AC90: AG-AC90 MTF50 Resolution

T

Tom Roper

Veteran
Panasonic AG-AC90 Resolution MTF50 Tests

I just completed resolution testing using the Imatest software. MTF50 resolution was measured at the center of the image at three different aperture settings, for two different zoom lengths. The readings were taken in both horizontal and vertical directions, the high and the low readings for each were thrown out, and the rest were averaged. Numbers thus obtained are usually lower than those reported using visual interpretation of ISO12233 charts.

The camera was tripod mounted. Each time a new measurement was taken the image was refocused and settings locked before the capture. Stabilization was off. The recording format was 1920x1080/p60 AVCHD.

The software normalizes the MTF reading to a standard 2 pixel radius for sharpening so that effects from in-camera sharpening are negated. Measurements are given in (LW/PH) line widths per picture height, sometimes referred to as "TV Lines." Readings above nyquist frequency are invalid or represent false detail or aliasing. A perfect MTF50 score is 1080 horizontal, 1080 vertical.

MTF50 result for Panasonic AG-AC90, overall score:

Horizontal: 1048 (LW/PH)
Vertical: 868 (LW/PH)

Comments:
The numbers are about on par with best HD cameras I have owned and tested including the Sony EX1, the best being the Sony PMW350k which is slightly higher on the vertical, and far above other cameras including the Canon 5DmkII, XH-A1, HV10, Sony Z1.

The tests were conducted in accordance with my usual methods from 7ft. It is not possible for me to get far away enough to test telephoto performance. The two tested zoom ranges were Z00 (fully wide), and Z13. The apertures chosen were f/2.8, f6.4 and open. Perhaps the best all around performance was at f/6.4, approximately 950-1000 tv lines both horizontal and vertical. The poorest averages about 825 lines vertical at aperture fully open. The variances from f/2.8-6.4 are unremarkable. The best vertical is at Z13, f/6.4, 1036 lines, and the worst horizontal 969 lines Z13, f/2.8. The best horizontal readings were above the nyquist at 1080 lines, and occurred at several settings.
 
Hi Tom,
thank you for carrying out and publishing those tests. It certainly indicates that the 90 is a class act at any price, at least regarding its lens performance.(data,data,data)

Have you carried out the same test on the 130/160/A? I realise they are very different cameras but a comparison on similar parameters would be interesting.
Thanks for the info and a really valuable contribution to the forum.
Cheers!
Rod
 
Thanks Tom. I was shooting with mine the past couple days and getting some very lovely images. It would be interesting to run that test on an Iphone 5.
 
"X900 price is still on sales level, and internally it is almost the same as AC90"
Click to expand...

Hmmm... I don't understand why he keeps saying that, but much better OIS, better gain, BSI sensor, pixel shifting, many more settings... No, I wouldn't say they are "almost the same"! Not to mention the external differences such as the 3 rings, dual SC card, XLR input, etc.
 
The X900 uses pixel shift as well and also has a 5 axis stabilizer. When shooting in 1080 60p, it wouldn't surprise me at all for the resolution to be about the same but I don't know if the chips of the X900 are BSI so the low light capabilities of the AC90 might be a little better. Heck since the chips are newer and in a more expensive camera body, it wouldn't surprise me if the resolution is higher as well. Who knows. Without question, the 1080 24p mode is way better on the AC90 since not only are you starting with a higher bit rate, it's also native 24p from the very beginning.
 
Last edited:
Even though the chip size is different, it looks like the effective pixels are the same as is the size of each individual photo site appears close, as well the lens focal length and zoom. It would not surprise me if the measured resolution was the same.
 
You must log in or register to reply here.
Back
Top