01 underlined (or some other letter/number) at the end of the mould number. Mabye B=Broadway? The only AC ones I've noticed so far are the alien heads (J01) which are perfectly fine, but then they do not appear to be ABS.

bmwlego · May 2011

While looking at the Build a Minifig bar at the LEGO Store I came across Series 1 CMF parts (no news here as they've been there for a bit now) but the stark contrast in the color of the CMF heads when seen alongside regular LEGO heads was easily noticeable.

^^I guess this would fall under the label of a color quality issue. I hadn't noticed the difference in color when looking only at my CMFs but mixed in with regular LEGO parts the difference in color screams at you.

yys4u · May 2011

@bmwlego I completely agree! I've been saying for a long time that the color of the heads and hands are for me the most disappointing thing. I noticed this even more with the Dragon Knights Kingdoms battle pack, the yellow is HORRIBLE!

bmwlego · May 2011

^I'm sure that there are countless others who have noticed this and I wonder if this is something that bothers LEGO as well.

LEGO_Nabii · May 2011

According to the LEGO Right Quality guy (Bjarke Schønwandt) at the Skærbæk LEGO fan event in Denmark last year the problem is standard production yellow has slightly drifted towards orange and the China factory slightly towards green, though both are still within LEGO tolerances when brought together the difference is blatantly obvious. (I’ve tried comparing with 20 and 30 year old hands and I can’t see the orange shift myself, but I’m told it is at least as bad as the green shift in some of the China produced hands. (Human eyes are not great at seeing reds I guess))
Apparently this is being realigned but we produced tens of millions of yellow hands and it’ll take some time to use them all up.

The translucency of parts is normally not due to the plastic (which is identical) but the amount of colourant added, I’m not sure if a better colourant is being used that requires less to be added and produces the translucency, if the mix is not right, or something else. I’ll ask next time I get a chance, though unless it’s made public like the stuff above was I might not be able to post anything about it!

bmwlego · May 2011

^Thank you for that info. I'm sure others here appreciate the information and insights you provide from a LEGO designer point of view and experience. It is obvious that you care deeply about what you do for a living and that you balance walking the line between fan and employee.

jwsmart · May 2011

When I first read this thread, I picked up a S3 minifig, a Space Police 3 minifig, and a Unitron minifig (1995). Under the fairly poor fluorescent light, the difference between the S3 and the Unitron was much more noticeable than the difference between the SP3 and the Unitron.

When I get a chance, I'll do a bit more scientific comparison, like take a picture of all three heads next to each other in natural light. (if the sun ever comes out while I'm not at work)

jwsmart · May 2011

OK, I've had a chance to take some pictures, and have some interesting results:
Here's the setup:

And the composite image:

From Left to right, the heads are as follows:
Solid Stud head: http://www.bricklink.com/catalogItem.asp?P=3626ap01
Unitron Head: http://www.bricklink.com/catalogItem.asp?P=3626bpx157
SP3 Head: http://www.bricklink.com/catalogItem.asp?P=3626bpb358
CMF Cyborg Head: http://www.bricklink.com/catalogItem.asp?P=3626bpb506

I've used the GIMP color picker tool to take the average RGB values for a 50px by 50px square, directly under the bevel on each head.

-Head Description-	-Year Produced-	-% Red-	-% Green-	-% Blue-
Solid Stud	1991	90	72	10
Unitron	1994	91	72	10
SP III	2010	91	71	11
CMF	2010	89	71	14

So, there we have it. It looks like the Broadway manufactured heads are a bit more blue, and a bit less red than the TLG made heads of 2010, and both of them have diverged from the solid stud heads from 1991. Unfortunately, since all of my LEGO heads that are older than about 1990 are second hand, and usually part of larger lots, I don't have anything that I'm sure is from the late 70's, early 80's...

ThisIsMyCup · May 2011

i think the best example of the poor color quality coming from china is the girl lifeguard in series 2 i think. there are 5 different shades of yellow present. the hair is platinum yellow, the hands, the head, the legs are all different shades...and the the yellow printing on the torso. the torso is the absolute worst. i dont see why they didnt cast it in yellow and did red printing instead.

jgadget · May 2011

^ My version of that figure looks pretty good to me.
I can't see much, if any, variation in the yellow. Certainly no more than the variation I've seen in yellow bricks.

As for the printing on the torso, if they printed red over the yellow, then she'd have nothing holding her outfit on at the sides.
I don't recall seeing minifigs with printing on the sides of the torso, except for the coating on the chromed versions, and we know how well that works.

bmwlego · May 2011

^^^wow Jwsmart! That is a good way of quantifying the color differences.

brickmatic · May 2011

Although, guys, keep in mind that lighting affects perceived color. Remember that the color you see if a result of what light is reflected by the object. So if your light source has a color cast, it can affect the perceived color. Also keep in mind the human brain compensates for the color cast of the lighting source.

Example, take a photo under florescent lighting and you'll often get a very yellow looking picture, even if the colors in the room when you take the picture appear mostly normal to you.

@jwsmart Although your pictures provide a rough measure of the color differences, unless you corrected for the color cast of the light source and the color cast of your camera, you did not capture actual color values of the objects. This is what controlling for color spaces and color calibration are all about, in order to get accurate color information.

Also, as humans we're more sensitive to certain colors in context. Green casts on skin make you look sickly and unnatural. Red casts on skin make you look blushed, but still rather natural. I'm not surprised that the orange shift would be less noticed than the green shift on a minifig.

madforLEGO · May 2011

i think the best example of the poor color quality coming from china is the girl lifeguard in series 2 i think. there are 5 different shades of yellow present. the hair is platinum yellow, the hands, the head, the legs are all different shades...and the the yellow printing on the torso. the torso is the absolute worst. i dont see why they didnt cast it in yellow and did red printing instead.

Maybe she had a tan from being out in the sun... Seriously, maybe it is my eyes.. maybe the photos (Thank you for taking the time jwsmart) but I do not see the difference.. If anything the 2 second head from the left seems to be darker than the rest... that may have been a lighting or shading issue though

jwsmart · May 2011

...your pictures provide a rough measure of the color differences...

Thanks, that's exactly what I was setting out to do, and I think my ground-breakingly non-specific conclusions are completely supported by my non-accurate (but reasonably precise) color measurements.

@madforlegos - after taking the photos, I didn't see much of a difference either, which is why I broke out my favorite (least expensive) photoshop replacement to see if my eyes were playing tricks on me.

brickmatic · May 2011

^ Lol. I was just pointing this out because some people who read your results might confuse your precision with accuracy. Although, I must ask, what's the point of precision without accuracy?

Also, I hate to point this out but not all of your conclusions are completely supported by your measurements. You say both of them have diverged from the solid stud heads from 1991. However, your data seems to indicate that only the Broadway head has. The other measurements are pretty much in agreement with one another.

jwsmart · May 2011

Uh, precision allows for determining relative differences quickly and easily. I didn't have to accurately calibrate my colors because I don't care how closely the LEGO head matches "absolute yellow". Even if I did care, of what use would that information be to anyone who can't be as accurate in their measurements?

Also, are you reading the same numbers I am?
I mean, you could say that 90 is pretty much the same number as 91. I wouldn't, but you're more than welcome to - that's why I posted my data and analysis, so that if someone thought my conclusion was bogus, they could come to their own.

I would love to see someone who has some known heads from different time periods do the same comparison, as would I like to see someone trying to recreate my little experiment and seeing if their data supported the same conclusions about relative color. (You couldn't directly compare my photos with someone else' stuff, because of the already-discussed-to-death accuracy issue, but a repeat of the side-by-side comparison might be interesting.)

brickmatic · May 2011

Uh, precision allows for determining relative differences quickly and easily.

Precision is not what allows you to determine differences. Measurement itself, regardless of precision, lets you do that. More precise measurements simply detect finer differences, which isn't always necessary. Furthermore, your inaccuracy might skew the scale of the relative differences. In other words, you might think the relative difference is larger or smaller than it actually is. Also, the inaccuracy may be distributed unevenly through the color spectrum. In other words the biases on different color components might be different. So the relative difference between reds might be incomparable to the relative difference between greens, etc.

Also, are you reading the same numbers I am? I mean, you could say that 90 is pretty much the same number as 91.

Yes I am. I'm not saying 90 is pretty much the same as 91. They are different. I am saying that difference is likely insignificant. There will always be some measurement error. How precise is your camera in capturing color information (as opposed to GIMP measuring color information in a photo)? Oh, and how precise is your placement of the color picker in GIMP to the relative position of the head? I know you're averaging over an area, but even so sampling the center of the head in the photo will yield a different result than sampling to the left, for instance. So variances in the color over an area of the picture yield another source of measurement error. That, and there is an allowable and actual tolerance variance between heads within a production year. So there is also sampling error since you only sampled one head per production year. Oh, and let's not forget the possibility of color fade over time. That would introduce yet another source of error.

Considering all these potential sources of error, I'd say values within such a close range do not indicate significant differences thus indicating the color of the heads have not changed.

Enbric · May 2011

While I tend to agree with brickmatic in regard to the percent error ruling out one or two point differences in making a sound scientific judgement, I do respect the work from a photographic standpoint that you, jwsmart, have put into examining the colors. Producing such a well-lit, quality photo without color balancing (I'm assuming there's no post-processing, that would defeat the experiment) is no easy feat.

So until someone breaks out some real color calibration equipment, I think we should realize the intent of jwsmart's photo, that is to examine color differences on an amateur level, and not make real scientific conclusions.

Anyways, if I have some time on my hands anytime soon I'll take you up on your offer for another side-by-side test in my own setup. I'd be more inclined to make an opinion if we had more than one image to compare, as you pointed out.

jwsmart · May 2011

^ "(I'm assuming there's no post-processing, that would defeat the experiment)"

Uh, that's a bad assumption. I did correct for white levels in the photo - which is why the white background is white. We can add that to the list of potential sources for error.

Hey @brickmatic, would you just give it a rest? We disagree. I get that. You may be right. I get that too. Or it could be that the big pile of potential sources for error you supplied adds up to less than 1% on the silly color picker. In which case, I'm right.

The best way forward would be to take more pictures. It sounds like you could do a much better job of this than I could, so stop talking about what and awful job I've done and go do!

brickmatic · May 2011

@jwsmart Sorry if I came across a bit harsh; just think of it as peer review. Back in school, a proper lab report required section on methodology as well as a list of potential sources of error in the conclusion. This is because it is easy to get bad results that at first glance look good. There are lots of papers published in scientific journals that are crap experiments because the authors made a small mistake here or there, like not perfectly randomizing their test subjects into the treatment and control groups.

I applaud your effort in trying to get a measured difference in color. I think two improvements are needed in the experimental design. 1) The camera needs to be color calibrated: "The camera calibration needs a known calibration target to be photographed and the resulting output from the camera to be converted to color values. A correction profile can then be built using the difference between the camera result values and the known reference values." 2) The sample size need to be greater: "The sample size is an important feature of any empirical study in which the goal is to make inferences about a population from a sample." Just to be clear, originally there was one sample (head) per test condition (production year).

That said, I'm not in a position to do the experiment. Despite my grasp of the theoretical considerations, I don't have the equipment or the practical experience to properly set up the experimental conditions I outlined above.

Enbric · May 2011

I've recreated the project using my own setup, here are the results...

Minifigs from left to right:
2009 Male Surfer from 7639
1999 Dutch Vander from 7150
2010 Cheerleader from CM1

All the minifigs are new and have been kept out of sunlight or extended indoor light. RGB was taken using the color sampler tool for a 25 pixel area square in between the eyes of the heads.

In brief, my setup was a Canon DSLR mounted on a tripod with a bounced speedlite, a Tamron macro lens, and two diffused fluorescent lights on either side. I've spared the real details since, as we have mentioned, I'm not writing a lab report here.

No post-processing (except for crop)

209 172 22 - Male Surfer
208 173 21 - Dutch Vander
199 170 27 - Cheerleader

Post-processing (Photoshop CS5)

255 210 13 - Male Surfer
255 213 12 - Dutch Vander
253 207 17 - Cheerleader

My numbers varied more but the results were similar to jwsmart, the collectible minifig was more blue and less red. I similarly don't have older minifig heads, but I still think it's interesting that we have similar results.

I won't bother listing sources of error, I think we've covered that. If I had more time, I would mix up the position of the figs, because in both mine and jwsmart's trials the CF was on the left...

Enbric · May 2011

I've uploaded the pictures to Flickr instead seeing as they don't appear to be working here...

No processing

Processing

brickmatic · May 2011

You know, there is an easier way. Just look at the LEGO. Does it look more blue to you? More orange? I don't see the need to measure the differences. And if you do see a need to measure the differences, you should do so properly. This whole process is like using a ruler made out of rubber and reporting the measurements to the one hundredth inch.

Enbric · May 2011

I understand this isn't really accurate, I already agreed with you on that. No one is going to really publish any of the work put forth here, it's just for fun. Obviously if there was something seriously wrong with the color of the CFs, it wouldn't have made it into a mass public release.

Clearly none of us have the equipment to "do so properly." I don't see a real need to measure the differences, I don't really care what color my cheerleader is as long as it's pretty much yellow, which it is. However, it's still hard for me to say it isn't interesting that two people conducting the experiment differently came across similar ends.

As for whether those similar ends have any meaning to a kid or an adult collector, the answer is no.

And that's not to say I disagree with the fact that more trials and more gear would be required for anything to really be said, but whose really saying anything anyway. I've done plenty of lab reports and I have no intention of writing another on this.

I don't mean to argue and I don't want to either, take whatever you want from the pictures. If they don't mean anything to you, that's fine. They don't mean much to me, either.

jwsmart · May 2011

Eh, I see those numbers and read "confirmation". I thought the idea that @LEGO_Nabii mentioned, specifically that both were still in tolerance, but had drifted in opposite directions was interesting, and wanted to see if I could objectively measure it.
I think I did that, and @Enbric did too. I've satisfied my curiosity.

Besides, I've always liked rubber rulers. I tend to injure myself much less whilst using them.

brickmatic · May 2011

^ It's almost as if you were saying your numbers confirmed that there were color drifts in opposite directions.

Alien Conquest Quality

Comments

Shopping at LEGO.com or Amazon?

Howdy, Stranger!

Quick Links

Categories