NEWSFLASH: update on Jan 7th, 12th, 19th, and Feb 7th (see end of article)
Today I got an email from a LAStools user at NOAA pointing out a new entry in the ArcGIS 10.2 documentation of ESRI that mentions a *.zlas format for the first time. This may have been an oversight at ESRI since there was no press release, blog post, etc preceding this documentation update (that happened 11 days ago). A screenshot of the entry can be found below.
I have heard about LAS compression by ESRI since Gene mentioned it in a blog entry after ESRI’s 3D Mapping and LiDAR Forum. Back then I throught they were talking about LAZ and that our 1.5 years of talking about including support for LASzip-compressed LiDAR into ArcGIS were finally getting somewhere. But turns out they have been doing their own thing. Here some rumors I have heard about ESRI’s new *.zlas format:
- similar compression rates as LAZ
- includes spatial indexing
- (maybe) re-orders points during compression
- performance is like laszip.exe or better
- will be available in ArcGIS 10.2.1
- can be used without the LAS Dataset
- „free“ Windows executable will be available soon
- development libraries with API will follow
- ESRI has been giving data providers heads up that clients may soon demand this format
My first thought was that this might be a reengineered version of the LizardTech LiDAR CompressorTM but it is not. This seems to be ESRI’s own development. Does anyone have more details on this?
What was their motivation? Is LAZ too slow for them? I would have happily adressed whatever LASzip was lacking as (compatible) extensions to the LAZ format – which has become de-facto standard for LiDAR compression and is open source. But instead they invested serious money and man-power into creating an entirely new format. Anyone want to speculate why …?
UPDATE (January 7th): It is official now. Apparently, Gene who has mentioned our rumors on his blog just received heads up from Clayton Crawford at ESRI that the LAS Optimizer is available for download here. The EzLAS.zip file contains a PDF with interesting details. Below a snapshot of the GUI and what the website says:
- The standalone executable does not require an ArcGIS install or license.
- The same executable is used for both compression and decompression.
- The download zip file contains more information and help in an included pdf document.“
Apparently, ArcGIS 10.2.1 is available for general release on January 7th and will support direct read of optimized LAS (*.zlas) via the LAS dataset. Now it’s your turn to try out what ESRI has cooked up and comment … (-:
UPDATE (January 9th): I was told that ESRI will be releasing an „official statement“ soon explaining why they have developed their own LiDAR compression format. And they really should do so. I have received (and continue to receive) a fair number of „off-the-record“ emails from people across the industry expressing feelings that range between „disappointment“, „anger“, and „disgust“ over what is seen as an attempt to sabotage our multi-year effort of creating an open, free, and efficient compressed LiDAR exchange format … time for some xkcd humor.
UPDATE (January 12th): This might just be that „official statement„. Seems someone is really trying hard to avoid using the word LASzip … (-;
UPDATE (January 19th): The story has been picked up by a number of blogs like Paul Ramsey’s „LiDAR format wars„, James Fee’s „LAS, LAZ, LASzip, zLAS and You„, and Randal Hale’s „LiDAR and your software„.
UPDATE (February 7th): The front-lines harden as an unlikely coalition of open source knights, laser guardians, imperial agencies, and competing thugs forms a rebel movement against the approaching Desri Star who threatens the free world announcing that the dArcG is going to spawn „parallelized LAZ clones„. Encrypted instructions to Jedis spread like „Point Clouds on the Horizon“ „Towards an Open Future“ as the insurgency prepares a better future for compression, free of proprietary oppression. The clone wars might be starting soon. Will the FOSS be with LAZ? (-;
Hi Martin,
We just found out about this yesterday ourselves. At first I thought it was a typo and that LASzip was being used, but your blog entry here means…yep, another format for us to support. I’m sure more details will be forthcoming at the Esri Developer Summit in a few weeks time…if not before.
Dale
My first reaction is the world needs another undocmented proprietary data format like another hole in the head. Why not just document the format instead of a limited functionality API ( just going by past experience with file geodatabase)?
One reply in the LinkedIn group GoGeomatics Canada was: „I guess it is the unfortunate but still common attitude of market leading vendors: Control the data formats, control the (software) …“. Replies on the LAStools facebook page were: „Another obscure proprietary format?“ and „ESRI could better spend its time in making arcgis more… I dunno… Stable, usable, multiplatform, less sucky in general.“ and „They must intend on squashing LAZ like a bug by creating their own format. 🙂 As I think about Martin Isenburg’s question about LAZ being too slow for ESRI….. Judging by the speed at which ArcMap loads, speed isn’t the issue.“ Similar sentiments are echoed in the LAStools user forum: „This is a bad news. We need standard formats (open format) not another formats.“ and „Perfectly agree with Jorge! I guess ESRI wants complete control over the formats it uses.“ and „I also agree. Open formats are particularly important for archive and I’ll continue to send data for NOAA archive in open formats such as LAS and LAZ.“ and „I agree, I find it amazing how many companies still push proprietary formats for no good reason at all, but I guess it comes down to controlling/keeping their market hold. Could be worse, they could have taken a proposed new variant of an open format and bastardised it for their own needs … I’ll avoid naming names; you know who you are! ;-)“ and „I agree as well. This is a redundant format, ESRI should just adopt the LAZ format that people are already successfully using.“ but there is also this more balanced view:
„Initially I was dismayed by the news. The State of Minnesota has been using LAZ compression to store LiDAR data for several years. We won’t be changing how we store the data either as it makes no sense at all to go to a proprietary format however we have a lot of stakeholders that are heavy users of ESRI software.
After pondering this notion for a while I csme to the conclusion that it’s fine for ESRI to develop their own proprietary compression and indexing schemes. IF (and that’s a big if) the products prove to be better and faster in their software then I’ll welcome using it.
Competition is one of the things that drives innovation and improvement and I would expect Martin would likely not sit still if LASTOOLs compression and indexing were outdone.
I for one am not going to stop putting pressure on ESRI to read LAZ compressed data. I’m going to continue to send them e-mails and press for them to write into their software the ability to ingest LAZ compressed LiDAR data. They support all kinds of vector and raster formats and I see no reason for them not to support LAZ format if their users continue to demand it.“
Pingback: clone wars and drone fights | rapidlasso GmbH
Pingback: new “LAStools Production” and “LAStools Pipelines” for ArcGIS | rapidlasso GmbH
Pingback: Realizing Open LiDAR with ArcGIS online and LASzip | rapidlasso GmbH
Pingback: Keeping ESRI Honest | rapidlasso GmbH
Pingback: The LAS format, the ASPRS, and the “LAZ clone” by ESRI | rapidlasso GmbH
Pingback: The dArc Force Awakens: ESRI escalates LiDAR format war | rapidlasso GmbH