21 replies to this topic

[mika]

Dear CartoTalkers,

I've just finished my latest project called 'The Ships Atlas' and I am looking for some feedback on the cartographic level it represents... But lets get started.

The atlas is meant to be a 'port finder' for people within the maritime industry. A potential reader (an office worker, ship's master, etc.) should be able to find a place he's looking for and get some basic information about it - port location, size, cargo handling facilities, drydock availability and airport availability. In addition, some ports have extra information in the index, but that's not a part of this presentation.

Maps were created using manifold and CorelDRAW with quite a bit of scripting in vba and vb.net to have as much as possible done automatically (or semiautomatically...)

The mapping part of the atlas is composed of 74 A3 plates incorporating following sections (just to warn those still using dial-up connections - files are heavyish, around 1.5MB each, so...):

1. Example of the Map Indexes:
 P2._Asia_and_Australia_and_Oceania_v3.jpg   1.06MB   267 downloads

2. Example of the Key to Map Symbols (an extract of the legend used for a bookmark):
 B1._Bookmark_Front.jpg   229.8K   286 downloads

3. Example of the thematic World Maps:
 07._Piracy_Incidents.jpg   1.43MB   237 downloads

4. Example of the Ocean Maps:
 13._Atlantic_and_Indian_Oceans.jpg   1.88MB   202 downloads

5. Example of the Political Maps (preceding each continent section):
 14._Europe___Political.jpg   1.66MB   183 downloads

6a. Example of the 'actual' atlas maps 1:
 67._Caribbean_Sea.jpg   1.57MB   213 downloads

6b. Example of the 'actual' atlas maps 2:
 68._Coastal_Areas_of_the_Caribbean.jpg   1.59MB   223 downloads

Although the atlas is already gone to printers, any comments you may have are highly appreciated. I would be great to know what carto professionals think about it and perhaps incorporate some of the bright ideas i may gain in the next edition.

Regards
Dominik

[MapMedia]

Mika, this is both fantastic and impressive. You must have enjoyed this project too.
Projections, symbology, and data seem fine. My only comment has to do with choice of colors (charts, and piracy incident color-ramp, or lack thereof, and intl border bands (only on shared border, not land/sea).
Thank you for sharing. This is a great collection. Where & when is it available?

[Hans van der Maarel]

Very impressive, Mika! I can't really comment on every map in detail (need to study them a bit more but I'm short on time) but here's 2 things that did come to mind when I looked at them:

• You're using a white outline to separate texts from the background. However, in some cases the white seems a bit 'too much'. I've experimented with using a partly transparent outline and that seemed to work well. The trick is to do it so that you don't see the outline, but do notice the effect it has
• I see some grid effects in the bathymetry on #4: faint light/white lines at 5x5 degree intervals. Looks like artifacts from the bathymetry gridding. Did you do that in Manifold? What did you use as source data?

[ELeFevre]

Mika,
These are beautiful maps! My only suggestion would be adjust (decrease) the outer-glow on the labels on the "actual atlas" pages. Particularily the glow/stroke behind the red ocean-labels creatd a blurry effect rather than making the label stand out. I'm sure I'll have additional comments when I get an extra minute or two to look more closely. Excellent work!

[frax]

Mika, looks beautiful!

I have a question for you - I noticed that you have used shaded relief on land, but not at sea - what was your decision for that based on? Land gets a bit dull without it, so I can see it, but I would have thought it would be useful, together with the hypsometric tints, to communicate the bathymetry.

[mika]

Dear all thanks for the answer! And just a few clarifications on the comments you have:

Chris - I did enjoy the project :-)
1. Colours on the 'Piracy Incidents' might be out of place as you say, though they looked fine on proofs so that's why I decided to go with them. The colour balance on the actual proofs was much better. They were better contrasted too. I might think of another colour scheme whenever I've got some ideas. At the moment, I must admit, I am quite exhausted and happy it's out of the way...
2. Water borders - To be honest, I didn't manage to find a decent data for the whole world and didn't have enough time to source and trace it myself. It's a 'to do' for the next edition though
3. And the availability... It should be on sale early December, so if you are interested in, please visit www.portinfo.co.uk then, to get some more info on pricing, shipping, etc.

Hans
1. Yes I use 'halo' effect on labels, so as you say they can be lifted-up optically. You also view the file in serious zoom, so perhaps that's why they look like they stand out too much. I think the appearance on the attached jpegs isn't the best due to compression and rgb colour scheme issues - the actual size of the 'halo' is 0.15mm and the colour is 10% Yellow (still whitish...) to tune them down a bit. They look fine on the wet proofs so I hope they will be fine after all. We'll see...
Your idea of having 'halos' semi-transparent is very interesting. I do have to give it a try whenever I've got a spare minute.
2. Yes, the data is indeed grided at 5°x5° interval. I had to do that so Corel could handle curves correctly - it just doesn't seem to like too complex curves at all (It will display curves made of around 15k nodes, but will have problems with filling them properly). The bathymetry lines where generated in Manifold from ETOPO2 and I think their 'quality' serves the purpose. Once again the griding effect luckily is not visible on the wet proofs so I believe I don't have to worry too much as yet.

Erin
1. As per above answer ;-) If you would like to see how it looks like in the real size and colour I may post a small true scale cmyk tiff. I'd be honoured if you were willing to waste some time on that, so please let me know if you wish to do so...

frax
1. Ah yes, shaded relief on seas... I did think about it but... Basically, the atlas is used in offices and on board, where the light availability might be a problem (especially on board), so I decided to opt out, as it could worsen the readability. In dark environment relief unfortunately looks like smudging...
Although I did give it a try, and my sample maps looked fine. I was just afraid of making them too 'sweet' you see :-)
If you look at the bathymetric scale, it gives an impression as it's extremely detailed, but in fact after printing you get almost smooth colour change along the scale. It's cheating I have to admit - making users 'believe' it's detailed while they cannot visually extract the data. Therefore in this particular case I treat bathymetry strictly as background only.

Regards
Dominik

[Hans van der Maarel]

Hello Dominik,

2. Yes, the data is indeed grided at 5°x5° interval. I had to do that so Corel could handle curves correctly - it just doesn't seem to like too complex curves at all (It will display curves made of around 15k nodes, but will have problems with filling them properly). The bathymetry lines where generated in Manifold from ETOPO2 and I think their 'quality' serves the purpose. Once again the griding effect luckily is not visible on the wet proofs so I believe I don't have to worry too much as yet.

Okay, if it's not visible on the printing proofs, you're probabely okay.

I'm a bit confused though. Did you use a raster background or vector? From your description it appears to be vector, but possibly raster could give you better (visual) performance.

[gp1]

Hi Dominik, an atlas is a huge undertaking and I admire anyone who takes on such a task.

A few comments:

- the red country labels are difficult on the eyes (well, mine at least)
- what data was used for the relief?
- what data was used for inland water (such as the Aral Sea which seems to be dated) (and other lakes are missing fills)
- maybe someone can suggest an alternate labeling scheme for the red map coordinates and the lat/long labels; they seem to be very tight in spots
- and how did you approach the issues of language & diacritics? (for example: Maracaibo Lake on the same map spread as Golfo de Coro)...I presume diacritics were disregarded for this product

[mika]

Hans:

I'm a bit confused though. Did you use a raster background or vector? From your description it appears to be vector, but possibly raster could give you better (visual) performance.

Well, I didn't think about using a raster background to be honest, though I may consider doing so in the future. It's just that using vector layers seemed a bit more sensible to me:
1. Easier to change a projection - at least in manifold 6.5 a raster displayed in a different projection on a map would export as hundreds of parallel lines. You can always change it's projection first, and that's what I did for the relief rasters, though it's pretty time consuming and labour intensive
2. Easier to handle - I would have to generate quite a few rasters for the maps as opposed to using only one bathymetry vector,
3. It's certainly easier to change colours in a vector document than in a raster document. I used some scripting for applying colours, so in a case I wanted to change them all it would be a matter of slight code adjustments and batch processing of the required files

So I think that's what made me go for vector.

gp1:

- the red country labels are difficult on the eyes (well, mine at least)

Well, yes, they might be, you're not the first one to say that I am afraid ... It looks like I'll have to take that seriously and think of some alternative colours before I post some examples of maps for the next edition... It won't be tomorrow luckily :-)

- what data was used for the relief?

bathymetry - ETOPO2'
world maps - SRTM30'' down sampled to 2'
ocean maps - as above
political - SRTM 30''
Standard maps - depending on scale SRTM 30'' to SRTM 3''

- what data was used for inland water (such as the Aral Sea which seems to be dated) (and other lakes are missing fills)

WDBII for most of the maps, updated where applicable, and for the largest scale maps VMAP0
World and political maps - lakes and coastline were traced on ETOPO2, rivers are from WDBII

The lakes without fills - could you please show me where? They might be really pale, but as far as I am concerned all should be closed and filled. I agree, the Aral sea could be less generalized on some maps as it's almost split now

- maybe someone can suggest an alternate labelling scheme for the red map coordinates and the lat/long labels; they seem to be very tight in spots

They are indeed. I believe I could drop some of the grid labels to make it clearer. Should help

- and how did you approach the issues of language & diacritics? (for example: Maracaibo Lake on the same map spread as Golfo de Coro)...I presume diacritics were disregarded for this product

I was afraid of that one... As you say diacritics were dropped, and that's the naming convention my company tries to follow. As per mixture of labels in English and local languages - I realize that the trend now is to use local names whenever possible (at least that's my impression) but:
1. In some cases places are better known under English names, at least within the shipping industry, so that's the reason they appear on maps even though they stand out
2. Our database with names certainly needs some standardisation. It will take a while, but until then it's 'always right' (unless it's something really important)

3. I am pretty sure I'll find some more mistakes myself as soon as I haven’t looked at the maps for a while. You know how it is - the more you work on a map the fewer mistakes you're able to spot. I don't doubt I'll have to put some effort in correcting and updating maps for the next edition.

[gp1]

Thanks for the replies

Regarding the lake fills: take a look at Lake Balkhash vs. the Aral Sea (both inland water fills). I see now that the Aral Sea may have been part of a compound path allowing the background ocean fill (darker) show through vs what I see in other inland lake fills like Lake Onega which is also lighter vs the Aral Sea.

Perhaps a slightly darker fill will contrast better with the non-subject background color of Asia in this case

[mika]

Regarding the lake fills: take a look at Lake Balkhash vs. the Aral Sea (both inland water fills).

Yes, i treated them as different features - lakes vs. seas. Perhaps i shouldn't have done so, but...
Regarding the fills - lakes are filled, though the fill is really faint... Goes onto my 'to do' list.

Thanks for the comments
Dom

[KennyRedman]

Those are awesome- I'd love to see the source file on something like this!

[araki5]

totally awesome! man, these are sharp! i didn't think manifold was up to the task, but I'm amazed!

good job!

[mika]

@ araki5: I don't think manifold is capable of creating similar layouts itself. Can't say for sure, as i didn't dig deep enough to find out, although I remember seeing recently some very nice maps prepared in manifold only. See http://forum.manifold.net/Site/ for some good examples.

@ KennyRedman: I am affraid I cannot post a file due to copyright issues... But if you're after an advice of any kind, go ahead, I'll see what I can do for you.

And a little description of the process - I worked in manifold 6.5 and CorelDRAW 12 and X3 (as x3 had some problems with importing ai files correctly). Data were prepared in manifold and then exported to ai - layouts of the maps had to match in both apps to avoid cropping data and so... Once i had all the ai files, corel would create an appropriate map layout and then import all the data - that was achieved automatically through scripting. After having a file ready to play with I had to do some data cropping, i.e. graticule, river shading using brushes (I mean by it that a river is thicker at its mouth than at its source) and eventually I had to rasterize the relief bitmap with the tint added (some ps printers don't like transparent objects). Colours (fills and strokes - thickness and patterns) were applied through scripts again, so it was easy to maintain a uniform colour scheme troughout the book. As I don't like gis like label placement, i had to adjust them manually so they could 'follow' cartographic grid where appropriate. And symbols... they were designed at the very begining and I used a symbol library to manage them. DRAW files had over 160 layers - that made it possible to access the needed objects without making vba or vb too confused and generate errors.

Once you know the workflow and have most of the stuff preprogrammed it's possible to prepare a map like that for the first proofreading within a day... Honestly :-)

[KennyRedman]

Thanks for that description, Mika, that's very insightful.

I wasn't expecting you to post any of your files, I was just thinking about the layers and the complexity of a project like that!