Topics

Monument Valley: Results


Dan Margulis
 

I’ve posted the results of our Monument study.

Reviewing: This is from the MIT study of 5,000 images and represents one of the most iconic scenes of the American west. The MIT study employed five intermediate retouchers to correct each image, and made their results public. Our entrants today include the one I judged did the best job, plus the averaged version of the five. I have never used this image in a class so I don’t know how well a more experienced group would have done.

We have a 30 entrants. When a person submitted two or versions, I chose the one I thought was better. Most people also submitted a list of their steps, thanks very much. I haven’t read these, because I’d rather get a sense of who was successful and who wasn’t before investigating why. 

The files don’t have people’s names on them, and were random-generator numbered from #801 to #830. As with past studies, we also have a “par” version, #831. To get it, I chose what I thought looked like the five best entrants, and averaged them, each one weighted 20%. This often creates a version that is superior to most if not all of its parents.

I’m going to have some things to say about this assortment, but as usual I’d like to open it up to group discussion first. What do the successful versions have in common? Meanwhile, if you’d like to know how your own version stacked up, download the par version and compare the two directly. Do you think you got the same kind of quality? If not, I hope you’ll find further discussion useful.

The folder is in the group Photos section, named Case Study: Monument Valley 
https://groups.io/g/colortheory/album?id=249516

Because some of us would like a closer look at these, I also have zipped all 28 and uploaded a 67 mb file to our Files section,
https://groups.io/g/colortheory/files/
Search for Monument-Valley_entries_062920.zip

I look forward to your comments.

Dan Margulis

P.S. The next case study is announced today, look for a separate post. The countdown continues, we have four case studies left, and this one could be the hardest of the four.


James Gray
 

I will stick my neck out.  I like 812, 824, and 829 the best.  I think that 812 is better than the par version.  I also like my version.  I am not saying what number it is yet.

James Gray

On Mon, Jul 6, 2020 at 5:22 AM Dan Margulis via groups.io <dmargulis=aol.com@groups.io> wrote:
I’ve posted the results of our Monument study.

Reviewing: This is from the MIT study of 5,000 images and represents one of the most iconic scenes of the American west. The MIT study employed five intermediate retouchers to correct each image, and made their results public. Our entrants today include the one I judged did the best job, plus the averaged version of the five. I have never used this image in a class so I don’t know how well a more experienced group would have done.

We have a 30 entrants. When a person submitted two or versions, I chose the one I thought was better. Most people also submitted a list of their steps, thanks very much. I haven’t read these, because I’d rather get a sense of who was successful and who wasn’t before investigating why. 

The files don’t have people’s names on them, and were random-generator numbered from #801 to #830. As with past studies, we also have a “par” version, #831. 


Gerald Bakker
 

This exercise is similar to the Cinque Terre image as it may again evoke discussions about "how colorful (or how contrasty) should it be".  I myself find the par version too much of both, but this is likely a matter of taste.

I found one of the crucial aspects of the successful versions a good color distinction between the foreground soil (orange) and the background rocks (red). Versions 811, 823 and 829 have this.
Another thing to consider is what to do with the harsh sun/shadow contrast. Should we keep/emphasize it or soften it to bring some detail back in the dark shadows? I prefer the latter, but after all there are not many shadow areas in the image, so the choice could be to keep them very dark and gain a bit of overall contrast.

Like James, I like my own version, but I am not afraid to tell which one it is: 810.
--
Gerald Bakker
http://geraldbakker.nl


John Furnes
 

I think to find the best colour is the most difficult. In my version – 819 – saturation has been overdone and it all became yellower than many of the others.

Also, I think, as this is in stark sunshine, the shadows would be darker.

If the PAR version was not there, I find #810, #811, #825 and #829 appealing to me.

I think the PAR version is a bit unrealistic, but it IS better than the rest.

 

If Dan would say something on what is most important in this image, and why – perhaps more than I would appreciate it.

 

John Furnes

 

 


Lee Varis
 

This image demonstrates the variety of personal preferences for how much color and contrast should be "forced" into the image! I prefer my own version (828) over just about every other version—I find the "par" version to be the best of the alternatives, but still too colorful, and contrasty! A lot of this is just personal preference. Many of the images posted look a bit too much like HDR renderings... Truthfully, after studying with Dan, I fell into the Lab saturation valley, and it took many years to climb out. Now with this image I may have swung a bit far the other way, but my version just looks more natural to me now, where 10 years ago I probably would serve up something more like 830 ! 


Rex Butcher
 

I submitted for the first time but how do I know which is my version?  My submitted file had the file name plus my initials, but the initials are not to be seen in the result numbering.
 
Toodle pip
 
Rex

--
What if the hokey cokey IS what it's all about??


Robert Wheeler
 

I see the tall mesa as the main subject of the image. The original has it very dull, with red areas not showing the variation in color our eyes would see and edge contrast being low. Good candidate for steepening the a* channel. When I did that, it helped the mesa, foreground dirt, and sky. Steepening the b* channel helped the sky a lot, it but made the rocks and dirt too orange and gave the plants a sickly off feeling. Sagebrush, even in bloom, tends to look dull grey rather than bright green and the yellow flowers would be partly washed out by the bright sun. I masked to b* change to affect only the sky. In retrospect, maybe I could have allowed the flowers on the plants to be a little more yellow (but not pushed so far as to distract from the mesa) and perhaps could have allowed the dirt to go a little more orange.

 

The foreground has stark shadows consistent with sun so bright I would need to squint even with sunglasses. That seems to call for a high L value and contrast. The white cumulus cloud on the right has detail to preserve. Going through all the entries and comparing the L values of the land (dirt centrally in several places and lighter part of mesa), I found a bright group with land L values in the high 70s to low 80s: 803, 807, 809, 810, 823. I find that aspect favorable. There also are some with L values in the low 60s to 50s: 801, 805, 808, 813, 814, 816, 819, 827, and 828. I found these uniformly less favorable to my eye. I found the exercise of looking at LAB values in each entry a good learning experience (luminance and color).

 

My entry is 803. I started with a darker sky action then converted to LAB and applied separate a* and b* channel steepening separately as described above. Instead of relying on an ersatz black channel, I made a black and white conversion in NIK Silver Effex Pro 2 to improve and reallocate luminance (and achieve greater local contrast in some areas without using forbidden sharpening methods). I added that as a layer in overlay mode at 15% opacity then auto toned and added a final curves layer with “auto” lending a tiny additional improvement. The sky ended up with blue areas so intense and dark that it was hard to look at the main subject, so my last edit was to add a layer with an earlier attempt that had a milder sky, masked to black, painted medium opacity white in various places to produce a better blended sky.

 

My favorites are 803, 807, 810, 823, 824, 829, and the PAR version. Will be interested to lean which images were combined for PAR.

 

Robert Wheeler


Robert Wheeler
 

Rex, I got caught with the same problem early on. Dan does shorten the submitted file name, but leaves some of what was submitted in place. When submitting, you can shorten the file name and add a personal identifier earlier in the string if you like. In my case, I ended up looking at the entries to narrow down the list of ones that could be mine, then opened mine in Photoshop, and added the candidates in a layer set to difference mode. The pair that ends up pure black (no difference) identified my entry. Tedious, but effective.

Robert Wheeler


Dan Margulis
 



On Jul 7, 2020, at 10:54 AM, Lee Varis <varis@...> wrote:

This image demonstrates the variety of personal preferences for how much color and contrast should be "forced" into the image! I prefer my own version (828) over just about every other version—I find the "par" version to be the best of the alternatives, but still too colorful, and contrasty! A lot of this is just personal preference. Many of the images posted look a bit too much like HDR renderings... Truthfully, after studying with Dan, I fell into the Lab saturation valley, and it took many years to climb out. Now with this image I may have swung a bit far the other way, but my version just looks more natural to me now, where 10 years ago I probably would serve up something more like 830 ! 

I agree with all this although I am not a big fan of #828. And I concur that the par version is overdone; I’ve added a new one called #831 conservative par for reasons explained below.

The first thing we notice about these results is that overall we seem to have done much better than in any previous case study except maybe Carnival, in the sense that very few of our entries are stinkers. Then again, it’s much harder to screw this image up than most of the others.

Why then is it part of the series, if it’s so easy? Well, for balance. The MIT set was 5,000 images and I chose 100 of them randomly. Why randomly? Because it would have been very easy to just cherry-pick ones that PPW would do well on and then declare to the world that PPW is the answer to everything except Covid-19.

Naturally, some of the 100 were tailor-made for PPW. Five to ten percent were large portraits, where MMM is very important for fleshtones. (I counted Veiled Bride as a medium-sized face). Another five to ten percent were scenes like this one. My LAB book was called the Canyon Conundrum because it is now well known that such scenes, which need lots of emphasis of subtle variations, give LAB a big advantage. And indeed, my PPW versions on this type of image were invariably much better than any of the MIT retouchers’ efforts.

Yours, too. The very best MIT result is #826. Their average is #801. We must have twenty versions better than either. Contrast that to the Colosseum study, where one of the MIT retouchers did roughly as well as any of us. But that’s as expected. The Colosseum exercise doesn’t feature anything that would give PPW an advantage: any sensible correction method works. The other case studies had some PPW-specific nuances. Cinque Terre, for example, had plenty of natural color variation so MMM wasn’t a big help, whereas in this Monument Valley scene it’s enormously useful to separate all the dull orange. Granted, H-K, which is not an everyday tool, was very helpful in Cinque Terre, but it is here as well.

So, this case study represents the images where PPW is at its most potent. Accepting that, this one for sure is all about personal preferences. As I discussed in the Chevreul book, when a scene contains elements that a camera doesn’t reproduce (like the vast scale of Monument Valley or Niagara Falls, or the clean air in one and the stinging mist in the other, or the silence in one and thunderous noise in the other) then it calls for enhancement of the simultaneous contrast effect. The only question is, how far to go. The person who did #830, which Lee seems to object to, admits that it was intended as “eye candy”; he’s famiiar with the scene and freely admits that his version is not realistic. Well, neither is Lee’s #828, for that matter. You wouldn’t see that much contrast in the real scene. It’s just a matter of degree.

This brings up the issue of the par version. With maybe fifteen really good possibilities, choosing only five is a hassle. And, I think, I was biased in favor of the most spectacular, so, yes, I chose #830 as one and a couple of others that were in the same ballpark. Consequently, the par version is rather loud.

So this morning I looked at the field again, and chose five different images to make a par from, but this time limiting myself to what I consider conservative treatments, my own included. So, that’s the “conservative par”. Personally I would split the difference between the two pars, but it certainly is to be expected that some people will prefer one version outright.

Dan


Robert Wheeler
 

In my opinion, the conservative PAR version represents a substantial improvement over the earlier PAR version, or at laest it moves well into territory fitting my personal preferences. Embryology and histology demonstrate the the retina is actually part of the brain rather than being a separate organ connected "to" the brain. The so-called optic "nerve" is like other within-brain tracts. So it is fair to assert that we literally see with our brains via a process affected by local physiology (e.g.: variations in receptor pigments and edge effects via retinal lateral connections) as well as by memory and expectations. No wonder that perceptions involve so much individual variation.

Robert Wheeler


James Gray
 

I am not so much afraid of telling which is mine.  I was wondering if anyone would praise it or pan it before I revealed which is mine.  It is interesting that those commenting have mentioned many they liked and some they disliked, but not mine.  My result was heavily influenced by my own photographs in Monument Valley as well as many other photos from Monument Valley taken by friends who were mostly shooting closer to sunrise or sunset.  I do tend to agree with Gerald that colorfulness, luminance contrast, and color variation may largely be a matter of taste.  However, I find the original too flat.  This may give it away, but I think color and luminance contrast would be the key to a good score in the competitions I enter often.

James Gray 

On Mon, Jul 6, 2020 at 2:02 PM Gerald Bakker <gc.bakker@...> wrote:
This exercise is similar to the Cinque Terre image as it may again evoke discussions about "how colorful (or how contrasty) should it be".  I myself find the par version too much of both, but this is likely a matter of taste.

I found one of the crucial aspects of the successful versions a good color distinction between the foreground soil (orange) and the background rocks (red). Versions 811, 823 and 829 have this.
Another thing to consider is what to do with the harsh sun/shadow contrast. Should we keep/emphasize it or soften it to bring some detail back in the dark shadows? I prefer the latter, but after all there are not many shadow areas in the image, so the choice could be to keep them very dark and gain a bit of overall contrast.

Like James, I like my own version, but I am not afraid to tell which one it is: 810.
--
Gerald Bakker
http://geraldbakker.nl


Gerald Bakker
 

On Tue, Jul 7, 2020 at 06:34 PM, Dan Margulis wrote:
This brings up the issue of the par version. With maybe fifteen really good possibilities, choosing only five is a hassle. And, I think, I was biased in favor of the most spectacular, so, yes, I chose #830 as one and a couple of others that were in the same ballpark. Consequently, the par version is rather loud.
 
So this morning I looked at the field again, and chose five different images to make a par from, but this time limiting myself to what I consider conservative treatments, my own included. So, that’s the “conservative par”. Personally I would split the difference between the two pars, but it certainly is to be expected that some people will prefer one version outright.
 
I definitely prefer the "conservative par" over the other one. I had the same idea to come up with an alternative par, constructed from my own 5 favorites. It would have been close to this one.
This new par happens to be rather close to my own version too (810). My rocks are less colorful but more pinkish, my sky is worse but my clouds are better.

Only now I notice that I completely missed the yellow flowers. These are worth emphasizing to give a nice touch to the dullish greens. One of the things that makes #811 so attractive.
--
Gerald Bakker
http://geraldbakker.nl


Thomas Hurd,MD
 

My entry was 827, and it was my personal favorite for eight hours. From midnight until 8 AM when I saw everyone else’s color.
I still learned a lot about scratching out contrast, especially in the absence of sharpening.
And although it appears very desaturated, if I place any of the more colorful versions and either blend them in color mode, or mine on theirs in luminosity it gets to be a pretty good image.
I had these objectives:
1. Bring out variation in the color of the butte face, so that the relative blueness of it was apparent.
2. Create contrast with the foreground and the sky.
3. Tone down the color in the foreground.
I tried to make the butte stand out more by making it overall darker than the sky and foreground. I blended the A channel onto the L channel in overlay mode. I got the result I wanted at the time, but I think I missed a chance for better contrast by not darkening the sky more.
I tried to tone down the greenery in front as well.
In the end, my submission needed better and more saturated color with the contrast.
A recurrent challenge I have is zeroing in on the appropriate color of the image. I knew the soil and rocks were reddish orange, but how far to take them down from there?
I enjoyed many of the entries: 
830 is a good blend of color, contrast and drama.  
828 and 829 are both enjoyable but are almost inverses if each other with 29’s contrast using a dark faced butte and 28 a lighter face. If you overlay either on the other at 50% the result seems identical.
824, 825 and 826 also excellent PPW versions
821 stands out with the spotlight effect on the butte.
817 and 818 are both good, but I like the sky in 817 better.
810, 811, and 812 are similar good colors with seemingly increasing saturation. But I think in 811 the yellow flowers drive the butte face more blue in my mind.
Like Gerald, I admired 811. I’d like to know the technique for bringing out that yellow. 
802-809 all seem to vary by taste without any glaring problem.
Both par versions are excellent to me and I favor the conservative version.
Thanks to all for sharing their images and their technique.

Tom Hurd

On Jul 7, 2020, at 2:18 PM, James Gray <James@...> wrote:


I am not so much afraid of telling which is mine.  I was wondering if anyone would praise it or pan it before I revealed which is mine.  It is interesting that those commenting have mentioned many they liked and some they disliked, but not mine.  My result was heavily influenced by my own photographs in Monument Valley as well as many other photos from Monument Valley taken by friends who were mostly shooting closer to sunrise or sunset.  I do tend to agree with Gerald that colorfulness, luminance contrast, and color variation may largely be a matter of taste.  However, I find the original too flat.  This may give it away, but I think color and luminance contrast would be the key to a good score in the competitions I enter often.

James Gray 

On Mon, Jul 6, 2020 at 2:02 PM Gerald Bakker <gc.bakker@...> wrote:
This exercise is similar to the Cinque Terre image as it may again evoke discussions about "how colorful (or how contrasty) should it be".  I myself find the par version too much of both, but this is likely a matter of taste.

I found one of the crucial aspects of the successful versions a good color distinction between the foreground soil (orange) and the background rocks (red). Versions 811, 823 and 829 have this.
Another thing to consider is what to do with the harsh sun/shadow contrast. Should we keep/emphasize it or soften it to bring some detail back in the dark shadows? I prefer the latter, but after all there are not many shadow areas in the image, so the choice could be to keep them very dark and gain a bit of overall contrast.

Like James, I like my own version, but I am not afraid to tell which one it is: 810.
--
Gerald Bakker
http://geraldbakker.nl


bill_iverson_washington
 

It's obviously a matter of taste how much to exaggerate colors (most of us tend to remember landscapes as having more intense colors than they actually did) and how much to separate colors to add drama.  I'm less certain about changing colors as a matter of taste or convenience.  The rabbitbrush plants that dominate this scene are a yellowish-green, not a dark bluish green (Google images of them).  Based on memory and several hundred images taken here in late August 2008, Monument Valley mesas and soil are not red; the a and b values tend to be about the same, with the b's slightly higher in probably the majority of cases.  I'm certainly not suggesting that those with redder mesas and soil, or bluish-green plants have sinned; this is also ultimately a matter of taste, I guess.  I'm also surprised that more people haven't commented on the sky.  This is not the most dramatic scene in Mounument Valley; some cloud definition helps the scene, but I found it mildly effortful (is that a word?) to work definition into the clouds while keeping the overall cloud feel very bright and light.  The sky itself should be a believable blue, but probably not an overly deep blue since this is a desert  at something close to midday.   In my submission, 825, I may have yielded to temptation in making the sky a bit too intense a blue.  Overall, I liked 802, 803, 807, 808, 809, 810, 811 (kudos for the only rabbitbrush plants with good yellow flowers), 824 and 825.  I found the par version of 831 I downloaded today too intense, but I'm not certain whether its the original or the conservative.


Thomas Hurd,MD
 

831 is the original. 832 is in photos (not the zoo file) and is the conservative version.

Tom

On Jul 7, 2020, at 3:34 PM, bill_iverson_washington <bill@...> wrote:

It's obviously a matter of taste how much to exaggerate colors (most of us tend to remember landscapes as having more intense colors than they actually did) and how much to separate colors to add drama.  I'm less certain about changing colors as a matter of taste or convenience.  The rabbitbrush plants that dominate this scene are a yellowish-green, not a dark bluish green (Google images of them).  Based on memory and several hundred images taken here in late August 2008, Monument Valley mesas and soil are not red; the a and b values tend to be about the same, with the b's slightly higher in probably the majority of cases.  I'm certainly not suggesting that those with redder mesas and soil, or bluish-green plants have sinned; this is also ultimately a matter of taste, I guess.  I'm also surprised that more people haven't commented on the sky.  This is not the most dramatic scene in Mounument Valley; some cloud definition helps the scene, but I found it mildly effortful (is that a word?) to work definition into the clouds while keeping the overall cloud feel very bright and light.  The sky itself should be a believable blue, but probably not an overly deep blue since this is a desert  at something close to midday.   In my submission, 825, I may have yielded to temptation in making the sky a bit too intense a blue.  Overall, I liked 802, 803, 807, 808, 809, 810, 811 (kudos for the only rabbitbrush plants with good yellow flowers), 824 and 825.  I found the par version of 831 I downloaded today too intense, but I'm not certain whether its the original or the conservative.


Thomas Hurd,MD
 

**not the zip file

On Jul 7, 2020, at 3:34 PM, bill_iverson_washington <bill@...> wrote:

It's obviously a matter of taste how much to exaggerate colors (most of us tend to remember landscapes as having more intense colors than they actually did) and how much to separate colors to add drama.  I'm less certain about changing colors as a matter of taste or convenience.  The rabbitbrush plants that dominate this scene are a yellowish-green, not a dark bluish green (Google images of them).  Based on memory and several hundred images taken here in late August 2008, Monument Valley mesas and soil are not red; the a and b values tend to be about the same, with the b's slightly higher in probably the majority of cases.  I'm certainly not suggesting that those with redder mesas and soil, or bluish-green plants have sinned; this is also ultimately a matter of taste, I guess.  I'm also surprised that more people haven't commented on the sky.  This is not the most dramatic scene in Mounument Valley; some cloud definition helps the scene, but I found it mildly effortful (is that a word?) to work definition into the clouds while keeping the overall cloud feel very bright and light.  The sky itself should be a believable blue, but probably not an overly deep blue since this is a desert  at something close to midday.   In my submission, 825, I may have yielded to temptation in making the sky a bit too intense a blue.  Overall, I liked 802, 803, 807, 808, 809, 810, 811 (kudos for the only rabbitbrush plants with good yellow flowers), 824 and 825.  I found the par version of 831 I downloaded today too intense, but I'm not certain whether its the original or the conservative.


jwlimages@...
 

On Tue, Jul 7, 2020 at 12:34 PM, bill_iverson_washington wrote:
It's obviously a matter of taste how much to exaggerate colors (most of us tend to remember landscapes as having more intense colors than they actually did) and how much to separate colors to add drama.  I'm less certain about changing colors as a matter of taste or convenience.
Agreed! I decided that for my entry I would allow some exaggeration of the color - I have shot a fair bit around southern Utah, and am used to seeing things in early morning or late afternoon light, so I rendered the color more like what I remember. Yes this was shot probably mid-day, sun high overhead - color should be much more "brownish" & contrast much flatter - but what's the fun in that?  ;-) I assume you are correct about the rabbitbrush plants as well, but I found that a bit of hue change made the reddish-browns appear even more vivid - maybe I took too much license there?

Looking through the group there are a lot of successful entries, with color ranging from probably faithful to fairly extreme like mine. A couple things I don't understand though - seems like many of these have skies that seem way too cyan-blue, lots of negative a* values even in areas that should be neutral-to-pinkish (slightly). It does make the brown of the dirt seem warmer, but doesn't look believable to me. And #830, mentioned above has a huge edge-darkening effect on the butte, which to my eye looks wildly unreal. A few others seem to have rendered the foreground lighter than the butte itself, which, again might be sort of realistic (from the high angle of the sun), but seems like a strange way to present a photo of the butte.

I think I prefer the original par image (no surprise there!) over the the "conservative" par one. Oh yeah, my entry is #821.

John Lund


Dan Margulis
 



On Jul 7, 2020, at 12:05 PM, Robert Wheeler <bwheeler350@...> wrote:

Rex, I got caught with the same problem early on. Dan does shorten the submitted file name, but leaves some of what was submitted in place. When submitting, you can shorten the file name and add a personal identifier earlier in the string if you like.

Good suggestion. I am real happy with the current situation where names are not revealed unless the individual decides to fess up publicly. So, I delete any part of the name that would serve to identify the source. Many members identify their files with their initials, which I delete. But something non-specific, like v5, or for that matter xxx, normally stays.

In my case, I ended up looking at the entries to narrow down the list of ones that could be mine, then opened mine in Photoshop, and added the candidates in a layer set to difference mode. The pair that ends up pure black (no difference) identified my entry. Tedious, but effective.

No need to go to all that effort. Just open all the suspects. With your own known original file active, simply Image: Apply Image one of the others, but don’t click OK—the preview will show instantaneously whether or not the images are identical, no need for Difference mode. If the first suspect isn’t a match, then change the dialog to the second, etc.

Dan


Gerald Bakker
 
Edited

On Wed, Jul 8, 2020 at 02:03 PM, Dan Margulis wrote:
No need to go to all that effort. Just open all the suspects. With your own known original file active, simply Image: Apply Image one of the others, but don’t click OK—the preview will show instantaneously whether or not the images are identical, no need for Difference mode. If the first suspect isn’t a match, then change the dialog to the second, etc.
 
I always identify my own submission by comparing it side-by-side with each entrant. Open a browser on one side of the monitor, a Photo viewer with my version (as on the local hard disk) on the other side and scroll through the list. Never any doubt which is mine. Only the carnival image was hard as the versions were so close to each other.
--
Gerald Bakker
http://geraldbakker.nl


Rick Gordon
 

What I've done for the last couple of weeks is to download all of the entries and:
  • Load them each into a separate layer with File > Scripts > Load Files into Stack.

  • Usually I convert the file to Lab at that point, since I find that most of my adjustments tend to be in Lab, and if I need another color space for something, I can create a smart object to do it in.

  • Locate mine and group it alone into a layer group, so that I can then I can add adjustments clipped to my own image.

  • By option-clicking between the various entries, I can see what I like from them, and then try to improve my own image, without affecting the others, since all of my adjustments are within a clipping mask.
Doing these exercises, and then trying an improved version after viewing all of the others as been a revelation — often humbling — but once having seen what cues I missed, the fixes are often surprisingly trivial.

My entry this time was 802, which got a better review than the previous disaster, but was much more improved after going through this process.

Rick Gordon

--------------------
On July 8, 2020 at 11:40:45 AM [-0700], Gerald Bakker wrote in an email entitled "Re: [colortheory] Monument Valley: Results":
I always identify my own submission by comparing it with each entrant. Open a browser on one side of the monitor, a Photo viewer with my version (as on the local hard disk) on the other side and scroll through the list. Never any doubt which is mine. Only the carnival image was hard as the versions were so close to each other.
___________________________________________
RICK GORDON
EMERALD VALLEY GRAPHICS AND CONSULTING
___________________________________________
WWW: http://www.shelterpub.com