PCI-1422 Image Distortion: Geometry Is Mismatched, Not PNG

After the acquisition geometry matches the PCI-1422 digital input and the Pulnix 9701 timing, circles remain circular in MAX, the LabVIEW 6.1 image, and the saved PNG. The PNG writer does not choose the capture dimensions; it stores the pixel array delivered by IMAQ. Follow the data from the camera to the frame grabber, through the camera file, into the IMAQ image, and finally to the viewer.

Where does the image geometry first change?

The camera sends a digital raster to the PCI-1422. IMAQ interprets its line and frame timing according to the selected camera file, then presents the acquired array to MAX or the LabVIEW VI. The VI creates an IMAQ image and writes that array to PNG. Because the image already appears flattened in MAX, the fault precedes the VI and PNG encoder.

Path point	Observed or configured geometry	Diagnostic meaning
Camera specification	768 x 484	Stated capture size of the Pulnix 9701
MAX camera properties	816 x 525 active image size	IMAQ is using geometry that differs from the stated camera size
MAX display	Flattened image	Distortion exists before LabVIEW processing
LabVIEW PNG	Flattened image	The file reproduces the acquired array rather than introducing the fault
TV monitor	Correct proportions	The scene and camera optics are not the primary suspects

Acquire a stationary circle or square in MAX and inspect it before opening the VI. If the dimensions are wrong there, stop at the acquisition configuration and do not modify PNG settings.

Does the physical acquisition path match the selected interface?

Layer one first. The installed board is a PCI-1422, and the intended path uses the camera's digital output. Advice for analog boards such as the PCI-1407, , , and does not define the correct setup for this path. Those analog configurations were associated with an expected image no larger than 640 x 480; the stated 768 x 484 capture applies to the digital-camera path through a PCI-1422 or .

Interface case	Camera-file direction	Geometry mentioned
Analog monochrome input	Standard RS-170; CCIR in Europe	Up to 640 x 480 for the listed analog boards
Analog color input	NTSC; PAL in Europe	Standard analog-video acquisition
Pulnix 9701 digital output to PCI-1422	Camera-specific digital configuration	Camera specification states 768 x 484

Trace the actual cable from the camera's digital output to the PCI-1422 and confirm that MAX is bound to that board and camera entry. The passing check is a live acquisition from the intended PCI-1422 channel, not from an analog or placeholder configuration.

Does 816 x 525 describe pixels being displayed or timing being acquired?

A digital frame contains active image data plus timing-defined regions that the acquisition configuration may count, crop, or offset. A camera file tells the frame grabber how to interpret line length, frame height, active width, active height, and starting position. The camera's advertised imaging area and a camera-file timing extent are therefore different concepts, but MAX labeling 816 x 525 as the active image size requires investigation because it does not match the stated 768 x 484 image.

Flattening occurs when the acquired array's width-to-height relationship does not represent the intended active raster, or when a viewer stretches the array to a different display rectangle. The relevant ratios are directly derived:

Geometry	Width-to-height ratio	Comparison
768 x 484	768 / 484 = 1.587	Camera specification
816 x 525	816 / 525 = 1.554	MAX property value

The ratios differ, so treating one raster as the other can alter apparent shape. Read the acquired image width and height from MAX or IMAQ rather than inferring them from the display-window size. The check passes only when the returned array dimensions and the configured active region agree with the intended camera output.

How should the PCI-1422 camera configuration be corrected?

The camera-specific file is the first configuration to audit. Its name alone does not prove that its active geometry matches the camera's selected output mode. A mismatched revision, an incorrect mode within the file, or an altered acquisition region can make the board interpret timing outside the intended 768 x 484 area.

1. Open the Pulnix 9701 entry associated with the PCI-1422 in MAX.
2. Record the current active width, active height, horizontal offset, vertical offset, and any camera mode exposed by the configuration. Do not use the display window's dimensions as acquisition values.
3. Compare those fields with the camera's documented digital output mode. For this installation, resolve why the properties report 816 x 525 while the camera specification reports 768 x 484.
4. Select the camera-file mode that matches the connected digital output. If the correct mode is unavailable or the properties cannot be edited, replace or repair the camera configuration through the installed IMAQ configuration mechanism rather than substituting an RS-170 file intended for analog input.
5. Acquire a new frame in MAX and read the returned array dimensions.

The reported message Could not open file indicates that the attempted configuration or display-setting operation did not load its target file. Treat that as a separate configuration-access failure: verify the selected file exists, can be opened by the installed software, and belongs to the intended camera entry. The passing check is a fresh MAX frame acquired with the corrected digital configuration and without that file-open error.

Can display depth or PNG encoding cause this symptom?

Display depth is worth checking when MAX shows corrupted colors or fails to render an image; 24-bit or 32-bit PC display settings were identified for that class of display problem. Color depth does not correct a width-to-height mismatch. Changing it cannot reconcile 816 x 525 with 768 x 484.

PNG records a rectangular pixel grid. It does not know the physical pitch or intended aspect of the source camera unless downstream software applies separate metadata or display rules. A viewer can still stretch a valid PNG to fit its window, so test both the stored pixel dimensions and the rendered proportions.

1. Save the unprocessed IMAQ image without resize, resample, or display-fit operations.
2. Read the PNG's stored width and height in the viewing application.
3. Display it at native size or with aspect-ratio locking enabled.
4. Compare the same target in MAX and the PNG viewer.

If MAX is distorted before saving, return to the camera file. If MAX is correct but only one viewer is distorted, correct that viewer's fit or aspect setting. The check passes when the PNG dimensions equal the IMAQ array dimensions and two independent native-size displays show the same proportions.

How is the correction verified end to end?

Use a target with known equal axes, such as a circle, and keep the camera and target fixed throughout the test. Measure in pixels rather than judging a resized window.

1. Acquire the target in MAX through the Pulnix 9701 digital connection and PCI-1422.
2. Confirm the live image uses the intended active raster and that the circle's horizontal and vertical diameters match within the measurement tolerance of the target and imaging setup.
3. Run the LabVIEW 6.1 VI without geometric processing between acquisition and PNG write.
4. Confirm the IMAQ array dimensions match the saved PNG dimensions.
5. Open the PNG at native size with aspect-ratio locking and repeat the horizontal-to-vertical diameter measurement.

The final acceptance check is agreement among the MAX acquisition, the IMAQ array, and the native-size PNG: identical pixel dimensions through the software path and no change from circles to ovals.

FAQ