GPS start up time and inserted SD card speeds

[+Forrest hunters]

I am starting this new topic to see how different gps speeds are affected due to SD cards and loaded information. Below is the original thread with a few other questions on it. In that thread you can read the info on my card and my start up times. Tomorrow when I'm on my computer and not on my crackberry here I will copy and paste it here again. I have a 4 gig class 4 card in my gps now with a start up time of 42 seconds. Rough map info is 1.4 gig on the card.

 

Any others with map segment and space on their units and what cards they are using would be great.

 

http://forums.Groundspeak.com/GC/index.php?showtopic=268322&st=0&gopid=4613600entry4613600

[+Bullygoat29]

I copied this over from the other thread. I was typing it while this thread was being created. To add, I just purchased a 16GB class 6 card from Newegg. Lets hope this speeds up my load times.

I've got 7.4GB of maps on my 8GB class 2 uSD card. It has a 1:15 boot time. It stays on the loading photos screen for almost all of that time. Swapping in a fully loaded 4GB class 4 card, load time drops to 32 seconds. Pulling the card all together, gets me a 12 second boot time. I believe that the class does make a difference. I just wish I knew why it spends so much time on photos and literally seconds on the maps. Doesn't matter which card it is. I'm going to try a 8GB or 16G higher class card and see what happens. New egg has got some pretty cheap cards with free shipping.

[+Forrest hunters]

Thanks bullygoat29 as I was sending you an email as you were copy and pasting. Do you have the same info on both class 2 and class 4 cards?

[+Bullygoat29]

Thanks bullygoat29 as I was sending you an email as you were copy and pasting. Do you have the same info on both class 2 and class 4 cards?

 

Almost the same data. The 4GB is obviously smaller so it doesn't have as much. They both have the same .kmz files, as much birdseye as I can fit and the 4GB has geotagged photos where the 8GB doesn't. I don't think the amount of data matters as much as the class of the card does. I've got a class 6 16GB card on the way and I'll load that to the gills and check startup time with that sometime next week.

[+Forrest hunters]

Below is what I have on my card and what the start up time was after installing each map. I did start it up and aquire signal and shut it down 3 times each before I recorded the time. Card is a Dane-Elec class 4 - 4 gig card.

 

michigan topos 14 seconds 146,536 kb 130 segments

other michigan topo 14 seconds 59,299 kb 44 segments

US states and counties 14 seconds 1,918 kb 2 segments

US P North Central 25 seconds 335,832 kb 330 segments

US P North East 32 seconds 282,872 kb 268 segments

NE Topo Part 1 46 seconds 642,656 kb 463 segments

 

Total 1,469,113 kb 1237 segments

[+dfx]

alright, here goes. Oregon 450 and I started off with an empty unit: No GPX files, no maps, no SD card, nothing.

 

Startup time: 11 secs

 

The I progressively added stuff to it and always rebooted twice in between:

 

13 archived GPX tracklogs, totalling 1 MB: 11 secs

7 birdseye maps, totalling 344 MB: 12 secs

416 MB of Garmin Topo Canada 4 (mostly Ontario): 13 secs

1.6 MB Ontario trails map: 13 secs

 

Then I put my 4 gig class 2 SD card in. Again it started out empty and I kept adding stuff to it:

 

empty: 13 secs

22 GPX files containing close to 5000 caches (most of which are in 5 of the files) plus their waypoints, totalling 34 MB: 18 secs

106 more MB of Garmin Topo Canada 4 (covering Nova Scotia and New Brunswick): 18 secs

41 birdseye maps, totalling 2.5 GB: 34 secs

 

and this is what I had when I tested in the other thread too.

[+Forrest hunters]

dxf

 

Awesome info. I wonder if it is the amount of gpsfiledepot maps that i have on the card itself. Looks like you have some on the device and some on the card. Also looks like your maps are all Garmin maps versus the free ones I have. maybe I should pony up and get the Garmin ones.

 

It also looks like you have a bit over 900MB of maps both on the unit and on the card. The birdseye look to take a bit of your time.

 

Do you think if I put a couple of the maps on the unit that would speed it up any?

[yogazoo]

dxf

 

Awesome info. I wonder if it is the amount of gpsfiledepot maps that i have on the card itself. Looks like you have some on the device and some on the card. Also looks like your maps are all Garmin maps versus the free ones I have. maybe I should pony up and get the Garmin ones.

 

It also looks like you have a bit over 900MB of maps both on the unit and on the card. The birdseye look to take a bit of your time.

 

Do you think if I put a couple of the maps on the unit that would speed it up any?

 

It would be nice to observe the results of an apples to apples test. 8gb "class 2" vs 8gb "class 6 or 10" with the same exact data. Me thinks we're splitting hairs a bit but, hey, faster is faster!

[+dfx]

It would be nice to observe the results of an apples to apples test. 8gb "class 2" vs 8gb "class 6 or 10" with the same exact data. Me thinks we're splitting hairs a bit but, hey, faster is faster!

 

Yeah I'd be curious too, but unfortuantely I don't have any faster SD cards available. Donations welcome via paypal however!

[+Forrest hunters]

I would not disagree that were are splitting hair, but hey speed is speed. Just trying to figure out if we can speed things up or not. Seems like others are getting a lot faster start time then I am.

 

Does anyone know what the segment size of the preloaded map data on the Oregon 450 is? Might try looading some maps onto the unit itself to see if that speeds anything up to.

 

Thanks again.

[yogazoo]

If my 2yr old goes to bed at a reasonable hour tonight I'll execute some tests. Otherwise the GPS world will have to wait until the weekend.

Edited February 10, 2011 by yogazoo

[yogazoo]

After researching the speed class of SD cards I have found some very interesting info. It appears as though the class or speed of an SD card often refers to it's write speed and not necessarily it's read speed. Most uses of the higher class of cards are for real-time HD video recording. The higher classes are necessary to write HD video in HD camcorders but not necessary to read it. Based upon this information and other reports, data, and testing charts I conclude that there is probably no real gain in screen redraw speeds or startup times using a higher class of card in a GPS unit.

 

It is my feeling (not scientific data) based on my experiences that the processor and other hardware in most GPS units is the limiting factor in GPS operation (i.e. screen redraws, starup times, etc.). A faster processor that could actually use the extra bandwidth of a high speed card would consume much more battery power and is why GPS's aren't usually cutting edge in the processor department (battery life is a much bigger deal than processing speeds. It's why battery life is typically included in GPS specs and processor info isn't). Essentially a faster card would be wasted on a GPS other than write speeds from the computer to the card. While I haven't tested this in any meaningful way, experiencing the limitations of GPS processing speeds I have concluded the information above and that it would probably be a waste to shell out for a higher speed class of card.

Edited February 14, 2011 by yogazoo

[+ecanderson]

After researching the speed class of SD cards I have found some very interesting info. It appears as though the class or speed of an SD card often refers to it's write speed and not necessarily it's read speed.

Not sure where your research took you, but that's not true. Especially for video, playback speed is also a major issue. The Class number represents both read and write speed minimums in Mb per second. Multiply by 8. For example, a compliant Class 2 card must be able to both read and write at 16Mb per second. A Class 4 card must be able to both read and write at twice that speed (32Mb per second).

 

I limped along with a Class 2 card stolen from my phone until I started loading Birdseye images on my 450, and finally bit the bullet. Believe me, the difference in screen update speed between that Class 2 and the Class 4 I replaced it with was considerable.

 

No one is suggesting that a top end card is needed, but the lowest end Class 2, while adequate for audio bandwidth, can't shuffle massive numbers of pixels around quickly enough for many devices. In the TomTom forum, we have discussed this at length since some users prefer to mount their map sets onto SD cards (on those units where the slot is available). Especially with a lot of POI categories selected, things can really get bogged down with a Class 2 on those units as well.

Edited February 14, 2011 by ecanderson

[+coggins]

I have filled up my 8GB card now that I have BirdsEye so I just ordered a 16 GB class 4 Kingston. Anyone use a 32GB on these units?

[+Forrest hunters]

It there a max speed that the Oregons can handle. Maybe I will go with a class 10 if they can handle it.

[+ecanderson]

It there a max speed that the Oregons can handle. Maybe I will go with a class 10 if they can handle it.

Given the processor speed and data requirements, I'd bet that Class 10 is total overkill. Wish I had a Class 6 to see if there was any improvement over my Class 4. Scrolling is smooth enough now with the 4, however.

[+Forrest hunters]

It there a max speed that the Oregons can handle. Maybe I will go with a class 10 if they can handle it.

Given the processor speed and data requirements, I'd bet that Class 10 is total overkill. Wish I had a Class 6 to see if there was any improvement over my Class 4. Scrolling is smooth enough now with the 4, however.

 

My scroll speed is sometimes slow especially when i am zoomed out to about 12 miles and want to pan from lower pennisula of Michigan to Upper pennisule.

[+ecanderson]

 

My scroll speed is sometimes slow especially when i am zoomed out to about 12 miles and want to pan from lower pennisula of Michigan to Upper pennisule.

That is more likely a function of the massive amount of scaling that must occur to take map data and sort out what to display at those zoom levels to make something useful of it. Picture the process of sorting out what data must be discarded in those views. At some point, as has been pointed out, you WILL wind up processor bound. Still, it's true that at that scale, you're having to move a truly massive amount of data around before it gets filtered for the display.

 

Does anyone out there have a decent range of uSD cards handy to do any experiments? I still have my slow Class 2 and the current Class 4, but I have no uSD cards faster than that, and they're not as easy to find as SD cards at those speeds. Again, most are using them in phones where the transfer rates aren't an issue.

[+Bullygoat29]

I've got a class 6 16GB uSD on the way. I'm planning on loading my 4GB class 4 and then copying that data to the 8GB class 2 and 16GB. Overall size shouldn't effect speed, just the data that's on it. I'm hoping to see a nice gain with the class 6.

[yogazoo]

Depending on what you read online, and how you parse the info, the argument for reading speeds on these cards can be unclear. I've seen tests where 2,4, and class 6 cards only improve by 1mb/s in succession interfaced through a card reader. And the slot in the battery compartment is essentially a card reader. I'm not saying there's no improvement, there may be but in there is something that makes me think there could be a placebo effect at work here.

 

The trickle of reports describing faster operation is good news though given the, at times, painfully slow redraw rates of my 450's display. However anecdotal reports are really all anyone can give (regarding screen redraws) since we're not engineers with piles of test equipment. I can say that having birdseye on my units internal memory vs external class 2 card has, appearantly, no effect on screen refresh rates. Maybe the internal memory is based on class 2 technology?

[snowfleurys]

That is more likely a function of the massive amount of scaling that must occur to take map data and sort out what to display at those zoom levels to make something useful of it. Picture the process of sorting out what data must be discarded in those views.

 

I do not think that is correct. When cgpsmapper compiles a .mp file to .img format it does one pass for each of the levels that the data is being defined at, and then combines it into the .img file. This creates a unique set of coordinates for each 'zoom display level range'. Note on the GPSr as you zoom in, the data will be 'enlarged', then the data will be displayed with more detail be it goes to a more detailed display level. At each 'zoom display level range' the GPSr is using that set of coordinaes which is defined to be used for that display zoom level; it does not have to discard anything, only scale it to the display window. Obviously at the more general display scale within any 'zoom display level range' more data will have to be scaled to displayed on the GPSr screen.

[savant9]

I have filled up my 8GB card now that I have BirdsEye so I just ordered a 16 GB class 4 Kingston. Anyone use a 32GB on these units?

 

Be aware that..... and I haven't really seen anything to document this....... You are limited to 200 birdseye files (similar to the 200 gpx limit). The GPS will just refuse to boot if there are more than 200 birdseye files, it will stay forever on the "Loading Photos" screen.

 

[+Bullygoat29]

 

Here's the data that I used, with the cards I used laid out in a fastest to slowest comparison. As you can see, there is a difference in speed depending on the class of card. The gap does close rather quickly though when you get to the class 6. I have a feeling you wouldn't see a speed increase by going to a class 8 or 10 in the Oregon 550. One thing I forgot to add to the chart is my startup time with no card was 11 seconds. I have a feeling even more data would slow startup times down and once I get my 16GB card fuller, I'll post my startup time then.

[+ecanderson]

The gap does close rather quickly though when you get to the class 6.

Looks like the difference between Class 4 and 6 isn't very large. That's pretty much what I'd both hoped and expected!

 

Since you seem to be the only one with a 2, 4 and 6 in hand, and have Birdseye images available (seems to present the worst case), I should ask whether you'd be willing to play guinea pig for map speed.

If you could, can you set up to a specific zoom level with Birdseye, and scroll about the screen with the 2, 4 and 6?

Further, if you scroll around a bit with the Topo24 at a couple of zoom levels (more realistic for many users), that would be great. I know the results are going to be subjective, but it would be great to hear your input.

[+Bullygoat29]

Well I tested scrolling on the screen between my class 2 and class 6 cards. The class 2 was faster but I believe that was because there was only a couple hundred MB of birdseye imagery on it. It takes forever to load this card, but with almost 8 GB of birdseye on my class 6 it only takes about 2-3 seconds for the screen to refresh. So I do believe it's faster but the amount of total data has an effect. My startup time is now 58 seconds with 10.1 GBs of data.

[+Forrest hunters]

I'm still confused why my class 4 is taking 46 seconds with 1.4GBs on it.

 

Bullygoat29 great feedback. I wonder if the manufacturer of the card makes a difference now. Maybe I will just pick up a 16GB card of class 6-10 and try it anyway.

 

Thanks again.

[+Bullygoat29]

I'm still confused why my class 4 is taking 46 seconds with 1.4GBs on it.

 

Bullygoat29 great feedback. I wonder if the manufacturer of the card makes a difference now. Maybe I will just pick up a 16GB card of class 6-10 and try it anyway.

 

Thanks again.

 

Where did your map files come from? Are they Garmin maps? If so, what versions are they?

[+Forrest hunters]

All my maps are from gpsfiledepot. What I have loaded is above in post #5.

[+Bullygoat29]

All my maps are from gpsfiledepot. What I have loaded is above in post #5.

 

Most of my maps are Garmin. I'm wondering if the have some kind of optimization that allows them to load slightly faster.

[yogazoo]

Having made more than a few maps sets for public use I have to say that creating a reasonable tile size and reasonable TRE size is key in GPS rendering of maps. I'd say that there's no proprietary optimization per-se but achieving appropriate tile and TRE size in the maps is something Garmin must have down pretty well. I'd say many map makers have it figured pretty good but there was a time when my tile sizes and TRE were maxed out, making for some really slow loading maps. Since then I've learned to tweak these settings and the GPS responds much better. - FWIW

[+Forrest hunters]

Okay this makes no sense at all. Today I downloaded OSM for Michigan and loaded it on my unit and the start up time went down to 10-12 seconds from the 45 plus. Don't understand why it was taking so long and then I add more and it speeds up. Oh well I like the quicker speed.