AZL November New Releases Part 1

[Scottl]

I think these are pretty amazing locomotives.  If I modeled in z, and was a steam era person, I'd be all over these!

[Sokramiketes]

Seriously....

And again we covered the CB&Q Mikado which is what appeared in the ad. See the below photo. Not all logos were centered.



And yes, I know the above photo it not an USRA Mikado... etc, etc. etc. Right or wrong. AZL has proven their point. Not all logos were centered. It is okay to admit this.

Rob

Uh, Rob, that logo is centered... Ignore the trucks, which arn't equidistant from the ends.   

[wcfn100]

This may require a lesson on how perspective works.


Jason

[Chris333]

Rob,
The point here is that nobody thinks you asked the factory to print them off center, we just think that's how you received them. It is/was a mistake somewhere along the line.

Plus most examples you showed are not USRA prototypes.

Anyways, yes they are nice locomotives and printing doesn't affect me.

[ztrack]

Mike… LOL. I would love to live in your world some times. The chute is the center point. I have photos of CB&Q tenders showing the logo over the chute. These I would call centered logos. The AZL logo matches the placement in relationship to the chute and rear truck. Same for the GN.

Chris, you are right. But this is not the factory as much as AZL making a decision to follow images that show off center logos on some tenders. Again is it right or wrong? We can show photos that illustrate both sides of the discussion. I think we can agree to disagree on this one since in some cases, both are accurate.

The fact is, these are phenomenal releases. They have taken Z to new levels. I plan to keep posting photos of the new releases as they are very relevant. It is important for folks to see the advances in Z and not get hung up on perceived imperfections.

Now please let the poor horse rest in peace. Look for our next Mikado announcement in about a week.

Rob

[peteski]

This may require a lesson on how perspective works.

+1000

[GaryHinshaw]

I should just keep my mouth shut, because I don't have a particular interest in this model one way or another, but I'm not sure I understand the definition of "centred" in this discussion.  Taking the Burlington photo as a concrete example, we can draw some simple lines and work out the effects of perspective:



The convergence of the red lines is quite minimal: the rear corner of the tender is 61 pixels tall and the front corner is 68, so the distance scale in the object plane only changes by 11% over the face of the tender.   In the image plane, the left edge of the logo is 69 pixels from the rear corner and the right edge is 88 pixels from the front corner.  If I make the extreme assumption that the distance scale in the rear section should be rescaled by 68/61 over its entire range, I get a corrected distance of 77 pixels, but this is an over-correction.  The right section should also be corrected upward, so it's true distance is over 88 pixels.

There is no way this particular logo is centred on that tender, at least given the way I have defined the ends of the tender.  What am I missing here?

[peteski]

I should also keep out of this since I don't have any stake in this model either.

While your calculations show the herald to be slightly towards the rear, the model's herald is placed much further towards the rear.  You know when something just doesn't look "right" to your eyes? Well, this is one of those instances.




Don't get me wrong - it looks like a really nice model. But the tender herald placement just looks off. This is TRW: We don't hold back here - we comment on all sorts of (even much lesser) model inaccuracies.

[Chris333]

What is missing is that isn't a USRA type locomotive or tender. A better comparision would be the GN 3200 photo because that is much closer to the models prototype.

[Chris333]

Funny thing is if the proto was offset and the model was centered we probably wouldn't care as much 

[wcfn100]

There is no way this particular logo is centred on that tender, at least given the way I have defined the ends of the tender.  What am I missing here?

Trying to do this on a perspective shot is futile.  It's hard enough straight on with just keystoning.  You don't know the lens or what type of lens distortion there may be. If there's any pincushioning going on, the rear of the tender will be taller than it should be in relation to the front which will flatten out your perspective lines and void any type of numbers you want to try and apply.

Besides all that, trying to define the 'corners' on the tender when they are founded off nearly impossible.  In the picture you marked up, I'd say the rear 'corner' is further back (the front one seems good to me).  I know there should be daylight between your line and the rounded tender edge.  And when using pixels as units, any error can make a big difference.

I will say after looking at all the proto pics, it did seem that much of it was slightly off to the rear from what I would think the 'center' would be.  But even with that, the AZL stuff doesn't look right.

Jason

[GaryHinshaw]

To demonstrate that I'm picking on everyone equally, I continue with my futile image analysis, this time on the GN units.  The model photo is straightforward because there is negligible perspective in the image:



The ratio of the front space to the rear space is 97.3/66.9 = 1.45, so the front space is 45% larger to within a few percent.  The prototype photo is trickier because of the perspective, but it is still a pretty undistorted image that is amenable to analysis:



This time I did a more careful integration of distance in the object plane (see the linked pdf file for my math).  I came up with a ratio of 63.4/52.4 = 1.21, so the front space is 21% larger.  Thus - as many have said - both logos are off centre, but the model logo is more so.

Take heart AZL, the only thing worse than being talked about is not being talked about.  This is still a very nice model, but I do have to agree with most that it would look better with the logo closer to centre.

[Sokramiketes]

I should just keep my mouth shut, because I don't have a particular interest in this model one way or another, but I'm not sure I understand the definition of "centred" in this discussion.  Taking the Burlington photo as a concrete example, we can draw some simple lines and work out the effects of perspective:



The convergence of the red lines is quite minimal: the rear corner of the tender is 61 pixels tall and the front corner is 68, so the distance scale in the object plane only changes by 11% over the face of the tender.   In the image plane, the left edge of the logo is 69 pixels from the rear corner and the right edge is 88 pixels from the front corner.  If I make the extreme assumption that the distance scale in the rear section should be rescaled by 68/61 over its entire range, I get a corrected distance of 77 pixels, but this is an over-correction.  The right section should also be corrected upward, so it's true distance is over 88 pixels.

There is no way this particular logo is centred on that tender, at least given the way I have defined the ends of the tender.  What am I missing here?

Gary- Why arn't the green lines vertical, or at least parallel to each other?

[GaryHinshaw]

I was just trying to trace the image.  It turns out that the front corner, as drawn, is exactly vertical, while the other 3 have their top end tilted one pixel to the left, an angle of about 1°.  This has a negligible effect on my conclusion.

FWIW, I went back and applied the perspective calculation from the GN image to this image and came up with the result that the front space is 20% larger than the rear (surprisingly close to the GN result).  I think the uncertainty in this estimate is a few percent.

-gfh

[ztrack]

Gary great job on this. I agree with your assessments. Heck you can't argue with the math! I will push this information to AZL for future runs. I appreciate putting real numbers behind the observations.

Rob