[orgopete]

I cannot say that I am older than discodermolide, but the first NMR I used was a Varian A60 and it could plot an integral. I would have bet all commercial NMRs could.

GCs were a lot more rudimentary and I triangulated, weighed, and used the integral gizmo that gave counts for area above baseline. After that, I used a digital plotter that calculated integrals. I was in synthesis, so areas really didn't have the same importance as analytical chemists.

[fledarmus]

My experience was the same as Orgopete's. The first NMR I used on a hands-on basis, back in 1984, was a Varian A60. These were scanning analog machines - the radio frequency signal projected through the sample would gradually scan from 10ppm to 0ppm (or whatever boundaries you set) and a chart printer indicated the absorption as the machine scanned the sample. To get better resolution, you just scanned slower. When you had scanned the sample once in a frequency readout mode, you went back and scanned it again in an integral mode. Essentially the same, but it didn't return to zero, it just kept accumulating signal. You would manually zero it after each set of peaks. I heard at the time that previous instruments lacked an integral mode and that you could Xerox the spectrum, zooming in on your peaks (there was one Xerox machine in the building that actually had a zoom function!) and then cut them out and weigh them to get the integral, but I never had to do that for an NMR.

Also, like Orgopete, I did have to do that for GC and later for LC traces which also had simple paper traces. It took much longer to get peak integration for those systems, and I never had access to a digital integrator. For simple traces where the peaks were reasonably triangular, however, it was faster and sufficiently accurate to use the height of the peak and width at half-height to calculate the peak area.

[Rutherford]

For simple traces where the peaks were reasonably triangular, however, it was faster and sufficiently accurate to use the height of the peak and width at half-height to calculate the peak area.

I am interested in this, could you explain it?

[discodermolide]

For simple traces where the peaks were reasonably triangular, however, it was faster and sufficiently accurate to use the height of the peak and width at half-height to calculate the peak area.

I am interested in this, could you explain it?

This is just the area of a triangle, half the base length times the height.

[fledarmus]

For simple traces where the peaks were reasonably triangular, however, it was faster and sufficiently accurate to use the height of the peak and width at half-height to calculate the peak area.

I am interested in this, could you explain it?

This is just the area of a triangle, half the base length times the height.

But since chromatographic peaks tend to have shallow slopes and shoulders at the base making it hard to determine base length, you usually use the form "width at half-height times the height". The width of a triangle measured at half its height is half of the width of the base of the triangle.

And since the peak is rounded at the top too, it can be a little tricky measuring the height as well. As I said, however, for simple peaks it was usually a close enough approximation.

[Rutherford]

Okay, I tried it and it was very tricky. Didn't give me the correct result, maybe because the picture I had was to small, so for precise measurement I need something that measures micrometers maybe.

[discodermolide]

I cannot say that I am older than discodermolide, but the first NMR I used was a Varian A60 and it could plot an integral. I would have bet all commercial NMRs could.

Dear Orgopete,
You make it sound like I am 100 years old, 

When I was an undergrad. The University was so poor it could not afford anything, we had to make it all ourselves. I think our physics/metalworking/electronic departments made the NMR machine although I can't be sure about that. I do remember getting a NMR copy that had no integration, it may have come from that machine!
So of cutting we went, although my scissors were not sharp and I ended up ripping the paper and having to estimate the area/weight.
The delights of Organic chemistry.

[AWK]

This brings up an interesting question though, how did you determine the integration in the old days when fancy software and powerful computers weren't around? For chromatograms, I've heard you literally cut out the peak and measure the weight against some standard, but surely that would be impractical for a reasonably complicated proton NMR.

I did personally integrated spectra (as operator) on Perkin-Elmer 60 MHz with solid magnet in 1972 yr. Integration device did not need computer if you do not use a Fourier transform.
Those were days!