I have used Oxford EBSD system. It is very stable, highly capable of doing everything that a current EBSD can do, and the user interface is very comfortable and easy. They have the post processing software Channel 5 HKL which is also easy to use and efficient. Here is the link:

http://www.oxford-instruments.com/products/microanalysis/ebsd

I agree with Soud, I also believe that Oxford is the best and its sof. is very user friendly.

Oxford Instrument's EBSD system is no doubt good. Although I have used FEI Nova EBSD system and it is  also  a good experience.

Thanks for the answers. Is there a way to batch process aquired data in OXFORD HKL? For example in EDAX OIM Analysis you could create a template with maps and charts and texture plots and then apply it to all datasets and export all data to jpg and txt files automatically. Does HKL have similar feature? 

It depends on the expectations you have. The software capabilities are certainly very rich at OIM of TSL. I only heard that speed and stability is not that good. Bruker is very easy to use, but the data processing is comparatively poor. However, if you can use MTEX (free Matlab toolbox) you have effectively the most powerful tools for data processing and analysis, it is under continuous development and you can directly ask the developer for help and specific tools. Oxfords Aztec is somehow in between. Great detector, quite simple operation, basic post-processing. The weak point is that the post processing is "effectively" still the same as about 5-10 years ago since still Channel5 is used. However, this is also comparable to EDAX TSL.

Bruker offers from the beginning as standard opportunity to save the raw patterns. I can only recommend this to do since it offers many opportunities to evaluate the indexing, phases, orientation solutions etc. I am asking you: Who is not saving the XRD-diffractograms or EDS-spectra and only uses the numbers derived from the raw data? This is nowadays not more state of the art only to save the "estimated" phase solutions and determined orientations. You can have a look at the linked presentation, e.g. what you can still do with the standard EBSD patterns after your EBSD mapping.

Finally, there is no straight answer. You have always to make compromises, but after some years I am still very satisfied with the combination Bruker aquisition and MTEX. If you don't want use MTEX , and you believe "black-box"-software, CrystAlign (Bruker) is not powerful enough to make you happy. If you are experienced enough and you understand all settings and parameters, use OIM of TSL. But you have to make compromises as well regarding the detectors sensitivity and software stability. Use Oxford, if you are a standard user and you don't want very specific data analysis. The detector is very good, and the software offers the basic tools, but not much more. Obviously, for many people this is OK.

Conference Paper EBSD: more reliable and precise

 MTEX uses scripts where you can do this very nicely and even after years :-).

As far as I know it, Oxford as well as Bruker don't support scripts. I am also not very sure whether this is in general very useful. No doubts, if you are doing always the same, and you know your microstructure, it will be fine. But in fact it is comparable like and EDS analysis always only looking for the same elements exclusvely, or in XRD looking only for the same phases. If you have such work it is absolutely fine, but I have to admit, that I have at least only similar materials. There are alway a potential of additional phases or precipitates, specific boundaries etc. If you use "blind" scripts, you will never realize these perhaps more importants propoerties of your microstructure. Automatics are only then useful if you know that nothing will change. But seriously, how often this really happens (except you are working in a steel company and always investigationg the same kind of steel)?

Thanks, I got the picture about post-processing software. What about hardware? Is there a difference between manufacturers? Like something to have in mind when buing a system. Camera or FSD detector performace maybe. And i read in one of the papers that Oxford system allows you to save only 40000 patterns (that is restrincting you to only like 200x200 pixels map if you want to save all patterns).

Well, the hardware is of course also different. Different size, speed, sensitivity, saving "strategy". Bruker has the highest-resolving image format:1600x1200. Usually you don't need this since for usual measurements images in the scale of 80x60 or (what i am using) 160x120 pixels. The detectors of Oxford and TSL are lightly lower in their CCD-Chip resolution, but this is commonly not the problem. More interesting is the sensitivity. Here, Oxford has perhaps the best (Nordlys Nano), followed by Bruker. But keep in mind: the best detector does not mean automatically, that the system can use it. Band detection and indexing procedure are actually more important. Here the best but also the by far slowest algorithm has TSL. Very fast and nearly comparable is Bruker. Oxford recently published a new algorithm which closes the gap to Bruker, but also there the speed is somehow slower. However, simply speed is not always the most important factor, since some people want also have a high orientation precision. here all systems are not that different. The Hough-related indexing has limits because of the use of only a few or more (Bruker) peaks. Therefore, the worse (but extremely fast) band detection of Bruker is compensated. Imagine you have 6 bad point to describe a circle, or you have 12 slightly worse point. The latter perhaps delivers a more reliable description of the "real" circle. in other words: 12 slightly less exact bands give often a more precise orientation than 6, especially if one of them is perhaps a "ghost" band, i.e. does not exist in a real pattern.

The FSD's are not necessary but may help. The images are, however, hard to interprete since they reflect channeling-in and not as commonly discussed channelling-out. The result: The visible contrast is not comparable to the EBSD map. The performance of the FSD detectors are not that different, or not that important that I would buy a system because of the FSD detectors. It should be integrated in the software which is the case for all systems (I don't know it exactly for  TSL). Color images are only available with the Bruker detector, since they offer 3 FSD diodes and not only 2 or even one.

Frankly speaking, if you are in science you need first a good acquisition system which , from my point of view, enables you to save ALL data. Then you can even reanalyse measurements after years with another, hopefully improved indexing or phase recognition software. So nothing is lost. If you missed one phase, you have either to remeasure it (Oxford), or you can reanalyse it (in practically the same speed as measured with EDAX) or extremely fast with Bruker. Especially for unknown phase you are actually lost with only band positions. You need to know the band wifth AND the intensities of the bands (or the pattern). Also here a reliable information you only get with CrystAlign. BUT: you have to pay a comparatively high "price" because of the limited (but again extremely fast) post processing. For scientific application acceptable, for industrial application only then when you don't need more that the tool available in the software. Or...as already mentioned...forget all these black-box systems and use MTEX. The software is written by a real mathematician (i.e. he really knows what he is doing which is not always the case for the commercial companies (sorry guys, I know you are doing your best!)) and you have practically access to everything. 

There certainly are hardware related differences in the performance but spec sheets only tell half the truth. If you are concerned about  FSD performance for instance, you'll see that TSL has only one diode with small active area on the Hikari and larger area on the DigiView cameras, whereas Oxford uses 4 diodes. However, the PRIAS option of TSL allows you to use the whole screen like some kind of FSD, which is even more versatile but also more complicated in terms of signal acquisition and interpretation of results...

Anyway, since all TLS, Oxford and Bruker perform reasonably well on most EBSD related tasks, the more important questions are certainly who is going to use the system and what are the main applications? It's a big difference if you just have a few expert users (who might like/use the options of the TSL software) or a large group of users with most of them having only little experience or no time to become expert users. In that case you should compare the software packages and think about which one is the easiest for new users.

Still one remark regarding the 40.000 patterns. The major reason is Windows since this option saves each single (processed, i.e. useless for anything else than indexing) pattern in a directory. This means, if you open this directory Windows react acain perhaps after 20min since it is busy reading all these files. there are many people who restart their computer since they think it is hanging. If you would save more patterns it becomes redicules slow. This option was also not originally implemented to save all patterns but only non-indexed. Sounds reasonable but what happens if you selected the wrong phase or you have a phase which can be indexed also by an already existing phase like magnetite and ferrite?

An important point if you are planning on purchasing an EBSD system, is your local technical support and maintenance team. The operating cost from one system to another may differ considerably depending if the vendor has good local coverage or not...

Dear all, thank you for the very useful discussion. How does it look from today's point of view? (June 2022). Are all considerations still up to date? Are there any reasonable alternatives to the Oxford and Bruker? Thank you!

Certainly EDAX with the only available direct electron detector. :-)

I agree with Gert, the EDAX Clarity is certainly a unique system but unfortunately also incredibly expensive. Paired with an EDS system it comes at a similar price tag as an entry level FEG SEM. It is also not really a direct competitor to CMOS based systems from Oxford, EDAX, Bruker, etc. which are also much faster and better suited for a lot of standard applications.

Personally, I doubt that there really is a best system for everyone. There are so many factors such as application portfolio, level of knowledge / training for EBSD users, etc. that should be taken into consideration.

For our group, a very specific software option (that you cannot mimic with MTEX) is far more important than hardware specs, which gives one supplier a clear edge over all competitors. However, the situation may be entirely different for other groups.

...and of course the compatibility to already existings systems in the own or collaborating institutes. To operate different systems increases to possibility to recognize "problems" (deviations, errors, uncertainties...) but it requires also a lot of additional time to find out from where they are coming. From my point of view, it would be better to get well running hardware and use software like MTEX or ATEX for data processing. The saved time could be used to improve or stabilize hardware-related challenges.

A car company also does not develop e.g. own MP3 players since they are part of the car interior.

Unique selling points of different EBSDs are exactly what I am trying to find out. 80% of the time we have routine measurements that any modern system can do. When characterising ultrananocrystalline surfaces at different temperatures, we would like to have speed and well-developed drift correction. And all of this on our old Leo-1550, in good condition.

For nanocrzstalline material you are very likely forced to use TKD because of the resolution limit of EBSD. No problem for a 1550 using 30kV. You only need to change the preparation in order to get "transparent" samples. TKD uses the same hardware.

For nano-crystalline applications (especially for older, low-current SEMs) you have to invest in a very sensitive EBSD detector. I would recommend one of the recent Oxford detectors. They use a fibre optics and don't loose any intensity from an additional lense. Also their IR filter is OK for high T. From my point of view, less meaningful is a drift compensation since this is not really working. If you have a drift you can only find this out with a post-acquisition of the image and image correlation. Any drift correction will then either measure the same again...or leave something out. Then it is better to reduce drift either by a gas injection system attached on the SEM, or a coating of the sample (e.g. Pd, Cr...). For TKD I havn't seen this until now. Most stable investigations are reported if you leave your sample inside the SE for a 30 min, over night, or even over weekend. Simply for "aclimatisation". Drift is an unpredictable phenomenon, disappears and re-appears again. I guess one can only try to reduce it as much as possible...or measure that fast that it s no more "visible".

Gert, thank you for the advice!