Difference in colour between 2012 and 2010 proc

Question

I know that newer 2012 process tries to save detail in the highlights (different tonal curve I guess), and has much better controls in Basic develop module so I was eager to edit my newest shots in LR4.
However, I wasn't so thrilled how it rendered bright pinkish colours.
Two examples below are cropped parts of a photograph, have default values, using Adobe Standard profile, from Canon7D raw file.
Image done in 2012 process, default values.
And this one is 2010:
It doesn't seem a big deal when viewed in the web browser on the white background, but the difference is very obvious when editing those  in LR. Notice that 2012 file has a "better" histogram with "safe" highlight zone.
However, it altered the colours as well and histogram shows that too. Pink blouse seem more washed out. I can assure you that process 2010 file is much more closer to the original scene, original colours.
So... can I do anything about this? Can it be corrected in final LR4 version? Sure I can raise the saturation in the pink channel, but that will raise other, well saturated, areas as well and I won't get the same result like in 2010 proc. file.

Answer 1

I've been testing the "2012 Process" using a Macbeth color checker image (raw). My first goal was to determine what settings in LR/2012 would best match CS5/ACR6/2010. I was not trying to preserve the 2010 process, just want to better understand the difference.
First pass just using the defaults in both versions resulted in differences of 3 to 5 points in the RGB values of the 24 patches of the colorchecker. On some patches, one of the RGB values might be the same, but at least one was different by 3 to 5 points. The 2012 version looked lighter overall, so I suspected that the difference was due to tones (as Jeff Schewe has suggested).
So by trial and error I kept reprocessing the LR4/2012 version, decreasing only the Blacks slider. When I got to a setting of Blacks -8, the black colorchecker patch finally matched all 3 RGB values. All the other patches were very close, with a few differences of only 1 point in RGB values, and usually in only 1 value (red or green or blue, but not all three)
However, there was 1 exception. The Blue value in the colorchecker's Yellow patch had a 7 point difference. All the other primary patches were exact matches.
My colorchecker test images were from a Canon 5D2, processed under the Adobe Standard profile, rendered as 16bit TIFs in Adobe98 colorspace, and stacked as Photoshop layers to read the RGB values.
So I suspect what many are perceiving as color differences are real differences, but are caused by the basic difference in tonal rendition between the 2012 and 2010 processes. For the OP, please reprocess your pinks in LR4 with some different black settings and see if you get closer to your original 2010 version. I'd be curious to hear the results.