Saturday, August 18, 2007

killer app

As you know, Ryan started his own mono-thematic blog on ACD NMR software. In his recent posts he repeatedly praises the "multiplet reporter". It is a simple little piece of software, a sort of translator, that creates a formatted list of chemical shifts and J couplings starting from a table of the same values. As far as I know, it doesn't perform all the possible checks. For example, let's say you have hydrogen A coupled with hydrogen B. If you state that A gives a doublet with J = 7.3 Hz and B another doublet with J = 7.4, the program will not complain (will your reviewer?). I have never used the ACD processor and I may be wrong on this point. I have examined other programs that also have a "multiplet reporter", namely MestreNova and "iNMR reader". [The latter costs 30 or 40 times less than Ryan's product and has far better graphics].
The principle, however, is the same. Ryan says that the multiplet reporter is a killer application; he also gave me a lesson on the meaning of the expression "killer application".
I have a theory about it. They say: theories come and go, facts remain. In my theory what gets killed is the spectrometer. Most of the people don't like off-line processing because:
1) Why buying another program when we already paid a lot for the spectrometer software and its upgrades?
2) They want to seat in front of the spectrometer. It's the most coveted toy in the department and everybody's still a child.
I loved sitting at the spectrometer even when the monitor was placed on top of the electronic console and the noise of the fans was too near to my hears. I had to struggle against their hypnotic power. You know, we Italians like pasta and it's difficult to remain awake after lunch, almost impossible when the fans are going and each Fast Fourier Transform takes from 20 to 30 seconds...
Afterwards I had my own NMR software to use, and it was more funny to use a PC or the Mac. For the rest of us, who don't usually write their own NMR software, the spectrometer fascination remains intact.
The multiplet reporter is the bridge between NMR software and MS Office (or Open Office for some users). It only makes sense if you copy and paste the text into a report (paper, patent, thesis, etc...). Who can use a spectrometer to write a paper? You have no excuse, you can't do it. The bottom line is: we finally have something to convince people to process NMR spectra away from the spectrometer.
To further explore the subject Ryan points to a movie that amply demonstrates how ugly the aliased graphics of ACD are by today's standards. I have no movie. You can get iNMR reader 2.3 from here (requires Mac OS 10.4) or see the "J Manager" in picture or wait for iNMR 2.3 (to be released on Monday).

2 Comments:

At 3:39 PM, Blogger ryan said...

Giuseppe,

Thanks for your comments and pointing some of the readers of your own monothematic blog to mine at http://acdlabs.typepad.com/

First of all, perhaps the most important comment in your post is, "I have never used the ACD processor"

While you are free to talk about it as much as you like, because you haven't actually used it, your comments lack credibility.

As far as your price "assumption" is concerned: Our software is 30 to 40 times more expensive than yours? Well what are we comparing here first and foremost? In your comparison you choose to compare the price of our full NMR processor which contains all of our 1D NMR processing functionality (there is no lite version) through 8-9 years of development, and a terrific drawing package in ACD/ChemSketch to your iNMR reader, which appears to be a second tier processing product compared to your much more expensive iNMR application. Furthermore, purchasing software is not all about the theory that ”cheaper is better” There are many, many things to consider when comparing software packages, such as available features and functionality, ease of use, reliability, operating system, etc. If the only difference between iNMR and iNMR reader is saving and batch processing, why do you charge 10 times the amount to have this functionality included? In addition to that, everyone is different, and everyone has different desires, expectations, and preferences for a specific software package. I would encourage people to just try many different packages and see which one best fits their needs.

As far as your "theory" is concerned, I think it is also wrong.

First of all, you make the comment that, "most people don't like off-line processing." This is simply not true. You should try and convince the many, many end-users and customers who have deployed off-line processing over the years that they either don't like it, or have wasted their money.

As far as wanting to sit in the front of the spectrometer, this is certainly true for many (but not all) NMR spectroscopists. Many like to do this and are also very comfortable in using the instrument vendor software because they have been using it for their entire career. That being said, there are many who prefer to process their own data in the comfort (and quiet) of their own lab or office. In addition to this, the era of open access instrumentation has significantly changed things in the offline NMR processing world. Instead of one person accessing the NMR instrument and software at all times, you now have (in some organizations) hundreds. For this reason, sitting in front of the spectrometer to use the instrument's software is not feasible.

Simply providing chemists and NMR users in an open access environment with offline desktop processing can result in a huge boost in productivity. Here's a few examples:

http://www.acdlabs.com/clients/2007_08_manu.html
http://www.acdlabs.com/clients/pr_scynexis0406.html
http://www.acdlabs.com/clients/pr_kalexsyn0506.html
http://www.acdlabs.com/clients/pr_bristol0806.html
http://www.acdlabs.com/clients/pr_nmr_1206.html

You've drawn from your experience as an NMR spectroscopist, now allow me to share my experience as a chemist in both academia and industry. Both are users of NMR software.

Both in graduate school and in the pharmaceutical industry, I worked in an environment that had only one NMR spectrometer to be shared by over 100 different users in each case. Over these years I can't count the number of times I went to the instrument room to process my files, only to be faced with a long lineup of other users waiting to use the instrument's software. I vividly remember the days I had to stay late just because I needed to process a data file and interpret a spectra to make a decision before the end of the day. Even worse, were the occasions where there was only one person in front of me, but they were the type of people that would spend an hour in front of the computer trying to process one NMR data file.

There you have it, two different people, and two completely different experiences. I think in your post you have made the dangerous assumption that your experiences are consistent with all end-users out there to the point that you can build a "theory" around it. Not everyone is fascinated with the NMR spectrometer.

I don't understand the point you are trying to make when you say, "Who can use an NMR Spectrometer to write a paper?" Where does this question come from? Of course it only makes sense to copy and paste a report from the NMR software into another application. The NMR software is not a word processor or an ELN. That's why the functionality that allows easy copying and pasting into other applications is there, and is simple to use. However, we also provide the functionality to create reports in ChemSketch that include structures, spectra, parameters, tables, multiplet reports, etc. that can be easily exported into Word or PDF, no copy and paste needed.

Finally, you have an obsession with graphics and you keep pointing to the superiority of the graphics in your own software over ACD/Labs. Your software looks very nice, I won't argue that. For years our focus has been on improving and extending the functionality in our products, improving ease-of-use, adding exciting new features and new technology. We can't do everything at once, and we have to set priorities.

Since you haven't used our software, the only critique you can make is in regards to graphics. Certainly this is an important point, but then again, I think that most people are more concerned with functionality and performance of a software product.

Sorry Giuseppe, we disagree on this one.

 
At 4:24 PM, Blogger old swan said...

If you talk about usability, drawing routines come first and foremost. What every user of any NMR software wants to do is: to observe the spectrum. In 2005, for the first time, a spectrum looked better on the screen than on paper and the reason was that iNMR:
1) implemented anti-aliasing;
2) Exploited the _whole_ screen.
You are a Wikipedia user. At the page:
http://en.wikipedia.org/wiki/Anti-aliasing#History
you can read that the technique is more than 30 years old. Today all graphics cards implement it and I am sure that soon or later even ACD will adopt antialiasing. After the first hour using it, you will agree with me 100%.

 

Post a Comment

<< Home